diff options
154 files changed, 4286 insertions, 3211 deletions
diff --git a/arch/alpha/include/asm/scatterlist.h b/arch/alpha/include/asm/scatterlist.h index 5728c52..017d747 100644 --- a/arch/alpha/include/asm/scatterlist.h +++ b/arch/alpha/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (~0UL) - #endif /* !(_ALPHA_SCATTERLIST_H) */ diff --git a/arch/avr32/include/asm/scatterlist.h b/arch/avr32/include/asm/scatterlist.h index 06394e5..a5902d9 100644 --- a/arch/avr32/include/asm/scatterlist.h +++ b/arch/avr32/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0xffffffff) - #endif /* __ASM_AVR32_SCATTERLIST_H */ diff --git a/arch/blackfin/include/asm/scatterlist.h b/arch/blackfin/include/asm/scatterlist.h index 64d41d3..d177a15 100644 --- a/arch/blackfin/include/asm/scatterlist.h +++ b/arch/blackfin/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0xffffffff) - #endif /* !(_BLACKFIN_SCATTERLIST_H) */ diff --git a/arch/cris/include/asm/scatterlist.h b/arch/cris/include/asm/scatterlist.h index 249a784..f11f8f4 100644 --- a/arch/cris/include/asm/scatterlist.h +++ b/arch/cris/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0x1fffffff) - #endif /* !(__ASM_CRIS_SCATTERLIST_H) */ diff --git a/arch/frv/include/asm/scatterlist.h b/arch/frv/include/asm/scatterlist.h index 1614bfd..0e5eb30 100644 --- a/arch/frv/include/asm/scatterlist.h +++ b/arch/frv/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0xffffffffUL) - #endif /* !_ASM_SCATTERLIST_H */ diff --git a/arch/h8300/include/asm/scatterlist.h b/arch/h8300/include/asm/scatterlist.h index de08a4a..82130ed 100644 --- a/arch/h8300/include/asm/scatterlist.h +++ b/arch/h8300/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0xffffffff) - #endif /* !(_H8300_SCATTERLIST_H) */ diff --git a/arch/ia64/include/asm/scatterlist.h b/arch/ia64/include/asm/scatterlist.h index f299a4f..08fd93b 100644 --- a/arch/ia64/include/asm/scatterlist.h +++ b/arch/ia64/include/asm/scatterlist.h @@ -2,15 +2,6 @@ #define _ASM_IA64_SCATTERLIST_H #include <asm-generic/scatterlist.h> -/* - * It used to be that ISA_DMA_THRESHOLD had something to do with the - * DMA-limits of ISA-devices. Nowadays, its only remaining use (apart - * from the aha1542.c driver, which isn't 64-bit clean anyhow) is to - * tell the block-layer (via BLK_BOUNCE_ISA) what the max. physical - * address of a page is that is allocated with GFP_DMA. On IA-64, - * that's 4GB - 1. - */ -#define ISA_DMA_THRESHOLD 0xffffffff #define ARCH_HAS_SG_CHAIN #endif /* _ASM_IA64_SCATTERLIST_H */ diff --git a/arch/m32r/include/asm/scatterlist.h b/arch/m32r/include/asm/scatterlist.h index aeeddd8..7370b8b 100644 --- a/arch/m32r/include/asm/scatterlist.h +++ b/arch/m32r/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0x1fffffff) - #endif /* _ASM_M32R_SCATTERLIST_H */ diff --git a/arch/m68k/include/asm/scatterlist.h b/arch/m68k/include/asm/scatterlist.h index 175da06..3125054 100644 --- a/arch/m68k/include/asm/scatterlist.h +++ b/arch/m68k/include/asm/scatterlist.h @@ -3,7 +3,4 @@ #include <asm-generic/scatterlist.h> -/* This is bogus and should go away. */ -#define ISA_DMA_THRESHOLD (0x00ffffff) - #endif /* !(_M68K_SCATTERLIST_H) */ diff --git a/arch/microblaze/include/asm/scatterlist.h b/arch/microblaze/include/asm/scatterlist.h index dc4a890..35d786f 100644 --- a/arch/microblaze/include/asm/scatterlist.h +++ b/arch/microblaze/include/asm/scatterlist.h @@ -1,3 +1 @@ #include <asm-generic/scatterlist.h> - -#define ISA_DMA_THRESHOLD (~0UL) diff --git a/arch/mips/include/asm/scatterlist.h b/arch/mips/include/asm/scatterlist.h index 9af65e7..7ee0e64 100644 --- a/arch/mips/include/asm/scatterlist.h +++ b/arch/mips/include/asm/scatterlist.h @@ -3,6 +3,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0x00ffffffUL) - #endif /* __ASM_SCATTERLIST_H */ diff --git a/arch/mn10300/include/asm/scatterlist.h b/arch/mn10300/include/asm/scatterlist.h index 7bd00b9..7baa400 100644 --- a/arch/mn10300/include/asm/scatterlist.h +++ b/arch/mn10300/include/asm/scatterlist.h @@ -13,6 +13,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0x00ffffff) - #endif /* _ASM_SCATTERLIST_H */ diff --git a/arch/parisc/include/asm/scatterlist.h b/arch/parisc/include/asm/scatterlist.h index 2c3b79b..8bf1f0d 100644 --- a/arch/parisc/include/asm/scatterlist.h +++ b/arch/parisc/include/asm/scatterlist.h @@ -5,7 +5,6 @@ #include <asm/types.h> #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (~0UL) #define sg_virt_addr(sg) ((unsigned long)sg_virt(sg)) #endif /* _ASM_PARISC_SCATTERLIST_H */ diff --git a/arch/powerpc/include/asm/scatterlist.h b/arch/powerpc/include/asm/scatterlist.h index 34cc78f..de1f620 100644 --- a/arch/powerpc/include/asm/scatterlist.h +++ b/arch/powerpc/include/asm/scatterlist.h @@ -12,9 +12,6 @@ #include <asm/dma.h> #include <asm-generic/scatterlist.h> -#ifdef __powerpc64__ -#define ISA_DMA_THRESHOLD (~0UL) -#endif #define ARCH_HAS_SG_CHAIN #endif /* _ASM_POWERPC_SCATTERLIST_H */ diff --git a/arch/s390/include/asm/scatterlist.h b/arch/s390/include/asm/scatterlist.h index be44d94..35d786f 100644 --- a/arch/s390/include/asm/scatterlist.h +++ b/arch/s390/include/asm/scatterlist.h @@ -1,3 +1 @@ -#define ISA_DMA_THRESHOLD (~0UL) - #include <asm-generic/scatterlist.h> diff --git a/arch/score/include/asm/scatterlist.h b/arch/score/include/asm/scatterlist.h index 4fa1a66..9f533b8 100644 --- a/arch/score/include/asm/scatterlist.h +++ b/arch/score/include/asm/scatterlist.h @@ -1,8 +1,6 @@ #ifndef _ASM_SCORE_SCATTERLIST_H #define _ASM_SCORE_SCATTERLIST_H -#define ISA_DMA_THRESHOLD (~0UL) - #include <asm-generic/scatterlist.h> #endif /* _ASM_SCORE_SCATTERLIST_H */ diff --git a/arch/sh/include/asm/scatterlist.h b/arch/sh/include/asm/scatterlist.h index e38d1d4..98dfc35 100644 --- a/arch/sh/include/asm/scatterlist.h +++ b/arch/sh/include/asm/scatterlist.h @@ -1,8 +1,6 @@ #ifndef __ASM_SH_SCATTERLIST_H #define __ASM_SH_SCATTERLIST_H -#define ISA_DMA_THRESHOLD phys_addr_mask() - #include <asm-generic/scatterlist.h> #endif /* __ASM_SH_SCATTERLIST_H */ diff --git a/arch/sparc/include/asm/scatterlist.h b/arch/sparc/include/asm/scatterlist.h index 433e45f..92bb638 100644 --- a/arch/sparc/include/asm/scatterlist.h +++ b/arch/sparc/include/asm/scatterlist.h @@ -3,7 +3,6 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (~0UL) #define ARCH_HAS_SG_CHAIN #endif /* !(_SPARC_SCATTERLIST_H) */ diff --git a/arch/um/drivers/ubd_kern.c b/arch/um/drivers/ubd_kern.c index da992a3..1bcd208 100644 --- a/arch/um/drivers/ubd_kern.c +++ b/arch/um/drivers/ubd_kern.c @@ -33,6 +33,7 @@ #include "linux/mm.h" #include "linux/slab.h" #include "linux/vmalloc.h" +#include "linux/smp_lock.h" #include "linux/blkpg.h" #include "linux/genhd.h" #include "linux/spinlock.h" @@ -1098,6 +1099,7 @@ static int ubd_open(struct block_device *bdev, fmode_t mode) struct ubd *ubd_dev = disk->private_data; int err = 0; + lock_kernel(); if(ubd_dev->count == 0){ err = ubd_open_dev(ubd_dev); if(err){ @@ -1115,7 +1117,8 @@ static int ubd_open(struct block_device *bdev, fmode_t mode) if(--ubd_dev->count == 0) ubd_close_dev(ubd_dev); err = -EROFS; }*/ - out: +out: + unlock_kernel(); return err; } @@ -1123,8 +1126,10 @@ static int ubd_release(struct gendisk *disk, fmode_t mode) { struct ubd *ubd_dev = disk->private_data; + lock_kernel(); if(--ubd_dev->count == 0) ubd_close_dev(ubd_dev); + unlock_kernel(); return 0; } diff --git a/arch/x86/include/asm/scatterlist.h b/arch/x86/include/asm/scatterlist.h index fb0b187..4240878 100644 --- a/arch/x86/include/asm/scatterlist.h +++ b/arch/x86/include/asm/scatterlist.h @@ -3,7 +3,6 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (0x00ffffff) #define ARCH_HAS_SG_CHAIN #endif /* _ASM_X86_SCATTERLIST_H */ diff --git a/arch/xtensa/include/asm/scatterlist.h b/arch/xtensa/include/asm/scatterlist.h index b1f9fdc..a0421a6 100644 --- a/arch/xtensa/include/asm/scatterlist.h +++ b/arch/xtensa/include/asm/scatterlist.h @@ -13,6 +13,4 @@ #include <asm-generic/scatterlist.h> -#define ISA_DMA_THRESHOLD (~0UL) - #endif /* _XTENSA_SCATTERLIST_H */ diff --git a/block/blk-barrier.c b/block/blk-barrier.c index 0d710c9..f0faefc 100644 --- a/block/blk-barrier.c +++ b/block/blk-barrier.c @@ -13,7 +13,6 @@ * blk_queue_ordered - does this queue support ordered writes * @q: the request queue * @ordered: one of QUEUE_ORDERED_* - * @prepare_flush_fn: rq setup helper for cache flush ordered writes * * Description: * For journalled file systems, doing ordered writes on a commit @@ -22,15 +21,8 @@ * feature should call this function and indicate so. * **/ -int blk_queue_ordered(struct request_queue *q, unsigned ordered, - prepare_flush_fn *prepare_flush_fn) +int blk_queue_ordered(struct request_queue *q, unsigned ordered) { - if (!prepare_flush_fn && (ordered & (QUEUE_ORDERED_DO_PREFLUSH | - QUEUE_ORDERED_DO_POSTFLUSH))) { - printk(KERN_ERR "%s: prepare_flush_fn required\n", __func__); - return -EINVAL; - } - if (ordered != QUEUE_ORDERED_NONE && ordered != QUEUE_ORDERED_DRAIN && ordered != QUEUE_ORDERED_DRAIN_FLUSH && @@ -44,7 +36,6 @@ int blk_queue_ordered(struct request_queue *q, unsigned ordered, q->ordered = ordered; q->next_ordered = ordered; - q->prepare_flush_fn = prepare_flush_fn; return 0; } @@ -79,7 +70,7 @@ unsigned blk_ordered_req_seq(struct request *rq) * * http://thread.gmane.org/gmane.linux.kernel/537473 */ - if (!blk_fs_request(rq)) + if (rq->cmd_type != REQ_TYPE_FS) return QUEUE_ORDSEQ_DRAIN; if ((rq->cmd_flags & REQ_ORDERED_COLOR) == @@ -143,10 +134,10 @@ static void queue_flush(struct request_queue *q, unsigned which) } blk_rq_init(q, rq); - rq->cmd_flags = REQ_HARDBARRIER; - rq->rq_disk = q->bar_rq.rq_disk; + rq->cmd_type = REQ_TYPE_FS; + rq->cmd_flags = REQ_HARDBARRIER | REQ_FLUSH; + rq->rq_disk = q->orig_bar_rq->rq_disk; rq->end_io = end_io; - q->prepare_flush_fn(q, rq); elv_insert(q, rq, ELEVATOR_INSERT_FRONT); } @@ -203,7 +194,7 @@ static inline bool start_ordered(struct request_queue *q, struct request **rqp) /* initialize proxy request and queue it */ blk_rq_init(q, rq); if (bio_data_dir(q->orig_bar_rq->bio) == WRITE) - rq->cmd_flags |= REQ_RW; + rq->cmd_flags |= REQ_WRITE; if (q->ordered & QUEUE_ORDERED_DO_FUA) rq->cmd_flags |= REQ_FUA; init_request_from_bio(rq, q->orig_bar_rq->bio); @@ -236,7 +227,8 @@ static inline bool start_ordered(struct request_queue *q, struct request **rqp) bool blk_do_ordered(struct request_queue *q, struct request **rqp) { struct request *rq = *rqp; - const int is_barrier = blk_fs_request(rq) && blk_barrier_rq(rq); + const int is_barrier = rq->cmd_type == REQ_TYPE_FS && + (rq->cmd_flags & REQ_HARDBARRIER); if (!q->ordseq) { if (!is_barrier) @@ -261,7 +253,7 @@ bool blk_do_ordered(struct request_queue *q, struct request **rqp) */ /* Special requests are not subject to ordering rules. */ - if (!blk_fs_request(rq) && + if (rq->cmd_type != REQ_TYPE_FS && rq != &q->pre_flush_rq && rq != &q->post_flush_rq) return true; @@ -319,6 +311,15 @@ int blkdev_issue_flush(struct block_device *bdev, gfp_t gfp_mask, if (!q) return -ENXIO; + /* + * some block devices may not have their queue correctly set up here + * (e.g. loop device without a backing file) and so issuing a flush + * here will panic. Ensure there is a request function before issuing + * the barrier. + */ + if (!q->make_request_fn) + return -ENXIO; + bio = bio_alloc(gfp_mask, 0); bio->bi_end_io = bio_end_empty_barrier; bio->bi_bdev = bdev; diff --git a/block/blk-core.c b/block/blk-core.c index f0640d7..7da630e 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -184,7 +184,7 @@ void blk_dump_rq_flags(struct request *rq, char *msg) printk(KERN_INFO " bio %p, biotail %p, buffer %p, len %u\n", rq->bio, rq->biotail, rq->buffer, blk_rq_bytes(rq)); - if (blk_pc_request(rq)) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { printk(KERN_INFO " cdb: "); for (bit = 0; bit < BLK_MAX_CDB; bit++) printk("%02x ", rq->cmd[bit]); @@ -608,6 +608,7 @@ blk_init_allocated_queue_node(struct request_queue *q, request_fn_proc *rfn, q->request_fn = rfn; q->prep_rq_fn = NULL; + q->unprep_rq_fn = NULL; q->unplug_fn = generic_unplug_device; q->queue_flags = QUEUE_FLAG_DEFAULT; q->queue_lock = lock; @@ -1135,30 +1136,46 @@ void blk_put_request(struct request *req) } EXPORT_SYMBOL(blk_put_request); +/** + * blk_add_request_payload - add a payload to a request + * @rq: request to update + * @page: page backing the payload + * @len: length of the payload. + * + * This allows to later add a payload to an already submitted request by + * a block driver. The driver needs to take care of freeing the payload + * itself. + * + * Note that this is a quite horrible hack and nothing but handling of + * discard requests should ever use it. + */ +void blk_add_request_payload(struct request *rq, struct page *page, + unsigned int len) +{ + struct bio *bio = rq->bio; + + bio->bi_io_vec->bv_page = page; + bio->bi_io_vec->bv_offset = 0; + bio->bi_io_vec->bv_len = len; + + bio->bi_size = len; + bio->bi_vcnt = 1; + bio->bi_phys_segments = 1; + + rq->__data_len = rq->resid_len = len; + rq->nr_phys_segments = 1; + rq->buffer = bio_data(bio); +} +EXPORT_SYMBOL_GPL(blk_add_request_payload); + void init_request_from_bio(struct request *req, struct bio *bio) { req->cpu = bio->bi_comp_cpu; req->cmd_type = REQ_TYPE_FS; - /* - * Inherit FAILFAST from bio (for read-ahead, and explicit - * FAILFAST). FAILFAST flags are identical for req and bio. - */ - if (bio_rw_flagged(bio, BIO_RW_AHEAD)) + req->cmd_flags |= bio->bi_rw & REQ_COMMON_MASK; + if (bio->bi_rw & REQ_RAHEAD) req->cmd_flags |= REQ_FAILFAST_MASK; - else - req->cmd_flags |= bio->bi_rw & REQ_FAILFAST_MASK; - - if (bio_rw_flagged(bio, BIO_RW_DISCARD)) - req->cmd_flags |= REQ_DISCARD; - if (bio_rw_flagged(bio, BIO_RW_BARRIER)) - req->cmd_flags |= REQ_HARDBARRIER; - if (bio_rw_flagged(bio, BIO_RW_SYNCIO)) - req->cmd_flags |= REQ_RW_SYNC; - if (bio_rw_flagged(bio, BIO_RW_META)) - req->cmd_flags |= REQ_RW_META; - if (bio_rw_flagged(bio, BIO_RW_NOIDLE)) - req->cmd_flags |= REQ_NOIDLE; req->errors = 0; req->__sector = bio->bi_sector; @@ -1181,12 +1198,12 @@ static int __make_request(struct request_queue *q, struct bio *bio) int el_ret; unsigned int bytes = bio->bi_size; const unsigned short prio = bio_prio(bio); - const bool sync = bio_rw_flagged(bio, BIO_RW_SYNCIO); - const bool unplug = bio_rw_flagged(bio, BIO_RW_UNPLUG); + const bool sync = (bio->bi_rw & REQ_SYNC); + const bool unplug = (bio->bi_rw & REQ_UNPLUG); const unsigned int ff = bio->bi_rw & REQ_FAILFAST_MASK; int rw_flags; - if (bio_rw_flagged(bio, BIO_RW_BARRIER) && + if ((bio->bi_rw & REQ_HARDBARRIER) && (q->next_ordered == QUEUE_ORDERED_NONE)) { bio_endio(bio, -EOPNOTSUPP); return 0; @@ -1200,7 +1217,7 @@ static int __make_request(struct request_queue *q, struct bio *bio) spin_lock_irq(q->queue_lock); - if (unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER)) || elv_queue_empty(q)) + if (unlikely((bio->bi_rw & REQ_HARDBARRIER)) || elv_queue_empty(q)) goto get_rq; el_ret = elv_merge(q, &req, bio); @@ -1275,7 +1292,7 @@ get_rq: */ rw_flags = bio_data_dir(bio); if (sync) - rw_flags |= REQ_RW_SYNC; + rw_flags |= REQ_SYNC; /* * Grab a free request. This is might sleep but can not fail. @@ -1464,7 +1481,7 @@ static inline void __generic_make_request(struct bio *bio) goto end_io; } - if (unlikely(!bio_rw_flagged(bio, BIO_RW_DISCARD) && + if (unlikely(!(bio->bi_rw & REQ_DISCARD) && nr_sectors > queue_max_hw_sectors(q))) { printk(KERN_ERR "bio too big device %s (%u > %u)\n", bdevname(bio->bi_bdev, b), @@ -1497,8 +1514,7 @@ static inline void __generic_make_request(struct bio *bio) if (bio_check_eod(bio, nr_sectors)) goto end_io; - if (bio_rw_flagged(bio, BIO_RW_DISCARD) && - !blk_queue_discard(q)) { + if ((bio->bi_rw & REQ_DISCARD) && !blk_queue_discard(q)) { err = -EOPNOTSUPP; goto end_io; } @@ -1583,7 +1599,7 @@ void submit_bio(int rw, struct bio *bio) * If it's a regular read/write or a barrier with data attached, * go through the normal accounting stuff before submission. */ - if (bio_has_data(bio) && !(rw & (1 << BIO_RW_DISCARD))) { + if (bio_has_data(bio) && !(rw & REQ_DISCARD)) { if (rw & WRITE) { count_vm_events(PGPGOUT, count); } else { @@ -1628,6 +1644,9 @@ EXPORT_SYMBOL(submit_bio); */ int blk_rq_check_limits(struct request_queue *q, struct request *rq) { + if (rq->cmd_flags & REQ_DISCARD) + return 0; + if (blk_rq_sectors(rq) > queue_max_sectors(q) || blk_rq_bytes(rq) > queue_max_hw_sectors(q) << 9) { printk(KERN_ERR "%s: over max size limit.\n", __func__); @@ -1796,7 +1815,7 @@ struct request *blk_peek_request(struct request_queue *q) * sees this request (possibly after * requeueing). Notify IO scheduler. */ - if (blk_sorted_rq(rq)) + if (rq->cmd_flags & REQ_SORTED) elv_activate_rq(q, rq); /* @@ -1984,10 +2003,11 @@ bool blk_update_request(struct request *req, int error, unsigned int nr_bytes) * TODO: tj: This is too subtle. It would be better to let * low level drivers do what they see fit. */ - if (blk_fs_request(req)) + if (req->cmd_type == REQ_TYPE_FS) req->errors = 0; - if (error && (blk_fs_request(req) && !(req->cmd_flags & REQ_QUIET))) { + if (error && req->cmd_type == REQ_TYPE_FS && + !(req->cmd_flags & REQ_QUIET)) { printk(KERN_ERR "end_request: I/O error, dev %s, sector %llu\n", req->rq_disk ? req->rq_disk->disk_name : "?", (unsigned long long)blk_rq_pos(req)); @@ -2074,7 +2094,7 @@ bool blk_update_request(struct request *req, int error, unsigned int nr_bytes) req->buffer = bio_data(req->bio); /* update sector only for requests with clear definition of sector */ - if (blk_fs_request(req) || blk_discard_rq(req)) + if (req->cmd_type == REQ_TYPE_FS || (req->cmd_flags & REQ_DISCARD)) req->__sector += total_bytes >> 9; /* mixed attributes always follow the first bio */ @@ -2111,11 +2131,32 @@ static bool blk_update_bidi_request(struct request *rq, int error, blk_update_request(rq->next_rq, error, bidi_bytes)) return true; - add_disk_randomness(rq->rq_disk); + if (blk_queue_add_random(rq->q)) + add_disk_randomness(rq->rq_disk); return false; } +/** + * blk_unprep_request - unprepare a request + * @req: the request + * + * This function makes a request ready for complete resubmission (or + * completion). It happens only after all error handling is complete, + * so represents the appropriate moment to deallocate any resources + * that were allocated to the request in the prep_rq_fn. The queue + * lock is held when calling this. + */ +void blk_unprep_request(struct request *req) +{ + struct request_queue *q = req->q; + + req->cmd_flags &= ~REQ_DONTPREP; + if (q->unprep_rq_fn) + q->unprep_rq_fn(q, req); +} +EXPORT_SYMBOL_GPL(blk_unprep_request); + /* * queue lock must be held */ @@ -2126,11 +2167,15 @@ static void blk_finish_request(struct request *req, int error) BUG_ON(blk_queued_rq(req)); - if (unlikely(laptop_mode) && blk_fs_request(req)) + if (unlikely(laptop_mode) && req->cmd_type == REQ_TYPE_FS) laptop_io_completion(&req->q->backing_dev_info); blk_delete_timer(req); + if (req->cmd_flags & REQ_DONTPREP) + blk_unprep_request(req); + + blk_account_io_done(req); if (req->end_io) @@ -2363,7 +2408,7 @@ void blk_rq_bio_prep(struct request_queue *q, struct request *rq, struct bio *bio) { /* Bit 0 (R/W) is identical in rq->cmd_flags and bio->bi_rw */ - rq->cmd_flags |= bio->bi_rw & REQ_RW; + rq->cmd_flags |= bio->bi_rw & REQ_WRITE; if (bio_has_data(bio)) { rq->nr_phys_segments = bio_phys_segments(q, bio); @@ -2450,6 +2495,8 @@ static void __blk_rq_prep_clone(struct request *dst, struct request *src) { dst->cpu = src->cpu; dst->cmd_flags = (rq_data_dir(src) | REQ_NOMERGE); + if (src->cmd_flags & REQ_DISCARD) + dst->cmd_flags |= REQ_DISCARD; dst->cmd_type = src->cmd_type; dst->__sector = blk_rq_pos(src); dst->__data_len = blk_rq_bytes(src); diff --git a/block/blk-exec.c b/block/blk-exec.c index 49557e9..e1672f1 100644 --- a/block/blk-exec.c +++ b/block/blk-exec.c @@ -57,7 +57,7 @@ void blk_execute_rq_nowait(struct request_queue *q, struct gendisk *bd_disk, __elv_add_request(q, rq, where, 1); __generic_unplug_device(q); /* the queue is stopped so it won't be plugged+unplugged */ - if (blk_pm_resume_request(rq)) + if (rq->cmd_type == REQ_TYPE_PM_RESUME) q->request_fn(q); spin_unlock_irq(q->queue_lock); } diff --git a/block/blk-lib.c b/block/blk-lib.c index d0216b9..c1fc55a 100644 --- a/block/blk-lib.c +++ b/block/blk-lib.c @@ -19,7 +19,6 @@ static void blkdev_discard_end_io(struct bio *bio, int err) if (bio->bi_private) complete(bio->bi_private); - __free_page(bio_page(bio)); bio_put(bio); } @@ -42,8 +41,8 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, struct request_queue *q = bdev_get_queue(bdev); int type = flags & BLKDEV_IFL_BARRIER ? DISCARD_BARRIER : DISCARD_NOBARRIER; + unsigned int max_discard_sectors; struct bio *bio; - struct page *page; int ret = 0; if (!q) @@ -52,36 +51,30 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, if (!blk_queue_discard(q)) return -EOPNOTSUPP; - while (nr_sects && !ret) { - unsigned int sector_size = q->limits.logical_block_size; - unsigned int max_discard_sectors = - min(q->limits.max_discard_sectors, UINT_MAX >> 9); + /* + * Ensure that max_discard_sectors is of the proper + * granularity + */ + max_discard_sectors = min(q->limits.max_discard_sectors, UINT_MAX >> 9); + if (q->limits.discard_granularity) { + unsigned int disc_sects = q->limits.discard_granularity >> 9; + max_discard_sectors &= ~(disc_sects - 1); + } + + while (nr_sects && !ret) { bio = bio_alloc(gfp_mask, 1); - if (!bio) - goto out; + if (!bio) { + ret = -ENOMEM; + break; + } + bio->bi_sector = sector; bio->bi_end_io = blkdev_discard_end_io; bio->bi_bdev = bdev; if (flags & BLKDEV_IFL_WAIT) bio->bi_private = &wait; - /* - * Add a zeroed one-sector payload as that's what - * our current implementations need. If we'll ever need - * more the interface will need revisiting. - */ - page = alloc_page(gfp_mask | __GFP_ZERO); - if (!page) - goto out_free_bio; - if (bio_add_pc_page(q, bio, page, sector_size, 0) < sector_size) - goto out_free_page; - - /* - * And override the bio size - the way discard works we - * touch many more blocks on disk than the actual payload - * length. - */ if (nr_sects > max_discard_sectors) { bio->bi_size = max_discard_sectors << 9; nr_sects -= max_discard_sectors; @@ -103,13 +96,8 @@ int blkdev_issue_discard(struct block_device *bdev, sector_t sector, ret = -EIO; bio_put(bio); } + return ret; -out_free_page: - __free_page(page); -out_free_bio: - bio_put(bio); -out: - return -ENOMEM; } EXPORT_SYMBOL(blkdev_issue_discard); @@ -157,7 +145,7 @@ static void bio_batch_end_io(struct bio *bio, int err) int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector, sector_t nr_sects, gfp_t gfp_mask, unsigned long flags) { - int ret = 0; + int ret; struct bio *bio; struct bio_batch bb; unsigned int sz, issued = 0; @@ -175,11 +163,14 @@ int blkdev_issue_zeroout(struct block_device *bdev, sector_t sector, return ret; } submit: + ret = 0; while (nr_sects != 0) { bio = bio_alloc(gfp_mask, min(nr_sects, (sector_t)BIO_MAX_PAGES)); - if (!bio) + if (!bio) { + ret = -ENOMEM; break; + } bio->bi_sector = sector; bio->bi_bdev = bdev; @@ -198,6 +189,7 @@ submit: if (ret < (sz << 9)) break; } + ret = 0; issued++; submit_bio(WRITE, bio); } diff --git a/block/blk-map.c b/block/blk-map.c index 9083cf0..c65d759 100644 --- a/block/blk-map.c +++ b/block/blk-map.c @@ -307,7 +307,7 @@ int blk_rq_map_kern(struct request_queue *q, struct request *rq, void *kbuf, return PTR_ERR(bio); if (rq_data_dir(rq) == WRITE) - bio->bi_rw |= (1 << BIO_RW); + bio->bi_rw |= (1 << REQ_WRITE); if (do_copy) rq->cmd_flags |= REQ_COPY_USER; diff --git a/block/blk-merge.c b/block/blk-merge.c index 5e7dc99..3b0cd42 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -12,7 +12,6 @@ static unsigned int __blk_recalc_rq_segments(struct request_queue *q, struct bio *bio) { - unsigned int phys_size; struct bio_vec *bv, *bvprv = NULL; int cluster, i, high, highprv = 1; unsigned int seg_size, nr_phys_segs; @@ -24,7 +23,7 @@ static unsigned int __blk_recalc_rq_segments(struct request_queue *q, fbio = bio; cluster = test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags); seg_size = 0; - phys_size = nr_phys_segs = 0; + nr_phys_segs = 0; for_each_bio(bio) { bio_for_each_segment(bv, bio, i) { /* @@ -180,7 +179,7 @@ new_segment: } if (q->dma_drain_size && q->dma_drain_needed(rq)) { - if (rq->cmd_flags & REQ_RW) + if (rq->cmd_flags & REQ_WRITE) memset(q->dma_drain_buffer, 0, q->dma_drain_size); sg->page_link &= ~0x02; @@ -226,7 +225,7 @@ int ll_back_merge_fn(struct request_queue *q, struct request *req, { unsigned short max_sectors; - if (unlikely(blk_pc_request(req))) + if (unlikely(req->cmd_type == REQ_TYPE_BLOCK_PC)) max_sectors = queue_max_hw_sectors(q); else max_sectors = queue_max_sectors(q); @@ -250,7 +249,7 @@ int ll_front_merge_fn(struct request_queue *q, struct request *req, { unsigned short max_sectors; - if (unlikely(blk_pc_request(req))) + if (unlikely(req->cmd_type == REQ_TYPE_BLOCK_PC)) max_sectors = queue_max_hw_sectors(q); else max_sectors = queue_max_sectors(q); diff --git a/block/blk-settings.c b/block/blk-settings.c index f5ed5a1..a234f4b 100644 --- a/block/blk-settings.c +++ b/block/blk-settings.c @@ -37,6 +37,23 @@ void blk_queue_prep_rq(struct request_queue *q, prep_rq_fn *pfn) EXPORT_SYMBOL(blk_queue_prep_rq); /** + * blk_queue_unprep_rq - set an unprepare_request function for queue + * @q: queue + * @ufn: unprepare_request function + * + * It's possible for a queue to register an unprepare_request callback + * which is invoked before the request is finally completed. The goal + * of the function is to deallocate any data that was allocated in the + * prepare_request callback. + * + */ +void blk_queue_unprep_rq(struct request_queue *q, unprep_rq_fn *ufn) +{ + q->unprep_rq_fn = ufn; +} +EXPORT_SYMBOL(blk_queue_unprep_rq); + +/** * blk_queue_merge_bvec - set a merge_bvec function for queue * @q: queue * @mbfn: merge_bvec_fn diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 306759b..001ab18 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -180,26 +180,36 @@ static ssize_t queue_max_hw_sectors_show(struct request_queue *q, char *page) return queue_var_show(max_hw_sectors_kb, (page)); } -static ssize_t queue_nonrot_show(struct request_queue *q, char *page) -{ - return queue_var_show(!blk_queue_nonrot(q), page); +#define QUEUE_SYSFS_BIT_FNS(name, flag, neg) \ +static ssize_t \ +queue_show_##name(struct request_queue *q, char *page) \ +{ \ + int bit; \ + bit = test_bit(QUEUE_FLAG_##flag, &q->queue_flags); \ + return queue_var_show(neg ? !bit : bit, page); \ +} \ +static ssize_t \ +queue_store_##name(struct request_queue *q, const char *page, size_t count) \ +{ \ + unsigned long val; \ + ssize_t ret; \ + ret = queue_var_store(&val, page, count); \ + if (neg) \ + val = !val; \ + \ + spin_lock_irq(q->queue_lock); \ + if (val) \ + queue_flag_set(QUEUE_FLAG_##flag, q); \ + else \ + queue_flag_clear(QUEUE_FLAG_##flag, q); \ + spin_unlock_irq(q->queue_lock); \ + return ret; \ } -static ssize_t queue_nonrot_store(struct request_queue *q, const char *page, - size_t count) -{ - unsigned long nm; - ssize_t ret = queue_var_store(&nm, page, count); - - spin_lock_irq(q->queue_lock); - if (nm) - queue_flag_clear(QUEUE_FLAG_NONROT, q); - else - queue_flag_set(QUEUE_FLAG_NONROT, q); - spin_unlock_irq(q->queue_lock); - - return ret; -} +QUEUE_SYSFS_BIT_FNS(nonrot, NONROT, 1); +QUEUE_SYSFS_BIT_FNS(random, ADD_RANDOM, 0); +QUEUE_SYSFS_BIT_FNS(iostats, IO_STAT, 0); +#undef QUEUE_SYSFS_BIT_FNS static ssize_t queue_nomerges_show(struct request_queue *q, char *page) { @@ -250,27 +260,6 @@ queue_rq_affinity_store(struct request_queue *q, const char *page, size_t count) return ret; } -static ssize_t queue_iostats_show(struct request_queue *q, char *page) -{ - return queue_var_show(blk_queue_io_stat(q), page); -} - -static ssize_t queue_iostats_store(struct request_queue *q, const char *page, - size_t count) -{ - unsigned long stats; - ssize_t ret = queue_var_store(&stats, page, count); - - spin_lock_irq(q->queue_lock); - if (stats) - queue_flag_set(QUEUE_FLAG_IO_STAT, q); - else - queue_flag_clear(QUEUE_FLAG_IO_STAT, q); - spin_unlock_irq(q->queue_lock); - - return ret; -} - static struct queue_sysfs_entry queue_requests_entry = { .attr = {.name = "nr_requests", .mode = S_IRUGO | S_IWUSR }, .show = queue_requests_show, @@ -352,8 +341,8 @@ static struct queue_sysfs_entry queue_discard_zeroes_data_entry = { static struct queue_sysfs_entry queue_nonrot_entry = { .attr = {.name = "rotational", .mode = S_IRUGO | S_IWUSR }, - .show = queue_nonrot_show, - .store = queue_nonrot_store, + .show = queue_show_nonrot, + .store = queue_store_nonrot, }; static struct queue_sysfs_entry queue_nomerges_entry = { @@ -370,8 +359,14 @@ static struct queue_sysfs_entry queue_rq_affinity_entry = { static struct queue_sysfs_entry queue_iostats_entry = { .attr = {.name = "iostats", .mode = S_IRUGO | S_IWUSR }, - .show = queue_iostats_show, - .store = queue_iostats_store, + .show = queue_show_iostats, + .store = queue_store_iostats, +}; + +static struct queue_sysfs_entry queue_random_entry = { + .attr = {.name = "add_random", .mode = S_IRUGO | S_IWUSR }, + .show = queue_show_random, + .store = queue_store_random, }; static struct attribute *default_attrs[] = { @@ -394,6 +389,7 @@ static struct attribute *default_attrs[] = { &queue_nomerges_entry.attr, &queue_rq_affinity_entry.attr, &queue_iostats_entry.attr, + &queue_random_entry.attr, NULL, }; diff --git a/block/blk.h b/block/blk.h index 5ee3d7e..6e7dc87 100644 --- a/block/blk.h +++ b/block/blk.h @@ -161,8 +161,10 @@ static inline int blk_cpu_to_group(int cpu) */ static inline int blk_do_io_stat(struct request *rq) { - return rq->rq_disk && blk_rq_io_stat(rq) && - (blk_fs_request(rq) || blk_discard_rq(rq)); + return rq->rq_disk && + (rq->cmd_flags & REQ_IO_STAT) && + (rq->cmd_type == REQ_TYPE_FS || + (rq->cmd_flags & REQ_DISCARD)); } #endif diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 7982b83..eb4086f 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -458,7 +458,7 @@ static inline struct cfq_data *cic_to_cfqd(struct cfq_io_context *cic) */ static inline bool cfq_bio_sync(struct bio *bio) { - return bio_data_dir(bio) == READ || bio_rw_flagged(bio, BIO_RW_SYNCIO); + return bio_data_dir(bio) == READ || (bio->bi_rw & REQ_SYNC); } /* @@ -646,9 +646,10 @@ cfq_choose_req(struct cfq_data *cfqd, struct request *rq1, struct request *rq2, return rq1; else if (rq_is_sync(rq2) && !rq_is_sync(rq1)) return rq2; - if (rq_is_meta(rq1) && !rq_is_meta(rq2)) + if ((rq1->cmd_flags & REQ_META) && !(rq2->cmd_flags & REQ_META)) return rq1; - else if (rq_is_meta(rq2) && !rq_is_meta(rq1)) + else if ((rq2->cmd_flags & REQ_META) && + !(rq1->cmd_flags & REQ_META)) return rq2; s1 = blk_rq_pos(rq1); @@ -1484,7 +1485,7 @@ static void cfq_remove_request(struct request *rq) cfqq->cfqd->rq_queued--; cfq_blkiocg_update_io_remove_stats(&(RQ_CFQG(rq))->blkg, rq_data_dir(rq), rq_is_sync(rq)); - if (rq_is_meta(rq)) { + if (rq->cmd_flags & REQ_META) { WARN_ON(!cfqq->meta_pending); cfqq->meta_pending--; } @@ -3176,7 +3177,7 @@ cfq_should_preempt(struct cfq_data *cfqd, struct cfq_queue *new_cfqq, * So both queues are sync. Let the new request get disk time if * it's a metadata request and the current queue is doing regular IO. */ - if (rq_is_meta(rq) && !cfqq->meta_pending) + if ((rq->cmd_flags & REQ_META) && !cfqq->meta_pending) return true; /* @@ -3230,7 +3231,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq, struct cfq_io_context *cic = RQ_CIC(rq); cfqd->rq_queued++; - if (rq_is_meta(rq)) + if (rq->cmd_flags & REQ_META) cfqq->meta_pending++; cfq_update_io_thinktime(cfqd, cic); @@ -3365,7 +3366,8 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq) unsigned long now; now = jiffies; - cfq_log_cfqq(cfqd, cfqq, "complete rqnoidle %d", !!rq_noidle(rq)); + cfq_log_cfqq(cfqd, cfqq, "complete rqnoidle %d", + !!(rq->cmd_flags & REQ_NOIDLE)); cfq_update_hw_tag(cfqd); @@ -3419,11 +3421,12 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq) cfq_slice_expired(cfqd, 1); else if (sync && cfqq_empty && !cfq_close_cooperator(cfqd, cfqq)) { - cfqd->noidle_tree_requires_idle |= !rq_noidle(rq); + cfqd->noidle_tree_requires_idle |= + !(rq->cmd_flags & REQ_NOIDLE); /* * Idling is enabled for SYNC_WORKLOAD. * SYNC_NOIDLE_WORKLOAD idles at the end of the tree - * only if we processed at least one !rq_noidle request + * only if we processed at least one !REQ_NOIDLE request */ if (cfqd->serving_type == SYNC_WORKLOAD || cfqd->noidle_tree_requires_idle diff --git a/block/compat_ioctl.c b/block/compat_ioctl.c index f26051f..d530856 100644 --- a/block/compat_ioctl.c +++ b/block/compat_ioctl.c @@ -535,56 +535,6 @@ out: return err; } -struct compat_blk_user_trace_setup { - char name[32]; - u16 act_mask; - u32 buf_size; - u32 buf_nr; - compat_u64 start_lba; - compat_u64 end_lba; - u32 pid; -}; -#define BLKTRACESETUP32 _IOWR(0x12, 115, struct compat_blk_user_trace_setup) - -static int compat_blk_trace_setup(struct block_device *bdev, char __user *arg) -{ - struct blk_user_trace_setup buts; - struct compat_blk_user_trace_setup cbuts; - struct request_queue *q; - char b[BDEVNAME_SIZE]; - int ret; - - q = bdev_get_queue(bdev); - if (!q) - return -ENXIO; - - if (copy_from_user(&cbuts, arg, sizeof(cbuts))) - return -EFAULT; - - bdevname(bdev, b); - - buts = (struct blk_user_trace_setup) { - .act_mask = cbuts.act_mask, - .buf_size = cbuts.buf_size, - .buf_nr = cbuts.buf_nr, - .start_lba = cbuts.start_lba, - .end_lba = cbuts.end_lba, - .pid = cbuts.pid, - }; - memcpy(&buts.name, &cbuts.name, 32); - - mutex_lock(&bdev->bd_mutex); - ret = do_blk_trace_setup(q, b, bdev->bd_dev, bdev, &buts); - mutex_unlock(&bdev->bd_mutex); - if (ret) - return ret; - - if (copy_to_user(arg, &buts.name, 32)) - return -EFAULT; - - return 0; -} - static int compat_blkdev_driver_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg) { @@ -802,16 +752,10 @@ long compat_blkdev_ioctl(struct file *file, unsigned cmd, unsigned long arg) return compat_put_u64(arg, bdev->bd_inode->i_size); case BLKTRACESETUP32: - lock_kernel(); - ret = compat_blk_trace_setup(bdev, compat_ptr(arg)); - unlock_kernel(); - return ret; case BLKTRACESTART: /* compatible */ case BLKTRACESTOP: /* compatible */ case BLKTRACETEARDOWN: /* compatible */ - lock_kernel(); ret = blk_trace_ioctl(bdev, cmd, compat_ptr(arg)); - unlock_kernel(); return ret; default: if (disk->fops->compat_ioctl) diff --git a/block/elevator.c b/block/elevator.c index 923a913..816a7c8 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -79,8 +79,7 @@ int elv_rq_merge_ok(struct request *rq, struct bio *bio) /* * Don't merge file system requests and discard requests */ - if (bio_rw_flagged(bio, BIO_RW_DISCARD) != - bio_rw_flagged(rq->bio, BIO_RW_DISCARD)) + if ((bio->bi_rw & REQ_DISCARD) != (rq->bio->bi_rw & REQ_DISCARD)) return 0; /* @@ -428,7 +427,8 @@ void elv_dispatch_sort(struct request_queue *q, struct request *rq) list_for_each_prev(entry, &q->queue_head) { struct request *pos = list_entry_rq(entry); - if (blk_discard_rq(rq) != blk_discard_rq(pos)) + if ((rq->cmd_flags & REQ_DISCARD) != + (pos->cmd_flags & REQ_DISCARD)) break; if (rq_data_dir(rq) != rq_data_dir(pos)) break; @@ -558,7 +558,7 @@ void elv_requeue_request(struct request_queue *q, struct request *rq) */ if (blk_account_rq(rq)) { q->in_flight[rq_is_sync(rq)]--; - if (blk_sorted_rq(rq)) + if (rq->cmd_flags & REQ_SORTED) elv_deactivate_rq(q, rq); } @@ -644,7 +644,8 @@ void elv_insert(struct request_queue *q, struct request *rq, int where) break; case ELEVATOR_INSERT_SORT: - BUG_ON(!blk_fs_request(rq) && !blk_discard_rq(rq)); + BUG_ON(rq->cmd_type != REQ_TYPE_FS && + !(rq->cmd_flags & REQ_DISCARD)); rq->cmd_flags |= REQ_SORTED; q->nr_sorted++; if (rq_mergeable(rq)) { @@ -716,7 +717,7 @@ void __elv_add_request(struct request_queue *q, struct request *rq, int where, /* * toggle ordered color */ - if (blk_barrier_rq(rq)) + if (rq->cmd_flags & REQ_HARDBARRIER) q->ordcolor ^= 1; /* @@ -729,7 +730,8 @@ void __elv_add_request(struct request_queue *q, struct request *rq, int where, * this request is scheduling boundary, update * end_sector */ - if (blk_fs_request(rq) || blk_discard_rq(rq)) { + if (rq->cmd_type == REQ_TYPE_FS || + (rq->cmd_flags & REQ_DISCARD)) { q->end_sector = rq_end_sector(rq); q->boundary_rq = rq; } @@ -843,7 +845,8 @@ void elv_completed_request(struct request_queue *q, struct request *rq) */ if (blk_account_rq(rq)) { q->in_flight[rq_is_sync(rq)]--; - if (blk_sorted_rq(rq) && e->ops->elevator_completed_req_fn) + if ((rq->cmd_flags & REQ_SORTED) && + e->ops->elevator_completed_req_fn) e->ops->elevator_completed_req_fn(q, rq); } diff --git a/block/ioctl.c b/block/ioctl.c index e8eb679..09fd7f1 100644 --- a/block/ioctl.c +++ b/block/ioctl.c @@ -163,18 +163,10 @@ int __blkdev_driver_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg) { struct gendisk *disk = bdev->bd_disk; - int ret; if (disk->fops->ioctl) return disk->fops->ioctl(bdev, mode, cmd, arg); - if (disk->fops->locked_ioctl) { - lock_kernel(); - ret = disk->fops->locked_ioctl(bdev, mode, cmd, arg); - unlock_kernel(); - return ret; - } - return -ENOTTY; } /* @@ -185,8 +177,7 @@ int __blkdev_driver_ioctl(struct block_device *bdev, fmode_t mode, EXPORT_SYMBOL_GPL(__blkdev_driver_ioctl); /* - * always keep this in sync with compat_blkdev_ioctl() and - * compat_blkdev_locked_ioctl() + * always keep this in sync with compat_blkdev_ioctl() */ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg) @@ -206,10 +197,8 @@ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, if (ret != -EINVAL && ret != -ENOTTY) return ret; - lock_kernel(); fsync_bdev(bdev); invalidate_bdev(bdev); - unlock_kernel(); return 0; case BLKROSET: @@ -221,9 +210,7 @@ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, return -EACCES; if (get_user(n, (int __user *)(arg))) return -EFAULT; - lock_kernel(); set_device_ro(bdev, n); - unlock_kernel(); return 0; case BLKDISCARD: { @@ -309,14 +296,10 @@ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, bd_release(bdev); return ret; case BLKPG: - lock_kernel(); ret = blkpg_ioctl(bdev, (struct blkpg_ioctl_arg __user *) arg); - unlock_kernel(); break; case BLKRRPART: - lock_kernel(); ret = blkdev_reread_part(bdev); - unlock_kernel(); break; case BLKGETSIZE: size = bdev->bd_inode->i_size; @@ -329,9 +312,7 @@ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, case BLKTRACESTOP: case BLKTRACESETUP: case BLKTRACETEARDOWN: - lock_kernel(); ret = blk_trace_ioctl(bdev, cmd, (char __user *) arg); - unlock_kernel(); break; default: ret = __blkdev_driver_ioctl(bdev, mode, cmd, arg); diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c index d75c9c4..a89172c 100644 --- a/drivers/ata/libata-scsi.c +++ b/drivers/ata/libata-scsi.c @@ -1111,10 +1111,10 @@ static void ata_scsi_sdev_config(struct scsi_device *sdev) */ static int atapi_drain_needed(struct request *rq) { - if (likely(!blk_pc_request(rq))) + if (likely(rq->cmd_type != REQ_TYPE_BLOCK_PC)) return 0; - if (!blk_rq_bytes(rq) || (rq->cmd_flags & REQ_RW)) + if (!blk_rq_bytes(rq) || (rq->cmd_flags & REQ_WRITE)) return 0; return atapi_cmd_type(rq->cmd[0]) == ATAPI_MISC; diff --git a/drivers/block/DAC960.c b/drivers/block/DAC960.c index c5f22bb..4e2c367 100644 --- a/drivers/block/DAC960.c +++ b/drivers/block/DAC960.c @@ -79,23 +79,28 @@ static int DAC960_open(struct block_device *bdev, fmode_t mode) struct gendisk *disk = bdev->bd_disk; DAC960_Controller_T *p = disk->queue->queuedata; int drive_nr = (long)disk->private_data; + int ret = -ENXIO; + lock_kernel(); if (p->FirmwareType == DAC960_V1_Controller) { if (p->V1.LogicalDriveInformation[drive_nr]. LogicalDriveState == DAC960_V1_LogicalDrive_Offline) - return -ENXIO; + goto out; } else { DAC960_V2_LogicalDeviceInfo_T *i = p->V2.LogicalDeviceInformation[drive_nr]; if (!i || i->LogicalDeviceState == DAC960_V2_LogicalDevice_Offline) - return -ENXIO; + goto out; } check_disk_change(bdev); if (!get_capacity(p->disks[drive_nr])) - return -ENXIO; - return 0; + goto out; + ret = 0; +out: + unlock_kernel(); + return ret; } static int DAC960_getgeo(struct block_device *bdev, struct hd_geometry *geo) diff --git a/drivers/block/amiflop.c b/drivers/block/amiflop.c index 832798a..76f114f 100644 --- a/drivers/block/amiflop.c +++ b/drivers/block/amiflop.c @@ -60,6 +60,7 @@ #include <linux/hdreg.h> #include <linux/delay.h> #include <linux/init.h> +#include <linux/smp_lock.h> #include <linux/amifdreg.h> #include <linux/amifd.h> #include <linux/buffer_head.h> @@ -1423,7 +1424,7 @@ static int fd_getgeo(struct block_device *bdev, struct hd_geometry *geo) return 0; } -static int fd_ioctl(struct block_device *bdev, fmode_t mode, +static int fd_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long param) { struct amiga_floppy_struct *p = bdev->bd_disk->private_data; @@ -1500,6 +1501,18 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode, return 0; } +static int fd_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long param) +{ + int ret; + + lock_kernel(); + ret = fd_locked_ioctl(bdev, mode, cmd, param); + unlock_kernel(); + + return ret; +} + static void fd_probe(int dev) { unsigned long code; @@ -1542,10 +1555,13 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) int old_dev; unsigned long flags; + lock_kernel(); old_dev = fd_device[drive]; - if (fd_ref[drive] && old_dev != system) + if (fd_ref[drive] && old_dev != system) { + unlock_kernel(); return -EBUSY; + } if (mode & (FMODE_READ|FMODE_WRITE)) { check_disk_change(bdev); @@ -1558,8 +1574,10 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) fd_deselect (drive); rel_fdc(); - if (wrprot) + if (wrprot) { + unlock_kernel(); return -EROFS; + } } } @@ -1576,6 +1594,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) printk(KERN_INFO "fd%d: accessing %s-disk with %s-layout\n",drive, unit[drive].type->name, data_types[system].name); + unlock_kernel(); return 0; } @@ -1584,6 +1603,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode) struct amiga_floppy_struct *p = disk->private_data; int drive = p - unit; + lock_kernel(); if (unit[drive].dirty == 1) { del_timer (flush_track_timer + drive); non_int_flush_track (drive); @@ -1597,6 +1617,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode) /* the mod_use counter is handled this way */ floppy_off (drive | 0x40000000); #endif + unlock_kernel(); return 0; } @@ -1638,7 +1659,7 @@ static const struct block_device_operations floppy_fops = { .owner = THIS_MODULE, .open = floppy_open, .release = floppy_release, - .locked_ioctl = fd_ioctl, + .ioctl = fd_ioctl, .getgeo = fd_getgeo, .media_changed = amiga_floppy_change, }; diff --git a/drivers/block/aoe/aoeblk.c b/drivers/block/aoe/aoeblk.c index 035cefe..a946929 100644 --- a/drivers/block/aoe/aoeblk.c +++ b/drivers/block/aoe/aoeblk.c @@ -12,6 +12,7 @@ #include <linux/slab.h> #include <linux/genhd.h> #include <linux/netdevice.h> +#include <linux/smp_lock.h> #include "aoe.h" static struct kmem_cache *buf_pool_cache; @@ -124,13 +125,16 @@ aoeblk_open(struct block_device *bdev, fmode_t mode) struct aoedev *d = bdev->bd_disk->private_data; ulong flags; + lock_kernel(); spin_lock_irqsave(&d->lock, flags); if (d->flags & DEVFL_UP) { d->nopen++; spin_unlock_irqrestore(&d->lock, flags); + unlock_kernel(); return 0; } spin_unlock_irqrestore(&d->lock, flags); + unlock_kernel(); return -ENODEV; } @@ -173,7 +177,7 @@ aoeblk_make_request(struct request_queue *q, struct bio *bio) BUG(); bio_endio(bio, -ENXIO); return 0; - } else if (bio_rw_flagged(bio, BIO_RW_BARRIER)) { + } else if (bio->bi_rw & REQ_HARDBARRIER) { bio_endio(bio, -EOPNOTSUPP); return 0; } else if (bio->bi_io_vec == NULL) { diff --git a/drivers/block/ataflop.c b/drivers/block/ataflop.c index e35cf59..aceb964 100644 --- a/drivers/block/ataflop.c +++ b/drivers/block/ataflop.c @@ -67,6 +67,7 @@ #include <linux/delay.h> #include <linux/init.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <asm/atafd.h> #include <asm/atafdreg.h> @@ -359,7 +360,7 @@ static void finish_fdc( void ); static void finish_fdc_done( int dummy ); static void setup_req_params( int drive ); static void redo_fd_request( void); -static int fd_ioctl(struct block_device *bdev, fmode_t mode, unsigned int +static int fd_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long param); static void fd_probe( int drive ); static int fd_test_drive_present( int drive ); @@ -1480,7 +1481,7 @@ void do_fd_request(struct request_queue * q) atari_enable_irq( IRQ_MFP_FDC ); } -static int fd_ioctl(struct block_device *bdev, fmode_t mode, +static int fd_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long param) { struct gendisk *disk = bdev->bd_disk; @@ -1665,6 +1666,17 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode, } } +static int fd_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long arg) +{ + int ret; + + lock_kernel(); + ret = fd_locked_ioctl(bdev, mode, cmd, arg); + unlock_kernel(); + + return ret; +} /* Initialize the 'unit' variable for drive 'drive' */ @@ -1838,24 +1850,36 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) return 0; } +static int floppy_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = floppy_open(bdev, mode); + unlock_kernel(); + + return ret; +} static int floppy_release(struct gendisk *disk, fmode_t mode) { struct atari_floppy_struct *p = disk->private_data; + lock_kernel(); if (p->ref < 0) p->ref = 0; else if (!p->ref--) { printk(KERN_ERR "floppy_release with fd_ref == 0"); p->ref = 0; } + unlock_kernel(); return 0; } static const struct block_device_operations floppy_fops = { .owner = THIS_MODULE, - .open = floppy_open, + .open = floppy_unlocked_open, .release = floppy_release, - .locked_ioctl = fd_ioctl, + .ioctl = fd_ioctl, .media_changed = check_floppy_change, .revalidate_disk= floppy_revalidate, }; diff --git a/drivers/block/brd.c b/drivers/block/brd.c index f1bf79d..1c7f637 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -15,6 +15,7 @@ #include <linux/blkdev.h> #include <linux/bio.h> #include <linux/highmem.h> +#include <linux/smp_lock.h> #include <linux/radix-tree.h> #include <linux/buffer_head.h> /* invalidate_bh_lrus() */ #include <linux/slab.h> @@ -340,7 +341,7 @@ static int brd_make_request(struct request_queue *q, struct bio *bio) get_capacity(bdev->bd_disk)) goto out; - if (unlikely(bio_rw_flagged(bio, BIO_RW_DISCARD))) { + if (unlikely(bio->bi_rw & REQ_DISCARD)) { err = 0; discard_from_brd(brd, sector, bio->bi_size); goto out; @@ -401,6 +402,7 @@ static int brd_ioctl(struct block_device *bdev, fmode_t mode, * ram device BLKFLSBUF has special semantics, we want to actually * release and destroy the ramdisk data. */ + lock_kernel(); mutex_lock(&bdev->bd_mutex); error = -EBUSY; if (bdev->bd_openers <= 1) { @@ -417,13 +419,14 @@ static int brd_ioctl(struct block_device *bdev, fmode_t mode, error = 0; } mutex_unlock(&bdev->bd_mutex); + unlock_kernel(); return error; } static const struct block_device_operations brd_fops = { .owner = THIS_MODULE, - .locked_ioctl = brd_ioctl, + .ioctl = brd_ioctl, #ifdef CONFIG_BLK_DEV_XIP .direct_access = brd_direct_access, #endif @@ -479,7 +482,7 @@ static struct brd_device *brd_alloc(int i) if (!brd->brd_queue) goto out_free_dev; blk_queue_make_request(brd->brd_queue, brd_make_request); - blk_queue_ordered(brd->brd_queue, QUEUE_ORDERED_TAG, NULL); + blk_queue_ordered(brd->brd_queue, QUEUE_ORDERED_TAG); blk_queue_max_hw_sectors(brd->brd_queue, 1024); blk_queue_bounce_limit(brd->brd_queue, BLK_BOUNCE_ANY); diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index e1e7143..31064df 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -56,16 +56,14 @@ #include <linux/kthread.h> #define CCISS_DRIVER_VERSION(maj,min,submin) ((maj<<16)|(min<<8)|(submin)) -#define DRIVER_NAME "HP CISS Driver (v 3.6.20)" -#define DRIVER_VERSION CCISS_DRIVER_VERSION(3, 6, 20) +#define DRIVER_NAME "HP CISS Driver (v 3.6.26)" +#define DRIVER_VERSION CCISS_DRIVER_VERSION(3, 6, 26) /* Embedded module documentation macros - see modules.h */ MODULE_AUTHOR("Hewlett-Packard Company"); MODULE_DESCRIPTION("Driver for HP Smart Array Controllers"); -MODULE_SUPPORTED_DEVICE("HP SA5i SA5i+ SA532 SA5300 SA5312 SA641 SA642 SA6400" - " SA6i P600 P800 P400 P400i E200 E200i E500 P700m" - " Smart Array G2 Series SAS/SATA Controllers"); -MODULE_VERSION("3.6.20"); +MODULE_SUPPORTED_DEVICE("HP Smart Array Controllers"); +MODULE_VERSION("3.6.26"); MODULE_LICENSE("GPL"); static int cciss_allow_hpsa; @@ -107,6 +105,11 @@ static const struct pci_device_id cciss_pci_device_id[] = { {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x3249}, {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x324A}, {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x324B}, + {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x3250}, + {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x3251}, + {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x3252}, + {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x3253}, + {PCI_VENDOR_ID_HP, PCI_DEVICE_ID_HP_CISSE, 0x103C, 0x3254}, {0,} }; @@ -146,6 +149,11 @@ static struct board_type products[] = { {0x3249103C, "Smart Array P812", &SA5_access}, {0x324A103C, "Smart Array P712m", &SA5_access}, {0x324B103C, "Smart Array P711m", &SA5_access}, + {0x3250103C, "Smart Array", &SA5_access}, + {0x3251103C, "Smart Array", &SA5_access}, + {0x3252103C, "Smart Array", &SA5_access}, + {0x3253103C, "Smart Array", &SA5_access}, + {0x3254103C, "Smart Array", &SA5_access}, }; /* How long to wait (in milliseconds) for board to go into simple mode */ @@ -167,9 +175,13 @@ static DEFINE_MUTEX(scan_mutex); static LIST_HEAD(scan_q); static void do_cciss_request(struct request_queue *q); -static irqreturn_t do_cciss_intr(int irq, void *dev_id); +static irqreturn_t do_cciss_intx(int irq, void *dev_id); +static irqreturn_t do_cciss_msix_intr(int irq, void *dev_id); static int cciss_open(struct block_device *bdev, fmode_t mode); +static int cciss_unlocked_open(struct block_device *bdev, fmode_t mode); static int cciss_release(struct gendisk *disk, fmode_t mode); +static int do_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long arg); static int cciss_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg); static int cciss_getgeo(struct block_device *bdev, struct hd_geometry *geo); @@ -179,25 +191,23 @@ static int rebuild_lun_table(ctlr_info_t *h, int first_time, int via_ioctl); static int deregister_disk(ctlr_info_t *h, int drv_index, int clear_all, int via_ioctl); -static void cciss_read_capacity(int ctlr, int logvol, +static void cciss_read_capacity(ctlr_info_t *h, int logvol, sector_t *total_size, unsigned int *block_size); -static void cciss_read_capacity_16(int ctlr, int logvol, +static void cciss_read_capacity_16(ctlr_info_t *h, int logvol, sector_t *total_size, unsigned int *block_size); -static void cciss_geometry_inquiry(int ctlr, int logvol, +static void cciss_geometry_inquiry(ctlr_info_t *h, int logvol, sector_t total_size, unsigned int block_size, InquiryData_struct *inq_buff, drive_info_struct *drv); -static void __devinit cciss_interrupt_mode(ctlr_info_t *, struct pci_dev *, - __u32); +static void __devinit cciss_interrupt_mode(ctlr_info_t *); static void start_io(ctlr_info_t *h); -static int sendcmd_withirq(__u8 cmd, int ctlr, void *buff, size_t size, +static int sendcmd_withirq(ctlr_info_t *h, __u8 cmd, void *buff, size_t size, __u8 page_code, unsigned char scsi3addr[], int cmd_type); static int sendcmd_withirq_core(ctlr_info_t *h, CommandList_struct *c, int attempt_retry); static int process_sendcmd_error(ctlr_info_t *h, CommandList_struct *c); -static void fail_all_cmds(unsigned long ctlr); static int add_to_scan_list(struct ctlr_info *h); static int scan_thread(void *data); static int check_for_unit_attention(ctlr_info_t *h, CommandList_struct *c); @@ -205,11 +215,23 @@ static void cciss_hba_release(struct device *dev); static void cciss_device_release(struct device *dev); static void cciss_free_gendisk(ctlr_info_t *h, int drv_index); static void cciss_free_drive_info(ctlr_info_t *h, int drv_index); +static inline u32 next_command(ctlr_info_t *h); +static int __devinit cciss_find_cfg_addrs(struct pci_dev *pdev, + void __iomem *vaddr, u32 *cfg_base_addr, u64 *cfg_base_addr_index, + u64 *cfg_offset); +static int __devinit cciss_pci_find_memory_BAR(struct pci_dev *pdev, + unsigned long *memory_bar); + + +/* performant mode helper functions */ +static void calc_bucket_map(int *bucket, int num_buckets, int nsgs, + int *bucket_map); +static void cciss_put_controller_into_performant_mode(ctlr_info_t *h); #ifdef CONFIG_PROC_FS -static void cciss_procinit(int i); +static void cciss_procinit(ctlr_info_t *h); #else -static void cciss_procinit(int i) +static void cciss_procinit(ctlr_info_t *h) { } #endif /* CONFIG_PROC_FS */ @@ -221,9 +243,9 @@ static int cciss_compat_ioctl(struct block_device *, fmode_t, static const struct block_device_operations cciss_fops = { .owner = THIS_MODULE, - .open = cciss_open, + .open = cciss_unlocked_open, .release = cciss_release, - .locked_ioctl = cciss_ioctl, + .ioctl = do_ioctl, .getgeo = cciss_getgeo, #ifdef CONFIG_COMPAT .compat_ioctl = cciss_compat_ioctl, @@ -231,6 +253,16 @@ static const struct block_device_operations cciss_fops = { .revalidate_disk = cciss_revalidate, }; +/* set_performant_mode: Modify the tag for cciss performant + * set bit 0 for pull model, bits 3-1 for block fetch + * register number + */ +static void set_performant_mode(ctlr_info_t *h, CommandList_struct *c) +{ + if (likely(h->transMethod == CFGTBL_Trans_Performant)) + c->busaddr |= 1 | (h->blockFetchTable[c->Header.SGList] << 1); +} + /* * Enqueuing and dequeuing functions for cmdlists. */ @@ -257,6 +289,18 @@ static inline void removeQ(CommandList_struct *c) hlist_del_init(&c->list); } +static void enqueue_cmd_and_start_io(ctlr_info_t *h, + CommandList_struct *c) +{ + unsigned long flags; + set_performant_mode(h, c); + spin_lock_irqsave(&h->lock, flags); + addQ(&h->reqQ, c); + h->Qdepth++; + start_io(h); + spin_unlock_irqrestore(&h->lock, flags); +} + static void cciss_free_sg_chain_blocks(SGDescriptor_struct **cmd_sg_list, int nr_cmds) { @@ -366,32 +410,31 @@ static void cciss_seq_show_header(struct seq_file *seq) h->product_name, (unsigned long)h->board_id, h->firm_ver[0], h->firm_ver[1], h->firm_ver[2], - h->firm_ver[3], (unsigned int)h->intr[SIMPLE_MODE_INT], + h->firm_ver[3], (unsigned int)h->intr[PERF_MODE_INT], h->num_luns, h->Qdepth, h->commands_outstanding, h->maxQsinceinit, h->max_outstanding, h->maxSG); #ifdef CONFIG_CISS_SCSI_TAPE - cciss_seq_tape_report(seq, h->ctlr); + cciss_seq_tape_report(seq, h); #endif /* CONFIG_CISS_SCSI_TAPE */ } static void *cciss_seq_start(struct seq_file *seq, loff_t *pos) { ctlr_info_t *h = seq->private; - unsigned ctlr = h->ctlr; unsigned long flags; /* prevent displaying bogus info during configuration * or deconfiguration of a logical volume */ - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) { - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return ERR_PTR(-EBUSY); } h->busy_configuring = 1; - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (*pos == 0) cciss_seq_show_header(seq); @@ -499,7 +542,7 @@ cciss_proc_write(struct file *file, const char __user *buf, struct seq_file *seq = file->private_data; ctlr_info_t *h = seq->private; - err = cciss_engage_scsi(h->ctlr); + err = cciss_engage_scsi(h); if (err == 0) err = length; } else @@ -522,7 +565,7 @@ static const struct file_operations cciss_proc_fops = { .write = cciss_proc_write, }; -static void __devinit cciss_procinit(int i) +static void __devinit cciss_procinit(ctlr_info_t *h) { struct proc_dir_entry *pde; @@ -530,9 +573,9 @@ static void __devinit cciss_procinit(int i) proc_cciss = proc_mkdir("driver/cciss", NULL); if (!proc_cciss) return; - pde = proc_create_data(hba[i]->devname, S_IWUSR | S_IRUSR | S_IRGRP | + pde = proc_create_data(h->devname, S_IWUSR | S_IRUSR | S_IRGRP | S_IROTH, proc_cciss, - &cciss_proc_fops, hba[i]); + &cciss_proc_fops, h); } #endif /* CONFIG_PROC_FS */ @@ -565,12 +608,12 @@ static ssize_t dev_show_unique_id(struct device *dev, unsigned long flags; int ret = 0; - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) ret = -EBUSY; else memcpy(sn, drv->serial_no, sizeof(sn)); - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (ret) return ret; @@ -595,12 +638,12 @@ static ssize_t dev_show_vendor(struct device *dev, unsigned long flags; int ret = 0; - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) ret = -EBUSY; else memcpy(vendor, drv->vendor, VENDOR_LEN + 1); - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (ret) return ret; @@ -619,12 +662,12 @@ static ssize_t dev_show_model(struct device *dev, unsigned long flags; int ret = 0; - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) ret = -EBUSY; else memcpy(model, drv->model, MODEL_LEN + 1); - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (ret) return ret; @@ -643,12 +686,12 @@ static ssize_t dev_show_rev(struct device *dev, unsigned long flags; int ret = 0; - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) ret = -EBUSY; else memcpy(rev, drv->rev, REV_LEN + 1); - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (ret) return ret; @@ -665,17 +708,17 @@ static ssize_t cciss_show_lunid(struct device *dev, unsigned long flags; unsigned char lunid[8]; - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) { - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return -EBUSY; } if (!drv->heads) { - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return -ENOTTY; } memcpy(lunid, drv->LunID, sizeof(lunid)); - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return snprintf(buf, 20, "0x%02x%02x%02x%02x%02x%02x%02x%02x\n", lunid[0], lunid[1], lunid[2], lunid[3], lunid[4], lunid[5], lunid[6], lunid[7]); @@ -690,13 +733,13 @@ static ssize_t cciss_show_raid_level(struct device *dev, int raid; unsigned long flags; - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) { - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return -EBUSY; } raid = drv->raid_level; - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (raid < 0 || raid > RAID_UNKNOWN) raid = RAID_UNKNOWN; @@ -713,13 +756,13 @@ static ssize_t cciss_show_usage_count(struct device *dev, unsigned long flags; int count; - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) { - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return -EBUSY; } count = drv->usage_count; - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return snprintf(buf, 20, "%d\n", count); } static DEVICE_ATTR(usage_count, S_IRUGO, cciss_show_usage_count, NULL); @@ -864,60 +907,70 @@ static void cciss_destroy_ld_sysfs_entry(struct ctlr_info *h, int drv_index, /* * For operations that cannot sleep, a command block is allocated at init, * and managed by cmd_alloc() and cmd_free() using a simple bitmap to track - * which ones are free or in use. For operations that can wait for kmalloc - * to possible sleep, this routine can be called with get_from_pool set to 0. - * cmd_free() MUST be called with a got_from_pool set to 0 if cmd_alloc was. + * which ones are free or in use. */ -static CommandList_struct *cmd_alloc(ctlr_info_t *h, int get_from_pool) +static CommandList_struct *cmd_alloc(ctlr_info_t *h) { CommandList_struct *c; int i; u64bit temp64; dma_addr_t cmd_dma_handle, err_dma_handle; - if (!get_from_pool) { - c = (CommandList_struct *) pci_alloc_consistent(h->pdev, - sizeof(CommandList_struct), &cmd_dma_handle); - if (c == NULL) + do { + i = find_first_zero_bit(h->cmd_pool_bits, h->nr_cmds); + if (i == h->nr_cmds) return NULL; - memset(c, 0, sizeof(CommandList_struct)); + } while (test_and_set_bit(i & (BITS_PER_LONG - 1), + h->cmd_pool_bits + (i / BITS_PER_LONG)) != 0); + c = h->cmd_pool + i; + memset(c, 0, sizeof(CommandList_struct)); + cmd_dma_handle = h->cmd_pool_dhandle + i * sizeof(CommandList_struct); + c->err_info = h->errinfo_pool + i; + memset(c->err_info, 0, sizeof(ErrorInfo_struct)); + err_dma_handle = h->errinfo_pool_dhandle + + i * sizeof(ErrorInfo_struct); + h->nr_allocs++; - c->cmdindex = -1; + c->cmdindex = i; - c->err_info = (ErrorInfo_struct *) - pci_alloc_consistent(h->pdev, sizeof(ErrorInfo_struct), - &err_dma_handle); + INIT_HLIST_NODE(&c->list); + c->busaddr = (__u32) cmd_dma_handle; + temp64.val = (__u64) err_dma_handle; + c->ErrDesc.Addr.lower = temp64.val32.lower; + c->ErrDesc.Addr.upper = temp64.val32.upper; + c->ErrDesc.Len = sizeof(ErrorInfo_struct); - if (c->err_info == NULL) { - pci_free_consistent(h->pdev, - sizeof(CommandList_struct), c, cmd_dma_handle); - return NULL; - } - memset(c->err_info, 0, sizeof(ErrorInfo_struct)); - } else { /* get it out of the controllers pool */ - - do { - i = find_first_zero_bit(h->cmd_pool_bits, h->nr_cmds); - if (i == h->nr_cmds) - return NULL; - } while (test_and_set_bit - (i & (BITS_PER_LONG - 1), - h->cmd_pool_bits + (i / BITS_PER_LONG)) != 0); -#ifdef CCISS_DEBUG - printk(KERN_DEBUG "cciss: using command buffer %d\n", i); -#endif - c = h->cmd_pool + i; - memset(c, 0, sizeof(CommandList_struct)); - cmd_dma_handle = h->cmd_pool_dhandle - + i * sizeof(CommandList_struct); - c->err_info = h->errinfo_pool + i; - memset(c->err_info, 0, sizeof(ErrorInfo_struct)); - err_dma_handle = h->errinfo_pool_dhandle - + i * sizeof(ErrorInfo_struct); - h->nr_allocs++; + c->ctlr = h->ctlr; + return c; +} - c->cmdindex = i; +/* allocate a command using pci_alloc_consistent, used for ioctls, + * etc., not for the main i/o path. + */ +static CommandList_struct *cmd_special_alloc(ctlr_info_t *h) +{ + CommandList_struct *c; + u64bit temp64; + dma_addr_t cmd_dma_handle, err_dma_handle; + + c = (CommandList_struct *) pci_alloc_consistent(h->pdev, + sizeof(CommandList_struct), &cmd_dma_handle); + if (c == NULL) + return NULL; + memset(c, 0, sizeof(CommandList_struct)); + + c->cmdindex = -1; + + c->err_info = (ErrorInfo_struct *) + pci_alloc_consistent(h->pdev, sizeof(ErrorInfo_struct), + &err_dma_handle); + + if (c->err_info == NULL) { + pci_free_consistent(h->pdev, + sizeof(CommandList_struct), c, cmd_dma_handle); + return NULL; } + memset(c->err_info, 0, sizeof(ErrorInfo_struct)); INIT_HLIST_NODE(&c->list); c->busaddr = (__u32) cmd_dma_handle; @@ -930,27 +983,26 @@ static CommandList_struct *cmd_alloc(ctlr_info_t *h, int get_from_pool) return c; } -/* - * Frees a command block that was previously allocated with cmd_alloc(). - */ -static void cmd_free(ctlr_info_t *h, CommandList_struct *c, int got_from_pool) +static void cmd_free(ctlr_info_t *h, CommandList_struct *c) { int i; + + i = c - h->cmd_pool; + clear_bit(i & (BITS_PER_LONG - 1), + h->cmd_pool_bits + (i / BITS_PER_LONG)); + h->nr_frees++; +} + +static void cmd_special_free(ctlr_info_t *h, CommandList_struct *c) +{ u64bit temp64; - if (!got_from_pool) { - temp64.val32.lower = c->ErrDesc.Addr.lower; - temp64.val32.upper = c->ErrDesc.Addr.upper; - pci_free_consistent(h->pdev, sizeof(ErrorInfo_struct), - c->err_info, (dma_addr_t) temp64.val); - pci_free_consistent(h->pdev, sizeof(CommandList_struct), - c, (dma_addr_t) c->busaddr); - } else { - i = c - h->cmd_pool; - clear_bit(i & (BITS_PER_LONG - 1), - h->cmd_pool_bits + (i / BITS_PER_LONG)); - h->nr_frees++; - } + temp64.val32.lower = c->ErrDesc.Addr.lower; + temp64.val32.upper = c->ErrDesc.Addr.upper; + pci_free_consistent(h->pdev, sizeof(ErrorInfo_struct), + c->err_info, (dma_addr_t) temp64.val); + pci_free_consistent(h->pdev, sizeof(CommandList_struct), + c, (dma_addr_t) c->busaddr); } static inline ctlr_info_t *get_host(struct gendisk *disk) @@ -968,13 +1020,10 @@ static inline drive_info_struct *get_drv(struct gendisk *disk) */ static int cciss_open(struct block_device *bdev, fmode_t mode) { - ctlr_info_t *host = get_host(bdev->bd_disk); + ctlr_info_t *h = get_host(bdev->bd_disk); drive_info_struct *drv = get_drv(bdev->bd_disk); -#ifdef CCISS_DEBUG - printk(KERN_DEBUG "cciss_open %s\n", bdev->bd_disk->disk_name); -#endif /* CCISS_DEBUG */ - + dev_dbg(&h->pdev->dev, "cciss_open %s\n", bdev->bd_disk->disk_name); if (drv->busy_configuring) return -EBUSY; /* @@ -1000,29 +1049,39 @@ static int cciss_open(struct block_device *bdev, fmode_t mode) return -EPERM; } drv->usage_count++; - host->usage_count++; + h->usage_count++; return 0; } +static int cciss_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = cciss_open(bdev, mode); + unlock_kernel(); + + return ret; +} + /* * Close. Sync first. */ static int cciss_release(struct gendisk *disk, fmode_t mode) { - ctlr_info_t *host = get_host(disk); - drive_info_struct *drv = get_drv(disk); - -#ifdef CCISS_DEBUG - printk(KERN_DEBUG "cciss_release %s\n", disk->disk_name); -#endif /* CCISS_DEBUG */ + ctlr_info_t *h; + drive_info_struct *drv; + lock_kernel(); + h = get_host(disk); + drv = get_drv(disk); + dev_dbg(&h->pdev->dev, "cciss_release %s\n", disk->disk_name); drv->usage_count--; - host->usage_count--; + h->usage_count--; + unlock_kernel(); return 0; } -#ifdef CONFIG_COMPAT - static int do_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg) { @@ -1033,6 +1092,8 @@ static int do_ioctl(struct block_device *bdev, fmode_t mode, return ret; } +#ifdef CONFIG_COMPAT + static int cciss_ioctl32_passthru(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg); static int cciss_ioctl32_big_passthru(struct block_device *bdev, fmode_t mode, @@ -1163,11 +1224,11 @@ static int cciss_getgeo(struct block_device *bdev, struct hd_geometry *geo) return 0; } -static void check_ioctl_unit_attention(ctlr_info_t *host, CommandList_struct *c) +static void check_ioctl_unit_attention(ctlr_info_t *h, CommandList_struct *c) { if (c->err_info->CommandStatus == CMD_TARGET_STATUS && c->err_info->ScsiStatus != SAM_STAT_CHECK_CONDITION) - (void)check_for_unit_attention(host, c); + (void)check_for_unit_attention(h, c); } /* * ioctl @@ -1176,15 +1237,12 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { struct gendisk *disk = bdev->bd_disk; - ctlr_info_t *host = get_host(disk); + ctlr_info_t *h = get_host(disk); drive_info_struct *drv = get_drv(disk); - int ctlr = host->ctlr; void __user *argp = (void __user *)arg; -#ifdef CCISS_DEBUG - printk(KERN_DEBUG "cciss_ioctl: Called with cmd=%x %lx\n", cmd, arg); -#endif /* CCISS_DEBUG */ - + dev_dbg(&h->pdev->dev, "cciss_ioctl: Called with cmd=%x %lx\n", + cmd, arg); switch (cmd) { case CCISS_GETPCIINFO: { @@ -1192,10 +1250,10 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, if (!arg) return -EINVAL; - pciinfo.domain = pci_domain_nr(host->pdev->bus); - pciinfo.bus = host->pdev->bus->number; - pciinfo.dev_fn = host->pdev->devfn; - pciinfo.board_id = host->board_id; + pciinfo.domain = pci_domain_nr(h->pdev->bus); + pciinfo.bus = h->pdev->bus->number; + pciinfo.dev_fn = h->pdev->devfn; + pciinfo.board_id = h->board_id; if (copy_to_user (argp, &pciinfo, sizeof(cciss_pci_info_struct))) return -EFAULT; @@ -1207,9 +1265,9 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, if (!arg) return -EINVAL; intinfo.delay = - readl(&host->cfgtable->HostWrite.CoalIntDelay); + readl(&h->cfgtable->HostWrite.CoalIntDelay); intinfo.count = - readl(&host->cfgtable->HostWrite.CoalIntCount); + readl(&h->cfgtable->HostWrite.CoalIntCount); if (copy_to_user (argp, &intinfo, sizeof(cciss_coalint_struct))) return -EFAULT; @@ -1229,26 +1287,23 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, (&intinfo, argp, sizeof(cciss_coalint_struct))) return -EFAULT; if ((intinfo.delay == 0) && (intinfo.count == 0)) - { -// printk("cciss_ioctl: delay and count cannot be 0\n"); return -EINVAL; - } - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); + spin_lock_irqsave(&h->lock, flags); /* Update the field, and then ring the doorbell */ writel(intinfo.delay, - &(host->cfgtable->HostWrite.CoalIntDelay)); + &(h->cfgtable->HostWrite.CoalIntDelay)); writel(intinfo.count, - &(host->cfgtable->HostWrite.CoalIntCount)); - writel(CFGTBL_ChangeReq, host->vaddr + SA5_DOORBELL); + &(h->cfgtable->HostWrite.CoalIntCount)); + writel(CFGTBL_ChangeReq, h->vaddr + SA5_DOORBELL); for (i = 0; i < MAX_IOCTL_CONFIG_WAIT; i++) { - if (!(readl(host->vaddr + SA5_DOORBELL) + if (!(readl(h->vaddr + SA5_DOORBELL) & CFGTBL_ChangeReq)) break; /* delay and try again */ udelay(1000); } - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (i >= MAX_IOCTL_CONFIG_WAIT) return -EAGAIN; return 0; @@ -1262,7 +1317,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, return -EINVAL; for (i = 0; i < 16; i++) NodeName[i] = - readb(&host->cfgtable->ServerName[i]); + readb(&h->cfgtable->ServerName[i]); if (copy_to_user(argp, NodeName, sizeof(NodeName_type))) return -EFAULT; return 0; @@ -1282,23 +1337,23 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, (NodeName, argp, sizeof(NodeName_type))) return -EFAULT; - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); + spin_lock_irqsave(&h->lock, flags); /* Update the field, and then ring the doorbell */ for (i = 0; i < 16; i++) writeb(NodeName[i], - &host->cfgtable->ServerName[i]); + &h->cfgtable->ServerName[i]); - writel(CFGTBL_ChangeReq, host->vaddr + SA5_DOORBELL); + writel(CFGTBL_ChangeReq, h->vaddr + SA5_DOORBELL); for (i = 0; i < MAX_IOCTL_CONFIG_WAIT; i++) { - if (!(readl(host->vaddr + SA5_DOORBELL) + if (!(readl(h->vaddr + SA5_DOORBELL) & CFGTBL_ChangeReq)) break; /* delay and try again */ udelay(1000); } - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); if (i >= MAX_IOCTL_CONFIG_WAIT) return -EAGAIN; return 0; @@ -1310,7 +1365,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, if (!arg) return -EINVAL; - heartbeat = readl(&host->cfgtable->HeartBeat); + heartbeat = readl(&h->cfgtable->HeartBeat); if (copy_to_user (argp, &heartbeat, sizeof(Heartbeat_type))) return -EFAULT; @@ -1322,7 +1377,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, if (!arg) return -EINVAL; - BusTypes = readl(&host->cfgtable->BusTypes); + BusTypes = readl(&h->cfgtable->BusTypes); if (copy_to_user (argp, &BusTypes, sizeof(BusTypes_type))) return -EFAULT; @@ -1334,7 +1389,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, if (!arg) return -EINVAL; - memcpy(firmware, host->firm_ver, 4); + memcpy(firmware, h->firm_ver, 4); if (copy_to_user (argp, firmware, sizeof(FirmwareVer_type))) @@ -1357,7 +1412,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, case CCISS_DEREGDISK: case CCISS_REGNEWD: case CCISS_REVALIDVOLS: - return rebuild_lun_table(host, 0, 1); + return rebuild_lun_table(h, 0, 1); case CCISS_GETLUNINFO:{ LogvolInfo_struct luninfo; @@ -1377,7 +1432,6 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, CommandList_struct *c; char *buff = NULL; u64bit temp64; - unsigned long flags; DECLARE_COMPLETION_ONSTACK(wait); if (!arg) @@ -1413,7 +1467,8 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, } else { memset(buff, 0, iocommand.buf_size); } - if ((c = cmd_alloc(host, 0)) == NULL) { + c = cmd_special_alloc(h); + if (!c) { kfree(buff); return -ENOMEM; } @@ -1439,7 +1494,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, /* Fill in the scatter gather information */ if (iocommand.buf_size > 0) { - temp64.val = pci_map_single(host->pdev, buff, + temp64.val = pci_map_single(h->pdev, buff, iocommand.buf_size, PCI_DMA_BIDIRECTIONAL); c->SG[0].Addr.lower = temp64.val32.lower; @@ -1449,30 +1504,24 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, } c->waiting = &wait; - /* Put the request on the tail of the request queue */ - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - addQ(&host->reqQ, c); - host->Qdepth++; - start_io(host); - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); - + enqueue_cmd_and_start_io(h, c); wait_for_completion(&wait); /* unlock the buffers from DMA */ temp64.val32.lower = c->SG[0].Addr.lower; temp64.val32.upper = c->SG[0].Addr.upper; - pci_unmap_single(host->pdev, (dma_addr_t) temp64.val, + pci_unmap_single(h->pdev, (dma_addr_t) temp64.val, iocommand.buf_size, PCI_DMA_BIDIRECTIONAL); - check_ioctl_unit_attention(host, c); + check_ioctl_unit_attention(h, c); /* Copy the error information out */ iocommand.error_info = *(c->err_info); if (copy_to_user (argp, &iocommand, sizeof(IOCTL_Command_struct))) { kfree(buff); - cmd_free(host, c, 0); + cmd_special_free(h, c); return -EFAULT; } @@ -1481,12 +1530,12 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, if (copy_to_user (iocommand.buf, buff, iocommand.buf_size)) { kfree(buff); - cmd_free(host, c, 0); + cmd_special_free(h, c); return -EFAULT; } } kfree(buff); - cmd_free(host, c, 0); + cmd_special_free(h, c); return 0; } case CCISS_BIG_PASSTHRU:{ @@ -1495,7 +1544,6 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, unsigned char **buff = NULL; int *buff_size = NULL; u64bit temp64; - unsigned long flags; BYTE sg_used = 0; int status = 0; int i; @@ -1569,7 +1617,8 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, data_ptr += sz; sg_used++; } - if ((c = cmd_alloc(host, 0)) == NULL) { + c = cmd_special_alloc(h); + if (!c) { status = -ENOMEM; goto cleanup1; } @@ -1590,7 +1639,7 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, if (ioc->buf_size > 0) { for (i = 0; i < sg_used; i++) { temp64.val = - pci_map_single(host->pdev, buff[i], + pci_map_single(h->pdev, buff[i], buff_size[i], PCI_DMA_BIDIRECTIONAL); c->SG[i].Addr.lower = @@ -1602,26 +1651,21 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, } } c->waiting = &wait; - /* Put the request on the tail of the request queue */ - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - addQ(&host->reqQ, c); - host->Qdepth++; - start_io(host); - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); + enqueue_cmd_and_start_io(h, c); wait_for_completion(&wait); /* unlock the buffers from DMA */ for (i = 0; i < sg_used; i++) { temp64.val32.lower = c->SG[i].Addr.lower; temp64.val32.upper = c->SG[i].Addr.upper; - pci_unmap_single(host->pdev, + pci_unmap_single(h->pdev, (dma_addr_t) temp64.val, buff_size[i], PCI_DMA_BIDIRECTIONAL); } - check_ioctl_unit_attention(host, c); + check_ioctl_unit_attention(h, c); /* Copy the error information out */ ioc->error_info = *(c->err_info); if (copy_to_user(argp, ioc, sizeof(*ioc))) { - cmd_free(host, c, 0); + cmd_special_free(h, c); status = -EFAULT; goto cleanup1; } @@ -1631,14 +1675,14 @@ static int cciss_ioctl(struct block_device *bdev, fmode_t mode, for (i = 0; i < sg_used; i++) { if (copy_to_user (ptr, buff[i], buff_size[i])) { - cmd_free(host, c, 0); + cmd_special_free(h, c); status = -EFAULT; goto cleanup1; } ptr += buff_size[i]; } } - cmd_free(host, c, 0); + cmd_special_free(h, c); status = 0; cleanup1: if (buff) { @@ -1726,26 +1770,26 @@ static void cciss_check_queues(ctlr_info_t *h) static void cciss_softirq_done(struct request *rq) { - CommandList_struct *cmd = rq->completion_data; - ctlr_info_t *h = hba[cmd->ctlr]; - SGDescriptor_struct *curr_sg = cmd->SG; - unsigned long flags; + CommandList_struct *c = rq->completion_data; + ctlr_info_t *h = hba[c->ctlr]; + SGDescriptor_struct *curr_sg = c->SG; u64bit temp64; + unsigned long flags; int i, ddir; int sg_index = 0; - if (cmd->Request.Type.Direction == XFER_READ) + if (c->Request.Type.Direction == XFER_READ) ddir = PCI_DMA_FROMDEVICE; else ddir = PCI_DMA_TODEVICE; /* command did not need to be retried */ /* unmap the DMA mapping for all the scatter gather elements */ - for (i = 0; i < cmd->Header.SGList; i++) { + for (i = 0; i < c->Header.SGList; i++) { if (curr_sg[sg_index].Ext == CCISS_SG_CHAIN) { - cciss_unmap_sg_chain_block(h, cmd); + cciss_unmap_sg_chain_block(h, c); /* Point to the next block */ - curr_sg = h->cmd_sg_list[cmd->cmdindex]; + curr_sg = h->cmd_sg_list[c->cmdindex]; sg_index = 0; } temp64.val32.lower = curr_sg[sg_index].Addr.lower; @@ -1755,18 +1799,16 @@ static void cciss_softirq_done(struct request *rq) ++sg_index; } -#ifdef CCISS_DEBUG - printk("Done with %p\n", rq); -#endif /* CCISS_DEBUG */ + dev_dbg(&h->pdev->dev, "Done with %p\n", rq); /* set the residual count for pc requests */ - if (blk_pc_request(rq)) - rq->resid_len = cmd->err_info->ResidualCnt; + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) + rq->resid_len = c->err_info->ResidualCnt; blk_end_request_all(rq, (rq->errors == 0) ? 0 : -EIO); spin_lock_irqsave(&h->lock, flags); - cmd_free(h, cmd, 1); + cmd_free(h, c); cciss_check_queues(h); spin_unlock_irqrestore(&h->lock, flags); } @@ -1782,7 +1824,7 @@ static inline void log_unit_to_scsi3addr(ctlr_info_t *h, * via the inquiry page 0. Model, vendor, and rev are set to empty strings if * they cannot be read. */ -static void cciss_get_device_descr(int ctlr, int logvol, +static void cciss_get_device_descr(ctlr_info_t *h, int logvol, char *vendor, char *model, char *rev) { int rc; @@ -1797,8 +1839,8 @@ static void cciss_get_device_descr(int ctlr, int logvol, if (!inq_buf) return; - log_unit_to_scsi3addr(hba[ctlr], scsi3addr, logvol); - rc = sendcmd_withirq(CISS_INQUIRY, ctlr, inq_buf, sizeof(*inq_buf), 0, + log_unit_to_scsi3addr(h, scsi3addr, logvol); + rc = sendcmd_withirq(h, CISS_INQUIRY, inq_buf, sizeof(*inq_buf), 0, scsi3addr, TYPE_CMD); if (rc == IO_OK) { memcpy(vendor, &inq_buf->data_byte[8], VENDOR_LEN); @@ -1818,7 +1860,7 @@ static void cciss_get_device_descr(int ctlr, int logvol, * number cannot be had, for whatever reason, 16 bytes of 0xff * are returned instead. */ -static void cciss_get_serial_no(int ctlr, int logvol, +static void cciss_get_serial_no(ctlr_info_t *h, int logvol, unsigned char *serial_no, int buflen) { #define PAGE_83_INQ_BYTES 64 @@ -1833,8 +1875,8 @@ static void cciss_get_serial_no(int ctlr, int logvol, if (!buf) return; memset(serial_no, 0, buflen); - log_unit_to_scsi3addr(hba[ctlr], scsi3addr, logvol); - rc = sendcmd_withirq(CISS_INQUIRY, ctlr, buf, + log_unit_to_scsi3addr(h, scsi3addr, logvol); + rc = sendcmd_withirq(h, CISS_INQUIRY, buf, PAGE_83_INQ_BYTES, 0x83, scsi3addr, TYPE_CMD); if (rc == IO_OK) memcpy(serial_no, &buf[8], buflen); @@ -1900,10 +1942,9 @@ init_queue_failure: * is also the controller node. Any changes to disk 0 will show up on * the next reboot. */ -static void cciss_update_drive_info(int ctlr, int drv_index, int first_time, - int via_ioctl) +static void cciss_update_drive_info(ctlr_info_t *h, int drv_index, + int first_time, int via_ioctl) { - ctlr_info_t *h = hba[ctlr]; struct gendisk *disk; InquiryData_struct *inq_buff = NULL; unsigned int block_size; @@ -1920,16 +1961,16 @@ static void cciss_update_drive_info(int ctlr, int drv_index, int first_time, /* testing to see if 16-byte CDBs are already being used */ if (h->cciss_read == CCISS_READ_16) { - cciss_read_capacity_16(h->ctlr, drv_index, + cciss_read_capacity_16(h, drv_index, &total_size, &block_size); } else { - cciss_read_capacity(ctlr, drv_index, &total_size, &block_size); + cciss_read_capacity(h, drv_index, &total_size, &block_size); /* if read_capacity returns all F's this volume is >2TB */ /* in size so we switch to 16-byte CDB's for all */ /* read/write ops */ if (total_size == 0xFFFFFFFFULL) { - cciss_read_capacity_16(ctlr, drv_index, + cciss_read_capacity_16(h, drv_index, &total_size, &block_size); h->cciss_read = CCISS_READ_16; h->cciss_write = CCISS_WRITE_16; @@ -1939,14 +1980,14 @@ static void cciss_update_drive_info(int ctlr, int drv_index, int first_time, } } - cciss_geometry_inquiry(ctlr, drv_index, total_size, block_size, + cciss_geometry_inquiry(h, drv_index, total_size, block_size, inq_buff, drvinfo); drvinfo->block_size = block_size; drvinfo->nr_blocks = total_size + 1; - cciss_get_device_descr(ctlr, drv_index, drvinfo->vendor, + cciss_get_device_descr(h, drv_index, drvinfo->vendor, drvinfo->model, drvinfo->rev); - cciss_get_serial_no(ctlr, drv_index, drvinfo->serial_no, + cciss_get_serial_no(h, drv_index, drvinfo->serial_no, sizeof(drvinfo->serial_no)); /* Save the lunid in case we deregister the disk, below. */ memcpy(drvinfo->LunID, h->drv[drv_index]->LunID, @@ -1971,10 +2012,10 @@ static void cciss_update_drive_info(int ctlr, int drv_index, int first_time, * (unless it's the first disk (for the controller node). */ if (h->drv[drv_index]->raid_level != -1 && drv_index != 0) { - printk(KERN_WARNING "disk %d has changed.\n", drv_index); - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + dev_warn(&h->pdev->dev, "disk %d has changed.\n", drv_index); + spin_lock_irqsave(&h->lock, flags); h->drv[drv_index]->busy_configuring = 1; - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); /* deregister_disk sets h->drv[drv_index]->queue = NULL * which keeps the interrupt handler from starting @@ -2024,8 +2065,8 @@ static void cciss_update_drive_info(int ctlr, int drv_index, int first_time, if (cciss_add_disk(h, disk, drv_index) != 0) { cciss_free_gendisk(h, drv_index); cciss_free_drive_info(h, drv_index); - printk(KERN_WARNING "cciss:%d could not update " - "disk %d\n", h->ctlr, drv_index); + dev_warn(&h->pdev->dev, "could not update disk %d\n", + drv_index); --h->num_luns; } } @@ -2035,7 +2076,7 @@ freeret: kfree(drvinfo); return; mem_msg: - printk(KERN_ERR "cciss: out of memory\n"); + dev_err(&h->pdev->dev, "out of memory\n"); goto freeret; } @@ -2127,9 +2168,9 @@ static int cciss_add_gendisk(ctlr_info_t *h, unsigned char lunid[], h->gendisk[drv_index] = alloc_disk(1 << NWD_SHIFT); if (!h->gendisk[drv_index]) { - printk(KERN_ERR "cciss%d: could not " - "allocate a new disk %d\n", - h->ctlr, drv_index); + dev_err(&h->pdev->dev, + "could not allocate a new disk %d\n", + drv_index); goto err_free_drive_info; } } @@ -2180,8 +2221,7 @@ static void cciss_add_controller_node(ctlr_info_t *h) cciss_free_gendisk(h, drv_index); cciss_free_drive_info(h, drv_index); error: - printk(KERN_WARNING "cciss%d: could not " - "add disk 0.\n", h->ctlr); + dev_warn(&h->pdev->dev, "could not add disk 0.\n"); return; } @@ -2196,7 +2236,6 @@ error: static int rebuild_lun_table(ctlr_info_t *h, int first_time, int via_ioctl) { - int ctlr = h->ctlr; int num_luns; ReportLunData_struct *ld_buff = NULL; int return_code; @@ -2211,27 +2250,27 @@ static int rebuild_lun_table(ctlr_info_t *h, int first_time, return -EPERM; /* Set busy_configuring flag for this operation */ - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); if (h->busy_configuring) { - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return -EBUSY; } h->busy_configuring = 1; - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); ld_buff = kzalloc(sizeof(ReportLunData_struct), GFP_KERNEL); if (ld_buff == NULL) goto mem_msg; - return_code = sendcmd_withirq(CISS_REPORT_LOG, ctlr, ld_buff, + return_code = sendcmd_withirq(h, CISS_REPORT_LOG, ld_buff, sizeof(ReportLunData_struct), 0, CTLR_LUNID, TYPE_CMD); if (return_code == IO_OK) listlength = be32_to_cpu(*(__be32 *) ld_buff->LUNListLength); else { /* reading number of logical volumes failed */ - printk(KERN_WARNING "cciss: report logical volume" - " command failed\n"); + dev_warn(&h->pdev->dev, + "report logical volume command failed\n"); listlength = 0; goto freeret; } @@ -2239,7 +2278,7 @@ static int rebuild_lun_table(ctlr_info_t *h, int first_time, num_luns = listlength / 8; /* 8 bytes per entry */ if (num_luns > CISS_MAX_LUN) { num_luns = CISS_MAX_LUN; - printk(KERN_WARNING "cciss: more luns configured" + dev_warn(&h->pdev->dev, "more luns configured" " on controller than can be handled by" " this driver.\n"); } @@ -2270,9 +2309,9 @@ static int rebuild_lun_table(ctlr_info_t *h, int first_time, } if (!drv_found) { /* Deregister it from the OS, it's gone. */ - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); h->drv[i]->busy_configuring = 1; - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); return_code = deregister_disk(h, i, 1, via_ioctl); if (h->drv[i] != NULL) h->drv[i]->busy_configuring = 0; @@ -2311,8 +2350,7 @@ static int rebuild_lun_table(ctlr_info_t *h, int first_time, if (drv_index == -1) goto freeret; } - cciss_update_drive_info(ctlr, drv_index, first_time, - via_ioctl); + cciss_update_drive_info(h, drv_index, first_time, via_ioctl); } /* end for */ freeret: @@ -2324,7 +2362,7 @@ freeret: */ return -1; mem_msg: - printk(KERN_ERR "cciss: out of memory\n"); + dev_err(&h->pdev->dev, "out of memory\n"); h->busy_configuring = 0; goto freeret; } @@ -2444,11 +2482,10 @@ static int deregister_disk(ctlr_info_t *h, int drv_index, return 0; } -static int fill_cmd(CommandList_struct *c, __u8 cmd, int ctlr, void *buff, +static int fill_cmd(ctlr_info_t *h, CommandList_struct *c, __u8 cmd, void *buff, size_t size, __u8 page_code, unsigned char *scsi3addr, int cmd_type) { - ctlr_info_t *h = hba[ctlr]; u64bit buff_dma_handle; int status = IO_OK; @@ -2532,8 +2569,7 @@ static int fill_cmd(CommandList_struct *c, __u8 cmd, int ctlr, void *buff, c->Request.Timeout = 0; break; default: - printk(KERN_WARNING - "cciss%d: Unknown Command 0x%c\n", ctlr, cmd); + dev_warn(&h->pdev->dev, "Unknown Command 0x%c\n", cmd); return IO_ERROR; } } else if (cmd_type == TYPE_MSG) { @@ -2565,13 +2601,12 @@ static int fill_cmd(CommandList_struct *c, __u8 cmd, int ctlr, void *buff, c->Request.CDB[0] = cmd; break; default: - printk(KERN_WARNING - "cciss%d: unknown message type %d\n", ctlr, cmd); + dev_warn(&h->pdev->dev, + "unknown message type %d\n", cmd); return IO_ERROR; } } else { - printk(KERN_WARNING - "cciss%d: unknown command type %d\n", ctlr, cmd_type); + dev_warn(&h->pdev->dev, "unknown command type %d\n", cmd_type); return IO_ERROR; } /* Fill in the scatter gather information */ @@ -2599,15 +2634,14 @@ static int check_target_status(ctlr_info_t *h, CommandList_struct *c) default: if (check_for_unit_attention(h, c)) return IO_NEEDS_RETRY; - printk(KERN_WARNING "cciss%d: cmd 0x%02x " + dev_warn(&h->pdev->dev, "cmd 0x%02x " "check condition, sense key = 0x%02x\n", - h->ctlr, c->Request.CDB[0], - c->err_info->SenseInfo[2]); + c->Request.CDB[0], c->err_info->SenseInfo[2]); } break; default: - printk(KERN_WARNING "cciss%d: cmd 0x%02x" - "scsi status = 0x%02x\n", h->ctlr, + dev_warn(&h->pdev->dev, "cmd 0x%02x" + "scsi status = 0x%02x\n", c->Request.CDB[0], c->err_info->ScsiStatus); break; } @@ -2630,43 +2664,42 @@ static int process_sendcmd_error(ctlr_info_t *h, CommandList_struct *c) /* expected for inquiry and report lun commands */ break; case CMD_INVALID: - printk(KERN_WARNING "cciss: cmd 0x%02x is " + dev_warn(&h->pdev->dev, "cmd 0x%02x is " "reported invalid\n", c->Request.CDB[0]); return_status = IO_ERROR; break; case CMD_PROTOCOL_ERR: - printk(KERN_WARNING "cciss: cmd 0x%02x has " - "protocol error \n", c->Request.CDB[0]); + dev_warn(&h->pdev->dev, "cmd 0x%02x has " + "protocol error\n", c->Request.CDB[0]); return_status = IO_ERROR; break; case CMD_HARDWARE_ERR: - printk(KERN_WARNING "cciss: cmd 0x%02x had " + dev_warn(&h->pdev->dev, "cmd 0x%02x had " " hardware error\n", c->Request.CDB[0]); return_status = IO_ERROR; break; case CMD_CONNECTION_LOST: - printk(KERN_WARNING "cciss: cmd 0x%02x had " + dev_warn(&h->pdev->dev, "cmd 0x%02x had " "connection lost\n", c->Request.CDB[0]); return_status = IO_ERROR; break; case CMD_ABORTED: - printk(KERN_WARNING "cciss: cmd 0x%02x was " + dev_warn(&h->pdev->dev, "cmd 0x%02x was " "aborted\n", c->Request.CDB[0]); return_status = IO_ERROR; break; case CMD_ABORT_FAILED: - printk(KERN_WARNING "cciss: cmd 0x%02x reports " + dev_warn(&h->pdev->dev, "cmd 0x%02x reports " "abort failed\n", c->Request.CDB[0]); return_status = IO_ERROR; break; case CMD_UNSOLICITED_ABORT: - printk(KERN_WARNING - "cciss%d: unsolicited abort 0x%02x\n", h->ctlr, + dev_warn(&h->pdev->dev, "unsolicited abort 0x%02x\n", c->Request.CDB[0]); return_status = IO_NEEDS_RETRY; break; default: - printk(KERN_WARNING "cciss: cmd 0x%02x returned " + dev_warn(&h->pdev->dev, "cmd 0x%02x returned " "unknown status %x\n", c->Request.CDB[0], c->err_info->CommandStatus); return_status = IO_ERROR; @@ -2679,17 +2712,11 @@ static int sendcmd_withirq_core(ctlr_info_t *h, CommandList_struct *c, { DECLARE_COMPLETION_ONSTACK(wait); u64bit buff_dma_handle; - unsigned long flags; int return_status = IO_OK; resend_cmd2: c->waiting = &wait; - /* Put the request on the tail of the queue and send it */ - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); - addQ(&h->reqQ, c); - h->Qdepth++; - start_io(h); - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); + enqueue_cmd_and_start_io(h, c); wait_for_completion(&wait); @@ -2700,7 +2727,7 @@ resend_cmd2: if (return_status == IO_NEEDS_RETRY && c->retry_count < MAX_CMD_RETRIES) { - printk(KERN_WARNING "cciss%d: retrying 0x%02x\n", h->ctlr, + dev_warn(&h->pdev->dev, "retrying 0x%02x\n", c->Request.CDB[0]); c->retry_count++; /* erase the old error information */ @@ -2719,27 +2746,26 @@ command_done: return return_status; } -static int sendcmd_withirq(__u8 cmd, int ctlr, void *buff, size_t size, +static int sendcmd_withirq(ctlr_info_t *h, __u8 cmd, void *buff, size_t size, __u8 page_code, unsigned char scsi3addr[], int cmd_type) { - ctlr_info_t *h = hba[ctlr]; CommandList_struct *c; int return_status; - c = cmd_alloc(h, 0); + c = cmd_special_alloc(h); if (!c) return -ENOMEM; - return_status = fill_cmd(c, cmd, ctlr, buff, size, page_code, + return_status = fill_cmd(h, c, cmd, buff, size, page_code, scsi3addr, cmd_type); if (return_status == IO_OK) return_status = sendcmd_withirq_core(h, c, 1); - cmd_free(h, c, 0); + cmd_special_free(h, c); return return_status; } -static void cciss_geometry_inquiry(int ctlr, int logvol, +static void cciss_geometry_inquiry(ctlr_info_t *h, int logvol, sector_t total_size, unsigned int block_size, InquiryData_struct *inq_buff, @@ -2750,13 +2776,13 @@ static void cciss_geometry_inquiry(int ctlr, int logvol, unsigned char scsi3addr[8]; memset(inq_buff, 0, sizeof(InquiryData_struct)); - log_unit_to_scsi3addr(hba[ctlr], scsi3addr, logvol); - return_code = sendcmd_withirq(CISS_INQUIRY, ctlr, inq_buff, + log_unit_to_scsi3addr(h, scsi3addr, logvol); + return_code = sendcmd_withirq(h, CISS_INQUIRY, inq_buff, sizeof(*inq_buff), 0xC1, scsi3addr, TYPE_CMD); if (return_code == IO_OK) { if (inq_buff->data_byte[8] == 0xFF) { - printk(KERN_WARNING - "cciss: reading geometry failed, volume " + dev_warn(&h->pdev->dev, + "reading geometry failed, volume " "does not support reading geometry\n"); drv->heads = 255; drv->sectors = 32; /* Sectors per track */ @@ -2780,12 +2806,12 @@ static void cciss_geometry_inquiry(int ctlr, int logvol, drv->cylinders = real_size; } } else { /* Get geometry failed */ - printk(KERN_WARNING "cciss: reading geometry failed\n"); + dev_warn(&h->pdev->dev, "reading geometry failed\n"); } } static void -cciss_read_capacity(int ctlr, int logvol, sector_t *total_size, +cciss_read_capacity(ctlr_info_t *h, int logvol, sector_t *total_size, unsigned int *block_size) { ReadCapdata_struct *buf; @@ -2794,25 +2820,25 @@ cciss_read_capacity(int ctlr, int logvol, sector_t *total_size, buf = kzalloc(sizeof(ReadCapdata_struct), GFP_KERNEL); if (!buf) { - printk(KERN_WARNING "cciss: out of memory\n"); + dev_warn(&h->pdev->dev, "out of memory\n"); return; } - log_unit_to_scsi3addr(hba[ctlr], scsi3addr, logvol); - return_code = sendcmd_withirq(CCISS_READ_CAPACITY, ctlr, buf, + log_unit_to_scsi3addr(h, scsi3addr, logvol); + return_code = sendcmd_withirq(h, CCISS_READ_CAPACITY, buf, sizeof(ReadCapdata_struct), 0, scsi3addr, TYPE_CMD); if (return_code == IO_OK) { *total_size = be32_to_cpu(*(__be32 *) buf->total_size); *block_size = be32_to_cpu(*(__be32 *) buf->block_size); } else { /* read capacity command failed */ - printk(KERN_WARNING "cciss: read capacity failed\n"); + dev_warn(&h->pdev->dev, "read capacity failed\n"); *total_size = 0; *block_size = BLOCK_SIZE; } kfree(buf); } -static void cciss_read_capacity_16(int ctlr, int logvol, +static void cciss_read_capacity_16(ctlr_info_t *h, int logvol, sector_t *total_size, unsigned int *block_size) { ReadCapdata_struct_16 *buf; @@ -2821,23 +2847,23 @@ static void cciss_read_capacity_16(int ctlr, int logvol, buf = kzalloc(sizeof(ReadCapdata_struct_16), GFP_KERNEL); if (!buf) { - printk(KERN_WARNING "cciss: out of memory\n"); + dev_warn(&h->pdev->dev, "out of memory\n"); return; } - log_unit_to_scsi3addr(hba[ctlr], scsi3addr, logvol); - return_code = sendcmd_withirq(CCISS_READ_CAPACITY_16, - ctlr, buf, sizeof(ReadCapdata_struct_16), + log_unit_to_scsi3addr(h, scsi3addr, logvol); + return_code = sendcmd_withirq(h, CCISS_READ_CAPACITY_16, + buf, sizeof(ReadCapdata_struct_16), 0, scsi3addr, TYPE_CMD); if (return_code == IO_OK) { *total_size = be64_to_cpu(*(__be64 *) buf->total_size); *block_size = be32_to_cpu(*(__be32 *) buf->block_size); } else { /* read capacity command failed */ - printk(KERN_WARNING "cciss: read capacity failed\n"); + dev_warn(&h->pdev->dev, "read capacity failed\n"); *total_size = 0; *block_size = BLOCK_SIZE; } - printk(KERN_INFO " blocks= %llu block_size= %d\n", + dev_info(&h->pdev->dev, " blocks= %llu block_size= %d\n", (unsigned long long)*total_size+1, *block_size); kfree(buf); } @@ -2865,17 +2891,17 @@ static int cciss_revalidate(struct gendisk *disk) inq_buff = kmalloc(sizeof(InquiryData_struct), GFP_KERNEL); if (inq_buff == NULL) { - printk(KERN_WARNING "cciss: out of memory\n"); + dev_warn(&h->pdev->dev, "out of memory\n"); return 1; } if (h->cciss_read == CCISS_READ_10) { - cciss_read_capacity(h->ctlr, logvol, + cciss_read_capacity(h, logvol, &total_size, &block_size); } else { - cciss_read_capacity_16(h->ctlr, logvol, + cciss_read_capacity_16(h, logvol, &total_size, &block_size); } - cciss_geometry_inquiry(h->ctlr, logvol, total_size, block_size, + cciss_geometry_inquiry(h, logvol, total_size, block_size, inq_buff, drv); blk_queue_logical_block_size(drv->queue, drv->block_size); @@ -2909,7 +2935,7 @@ static void start_io(ctlr_info_t *h) c = hlist_entry(h->reqQ.first, CommandList_struct, list); /* can't do anything if fifo is full */ if ((h->access.fifo_full(h))) { - printk(KERN_WARNING "cciss: fifo full\n"); + dev_warn(&h->pdev->dev, "fifo full\n"); break; } @@ -2925,7 +2951,7 @@ static void start_io(ctlr_info_t *h) } } -/* Assumes that CCISS_LOCK(h->ctlr) is held. */ +/* Assumes that h->lock is held. */ /* Zeros out the error record and then resends the command back */ /* to the controller */ static inline void resend_cciss_cmd(ctlr_info_t *h, CommandList_struct *c) @@ -2966,7 +2992,7 @@ static inline int evaluate_target_status(ctlr_info_t *h, driver_byte = DRIVER_OK; msg_byte = cmd->err_info->CommandStatus; /* correct? seems too device specific */ - if (blk_pc_request(cmd->rq)) + if (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) host_byte = DID_PASSTHROUGH; else host_byte = DID_OK; @@ -2975,8 +3001,8 @@ static inline int evaluate_target_status(ctlr_info_t *h, host_byte, driver_byte); if (cmd->err_info->ScsiStatus != SAM_STAT_CHECK_CONDITION) { - if (!blk_pc_request(cmd->rq)) - printk(KERN_WARNING "cciss: cmd %p " + if (cmd->rq->cmd_type != REQ_TYPE_BLOCK_PC) + dev_warn(&h->pdev->dev, "cmd %p " "has SCSI Status 0x%x\n", cmd, cmd->err_info->ScsiStatus); return error_value; @@ -2985,17 +3011,19 @@ static inline int evaluate_target_status(ctlr_info_t *h, /* check the sense key */ sense_key = 0xf & cmd->err_info->SenseInfo[2]; /* no status or recovered error */ - if (((sense_key == 0x0) || (sense_key == 0x1)) && !blk_pc_request(cmd->rq)) + if (((sense_key == 0x0) || (sense_key == 0x1)) && + (cmd->rq->cmd_type != REQ_TYPE_BLOCK_PC)) error_value = 0; if (check_for_unit_attention(h, cmd)) { - *retry_cmd = !blk_pc_request(cmd->rq); + *retry_cmd = !(cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC); return 0; } - if (!blk_pc_request(cmd->rq)) { /* Not SG_IO or similar? */ + /* Not SG_IO or similar? */ + if (cmd->rq->cmd_type != REQ_TYPE_BLOCK_PC) { if (error_value != 0) - printk(KERN_WARNING "cciss: cmd %p has CHECK CONDITION" + dev_warn(&h->pdev->dev, "cmd %p has CHECK CONDITION" " sense key = 0x%x\n", cmd, sense_key); return error_value; } @@ -3035,90 +3063,97 @@ static inline void complete_command(ctlr_info_t *h, CommandList_struct *cmd, rq->errors = evaluate_target_status(h, cmd, &retry_cmd); break; case CMD_DATA_UNDERRUN: - if (blk_fs_request(cmd->rq)) { - printk(KERN_WARNING "cciss: cmd %p has" + if (cmd->rq->cmd_type == REQ_TYPE_FS) { + dev_warn(&h->pdev->dev, "cmd %p has" " completed with data underrun " "reported\n", cmd); cmd->rq->resid_len = cmd->err_info->ResidualCnt; } break; case CMD_DATA_OVERRUN: - if (blk_fs_request(cmd->rq)) - printk(KERN_WARNING "cciss: cmd %p has" + if (cmd->rq->cmd_type == REQ_TYPE_FS) + dev_warn(&h->pdev->dev, "cciss: cmd %p has" " completed with data overrun " "reported\n", cmd); break; case CMD_INVALID: - printk(KERN_WARNING "cciss: cmd %p is " + dev_warn(&h->pdev->dev, "cciss: cmd %p is " "reported invalid\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ERROR); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ERROR); break; case CMD_PROTOCOL_ERR: - printk(KERN_WARNING "cciss: cmd %p has " - "protocol error \n", cmd); + dev_warn(&h->pdev->dev, "cciss: cmd %p has " + "protocol error\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ERROR); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ERROR); break; case CMD_HARDWARE_ERR: - printk(KERN_WARNING "cciss: cmd %p had " + dev_warn(&h->pdev->dev, "cciss: cmd %p had " " hardware error\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ERROR); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ERROR); break; case CMD_CONNECTION_LOST: - printk(KERN_WARNING "cciss: cmd %p had " + dev_warn(&h->pdev->dev, "cciss: cmd %p had " "connection lost\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ERROR); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ERROR); break; case CMD_ABORTED: - printk(KERN_WARNING "cciss: cmd %p was " + dev_warn(&h->pdev->dev, "cciss: cmd %p was " "aborted\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ABORT); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ABORT); break; case CMD_ABORT_FAILED: - printk(KERN_WARNING "cciss: cmd %p reports " + dev_warn(&h->pdev->dev, "cciss: cmd %p reports " "abort failed\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ERROR); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ERROR); break; case CMD_UNSOLICITED_ABORT: - printk(KERN_WARNING "cciss%d: unsolicited " + dev_warn(&h->pdev->dev, "cciss%d: unsolicited " "abort %p\n", h->ctlr, cmd); if (cmd->retry_count < MAX_CMD_RETRIES) { retry_cmd = 1; - printk(KERN_WARNING - "cciss%d: retrying %p\n", h->ctlr, cmd); + dev_warn(&h->pdev->dev, "retrying %p\n", cmd); cmd->retry_count++; } else - printk(KERN_WARNING - "cciss%d: %p retried too " - "many times\n", h->ctlr, cmd); + dev_warn(&h->pdev->dev, + "%p retried too many times\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ABORT); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ABORT); break; case CMD_TIMEOUT: - printk(KERN_WARNING "cciss: cmd %p timedout\n", cmd); + dev_warn(&h->pdev->dev, "cmd %p timedout\n", cmd); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ERROR); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ERROR); break; default: - printk(KERN_WARNING "cciss: cmd %p returned " + dev_warn(&h->pdev->dev, "cmd %p returned " "unknown status %x\n", cmd, cmd->err_info->CommandStatus); rq->errors = make_status_bytes(SAM_STAT_GOOD, cmd->err_info->CommandStatus, DRIVER_OK, - blk_pc_request(cmd->rq) ? DID_PASSTHROUGH : DID_ERROR); + (cmd->rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + DID_PASSTHROUGH : DID_ERROR); } after_error_processing: @@ -3132,6 +3167,34 @@ after_error_processing: blk_complete_request(cmd->rq); } +static inline u32 cciss_tag_contains_index(u32 tag) +{ +#define DIRECT_LOOKUP_BIT 0x10 + return tag & DIRECT_LOOKUP_BIT; +} + +static inline u32 cciss_tag_to_index(u32 tag) +{ +#define DIRECT_LOOKUP_SHIFT 5 + return tag >> DIRECT_LOOKUP_SHIFT; +} + +static inline u32 cciss_tag_discard_error_bits(u32 tag) +{ +#define CCISS_ERROR_BITS 0x03 + return tag & ~CCISS_ERROR_BITS; +} + +static inline void cciss_mark_tag_indexed(u32 *tag) +{ + *tag |= DIRECT_LOOKUP_BIT; +} + +static inline void cciss_set_tag_index(u32 *tag, u32 index) +{ + *tag |= (index << DIRECT_LOOKUP_SHIFT); +} + /* * Get a request and submit it to the controller. */ @@ -3163,7 +3226,8 @@ static void do_cciss_request(struct request_queue *q) BUG_ON(creq->nr_phys_segments > h->maxsgentries); - if ((c = cmd_alloc(h, 1)) == NULL) + c = cmd_alloc(h); + if (!c) goto full; blk_start_request(creq); @@ -3180,8 +3244,8 @@ static void do_cciss_request(struct request_queue *q) /* got command from pool, so use the command block index instead */ /* for direct lookups. */ /* The first 2 bits are reserved for controller error reporting. */ - c->Header.Tag.lower = (c->cmdindex << 3); - c->Header.Tag.lower |= 0x04; /* flag for direct lookup. */ + cciss_set_tag_index(&c->Header.Tag.lower, c->cmdindex); + cciss_mark_tag_indexed(&c->Header.Tag.lower); memcpy(&c->Header.LUN, drv->LunID, sizeof(drv->LunID)); c->Request.CDBLen = 10; /* 12 byte commands not in FW yet; */ c->Request.Type.Type = TYPE_CMD; /* It is a command. */ @@ -3192,11 +3256,8 @@ static void do_cciss_request(struct request_queue *q) c->Request.CDB[0] = (rq_data_dir(creq) == READ) ? h->cciss_read : h->cciss_write; start_blk = blk_rq_pos(creq); -#ifdef CCISS_DEBUG - printk(KERN_DEBUG "ciss: sector =%d nr_sectors=%d\n", + dev_dbg(&h->pdev->dev, "sector =%d nr_sectors=%d\n", (int)blk_rq_pos(creq), (int)blk_rq_sectors(creq)); -#endif /* CCISS_DEBUG */ - sg_init_table(tmp_sg, h->maxsgentries); seg = blk_rq_map_sg(q, creq, tmp_sg); @@ -3236,17 +3297,18 @@ static void do_cciss_request(struct request_queue *q) if (seg > h->maxSG) h->maxSG = seg; -#ifdef CCISS_DEBUG - printk(KERN_DEBUG "cciss: Submitting %ld sectors in %d segments " + dev_dbg(&h->pdev->dev, "Submitting %u sectors in %d segments " "chained[%d]\n", blk_rq_sectors(creq), seg, chained); -#endif /* CCISS_DEBUG */ - c->Header.SGList = c->Header.SGTotal = seg + chained; - if (seg > h->max_cmd_sgentries) + c->Header.SGTotal = seg + chained; + if (seg <= h->max_cmd_sgentries) + c->Header.SGList = c->Header.SGTotal; + else c->Header.SGList = h->max_cmd_sgentries; + set_performant_mode(h, c); - if (likely(blk_fs_request(creq))) { + if (likely(creq->cmd_type == REQ_TYPE_FS)) { if(h->cciss_read == CCISS_READ_10) { c->Request.CDB[1] = 0; c->Request.CDB[2] = (start_blk >> 24) & 0xff; /* MSB */ @@ -3276,11 +3338,12 @@ static void do_cciss_request(struct request_queue *q) c->Request.CDB[13]= blk_rq_sectors(creq) & 0xff; c->Request.CDB[14] = c->Request.CDB[15] = 0; } - } else if (blk_pc_request(creq)) { + } else if (creq->cmd_type == REQ_TYPE_BLOCK_PC) { c->Request.CDBLen = creq->cmd_len; memcpy(c->Request.CDB, creq->cmd, BLK_MAX_CDB); } else { - printk(KERN_WARNING "cciss%d: bad request type %d\n", h->ctlr, creq->cmd_type); + dev_warn(&h->pdev->dev, "bad request type %d\n", + creq->cmd_type); BUG(); } @@ -3313,72 +3376,131 @@ static inline int interrupt_pending(ctlr_info_t *h) static inline long interrupt_not_for_us(ctlr_info_t *h) { - return (((h->access.intr_pending(h) == 0) || - (h->interrupts_enabled == 0))); + return ((h->access.intr_pending(h) == 0) || + (h->interrupts_enabled == 0)); } -static irqreturn_t do_cciss_intr(int irq, void *dev_id) +static inline int bad_tag(ctlr_info_t *h, u32 tag_index, + u32 raw_tag) { - ctlr_info_t *h = dev_id; + if (unlikely(tag_index >= h->nr_cmds)) { + dev_warn(&h->pdev->dev, "bad tag 0x%08x ignored.\n", raw_tag); + return 1; + } + return 0; +} + +static inline void finish_cmd(ctlr_info_t *h, CommandList_struct *c, + u32 raw_tag) +{ + removeQ(c); + if (likely(c->cmd_type == CMD_RWREQ)) + complete_command(h, c, 0); + else if (c->cmd_type == CMD_IOCTL_PEND) + complete(c->waiting); +#ifdef CONFIG_CISS_SCSI_TAPE + else if (c->cmd_type == CMD_SCSI) + complete_scsi_command(c, 0, raw_tag); +#endif +} + +static inline u32 next_command(ctlr_info_t *h) +{ + u32 a; + + if (unlikely(h->transMethod != CFGTBL_Trans_Performant)) + return h->access.command_completed(h); + + if ((*(h->reply_pool_head) & 1) == (h->reply_pool_wraparound)) { + a = *(h->reply_pool_head); /* Next cmd in ring buffer */ + (h->reply_pool_head)++; + h->commands_outstanding--; + } else { + a = FIFO_EMPTY; + } + /* Check for wraparound */ + if (h->reply_pool_head == (h->reply_pool + h->max_commands)) { + h->reply_pool_head = h->reply_pool; + h->reply_pool_wraparound ^= 1; + } + return a; +} + +/* process completion of an indexed ("direct lookup") command */ +static inline u32 process_indexed_cmd(ctlr_info_t *h, u32 raw_tag) +{ + u32 tag_index; CommandList_struct *c; + + tag_index = cciss_tag_to_index(raw_tag); + if (bad_tag(h, tag_index, raw_tag)) + return next_command(h); + c = h->cmd_pool + tag_index; + finish_cmd(h, c, raw_tag); + return next_command(h); +} + +/* process completion of a non-indexed command */ +static inline u32 process_nonindexed_cmd(ctlr_info_t *h, u32 raw_tag) +{ + u32 tag; + CommandList_struct *c = NULL; + struct hlist_node *tmp; + __u32 busaddr_masked, tag_masked; + + tag = cciss_tag_discard_error_bits(raw_tag); + hlist_for_each_entry(c, tmp, &h->cmpQ, list) { + busaddr_masked = cciss_tag_discard_error_bits(c->busaddr); + tag_masked = cciss_tag_discard_error_bits(tag); + if (busaddr_masked == tag_masked) { + finish_cmd(h, c, raw_tag); + return next_command(h); + } + } + bad_tag(h, h->nr_cmds + 1, raw_tag); + return next_command(h); +} + +static irqreturn_t do_cciss_intx(int irq, void *dev_id) +{ + ctlr_info_t *h = dev_id; unsigned long flags; - __u32 a, a1, a2; + u32 raw_tag; if (interrupt_not_for_us(h)) return IRQ_NONE; - /* - * If there are completed commands in the completion queue, - * we had better do something about it. - */ - spin_lock_irqsave(CCISS_LOCK(h->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); while (interrupt_pending(h)) { - while ((a = get_next_completion(h)) != FIFO_EMPTY) { - a1 = a; - if ((a & 0x04)) { - a2 = (a >> 3); - if (a2 >= h->nr_cmds) { - printk(KERN_WARNING - "cciss: controller cciss%d failed, stopping.\n", - h->ctlr); - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); - fail_all_cmds(h->ctlr); - return IRQ_HANDLED; - } - - c = h->cmd_pool + a2; - a = c->busaddr; - - } else { - struct hlist_node *tmp; - - a &= ~3; - c = NULL; - hlist_for_each_entry(c, tmp, &h->cmpQ, list) { - if (c->busaddr == a) - break; - } - } - /* - * If we've found the command, take it off the - * completion Q and free it - */ - if (c && c->busaddr == a) { - removeQ(c); - if (c->cmd_type == CMD_RWREQ) { - complete_command(h, c, 0); - } else if (c->cmd_type == CMD_IOCTL_PEND) { - complete(c->waiting); - } -# ifdef CONFIG_CISS_SCSI_TAPE - else if (c->cmd_type == CMD_SCSI) - complete_scsi_command(c, 0, a1); -# endif - continue; - } + raw_tag = get_next_completion(h); + while (raw_tag != FIFO_EMPTY) { + if (cciss_tag_contains_index(raw_tag)) + raw_tag = process_indexed_cmd(h, raw_tag); + else + raw_tag = process_nonindexed_cmd(h, raw_tag); } } + spin_unlock_irqrestore(&h->lock, flags); + return IRQ_HANDLED; +} - spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); +/* Add a second interrupt handler for MSI/MSI-X mode. In this mode we never + * check the interrupt pending register because it is not set. + */ +static irqreturn_t do_cciss_msix_intr(int irq, void *dev_id) +{ + ctlr_info_t *h = dev_id; + unsigned long flags; + u32 raw_tag; + + spin_lock_irqsave(&h->lock, flags); + raw_tag = get_next_completion(h); + while (raw_tag != FIFO_EMPTY) { + if (cciss_tag_contains_index(raw_tag)) + raw_tag = process_indexed_cmd(h, raw_tag); + else + raw_tag = process_nonindexed_cmd(h, raw_tag); + } + spin_unlock_irqrestore(&h->lock, flags); return IRQ_HANDLED; } @@ -3510,18 +3632,17 @@ static int check_for_unit_attention(ctlr_info_t *h, CommandList_struct *c) switch (c->err_info->SenseInfo[12]) { case STATE_CHANGED: - printk(KERN_WARNING "cciss%d: a state change " - "detected, command retried\n", h->ctlr); + dev_warn(&h->pdev->dev, "a state change " + "detected, command retried\n"); return 1; break; case LUN_FAILED: - printk(KERN_WARNING "cciss%d: LUN failure " - "detected, action required\n", h->ctlr); + dev_warn(&h->pdev->dev, "LUN failure " + "detected, action required\n"); return 1; break; case REPORT_LUNS_CHANGED: - printk(KERN_WARNING "cciss%d: report LUN data " - "changed\n", h->ctlr); + dev_warn(&h->pdev->dev, "report LUN data changed\n"); /* * Here, we could call add_to_scan_list and wake up the scan thread, * except that it's quite likely that we will get more than one @@ -3541,19 +3662,18 @@ static int check_for_unit_attention(ctlr_info_t *h, CommandList_struct *c) return 1; break; case POWER_OR_RESET: - printk(KERN_WARNING "cciss%d: a power on " - "or device reset detected\n", h->ctlr); + dev_warn(&h->pdev->dev, + "a power on or device reset detected\n"); return 1; break; case UNIT_ATTENTION_CLEARED: - printk(KERN_WARNING "cciss%d: unit attention " - "cleared by another initiator\n", h->ctlr); + dev_warn(&h->pdev->dev, + "unit attention cleared by another initiator\n"); return 1; break; default: - printk(KERN_WARNING "cciss%d: unknown " - "unit attention detected\n", h->ctlr); - return 1; + dev_warn(&h->pdev->dev, "unknown unit attention detected\n"); + return 1; } } @@ -3562,39 +3682,41 @@ static int check_for_unit_attention(ctlr_info_t *h, CommandList_struct *c) * the io functions. * This is for debug only. */ -#ifdef CCISS_DEBUG -static void print_cfg_table(CfgTable_struct *tb) +static void print_cfg_table(ctlr_info_t *h) { int i; char temp_name[17]; + CfgTable_struct *tb = h->cfgtable; - printk("Controller Configuration information\n"); - printk("------------------------------------\n"); + dev_dbg(&h->pdev->dev, "Controller Configuration information\n"); + dev_dbg(&h->pdev->dev, "------------------------------------\n"); for (i = 0; i < 4; i++) temp_name[i] = readb(&(tb->Signature[i])); temp_name[4] = '\0'; - printk(" Signature = %s\n", temp_name); - printk(" Spec Number = %d\n", readl(&(tb->SpecValence))); - printk(" Transport methods supported = 0x%x\n", + dev_dbg(&h->pdev->dev, " Signature = %s\n", temp_name); + dev_dbg(&h->pdev->dev, " Spec Number = %d\n", + readl(&(tb->SpecValence))); + dev_dbg(&h->pdev->dev, " Transport methods supported = 0x%x\n", readl(&(tb->TransportSupport))); - printk(" Transport methods active = 0x%x\n", + dev_dbg(&h->pdev->dev, " Transport methods active = 0x%x\n", readl(&(tb->TransportActive))); - printk(" Requested transport Method = 0x%x\n", + dev_dbg(&h->pdev->dev, " Requested transport Method = 0x%x\n", readl(&(tb->HostWrite.TransportRequest))); - printk(" Coalesce Interrupt Delay = 0x%x\n", + dev_dbg(&h->pdev->dev, " Coalesce Interrupt Delay = 0x%x\n", readl(&(tb->HostWrite.CoalIntDelay))); - printk(" Coalesce Interrupt Count = 0x%x\n", + dev_dbg(&h->pdev->dev, " Coalesce Interrupt Count = 0x%x\n", readl(&(tb->HostWrite.CoalIntCount))); - printk(" Max outstanding commands = 0x%d\n", + dev_dbg(&h->pdev->dev, " Max outstanding commands = 0x%d\n", readl(&(tb->CmdsOutMax))); - printk(" Bus Types = 0x%x\n", readl(&(tb->BusTypes))); + dev_dbg(&h->pdev->dev, " Bus Types = 0x%x\n", + readl(&(tb->BusTypes))); for (i = 0; i < 16; i++) temp_name[i] = readb(&(tb->ServerName[i])); temp_name[16] = '\0'; - printk(" Server Name = %s\n", temp_name); - printk(" Heartbeat Counter = 0x%x\n\n\n", readl(&(tb->HeartBeat))); + dev_dbg(&h->pdev->dev, " Server Name = %s\n", temp_name); + dev_dbg(&h->pdev->dev, " Heartbeat Counter = 0x%x\n\n\n", + readl(&(tb->HeartBeat))); } -#endif /* CCISS_DEBUG */ static int find_PCI_BAR_index(struct pci_dev *pdev, unsigned long pci_bar_addr) { @@ -3618,7 +3740,7 @@ static int find_PCI_BAR_index(struct pci_dev *pdev, unsigned long pci_bar_addr) offset += 8; break; default: /* reserved in PCI 2.2 */ - printk(KERN_WARNING + dev_warn(&pdev->dev, "Base address is invalid\n"); return -1; break; @@ -3630,12 +3752,182 @@ static int find_PCI_BAR_index(struct pci_dev *pdev, unsigned long pci_bar_addr) return -1; } +/* Fill in bucket_map[], given nsgs (the max number of + * scatter gather elements supported) and bucket[], + * which is an array of 8 integers. The bucket[] array + * contains 8 different DMA transfer sizes (in 16 + * byte increments) which the controller uses to fetch + * commands. This function fills in bucket_map[], which + * maps a given number of scatter gather elements to one of + * the 8 DMA transfer sizes. The point of it is to allow the + * controller to only do as much DMA as needed to fetch the + * command, with the DMA transfer size encoded in the lower + * bits of the command address. + */ +static void calc_bucket_map(int bucket[], int num_buckets, + int nsgs, int *bucket_map) +{ + int i, j, b, size; + + /* even a command with 0 SGs requires 4 blocks */ +#define MINIMUM_TRANSFER_BLOCKS 4 +#define NUM_BUCKETS 8 + /* Note, bucket_map must have nsgs+1 entries. */ + for (i = 0; i <= nsgs; i++) { + /* Compute size of a command with i SG entries */ + size = i + MINIMUM_TRANSFER_BLOCKS; + b = num_buckets; /* Assume the biggest bucket */ + /* Find the bucket that is just big enough */ + for (j = 0; j < 8; j++) { + if (bucket[j] >= size) { + b = j; + break; + } + } + /* for a command with i SG entries, use bucket b. */ + bucket_map[i] = b; + } +} + +static void __devinit cciss_wait_for_mode_change_ack(ctlr_info_t *h) +{ + int i; + + /* under certain very rare conditions, this can take awhile. + * (e.g.: hot replace a failed 144GB drive in a RAID 5 set right + * as we enter this code.) */ + for (i = 0; i < MAX_CONFIG_WAIT; i++) { + if (!(readl(h->vaddr + SA5_DOORBELL) & CFGTBL_ChangeReq)) + break; + msleep(10); + } +} + +static __devinit void cciss_enter_performant_mode(ctlr_info_t *h) +{ + /* This is a bit complicated. There are 8 registers on + * the controller which we write to to tell it 8 different + * sizes of commands which there may be. It's a way of + * reducing the DMA done to fetch each command. Encoded into + * each command's tag are 3 bits which communicate to the controller + * which of the eight sizes that command fits within. The size of + * each command depends on how many scatter gather entries there are. + * Each SG entry requires 16 bytes. The eight registers are programmed + * with the number of 16-byte blocks a command of that size requires. + * The smallest command possible requires 5 such 16 byte blocks. + * the largest command possible requires MAXSGENTRIES + 4 16-byte + * blocks. Note, this only extends to the SG entries contained + * within the command block, and does not extend to chained blocks + * of SG elements. bft[] contains the eight values we write to + * the registers. They are not evenly distributed, but have more + * sizes for small commands, and fewer sizes for larger commands. + */ + __u32 trans_offset; + int bft[8] = { 5, 6, 8, 10, 12, 20, 28, MAXSGENTRIES + 4}; + /* + * 5 = 1 s/g entry or 4k + * 6 = 2 s/g entry or 8k + * 8 = 4 s/g entry or 16k + * 10 = 6 s/g entry or 24k + */ + unsigned long register_value; + BUILD_BUG_ON(28 > MAXSGENTRIES + 4); + + h->reply_pool_wraparound = 1; /* spec: init to 1 */ + + /* Controller spec: zero out this buffer. */ + memset(h->reply_pool, 0, h->max_commands * sizeof(__u64)); + h->reply_pool_head = h->reply_pool; + + trans_offset = readl(&(h->cfgtable->TransMethodOffset)); + calc_bucket_map(bft, ARRAY_SIZE(bft), h->maxsgentries, + h->blockFetchTable); + writel(bft[0], &h->transtable->BlockFetch0); + writel(bft[1], &h->transtable->BlockFetch1); + writel(bft[2], &h->transtable->BlockFetch2); + writel(bft[3], &h->transtable->BlockFetch3); + writel(bft[4], &h->transtable->BlockFetch4); + writel(bft[5], &h->transtable->BlockFetch5); + writel(bft[6], &h->transtable->BlockFetch6); + writel(bft[7], &h->transtable->BlockFetch7); + + /* size of controller ring buffer */ + writel(h->max_commands, &h->transtable->RepQSize); + writel(1, &h->transtable->RepQCount); + writel(0, &h->transtable->RepQCtrAddrLow32); + writel(0, &h->transtable->RepQCtrAddrHigh32); + writel(h->reply_pool_dhandle, &h->transtable->RepQAddr0Low32); + writel(0, &h->transtable->RepQAddr0High32); + writel(CFGTBL_Trans_Performant, + &(h->cfgtable->HostWrite.TransportRequest)); + + writel(CFGTBL_ChangeReq, h->vaddr + SA5_DOORBELL); + cciss_wait_for_mode_change_ack(h); + register_value = readl(&(h->cfgtable->TransportActive)); + if (!(register_value & CFGTBL_Trans_Performant)) + dev_warn(&h->pdev->dev, "cciss: unable to get board into" + " performant mode\n"); +} + +static void __devinit cciss_put_controller_into_performant_mode(ctlr_info_t *h) +{ + __u32 trans_support; + + dev_dbg(&h->pdev->dev, "Trying to put board into Performant mode\n"); + /* Attempt to put controller into performant mode if supported */ + /* Does board support performant mode? */ + trans_support = readl(&(h->cfgtable->TransportSupport)); + if (!(trans_support & PERFORMANT_MODE)) + return; + + dev_dbg(&h->pdev->dev, "Placing controller into performant mode\n"); + /* Performant mode demands commands on a 32 byte boundary + * pci_alloc_consistent aligns on page boundarys already. + * Just need to check if divisible by 32 + */ + if ((sizeof(CommandList_struct) % 32) != 0) { + dev_warn(&h->pdev->dev, "%s %d %s\n", + "cciss info: command size[", + (int)sizeof(CommandList_struct), + "] not divisible by 32, no performant mode..\n"); + return; + } + + /* Performant mode ring buffer and supporting data structures */ + h->reply_pool = (__u64 *)pci_alloc_consistent( + h->pdev, h->max_commands * sizeof(__u64), + &(h->reply_pool_dhandle)); + + /* Need a block fetch table for performant mode */ + h->blockFetchTable = kmalloc(((h->maxsgentries+1) * + sizeof(__u32)), GFP_KERNEL); + + if ((h->reply_pool == NULL) || (h->blockFetchTable == NULL)) + goto clean_up; + + cciss_enter_performant_mode(h); + + /* Change the access methods to the performant access methods */ + h->access = SA5_performant_access; + h->transMethod = CFGTBL_Trans_Performant; + + return; +clean_up: + kfree(h->blockFetchTable); + if (h->reply_pool) + pci_free_consistent(h->pdev, + h->max_commands * sizeof(__u64), + h->reply_pool, + h->reply_pool_dhandle); + return; + +} /* cciss_put_controller_into_performant_mode */ + /* If MSI/MSI-X is supported by the kernel we will try to enable it on * controllers that are capable. If not, we use IO-APIC mode. */ -static void __devinit cciss_interrupt_mode(ctlr_info_t *c, - struct pci_dev *pdev, __u32 board_id) +static void __devinit cciss_interrupt_mode(ctlr_info_t *h) { #ifdef CONFIG_PCI_MSI int err; @@ -3644,268 +3936,283 @@ static void __devinit cciss_interrupt_mode(ctlr_info_t *c, }; /* Some boards advertise MSI but don't really support it */ - if ((board_id == 0x40700E11) || - (board_id == 0x40800E11) || - (board_id == 0x40820E11) || (board_id == 0x40830E11)) + if ((h->board_id == 0x40700E11) || (h->board_id == 0x40800E11) || + (h->board_id == 0x40820E11) || (h->board_id == 0x40830E11)) goto default_int_mode; - if (pci_find_capability(pdev, PCI_CAP_ID_MSIX)) { - err = pci_enable_msix(pdev, cciss_msix_entries, 4); + if (pci_find_capability(h->pdev, PCI_CAP_ID_MSIX)) { + err = pci_enable_msix(h->pdev, cciss_msix_entries, 4); if (!err) { - c->intr[0] = cciss_msix_entries[0].vector; - c->intr[1] = cciss_msix_entries[1].vector; - c->intr[2] = cciss_msix_entries[2].vector; - c->intr[3] = cciss_msix_entries[3].vector; - c->msix_vector = 1; + h->intr[0] = cciss_msix_entries[0].vector; + h->intr[1] = cciss_msix_entries[1].vector; + h->intr[2] = cciss_msix_entries[2].vector; + h->intr[3] = cciss_msix_entries[3].vector; + h->msix_vector = 1; return; } if (err > 0) { - printk(KERN_WARNING "cciss: only %d MSI-X vectors " - "available\n", err); + dev_warn(&h->pdev->dev, + "only %d MSI-X vectors available\n", err); goto default_int_mode; } else { - printk(KERN_WARNING "cciss: MSI-X init failed %d\n", - err); + dev_warn(&h->pdev->dev, + "MSI-X init failed %d\n", err); goto default_int_mode; } } - if (pci_find_capability(pdev, PCI_CAP_ID_MSI)) { - if (!pci_enable_msi(pdev)) { - c->msi_vector = 1; - } else { - printk(KERN_WARNING "cciss: MSI init failed\n"); - } + if (pci_find_capability(h->pdev, PCI_CAP_ID_MSI)) { + if (!pci_enable_msi(h->pdev)) + h->msi_vector = 1; + else + dev_warn(&h->pdev->dev, "MSI init failed\n"); } default_int_mode: #endif /* CONFIG_PCI_MSI */ /* if we get here we're going to use the default interrupt mode */ - c->intr[SIMPLE_MODE_INT] = pdev->irq; + h->intr[PERF_MODE_INT] = h->pdev->irq; return; } -static int __devinit cciss_pci_init(ctlr_info_t *c, struct pci_dev *pdev) +static int __devinit cciss_lookup_board_id(struct pci_dev *pdev, u32 *board_id) { - ushort subsystem_vendor_id, subsystem_device_id, command; - __u32 board_id, scratchpad = 0; - __u64 cfg_offset; - __u32 cfg_base_addr; - __u64 cfg_base_addr_index; - int i, prod_index, err; + int i; + u32 subsystem_vendor_id, subsystem_device_id; subsystem_vendor_id = pdev->subsystem_vendor; subsystem_device_id = pdev->subsystem_device; - board_id = (((__u32) (subsystem_device_id << 16) & 0xffff0000) | - subsystem_vendor_id); + *board_id = ((subsystem_device_id << 16) & 0xffff0000) | + subsystem_vendor_id; for (i = 0; i < ARRAY_SIZE(products); i++) { /* Stand aside for hpsa driver on request */ if (cciss_allow_hpsa && products[i].board_id == HPSA_BOUNDARY) return -ENODEV; - if (board_id == products[i].board_id) - break; - } - prod_index = i; - if (prod_index == ARRAY_SIZE(products)) { - dev_warn(&pdev->dev, - "unrecognized board ID: 0x%08lx, ignoring.\n", - (unsigned long) board_id); - return -ENODEV; + if (*board_id == products[i].board_id) + return i; } + dev_warn(&pdev->dev, "unrecognized board ID: 0x%08x, ignoring.\n", + *board_id); + return -ENODEV; +} - /* check to see if controller has been disabled */ - /* BEFORE trying to enable it */ - (void)pci_read_config_word(pdev, PCI_COMMAND, &command); - if (!(command & 0x02)) { - printk(KERN_WARNING - "cciss: controller appears to be disabled\n"); - return -ENODEV; - } +static inline bool cciss_board_disabled(ctlr_info_t *h) +{ + u16 command; - err = pci_enable_device(pdev); - if (err) { - printk(KERN_ERR "cciss: Unable to Enable PCI device\n"); - return err; - } + (void) pci_read_config_word(h->pdev, PCI_COMMAND, &command); + return ((command & PCI_COMMAND_MEMORY) == 0); +} - err = pci_request_regions(pdev, "cciss"); - if (err) { - printk(KERN_ERR "cciss: Cannot obtain PCI resources, " - "aborting\n"); - return err; - } +static int __devinit cciss_pci_find_memory_BAR(struct pci_dev *pdev, + unsigned long *memory_bar) +{ + int i; -#ifdef CCISS_DEBUG - printk("command = %x\n", command); - printk("irq = %x\n", pdev->irq); - printk("board_id = %x\n", board_id); -#endif /* CCISS_DEBUG */ + for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) + if (pci_resource_flags(pdev, i) & IORESOURCE_MEM) { + /* addressing mode bits already removed */ + *memory_bar = pci_resource_start(pdev, i); + dev_dbg(&pdev->dev, "memory BAR = %lx\n", + *memory_bar); + return 0; + } + dev_warn(&pdev->dev, "no memory BAR found\n"); + return -ENODEV; +} -/* If the kernel supports MSI/MSI-X we will try to enable that functionality, - * else we use the IO-APIC interrupt assigned to us by system ROM. - */ - cciss_interrupt_mode(c, pdev, board_id); +static int __devinit cciss_wait_for_board_ready(ctlr_info_t *h) +{ + int i; + u32 scratchpad; - /* find the memory BAR */ - for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) { - if (pci_resource_flags(pdev, i) & IORESOURCE_MEM) - break; - } - if (i == DEVICE_COUNT_RESOURCE) { - printk(KERN_WARNING "cciss: No memory BAR found\n"); - err = -ENODEV; - goto err_out_free_res; + for (i = 0; i < CCISS_BOARD_READY_ITERATIONS; i++) { + scratchpad = readl(h->vaddr + SA5_SCRATCHPAD_OFFSET); + if (scratchpad == CCISS_FIRMWARE_READY) + return 0; + msleep(CCISS_BOARD_READY_POLL_INTERVAL_MSECS); } + dev_warn(&h->pdev->dev, "board not ready, timed out.\n"); + return -ENODEV; +} - c->paddr = pci_resource_start(pdev, i); /* addressing mode bits - * already removed - */ +static int __devinit cciss_find_cfg_addrs(struct pci_dev *pdev, + void __iomem *vaddr, u32 *cfg_base_addr, u64 *cfg_base_addr_index, + u64 *cfg_offset) +{ + *cfg_base_addr = readl(vaddr + SA5_CTCFG_OFFSET); + *cfg_offset = readl(vaddr + SA5_CTMEM_OFFSET); + *cfg_base_addr &= (u32) 0x0000ffff; + *cfg_base_addr_index = find_PCI_BAR_index(pdev, *cfg_base_addr); + if (*cfg_base_addr_index == -1) { + dev_warn(&pdev->dev, "cannot find cfg_base_addr_index, " + "*cfg_base_addr = 0x%08x\n", *cfg_base_addr); + return -ENODEV; + } + return 0; +} -#ifdef CCISS_DEBUG - printk("address 0 = %lx\n", c->paddr); -#endif /* CCISS_DEBUG */ - c->vaddr = remap_pci_mem(c->paddr, 0x250); +static int __devinit cciss_find_cfgtables(ctlr_info_t *h) +{ + u64 cfg_offset; + u32 cfg_base_addr; + u64 cfg_base_addr_index; + u32 trans_offset; + int rc; - /* Wait for the board to become ready. (PCI hotplug needs this.) - * We poll for up to 120 secs, once per 100ms. */ - for (i = 0; i < 1200; i++) { - scratchpad = readl(c->vaddr + SA5_SCRATCHPAD_OFFSET); - if (scratchpad == CCISS_FIRMWARE_READY) - break; - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(msecs_to_jiffies(100)); /* wait 100ms */ - } - if (scratchpad != CCISS_FIRMWARE_READY) { - printk(KERN_WARNING "cciss: Board not ready. Timed out.\n"); - err = -ENODEV; - goto err_out_free_res; - } + rc = cciss_find_cfg_addrs(h->pdev, h->vaddr, &cfg_base_addr, + &cfg_base_addr_index, &cfg_offset); + if (rc) + return rc; + h->cfgtable = remap_pci_mem(pci_resource_start(h->pdev, + cfg_base_addr_index) + cfg_offset, sizeof(h->cfgtable)); + if (!h->cfgtable) + return -ENOMEM; + /* Find performant mode table. */ + trans_offset = readl(&h->cfgtable->TransMethodOffset); + h->transtable = remap_pci_mem(pci_resource_start(h->pdev, + cfg_base_addr_index)+cfg_offset+trans_offset, + sizeof(*h->transtable)); + if (!h->transtable) + return -ENOMEM; + return 0; +} - /* get the address index number */ - cfg_base_addr = readl(c->vaddr + SA5_CTCFG_OFFSET); - cfg_base_addr &= (__u32) 0x0000ffff; -#ifdef CCISS_DEBUG - printk("cfg base address = %x\n", cfg_base_addr); -#endif /* CCISS_DEBUG */ - cfg_base_addr_index = find_PCI_BAR_index(pdev, cfg_base_addr); -#ifdef CCISS_DEBUG - printk("cfg base address index = %llx\n", - (unsigned long long)cfg_base_addr_index); -#endif /* CCISS_DEBUG */ - if (cfg_base_addr_index == -1) { - printk(KERN_WARNING "cciss: Cannot find cfg_base_addr_index\n"); - err = -ENODEV; - goto err_out_free_res; +static void __devinit cciss_get_max_perf_mode_cmds(struct ctlr_info *h) +{ + h->max_commands = readl(&(h->cfgtable->MaxPerformantModeCommands)); + if (h->max_commands < 16) { + dev_warn(&h->pdev->dev, "Controller reports " + "max supported commands of %d, an obvious lie. " + "Using 16. Ensure that firmware is up to date.\n", + h->max_commands); + h->max_commands = 16; } +} - cfg_offset = readl(c->vaddr + SA5_CTMEM_OFFSET); -#ifdef CCISS_DEBUG - printk("cfg offset = %llx\n", (unsigned long long)cfg_offset); -#endif /* CCISS_DEBUG */ - c->cfgtable = remap_pci_mem(pci_resource_start(pdev, - cfg_base_addr_index) + - cfg_offset, sizeof(CfgTable_struct)); - c->board_id = board_id; - -#ifdef CCISS_DEBUG - print_cfg_table(c->cfgtable); -#endif /* CCISS_DEBUG */ - - /* Some controllers support Zero Memory Raid (ZMR). - * When configured in ZMR mode the number of supported - * commands drops to 64. So instead of just setting an - * arbitrary value we make the driver a little smarter. - * We read the config table to tell us how many commands - * are supported on the controller then subtract 4 to - * leave a little room for ioctl calls. - */ - c->max_commands = readl(&(c->cfgtable->CmdsOutMax)); - c->maxsgentries = readl(&(c->cfgtable->MaxSGElements)); - +/* Interrogate the hardware for some limits: + * max commands, max SG elements without chaining, and with chaining, + * SG chain block size, etc. + */ +static void __devinit cciss_find_board_params(ctlr_info_t *h) +{ + cciss_get_max_perf_mode_cmds(h); + h->nr_cmds = h->max_commands - 4; /* Allow room for some ioctls */ + h->maxsgentries = readl(&(h->cfgtable->MaxSGElements)); /* - * Limit native command to 32 s/g elements to save dma'able memory. + * Limit in-command s/g elements to 32 save dma'able memory. * Howvever spec says if 0, use 31 */ - - c->max_cmd_sgentries = 31; - if (c->maxsgentries > 512) { - c->max_cmd_sgentries = 32; - c->chainsize = c->maxsgentries - c->max_cmd_sgentries + 1; - c->maxsgentries -= 1; /* account for chain pointer */ + h->max_cmd_sgentries = 31; + if (h->maxsgentries > 512) { + h->max_cmd_sgentries = 32; + h->chainsize = h->maxsgentries - h->max_cmd_sgentries + 1; + h->maxsgentries--; /* save one for chain pointer */ } else { - c->maxsgentries = 31; /* Default to traditional value */ - c->chainsize = 0; /* traditional */ + h->maxsgentries = 31; /* default to traditional values */ + h->chainsize = 0; } +} - c->product_name = products[prod_index].product_name; - c->access = *(products[prod_index].access); - c->nr_cmds = c->max_commands - 4; - if ((readb(&c->cfgtable->Signature[0]) != 'C') || - (readb(&c->cfgtable->Signature[1]) != 'I') || - (readb(&c->cfgtable->Signature[2]) != 'S') || - (readb(&c->cfgtable->Signature[3]) != 'S')) { - printk("Does not appear to be a valid CISS config table\n"); - err = -ENODEV; - goto err_out_free_res; +static inline bool CISS_signature_present(ctlr_info_t *h) +{ + if ((readb(&h->cfgtable->Signature[0]) != 'C') || + (readb(&h->cfgtable->Signature[1]) != 'I') || + (readb(&h->cfgtable->Signature[2]) != 'S') || + (readb(&h->cfgtable->Signature[3]) != 'S')) { + dev_warn(&h->pdev->dev, "not a valid CISS config table\n"); + return false; } + return true; +} + +/* Need to enable prefetch in the SCSI core for 6400 in x86 */ +static inline void cciss_enable_scsi_prefetch(ctlr_info_t *h) +{ #ifdef CONFIG_X86 - { - /* Need to enable prefetch in the SCSI core for 6400 in x86 */ - __u32 prefetch; - prefetch = readl(&(c->cfgtable->SCSI_Prefetch)); - prefetch |= 0x100; - writel(prefetch, &(c->cfgtable->SCSI_Prefetch)); - } + u32 prefetch; + + prefetch = readl(&(h->cfgtable->SCSI_Prefetch)); + prefetch |= 0x100; + writel(prefetch, &(h->cfgtable->SCSI_Prefetch)); #endif +} - /* Disabling DMA prefetch and refetch for the P600. - * An ASIC bug may result in accesses to invalid memory addresses. - * We've disabled prefetch for some time now. Testing with XEN - * kernels revealed a bug in the refetch if dom0 resides on a P600. - */ - if(board_id == 0x3225103C) { - __u32 dma_prefetch; - __u32 dma_refetch; - dma_prefetch = readl(c->vaddr + I2O_DMA1_CFG); - dma_prefetch |= 0x8000; - writel(dma_prefetch, c->vaddr + I2O_DMA1_CFG); - pci_read_config_dword(pdev, PCI_COMMAND_PARITY, &dma_refetch); - dma_refetch |= 0x1; - pci_write_config_dword(pdev, PCI_COMMAND_PARITY, dma_refetch); +/* Disable DMA prefetch for the P600. Otherwise an ASIC bug may result + * in a prefetch beyond physical memory. + */ +static inline void cciss_p600_dma_prefetch_quirk(ctlr_info_t *h) +{ + u32 dma_prefetch; + __u32 dma_refetch; + + if (h->board_id != 0x3225103C) + return; + dma_prefetch = readl(h->vaddr + I2O_DMA1_CFG); + dma_prefetch |= 0x8000; + writel(dma_prefetch, h->vaddr + I2O_DMA1_CFG); + pci_read_config_dword(h->pdev, PCI_COMMAND_PARITY, &dma_refetch); + dma_refetch |= 0x1; + pci_write_config_dword(h->pdev, PCI_COMMAND_PARITY, dma_refetch); +} + +static int __devinit cciss_pci_init(ctlr_info_t *h) +{ + int prod_index, err; + + prod_index = cciss_lookup_board_id(h->pdev, &h->board_id); + if (prod_index < 0) + return -ENODEV; + h->product_name = products[prod_index].product_name; + h->access = *(products[prod_index].access); + + if (cciss_board_disabled(h)) { + dev_warn(&h->pdev->dev, "controller appears to be disabled\n"); + return -ENODEV; + } + err = pci_enable_device(h->pdev); + if (err) { + dev_warn(&h->pdev->dev, "Unable to Enable PCI device\n"); + return err; } -#ifdef CCISS_DEBUG - printk("Trying to put board into Simple mode\n"); -#endif /* CCISS_DEBUG */ - c->max_commands = readl(&(c->cfgtable->CmdsOutMax)); - /* Update the field, and then ring the doorbell */ - writel(CFGTBL_Trans_Simple, &(c->cfgtable->HostWrite.TransportRequest)); - writel(CFGTBL_ChangeReq, c->vaddr + SA5_DOORBELL); + err = pci_request_regions(h->pdev, "cciss"); + if (err) { + dev_warn(&h->pdev->dev, + "Cannot obtain PCI resources, aborting\n"); + return err; + } - /* under certain very rare conditions, this can take awhile. - * (e.g.: hot replace a failed 144GB drive in a RAID 5 set right - * as we enter this code.) */ - for (i = 0; i < MAX_CONFIG_WAIT; i++) { - if (!(readl(c->vaddr + SA5_DOORBELL) & CFGTBL_ChangeReq)) - break; - /* delay and try again */ - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(msecs_to_jiffies(1)); + dev_dbg(&h->pdev->dev, "irq = %x\n", h->pdev->irq); + dev_dbg(&h->pdev->dev, "board_id = %x\n", h->board_id); + +/* If the kernel supports MSI/MSI-X we will try to enable that functionality, + * else we use the IO-APIC interrupt assigned to us by system ROM. + */ + cciss_interrupt_mode(h); + err = cciss_pci_find_memory_BAR(h->pdev, &h->paddr); + if (err) + goto err_out_free_res; + h->vaddr = remap_pci_mem(h->paddr, 0x250); + if (!h->vaddr) { + err = -ENOMEM; + goto err_out_free_res; } + err = cciss_wait_for_board_ready(h); + if (err) + goto err_out_free_res; + err = cciss_find_cfgtables(h); + if (err) + goto err_out_free_res; + print_cfg_table(h); + cciss_find_board_params(h); -#ifdef CCISS_DEBUG - printk(KERN_DEBUG "I counter got to %d %x\n", i, - readl(c->vaddr + SA5_DOORBELL)); -#endif /* CCISS_DEBUG */ -#ifdef CCISS_DEBUG - print_cfg_table(c->cfgtable); -#endif /* CCISS_DEBUG */ - - if (!(readl(&(c->cfgtable->TransportActive)) & CFGTBL_Trans_Simple)) { - printk(KERN_WARNING "cciss: unable to get board into" - " simple mode\n"); + if (!CISS_signature_present(h)) { err = -ENODEV; goto err_out_free_res; } + cciss_enable_scsi_prefetch(h); + cciss_p600_dma_prefetch_quirk(h); + cciss_put_controller_into_performant_mode(h); return 0; err_out_free_res: @@ -3913,42 +4220,47 @@ err_out_free_res: * Deliberately omit pci_disable_device(): it does something nasty to * Smart Array controllers that pci_enable_device does not undo */ - pci_release_regions(pdev); + if (h->transtable) + iounmap(h->transtable); + if (h->cfgtable) + iounmap(h->cfgtable); + if (h->vaddr) + iounmap(h->vaddr); + pci_release_regions(h->pdev); return err; } /* Function to find the first free pointer into our hba[] array * Returns -1 if no free entries are left. */ -static int alloc_cciss_hba(void) +static int alloc_cciss_hba(struct pci_dev *pdev) { int i; for (i = 0; i < MAX_CTLR; i++) { if (!hba[i]) { - ctlr_info_t *p; + ctlr_info_t *h; - p = kzalloc(sizeof(ctlr_info_t), GFP_KERNEL); - if (!p) + h = kzalloc(sizeof(ctlr_info_t), GFP_KERNEL); + if (!h) goto Enomem; - hba[i] = p; + hba[i] = h; return i; } } - printk(KERN_WARNING "cciss: This driver supports a maximum" + dev_warn(&pdev->dev, "This driver supports a maximum" " of %d controllers.\n", MAX_CTLR); return -1; Enomem: - printk(KERN_ERR "cciss: out of memory.\n"); + dev_warn(&pdev->dev, "out of memory.\n"); return -1; } -static void free_hba(int n) +static void free_hba(ctlr_info_t *h) { - ctlr_info_t *h = hba[n]; int i; - hba[n] = NULL; + hba[h->ctlr] = NULL; for (i = 0; i < h->highest_lun + 1; i++) if (h->gendisk[i] != NULL) put_disk(h->gendisk[i]); @@ -4028,7 +4340,8 @@ static __devinit int cciss_message(struct pci_dev *pdev, unsigned char opcode, u /* we leak the DMA buffer here ... no choice since the controller could still complete the command. */ if (i == 10) { - printk(KERN_ERR "cciss: controller message %02x:%02x timed out\n", + dev_err(&pdev->dev, + "controller message %02x:%02x timed out\n", opcode, type); return -ETIMEDOUT; } @@ -4036,12 +4349,12 @@ static __devinit int cciss_message(struct pci_dev *pdev, unsigned char opcode, u pci_free_consistent(pdev, cmd_sz, cmd, paddr64); if (tag & 2) { - printk(KERN_ERR "cciss: controller message %02x:%02x failed\n", + dev_err(&pdev->dev, "controller message %02x:%02x failed\n", opcode, type); return -EIO; } - printk(KERN_INFO "cciss: controller message %02x:%02x succeeded\n", + dev_info(&pdev->dev, "controller message %02x:%02x succeeded\n", opcode, type); return 0; } @@ -4062,7 +4375,7 @@ static __devinit int cciss_reset_msi(struct pci_dev *pdev) if (pos) { pci_read_config_word(pdev, msi_control_reg(pos), &control); if (control & PCI_MSI_FLAGS_ENABLE) { - printk(KERN_INFO "cciss: resetting MSI\n"); + dev_info(&pdev->dev, "resetting MSI\n"); pci_write_config_word(pdev, msi_control_reg(pos), control & ~PCI_MSI_FLAGS_ENABLE); } } @@ -4071,7 +4384,7 @@ static __devinit int cciss_reset_msi(struct pci_dev *pdev) if (pos) { pci_read_config_word(pdev, msi_control_reg(pos), &control); if (control & PCI_MSIX_FLAGS_ENABLE) { - printk(KERN_INFO "cciss: resetting MSI-X\n"); + dev_info(&pdev->dev, "resetting MSI-X\n"); pci_write_config_word(pdev, msi_control_reg(pos), control & ~PCI_MSIX_FLAGS_ENABLE); } } @@ -4079,68 +4392,144 @@ static __devinit int cciss_reset_msi(struct pci_dev *pdev) return 0; } -/* This does a hard reset of the controller using PCI power management - * states. */ -static __devinit int cciss_hard_reset_controller(struct pci_dev *pdev) +static int cciss_controller_hard_reset(struct pci_dev *pdev, + void * __iomem vaddr, bool use_doorbell) { - u16 pmcsr, saved_config_space[32]; - int i, pos; + u16 pmcsr; + int pos; - printk(KERN_INFO "cciss: using PCI PM to reset controller\n"); + if (use_doorbell) { + /* For everything after the P600, the PCI power state method + * of resetting the controller doesn't work, so we have this + * other way using the doorbell register. + */ + dev_info(&pdev->dev, "using doorbell to reset controller\n"); + writel(DOORBELL_CTLR_RESET, vaddr + SA5_DOORBELL); + msleep(1000); + } else { /* Try to do it the PCI power state way */ + + /* Quoting from the Open CISS Specification: "The Power + * Management Control/Status Register (CSR) controls the power + * state of the device. The normal operating state is D0, + * CSR=00h. The software off state is D3, CSR=03h. To reset + * the controller, place the interface device in D3 then to D0, + * this causes a secondary PCI reset which will reset the + * controller." */ + + pos = pci_find_capability(pdev, PCI_CAP_ID_PM); + if (pos == 0) { + dev_err(&pdev->dev, + "cciss_controller_hard_reset: " + "PCI PM not supported\n"); + return -ENODEV; + } + dev_info(&pdev->dev, "using PCI PM to reset controller\n"); + /* enter the D3hot power management state */ + pci_read_config_word(pdev, pos + PCI_PM_CTRL, &pmcsr); + pmcsr &= ~PCI_PM_CTRL_STATE_MASK; + pmcsr |= PCI_D3hot; + pci_write_config_word(pdev, pos + PCI_PM_CTRL, pmcsr); - /* This is very nearly the same thing as + msleep(500); - pci_save_state(pci_dev); - pci_set_power_state(pci_dev, PCI_D3hot); - pci_set_power_state(pci_dev, PCI_D0); - pci_restore_state(pci_dev); + /* enter the D0 power management state */ + pmcsr &= ~PCI_PM_CTRL_STATE_MASK; + pmcsr |= PCI_D0; + pci_write_config_word(pdev, pos + PCI_PM_CTRL, pmcsr); - but we can't use these nice canned kernel routines on - kexec, because they also check the MSI/MSI-X state in PCI - configuration space and do the wrong thing when it is - set/cleared. Also, the pci_save/restore_state functions - violate the ordering requirements for restoring the - configuration space from the CCISS document (see the - comment below). So we roll our own .... */ + msleep(500); + } + return 0; +} - for (i = 0; i < 32; i++) - pci_read_config_word(pdev, 2*i, &saved_config_space[i]); +/* This does a hard reset of the controller using PCI power management + * states or using the doorbell register. */ +static __devinit int cciss_kdump_hard_reset_controller(struct pci_dev *pdev) +{ + u16 saved_config_space[32]; + u64 cfg_offset; + u32 cfg_base_addr; + u64 cfg_base_addr_index; + void __iomem *vaddr; + unsigned long paddr; + u32 misc_fw_support, active_transport; + int rc, i; + CfgTable_struct __iomem *cfgtable; + bool use_doorbell; + u32 board_id; + + /* For controllers as old a the p600, this is very nearly + * the same thing as + * + * pci_save_state(pci_dev); + * pci_set_power_state(pci_dev, PCI_D3hot); + * pci_set_power_state(pci_dev, PCI_D0); + * pci_restore_state(pci_dev); + * + * but we can't use these nice canned kernel routines on + * kexec, because they also check the MSI/MSI-X state in PCI + * configuration space and do the wrong thing when it is + * set/cleared. Also, the pci_save/restore_state functions + * violate the ordering requirements for restoring the + * configuration space from the CCISS document (see the + * comment below). So we roll our own .... + * + * For controllers newer than the P600, the pci power state + * method of resetting doesn't work so we have another way + * using the doorbell register. + */ - pos = pci_find_capability(pdev, PCI_CAP_ID_PM); - if (pos == 0) { - printk(KERN_ERR "cciss_reset_controller: PCI PM not supported\n"); + /* Exclude 640x boards. These are two pci devices in one slot + * which share a battery backed cache module. One controls the + * cache, the other accesses the cache through the one that controls + * it. If we reset the one controlling the cache, the other will + * likely not be happy. Just forbid resetting this conjoined mess. + */ + cciss_lookup_board_id(pdev, &board_id); + if (board_id == 0x409C0E11 || board_id == 0x409D0E11) { + dev_warn(&pdev->dev, "Cannot reset Smart Array 640x " + "due to shared cache module."); return -ENODEV; } - /* Quoting from the Open CISS Specification: "The Power - * Management Control/Status Register (CSR) controls the power - * state of the device. The normal operating state is D0, - * CSR=00h. The software off state is D3, CSR=03h. To reset - * the controller, place the interface device in D3 then to - * D0, this causes a secondary PCI reset which will reset the - * controller." */ + for (i = 0; i < 32; i++) + pci_read_config_word(pdev, 2*i, &saved_config_space[i]); - /* enter the D3hot power management state */ - pci_read_config_word(pdev, pos + PCI_PM_CTRL, &pmcsr); - pmcsr &= ~PCI_PM_CTRL_STATE_MASK; - pmcsr |= PCI_D3hot; - pci_write_config_word(pdev, pos + PCI_PM_CTRL, pmcsr); + /* find the first memory BAR, so we can find the cfg table */ + rc = cciss_pci_find_memory_BAR(pdev, &paddr); + if (rc) + return rc; + vaddr = remap_pci_mem(paddr, 0x250); + if (!vaddr) + return -ENOMEM; - schedule_timeout_uninterruptible(HZ >> 1); + /* find cfgtable in order to check if reset via doorbell is supported */ + rc = cciss_find_cfg_addrs(pdev, vaddr, &cfg_base_addr, + &cfg_base_addr_index, &cfg_offset); + if (rc) + goto unmap_vaddr; + cfgtable = remap_pci_mem(pci_resource_start(pdev, + cfg_base_addr_index) + cfg_offset, sizeof(*cfgtable)); + if (!cfgtable) { + rc = -ENOMEM; + goto unmap_vaddr; + } - /* enter the D0 power management state */ - pmcsr &= ~PCI_PM_CTRL_STATE_MASK; - pmcsr |= PCI_D0; - pci_write_config_word(pdev, pos + PCI_PM_CTRL, pmcsr); + /* If reset via doorbell register is supported, use that. */ + misc_fw_support = readl(&cfgtable->misc_fw_support); + use_doorbell = misc_fw_support & MISC_FW_DOORBELL_RESET; - schedule_timeout_uninterruptible(HZ >> 1); + rc = cciss_controller_hard_reset(pdev, vaddr, use_doorbell); + if (rc) + goto unmap_cfgtable; /* Restore the PCI configuration space. The Open CISS * Specification says, "Restore the PCI Configuration * Registers, offsets 00h through 60h. It is important to * restore the command register, 16-bits at offset 04h, * last. Do not restore the configuration status register, - * 16-bits at offset 06h." Note that the offset is 2*i. */ + * 16-bits at offset 06h." Note that the offset is 2*i. + */ for (i = 0; i < 32; i++) { if (i == 2 || i == 3) continue; @@ -4149,6 +4538,63 @@ static __devinit int cciss_hard_reset_controller(struct pci_dev *pdev) wmb(); pci_write_config_word(pdev, 4, saved_config_space[2]); + /* Some devices (notably the HP Smart Array 5i Controller) + need a little pause here */ + msleep(CCISS_POST_RESET_PAUSE_MSECS); + + /* Controller should be in simple mode at this point. If it's not, + * It means we're on one of those controllers which doesn't support + * the doorbell reset method and on which the PCI power management reset + * method doesn't work (P800, for example.) + * In those cases, don't try to proceed, as it generally doesn't work. + */ + active_transport = readl(&cfgtable->TransportActive); + if (active_transport & PERFORMANT_MODE) { + dev_warn(&pdev->dev, "Unable to successfully reset controller," + " Ignoring controller.\n"); + rc = -ENODEV; + } + +unmap_cfgtable: + iounmap(cfgtable); + +unmap_vaddr: + iounmap(vaddr); + return rc; +} + +static __devinit int cciss_init_reset_devices(struct pci_dev *pdev) +{ + int rc, i; + + if (!reset_devices) + return 0; + + /* Reset the controller with a PCI power-cycle or via doorbell */ + rc = cciss_kdump_hard_reset_controller(pdev); + + /* -ENOTSUPP here means we cannot reset the controller + * but it's already (and still) up and running in + * "performant mode". Or, it might be 640x, which can't reset + * due to concerns about shared bbwc between 6402/6404 pair. + */ + if (rc == -ENOTSUPP) + return 0; /* just try to do the kdump anyhow. */ + if (rc) + return -ENODEV; + if (cciss_reset_msi(pdev)) + return -ENODEV; + + /* Now try to get the controller to respond to a no-op */ + for (i = 0; i < CCISS_POST_RESET_NOOP_RETRIES; i++) { + if (cciss_noop(pdev) == 0) + break; + else + dev_warn(&pdev->dev, "no-op failed%s\n", + (i < CCISS_POST_RESET_NOOP_RETRIES - 1 ? + "; re-trying" : "")); + msleep(CCISS_POST_RESET_NOOP_INTERVAL_MSECS); + } return 0; } @@ -4166,46 +4612,31 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, int rc; int dac, return_code; InquiryData_struct *inq_buff; + ctlr_info_t *h; - if (reset_devices) { - /* Reset the controller with a PCI power-cycle */ - if (cciss_hard_reset_controller(pdev) || cciss_reset_msi(pdev)) - return -ENODEV; - - /* Now try to get the controller to respond to a no-op. Some - devices (notably the HP Smart Array 5i Controller) need - up to 30 seconds to respond. */ - for (i=0; i<30; i++) { - if (cciss_noop(pdev) == 0) - break; - - schedule_timeout_uninterruptible(HZ); - } - if (i == 30) { - printk(KERN_ERR "cciss: controller seems dead\n"); - return -EBUSY; - } - } - - i = alloc_cciss_hba(); + rc = cciss_init_reset_devices(pdev); + if (rc) + return rc; + i = alloc_cciss_hba(pdev); if (i < 0) return -1; - hba[i]->busy_initializing = 1; - INIT_HLIST_HEAD(&hba[i]->cmpQ); - INIT_HLIST_HEAD(&hba[i]->reqQ); - mutex_init(&hba[i]->busy_shutting_down); + h = hba[i]; + h->pdev = pdev; + h->busy_initializing = 1; + INIT_HLIST_HEAD(&h->cmpQ); + INIT_HLIST_HEAD(&h->reqQ); + mutex_init(&h->busy_shutting_down); - if (cciss_pci_init(hba[i], pdev) != 0) + if (cciss_pci_init(h) != 0) goto clean_no_release_regions; - sprintf(hba[i]->devname, "cciss%d", i); - hba[i]->ctlr = i; - hba[i]->pdev = pdev; + sprintf(h->devname, "cciss%d", i); + h->ctlr = i; - init_completion(&hba[i]->scan_wait); + init_completion(&h->scan_wait); - if (cciss_create_hba_sysfs_entry(hba[i])) + if (cciss_create_hba_sysfs_entry(h)) goto clean0; /* configure PCI DMA stuff */ @@ -4214,7 +4645,7 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, else if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(32))) dac = 0; else { - printk(KERN_ERR "cciss: no suitable DMA available\n"); + dev_err(&h->pdev->dev, "no suitable DMA available\n"); goto clean1; } @@ -4224,151 +4655,161 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, * 8 controller support. */ if (i < MAX_CTLR_ORIG) - hba[i]->major = COMPAQ_CISS_MAJOR + i; - rc = register_blkdev(hba[i]->major, hba[i]->devname); + h->major = COMPAQ_CISS_MAJOR + i; + rc = register_blkdev(h->major, h->devname); if (rc == -EBUSY || rc == -EINVAL) { - printk(KERN_ERR - "cciss: Unable to get major number %d for %s " - "on hba %d\n", hba[i]->major, hba[i]->devname, i); + dev_err(&h->pdev->dev, + "Unable to get major number %d for %s " + "on hba %d\n", h->major, h->devname, i); goto clean1; } else { if (i >= MAX_CTLR_ORIG) - hba[i]->major = rc; + h->major = rc; } /* make sure the board interrupts are off */ - hba[i]->access.set_intr_mask(hba[i], CCISS_INTR_OFF); - if (request_irq(hba[i]->intr[SIMPLE_MODE_INT], do_cciss_intr, - IRQF_DISABLED | IRQF_SHARED, hba[i]->devname, hba[i])) { - printk(KERN_ERR "cciss: Unable to get irq %d for %s\n", - hba[i]->intr[SIMPLE_MODE_INT], hba[i]->devname); - goto clean2; + h->access.set_intr_mask(h, CCISS_INTR_OFF); + if (h->msi_vector || h->msix_vector) { + if (request_irq(h->intr[PERF_MODE_INT], + do_cciss_msix_intr, + IRQF_DISABLED, h->devname, h)) { + dev_err(&h->pdev->dev, "Unable to get irq %d for %s\n", + h->intr[PERF_MODE_INT], h->devname); + goto clean2; + } + } else { + if (request_irq(h->intr[PERF_MODE_INT], do_cciss_intx, + IRQF_DISABLED, h->devname, h)) { + dev_err(&h->pdev->dev, "Unable to get irq %d for %s\n", + h->intr[PERF_MODE_INT], h->devname); + goto clean2; + } } - printk(KERN_INFO "%s: <0x%x> at PCI %s IRQ %d%s using DAC\n", - hba[i]->devname, pdev->device, pci_name(pdev), - hba[i]->intr[SIMPLE_MODE_INT], dac ? "" : " not"); + dev_info(&h->pdev->dev, "%s: <0x%x> at PCI %s IRQ %d%s using DAC\n", + h->devname, pdev->device, pci_name(pdev), + h->intr[PERF_MODE_INT], dac ? "" : " not"); - hba[i]->cmd_pool_bits = - kmalloc(DIV_ROUND_UP(hba[i]->nr_cmds, BITS_PER_LONG) + h->cmd_pool_bits = + kmalloc(DIV_ROUND_UP(h->nr_cmds, BITS_PER_LONG) * sizeof(unsigned long), GFP_KERNEL); - hba[i]->cmd_pool = (CommandList_struct *) - pci_alloc_consistent(hba[i]->pdev, - hba[i]->nr_cmds * sizeof(CommandList_struct), - &(hba[i]->cmd_pool_dhandle)); - hba[i]->errinfo_pool = (ErrorInfo_struct *) - pci_alloc_consistent(hba[i]->pdev, - hba[i]->nr_cmds * sizeof(ErrorInfo_struct), - &(hba[i]->errinfo_pool_dhandle)); - if ((hba[i]->cmd_pool_bits == NULL) - || (hba[i]->cmd_pool == NULL) - || (hba[i]->errinfo_pool == NULL)) { - printk(KERN_ERR "cciss: out of memory"); + h->cmd_pool = (CommandList_struct *) + pci_alloc_consistent(h->pdev, + h->nr_cmds * sizeof(CommandList_struct), + &(h->cmd_pool_dhandle)); + h->errinfo_pool = (ErrorInfo_struct *) + pci_alloc_consistent(h->pdev, + h->nr_cmds * sizeof(ErrorInfo_struct), + &(h->errinfo_pool_dhandle)); + if ((h->cmd_pool_bits == NULL) + || (h->cmd_pool == NULL) + || (h->errinfo_pool == NULL)) { + dev_err(&h->pdev->dev, "out of memory"); goto clean4; } /* Need space for temp scatter list */ - hba[i]->scatter_list = kmalloc(hba[i]->max_commands * + h->scatter_list = kmalloc(h->max_commands * sizeof(struct scatterlist *), GFP_KERNEL); - for (k = 0; k < hba[i]->nr_cmds; k++) { - hba[i]->scatter_list[k] = kmalloc(sizeof(struct scatterlist) * - hba[i]->maxsgentries, + for (k = 0; k < h->nr_cmds; k++) { + h->scatter_list[k] = kmalloc(sizeof(struct scatterlist) * + h->maxsgentries, GFP_KERNEL); - if (hba[i]->scatter_list[k] == NULL) { - printk(KERN_ERR "cciss%d: could not allocate " - "s/g lists\n", i); + if (h->scatter_list[k] == NULL) { + dev_err(&h->pdev->dev, + "could not allocate s/g lists\n"); goto clean4; } } - hba[i]->cmd_sg_list = cciss_allocate_sg_chain_blocks(hba[i], - hba[i]->chainsize, hba[i]->nr_cmds); - if (!hba[i]->cmd_sg_list && hba[i]->chainsize > 0) + h->cmd_sg_list = cciss_allocate_sg_chain_blocks(h, + h->chainsize, h->nr_cmds); + if (!h->cmd_sg_list && h->chainsize > 0) goto clean4; - spin_lock_init(&hba[i]->lock); + spin_lock_init(&h->lock); /* Initialize the pdev driver private data. - have it point to hba[i]. */ - pci_set_drvdata(pdev, hba[i]); + have it point to h. */ + pci_set_drvdata(pdev, h); /* command and error info recs zeroed out before they are used */ - memset(hba[i]->cmd_pool_bits, 0, - DIV_ROUND_UP(hba[i]->nr_cmds, BITS_PER_LONG) + memset(h->cmd_pool_bits, 0, + DIV_ROUND_UP(h->nr_cmds, BITS_PER_LONG) * sizeof(unsigned long)); - hba[i]->num_luns = 0; - hba[i]->highest_lun = -1; + h->num_luns = 0; + h->highest_lun = -1; for (j = 0; j < CISS_MAX_LUN; j++) { - hba[i]->drv[j] = NULL; - hba[i]->gendisk[j] = NULL; + h->drv[j] = NULL; + h->gendisk[j] = NULL; } - cciss_scsi_setup(i); + cciss_scsi_setup(h); /* Turn the interrupts on so we can service requests */ - hba[i]->access.set_intr_mask(hba[i], CCISS_INTR_ON); + h->access.set_intr_mask(h, CCISS_INTR_ON); /* Get the firmware version */ inq_buff = kzalloc(sizeof(InquiryData_struct), GFP_KERNEL); if (inq_buff == NULL) { - printk(KERN_ERR "cciss: out of memory\n"); + dev_err(&h->pdev->dev, "out of memory\n"); goto clean4; } - return_code = sendcmd_withirq(CISS_INQUIRY, i, inq_buff, + return_code = sendcmd_withirq(h, CISS_INQUIRY, inq_buff, sizeof(InquiryData_struct), 0, CTLR_LUNID, TYPE_CMD); if (return_code == IO_OK) { - hba[i]->firm_ver[0] = inq_buff->data_byte[32]; - hba[i]->firm_ver[1] = inq_buff->data_byte[33]; - hba[i]->firm_ver[2] = inq_buff->data_byte[34]; - hba[i]->firm_ver[3] = inq_buff->data_byte[35]; + h->firm_ver[0] = inq_buff->data_byte[32]; + h->firm_ver[1] = inq_buff->data_byte[33]; + h->firm_ver[2] = inq_buff->data_byte[34]; + h->firm_ver[3] = inq_buff->data_byte[35]; } else { /* send command failed */ - printk(KERN_WARNING "cciss: unable to determine firmware" + dev_warn(&h->pdev->dev, "unable to determine firmware" " version of controller\n"); } kfree(inq_buff); - cciss_procinit(i); + cciss_procinit(h); - hba[i]->cciss_max_sectors = 8192; + h->cciss_max_sectors = 8192; - rebuild_lun_table(hba[i], 1, 0); - hba[i]->busy_initializing = 0; + rebuild_lun_table(h, 1, 0); + h->busy_initializing = 0; return 1; clean4: - kfree(hba[i]->cmd_pool_bits); + kfree(h->cmd_pool_bits); /* Free up sg elements */ - for (k = 0; k < hba[i]->nr_cmds; k++) - kfree(hba[i]->scatter_list[k]); - kfree(hba[i]->scatter_list); - cciss_free_sg_chain_blocks(hba[i]->cmd_sg_list, hba[i]->nr_cmds); - if (hba[i]->cmd_pool) - pci_free_consistent(hba[i]->pdev, - hba[i]->nr_cmds * sizeof(CommandList_struct), - hba[i]->cmd_pool, hba[i]->cmd_pool_dhandle); - if (hba[i]->errinfo_pool) - pci_free_consistent(hba[i]->pdev, - hba[i]->nr_cmds * sizeof(ErrorInfo_struct), - hba[i]->errinfo_pool, - hba[i]->errinfo_pool_dhandle); - free_irq(hba[i]->intr[SIMPLE_MODE_INT], hba[i]); + for (k = 0; k < h->nr_cmds; k++) + kfree(h->scatter_list[k]); + kfree(h->scatter_list); + cciss_free_sg_chain_blocks(h->cmd_sg_list, h->nr_cmds); + if (h->cmd_pool) + pci_free_consistent(h->pdev, + h->nr_cmds * sizeof(CommandList_struct), + h->cmd_pool, h->cmd_pool_dhandle); + if (h->errinfo_pool) + pci_free_consistent(h->pdev, + h->nr_cmds * sizeof(ErrorInfo_struct), + h->errinfo_pool, + h->errinfo_pool_dhandle); + free_irq(h->intr[PERF_MODE_INT], h); clean2: - unregister_blkdev(hba[i]->major, hba[i]->devname); + unregister_blkdev(h->major, h->devname); clean1: - cciss_destroy_hba_sysfs_entry(hba[i]); + cciss_destroy_hba_sysfs_entry(h); clean0: pci_release_regions(pdev); clean_no_release_regions: - hba[i]->busy_initializing = 0; + h->busy_initializing = 0; /* * Deliberately omit pci_disable_device(): it does something nasty to * Smart Array controllers that pci_enable_device does not undo */ pci_set_drvdata(pdev, NULL); - free_hba(i); + free_hba(h); return -1; } @@ -4381,55 +4822,51 @@ static void cciss_shutdown(struct pci_dev *pdev) h = pci_get_drvdata(pdev); flush_buf = kzalloc(4, GFP_KERNEL); if (!flush_buf) { - printk(KERN_WARNING - "cciss:%d cache not flushed, out of memory.\n", - h->ctlr); + dev_warn(&h->pdev->dev, "cache not flushed, out of memory.\n"); return; } /* write all data in the battery backed cache to disk */ memset(flush_buf, 0, 4); - return_code = sendcmd_withirq(CCISS_CACHE_FLUSH, h->ctlr, flush_buf, + return_code = sendcmd_withirq(h, CCISS_CACHE_FLUSH, flush_buf, 4, 0, CTLR_LUNID, TYPE_CMD); kfree(flush_buf); if (return_code != IO_OK) - printk(KERN_WARNING "cciss%d: Error flushing cache\n", - h->ctlr); + dev_warn(&h->pdev->dev, "Error flushing cache\n"); h->access.set_intr_mask(h, CCISS_INTR_OFF); - free_irq(h->intr[2], h); + free_irq(h->intr[PERF_MODE_INT], h); } static void __devexit cciss_remove_one(struct pci_dev *pdev) { - ctlr_info_t *tmp_ptr; + ctlr_info_t *h; int i, j; if (pci_get_drvdata(pdev) == NULL) { - printk(KERN_ERR "cciss: Unable to remove device \n"); + dev_err(&pdev->dev, "Unable to remove device\n"); return; } - tmp_ptr = pci_get_drvdata(pdev); - i = tmp_ptr->ctlr; + h = pci_get_drvdata(pdev); + i = h->ctlr; if (hba[i] == NULL) { - printk(KERN_ERR "cciss: device appears to " - "already be removed \n"); + dev_err(&pdev->dev, "device appears to already be removed\n"); return; } - mutex_lock(&hba[i]->busy_shutting_down); + mutex_lock(&h->busy_shutting_down); - remove_from_scan_list(hba[i]); - remove_proc_entry(hba[i]->devname, proc_cciss); - unregister_blkdev(hba[i]->major, hba[i]->devname); + remove_from_scan_list(h); + remove_proc_entry(h->devname, proc_cciss); + unregister_blkdev(h->major, h->devname); /* remove it from the disk list */ for (j = 0; j < CISS_MAX_LUN; j++) { - struct gendisk *disk = hba[i]->gendisk[j]; + struct gendisk *disk = h->gendisk[j]; if (disk) { struct request_queue *q = disk->queue; if (disk->flags & GENHD_FL_UP) { - cciss_destroy_ld_sysfs_entry(hba[i], j, 1); + cciss_destroy_ld_sysfs_entry(h, j, 1); del_gendisk(disk); } if (q) @@ -4438,39 +4875,41 @@ static void __devexit cciss_remove_one(struct pci_dev *pdev) } #ifdef CONFIG_CISS_SCSI_TAPE - cciss_unregister_scsi(i); /* unhook from SCSI subsystem */ + cciss_unregister_scsi(h); /* unhook from SCSI subsystem */ #endif cciss_shutdown(pdev); #ifdef CONFIG_PCI_MSI - if (hba[i]->msix_vector) - pci_disable_msix(hba[i]->pdev); - else if (hba[i]->msi_vector) - pci_disable_msi(hba[i]->pdev); + if (h->msix_vector) + pci_disable_msix(h->pdev); + else if (h->msi_vector) + pci_disable_msi(h->pdev); #endif /* CONFIG_PCI_MSI */ - iounmap(hba[i]->vaddr); + iounmap(h->transtable); + iounmap(h->cfgtable); + iounmap(h->vaddr); - pci_free_consistent(hba[i]->pdev, hba[i]->nr_cmds * sizeof(CommandList_struct), - hba[i]->cmd_pool, hba[i]->cmd_pool_dhandle); - pci_free_consistent(hba[i]->pdev, hba[i]->nr_cmds * sizeof(ErrorInfo_struct), - hba[i]->errinfo_pool, hba[i]->errinfo_pool_dhandle); - kfree(hba[i]->cmd_pool_bits); + pci_free_consistent(h->pdev, h->nr_cmds * sizeof(CommandList_struct), + h->cmd_pool, h->cmd_pool_dhandle); + pci_free_consistent(h->pdev, h->nr_cmds * sizeof(ErrorInfo_struct), + h->errinfo_pool, h->errinfo_pool_dhandle); + kfree(h->cmd_pool_bits); /* Free up sg elements */ - for (j = 0; j < hba[i]->nr_cmds; j++) - kfree(hba[i]->scatter_list[j]); - kfree(hba[i]->scatter_list); - cciss_free_sg_chain_blocks(hba[i]->cmd_sg_list, hba[i]->nr_cmds); + for (j = 0; j < h->nr_cmds; j++) + kfree(h->scatter_list[j]); + kfree(h->scatter_list); + cciss_free_sg_chain_blocks(h->cmd_sg_list, h->nr_cmds); /* * Deliberately omit pci_disable_device(): it does something nasty to * Smart Array controllers that pci_enable_device does not undo */ pci_release_regions(pdev); pci_set_drvdata(pdev, NULL); - cciss_destroy_hba_sysfs_entry(hba[i]); - mutex_unlock(&hba[i]->busy_shutting_down); - free_hba(i); + cciss_destroy_hba_sysfs_entry(h); + mutex_unlock(&h->busy_shutting_down); + free_hba(h); } static struct pci_driver cciss_pci_driver = { @@ -4495,7 +4934,6 @@ static int __init cciss_init(void) * array of them, the size must be a multiple of 8 bytes. */ BUILD_BUG_ON(sizeof(CommandList_struct) % COMMANDLIST_ALIGNMENT); - printk(KERN_INFO DRIVER_NAME "\n"); err = bus_register(&cciss_bus_type); @@ -4532,8 +4970,8 @@ static void __exit cciss_cleanup(void) /* double check that all controller entrys have been removed */ for (i = 0; i < MAX_CTLR; i++) { if (hba[i] != NULL) { - printk(KERN_WARNING "cciss: had to remove" - " controller %d\n", i); + dev_warn(&hba[i]->pdev->dev, + "had to remove controller\n"); cciss_remove_one(hba[i]->pdev); } } @@ -4542,46 +4980,5 @@ static void __exit cciss_cleanup(void) bus_unregister(&cciss_bus_type); } -static void fail_all_cmds(unsigned long ctlr) -{ - /* If we get here, the board is apparently dead. */ - ctlr_info_t *h = hba[ctlr]; - CommandList_struct *c; - unsigned long flags; - - printk(KERN_WARNING "cciss%d: controller not responding.\n", h->ctlr); - h->alive = 0; /* the controller apparently died... */ - - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - - pci_disable_device(h->pdev); /* Make sure it is really dead. */ - - /* move everything off the request queue onto the completed queue */ - while (!hlist_empty(&h->reqQ)) { - c = hlist_entry(h->reqQ.first, CommandList_struct, list); - removeQ(c); - h->Qdepth--; - addQ(&h->cmpQ, c); - } - - /* Now, fail everything on the completed queue with a HW error */ - while (!hlist_empty(&h->cmpQ)) { - c = hlist_entry(h->cmpQ.first, CommandList_struct, list); - removeQ(c); - if (c->cmd_type != CMD_MSG_STALE) - c->err_info->CommandStatus = CMD_HARDWARE_ERR; - if (c->cmd_type == CMD_RWREQ) { - complete_command(h, c, 0); - } else if (c->cmd_type == CMD_IOCTL_PEND) - complete(c->waiting); -#ifdef CONFIG_CISS_SCSI_TAPE - else if (c->cmd_type == CMD_SCSI) - complete_scsi_command(c, 0, 0); -#endif - } - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); - return; -} - module_init(cciss_init); module_exit(cciss_cleanup); diff --git a/drivers/block/cciss.h b/drivers/block/cciss.h index c5d4111..ae340ff 100644 --- a/drivers/block/cciss.h +++ b/drivers/block/cciss.h @@ -25,7 +25,7 @@ struct access_method { void (*submit_command)(ctlr_info_t *h, CommandList_struct *c); void (*set_intr_mask)(ctlr_info_t *h, unsigned long val); unsigned long (*fifo_full)(ctlr_info_t *h); - unsigned long (*intr_pending)(ctlr_info_t *h); + bool (*intr_pending)(ctlr_info_t *h); unsigned long (*command_completed)(ctlr_info_t *h); }; typedef struct _drive_info_struct @@ -85,8 +85,8 @@ struct ctlr_info int max_cmd_sgentries; SGDescriptor_struct **cmd_sg_list; -# define DOORBELL_INT 0 -# define PERF_MODE_INT 1 +# define PERF_MODE_INT 0 +# define DOORBELL_INT 1 # define SIMPLE_MODE_INT 2 # define MEMQ_MODE_INT 3 unsigned int intr[4]; @@ -137,10 +137,27 @@ struct ctlr_info struct list_head scan_list; struct completion scan_wait; struct device dev; + /* + * Performant mode tables. + */ + u32 trans_support; + u32 trans_offset; + struct TransTable_struct *transtable; + unsigned long transMethod; + + /* + * Performant mode completion buffer + */ + u64 *reply_pool; + dma_addr_t reply_pool_dhandle; + u64 *reply_pool_head; + size_t reply_pool_size; + unsigned char reply_pool_wraparound; + u32 *blockFetchTable; }; -/* Defining the diffent access_menthods */ -/* +/* Defining the diffent access_methods + * * Memory mapped FIFO interface (SMART 53xx cards) */ #define SA5_DOORBELL 0x20 @@ -159,19 +176,47 @@ struct ctlr_info #define SA5B_INTR_PENDING 0x04 #define FIFO_EMPTY 0xffffffff #define CCISS_FIRMWARE_READY 0xffff0000 /* value in scratchpad register */ +/* Perf. mode flags */ +#define SA5_PERF_INTR_PENDING 0x04 +#define SA5_PERF_INTR_OFF 0x05 +#define SA5_OUTDB_STATUS_PERF_BIT 0x01 +#define SA5_OUTDB_CLEAR_PERF_BIT 0x01 +#define SA5_OUTDB_CLEAR 0xA0 +#define SA5_OUTDB_CLEAR_PERF_BIT 0x01 +#define SA5_OUTDB_STATUS 0x9C + #define CISS_ERROR_BIT 0x02 #define CCISS_INTR_ON 1 #define CCISS_INTR_OFF 0 + + +/* CCISS_BOARD_READY_WAIT_SECS is how long to wait for a board + * to become ready, in seconds, before giving up on it. + * CCISS_BOARD_READY_POLL_INTERVAL_MSECS * is how long to wait + * between polling the board to see if it is ready, in + * milliseconds. CCISS_BOARD_READY_ITERATIONS is derived + * the above. + */ +#define CCISS_BOARD_READY_WAIT_SECS (120) +#define CCISS_BOARD_READY_POLL_INTERVAL_MSECS (100) +#define CCISS_BOARD_READY_ITERATIONS \ + ((CCISS_BOARD_READY_WAIT_SECS * 1000) / \ + CCISS_BOARD_READY_POLL_INTERVAL_MSECS) +#define CCISS_POST_RESET_PAUSE_MSECS (3000) +#define CCISS_POST_RESET_NOOP_INTERVAL_MSECS (1000) +#define CCISS_POST_RESET_NOOP_RETRIES (12) + /* Send the command to the hardware */ static void SA5_submit_command( ctlr_info_t *h, CommandList_struct *c) { #ifdef CCISS_DEBUG - printk("Sending %x - down to controller\n", c->busaddr ); -#endif /* CCISS_DEBUG */ + printk(KERN_WARNING "cciss%d: Sending %08x - down to controller\n", + h->ctlr, c->busaddr); +#endif /* CCISS_DEBUG */ writel(c->busaddr, h->vaddr + SA5_REQUEST_PORT_OFFSET); h->commands_outstanding++; if ( h->commands_outstanding > h->max_outstanding) @@ -214,6 +259,20 @@ static void SA5B_intr_mask(ctlr_info_t *h, unsigned long val) h->vaddr + SA5_REPLY_INTR_MASK_OFFSET); } } + +/* Performant mode intr_mask */ +static void SA5_performant_intr_mask(ctlr_info_t *h, unsigned long val) +{ + if (val) { /* turn on interrupts */ + h->interrupts_enabled = 1; + writel(0, h->vaddr + SA5_REPLY_INTR_MASK_OFFSET); + } else { + h->interrupts_enabled = 0; + writel(SA5_PERF_INTR_OFF, + h->vaddr + SA5_REPLY_INTR_MASK_OFFSET); + } +} + /* * Returns true if fifo is full. * @@ -250,10 +309,44 @@ static unsigned long SA5_completed(ctlr_info_t *h) return ( register_value); } + +/* Performant mode command completed */ +static unsigned long SA5_performant_completed(ctlr_info_t *h) +{ + unsigned long register_value = FIFO_EMPTY; + + /* flush the controller write of the reply queue by reading + * outbound doorbell status register. + */ + register_value = readl(h->vaddr + SA5_OUTDB_STATUS); + /* msi auto clears the interrupt pending bit. */ + if (!(h->msi_vector || h->msix_vector)) { + writel(SA5_OUTDB_CLEAR_PERF_BIT, h->vaddr + SA5_OUTDB_CLEAR); + /* Do a read in order to flush the write to the controller + * (as per spec.) + */ + register_value = readl(h->vaddr + SA5_OUTDB_STATUS); + } + + if ((*(h->reply_pool_head) & 1) == (h->reply_pool_wraparound)) { + register_value = *(h->reply_pool_head); + (h->reply_pool_head)++; + h->commands_outstanding--; + } else { + register_value = FIFO_EMPTY; + } + /* Check for wraparound */ + if (h->reply_pool_head == (h->reply_pool + h->max_commands)) { + h->reply_pool_head = h->reply_pool; + h->reply_pool_wraparound ^= 1; + } + + return register_value; +} /* * Returns true if an interrupt is pending.. */ -static unsigned long SA5_intr_pending(ctlr_info_t *h) +static bool SA5_intr_pending(ctlr_info_t *h) { unsigned long register_value = readl(h->vaddr + SA5_INTR_STATUS); @@ -268,7 +361,7 @@ static unsigned long SA5_intr_pending(ctlr_info_t *h) /* * Returns true if an interrupt is pending.. */ -static unsigned long SA5B_intr_pending(ctlr_info_t *h) +static bool SA5B_intr_pending(ctlr_info_t *h) { unsigned long register_value = readl(h->vaddr + SA5_INTR_STATUS); @@ -280,6 +373,20 @@ static unsigned long SA5B_intr_pending(ctlr_info_t *h) return 0 ; } +static bool SA5_performant_intr_pending(ctlr_info_t *h) +{ + unsigned long register_value = readl(h->vaddr + SA5_INTR_STATUS); + + if (!register_value) + return false; + + if (h->msi_vector || h->msix_vector) + return true; + + /* Read outbound doorbell to flush */ + register_value = readl(h->vaddr + SA5_OUTDB_STATUS); + return register_value & SA5_OUTDB_STATUS_PERF_BIT; +} static struct access_method SA5_access = { SA5_submit_command, @@ -297,6 +404,14 @@ static struct access_method SA5B_access = { SA5_completed, }; +static struct access_method SA5_performant_access = { + SA5_submit_command, + SA5_performant_intr_mask, + SA5_fifo_full, + SA5_performant_intr_pending, + SA5_performant_completed, +}; + struct board_type { __u32 board_id; char *product_name; @@ -304,6 +419,4 @@ struct board_type { int nr_cmds; /* Max cmds this kind of ctlr can handle. */ }; -#define CCISS_LOCK(i) (&hba[i]->lock) - #endif /* CCISS_H */ diff --git a/drivers/block/cciss_cmd.h b/drivers/block/cciss_cmd.h index e624ff9..eb060f1 100644 --- a/drivers/block/cciss_cmd.h +++ b/drivers/block/cciss_cmd.h @@ -52,8 +52,10 @@ /* Configuration Table */ #define CFGTBL_ChangeReq 0x00000001l #define CFGTBL_AccCmds 0x00000001l +#define DOORBELL_CTLR_RESET 0x00000004l #define CFGTBL_Trans_Simple 0x00000002l +#define CFGTBL_Trans_Performant 0x00000004l #define CFGTBL_BusType_Ultra2 0x00000001l #define CFGTBL_BusType_Ultra3 0x00000002l @@ -173,12 +175,15 @@ typedef struct _SGDescriptor_struct { * PAD_64 can be adjusted independently as needed for 32-bit * and 64-bits systems. */ -#define COMMANDLIST_ALIGNMENT (8) +#define COMMANDLIST_ALIGNMENT (32) #define IS_64_BIT ((sizeof(long) - 4)/4) #define IS_32_BIT (!IS_64_BIT) #define PAD_32 (0) #define PAD_64 (4) #define PADSIZE (IS_32_BIT * PAD_32 + IS_64_BIT * PAD_64) +#define DIRECT_LOOKUP_BIT 0x10 +#define DIRECT_LOOKUP_SHIFT 5 + typedef struct _CommandList_struct { CommandListHeader_struct Header; RequestBlock_struct Request; @@ -195,7 +200,7 @@ typedef struct _CommandList_struct { struct completion *waiting; int retry_count; void * scsi_cmd; - char pad[PADSIZE]; + char pad[PADSIZE]; } CommandList_struct; /* Configuration Table Structure */ @@ -209,12 +214,15 @@ typedef struct _HostWrite_struct { typedef struct _CfgTable_struct { BYTE Signature[4]; DWORD SpecValence; +#define SIMPLE_MODE 0x02 +#define PERFORMANT_MODE 0x04 +#define MEMQ_MODE 0x08 DWORD TransportSupport; DWORD TransportActive; HostWrite_struct HostWrite; DWORD CmdsOutMax; DWORD BusTypes; - DWORD Reserved; + DWORD TransMethodOffset; BYTE ServerName[16]; DWORD HeartBeat; DWORD SCSI_Prefetch; @@ -222,6 +230,28 @@ typedef struct _CfgTable_struct { DWORD MaxLogicalUnits; DWORD MaxPhysicalDrives; DWORD MaxPhysicalDrivesPerLogicalUnit; + DWORD MaxPerformantModeCommands; + u8 reserved[0x78 - 0x58]; + u32 misc_fw_support; /* offset 0x78 */ +#define MISC_FW_DOORBELL_RESET (0x02) } CfgTable_struct; + +struct TransTable_struct { + u32 BlockFetch0; + u32 BlockFetch1; + u32 BlockFetch2; + u32 BlockFetch3; + u32 BlockFetch4; + u32 BlockFetch5; + u32 BlockFetch6; + u32 BlockFetch7; + u32 RepQSize; + u32 RepQCount; + u32 RepQCtrAddrLow32; + u32 RepQCtrAddrHigh32; + u32 RepQAddr0Low32; + u32 RepQAddr0High32; +}; + #pragma pack() #endif /* CCISS_CMD_H */ diff --git a/drivers/block/cciss_scsi.c b/drivers/block/cciss_scsi.c index 72dae92..575495f 100644 --- a/drivers/block/cciss_scsi.c +++ b/drivers/block/cciss_scsi.c @@ -44,13 +44,15 @@ #define CCISS_ABORT_MSG 0x00 #define CCISS_RESET_MSG 0x01 -static int fill_cmd(CommandList_struct *c, __u8 cmd, int ctlr, void *buff, +static int fill_cmd(ctlr_info_t *h, CommandList_struct *c, __u8 cmd, void *buff, size_t size, __u8 page_code, unsigned char *scsi3addr, int cmd_type); -static CommandList_struct *cmd_alloc(ctlr_info_t *h, int get_from_pool); -static void cmd_free(ctlr_info_t *h, CommandList_struct *c, int got_from_pool); +static CommandList_struct *cmd_alloc(ctlr_info_t *h); +static CommandList_struct *cmd_special_alloc(ctlr_info_t *h); +static void cmd_free(ctlr_info_t *h, CommandList_struct *c); +static void cmd_special_free(ctlr_info_t *h, CommandList_struct *c); static int cciss_scsi_proc_info( struct Scsi_Host *sh, @@ -93,8 +95,8 @@ static struct scsi_host_template cciss_driver_template = { #pragma pack(1) -#define SCSI_PAD_32 0 -#define SCSI_PAD_64 0 +#define SCSI_PAD_32 8 +#define SCSI_PAD_64 8 struct cciss_scsi_cmd_stack_elem_t { CommandList_struct cmd; @@ -127,16 +129,16 @@ struct cciss_scsi_adapter_data_t { spinlock_t lock; // to protect ccissscsi[ctlr]; }; -#define CPQ_TAPE_LOCK(ctlr, flags) spin_lock_irqsave( \ - &hba[ctlr]->scsi_ctlr->lock, flags); -#define CPQ_TAPE_UNLOCK(ctlr, flags) spin_unlock_irqrestore( \ - &hba[ctlr]->scsi_ctlr->lock, flags); +#define CPQ_TAPE_LOCK(h, flags) spin_lock_irqsave( \ + &h->scsi_ctlr->lock, flags); +#define CPQ_TAPE_UNLOCK(h, flags) spin_unlock_irqrestore( \ + &h->scsi_ctlr->lock, flags); static CommandList_struct * scsi_cmd_alloc(ctlr_info_t *h) { /* assume only one process in here at a time, locking done by caller. */ - /* use CCISS_LOCK(ctlr) */ + /* use h->lock */ /* might be better to rewrite how we allocate scsi commands in a way that */ /* needs no locking at all. */ @@ -177,10 +179,10 @@ scsi_cmd_alloc(ctlr_info_t *h) } static void -scsi_cmd_free(ctlr_info_t *h, CommandList_struct *cmd) +scsi_cmd_free(ctlr_info_t *h, CommandList_struct *c) { /* assume only one process in here at a time, locking done by caller. */ - /* use CCISS_LOCK(ctlr) */ + /* use h->lock */ /* drop the free memory chunk on top of the stack. */ struct cciss_scsi_adapter_data_t *sa; @@ -190,22 +192,23 @@ scsi_cmd_free(ctlr_info_t *h, CommandList_struct *cmd) stk = &sa->cmd_stack; stk->top++; if (stk->top >= CMD_STACK_SIZE) { - printk("cciss: scsi_cmd_free called too many times.\n"); + dev_err(&h->pdev->dev, + "scsi_cmd_free called too many times.\n"); BUG(); } - stk->elem[stk->top] = (struct cciss_scsi_cmd_stack_elem_t *) cmd; + stk->elem[stk->top] = (struct cciss_scsi_cmd_stack_elem_t *) c; } static int -scsi_cmd_stack_setup(int ctlr, struct cciss_scsi_adapter_data_t *sa) +scsi_cmd_stack_setup(ctlr_info_t *h, struct cciss_scsi_adapter_data_t *sa) { int i; struct cciss_scsi_cmd_stack_t *stk; size_t size; - sa->cmd_sg_list = cciss_allocate_sg_chain_blocks(hba[ctlr], - hba[ctlr]->chainsize, CMD_STACK_SIZE); - if (!sa->cmd_sg_list && hba[ctlr]->chainsize > 0) + sa->cmd_sg_list = cciss_allocate_sg_chain_blocks(h, + h->chainsize, CMD_STACK_SIZE); + if (!sa->cmd_sg_list && h->chainsize > 0) return -ENOMEM; stk = &sa->cmd_stack; @@ -215,7 +218,7 @@ scsi_cmd_stack_setup(int ctlr, struct cciss_scsi_adapter_data_t *sa) BUILD_BUG_ON((sizeof(*stk->pool) % COMMANDLIST_ALIGNMENT) != 0); /* pci_alloc_consistent guarantees 32-bit DMA address will be used */ stk->pool = (struct cciss_scsi_cmd_stack_elem_t *) - pci_alloc_consistent(hba[ctlr]->pdev, size, &stk->cmd_pool_handle); + pci_alloc_consistent(h->pdev, size, &stk->cmd_pool_handle); if (stk->pool == NULL) { cciss_free_sg_chain_blocks(sa->cmd_sg_list, CMD_STACK_SIZE); @@ -234,23 +237,22 @@ scsi_cmd_stack_setup(int ctlr, struct cciss_scsi_adapter_data_t *sa) } static void -scsi_cmd_stack_free(int ctlr) +scsi_cmd_stack_free(ctlr_info_t *h) { struct cciss_scsi_adapter_data_t *sa; struct cciss_scsi_cmd_stack_t *stk; size_t size; - sa = hba[ctlr]->scsi_ctlr; + sa = h->scsi_ctlr; stk = &sa->cmd_stack; if (stk->top != CMD_STACK_SIZE-1) { - printk( "cciss: %d scsi commands are still outstanding.\n", + dev_warn(&h->pdev->dev, + "bug: %d scsi commands are still outstanding.\n", CMD_STACK_SIZE - stk->top); - // BUG(); - printk("WE HAVE A BUG HERE!!! stk=0x%p\n", stk); } size = sizeof(struct cciss_scsi_cmd_stack_elem_t) * CMD_STACK_SIZE; - pci_free_consistent(hba[ctlr]->pdev, size, stk->pool, stk->cmd_pool_handle); + pci_free_consistent(h->pdev, size, stk->pool, stk->cmd_pool_handle); stk->pool = NULL; cciss_free_sg_chain_blocks(sa->cmd_sg_list, CMD_STACK_SIZE); } @@ -342,20 +344,20 @@ print_cmd(CommandList_struct *cp) #endif static int -find_bus_target_lun(int ctlr, int *bus, int *target, int *lun) +find_bus_target_lun(ctlr_info_t *h, int *bus, int *target, int *lun) { /* finds an unused bus, target, lun for a new device */ - /* assumes hba[ctlr]->scsi_ctlr->lock is held */ + /* assumes h->scsi_ctlr->lock is held */ int i, found=0; unsigned char target_taken[CCISS_MAX_SCSI_DEVS_PER_HBA]; memset(&target_taken[0], 0, CCISS_MAX_SCSI_DEVS_PER_HBA); target_taken[SELF_SCSI_ID] = 1; - for (i=0;i<ccissscsi[ctlr].ndevices;i++) - target_taken[ccissscsi[ctlr].dev[i].target] = 1; + for (i = 0; i < ccissscsi[h->ctlr].ndevices; i++) + target_taken[ccissscsi[h->ctlr].dev[i].target] = 1; - for (i=0;i<CCISS_MAX_SCSI_DEVS_PER_HBA;i++) { + for (i = 0; i < CCISS_MAX_SCSI_DEVS_PER_HBA; i++) { if (!target_taken[i]) { *bus = 0; *target=i; *lun = 0; found=1; break; @@ -369,19 +371,19 @@ struct scsi2map { }; static int -cciss_scsi_add_entry(int ctlr, int hostno, +cciss_scsi_add_entry(ctlr_info_t *h, int hostno, struct cciss_scsi_dev_t *device, struct scsi2map *added, int *nadded) { - /* assumes hba[ctlr]->scsi_ctlr->lock is held */ - int n = ccissscsi[ctlr].ndevices; + /* assumes h->scsi_ctlr->lock is held */ + int n = ccissscsi[h->ctlr].ndevices; struct cciss_scsi_dev_t *sd; int i, bus, target, lun; unsigned char addr1[8], addr2[8]; if (n >= CCISS_MAX_SCSI_DEVS_PER_HBA) { - printk("cciss%d: Too many devices, " - "some will be inaccessible.\n", ctlr); + dev_warn(&h->pdev->dev, "Too many devices, " + "some will be inaccessible.\n"); return -1; } @@ -397,7 +399,7 @@ cciss_scsi_add_entry(int ctlr, int hostno, memcpy(addr1, device->scsi3addr, 8); addr1[4] = 0; for (i = 0; i < n; i++) { - sd = &ccissscsi[ctlr].dev[i]; + sd = &ccissscsi[h->ctlr].dev[i]; memcpy(addr2, sd->scsi3addr, 8); addr2[4] = 0; /* differ only in byte 4? */ @@ -410,9 +412,9 @@ cciss_scsi_add_entry(int ctlr, int hostno, } } - sd = &ccissscsi[ctlr].dev[n]; + sd = &ccissscsi[h->ctlr].dev[n]; if (lun == 0) { - if (find_bus_target_lun(ctlr, + if (find_bus_target_lun(h, &sd->bus, &sd->target, &sd->lun) != 0) return -1; } else { @@ -431,37 +433,37 @@ cciss_scsi_add_entry(int ctlr, int hostno, memcpy(sd->device_id, device->device_id, sizeof(sd->device_id)); sd->devtype = device->devtype; - ccissscsi[ctlr].ndevices++; + ccissscsi[h->ctlr].ndevices++; /* initially, (before registering with scsi layer) we don't know our hostno and we don't want to print anything first time anyway (the scsi layer's inquiries will show that info) */ if (hostno != -1) - printk("cciss%d: %s device c%db%dt%dl%d added.\n", - ctlr, scsi_device_type(sd->devtype), hostno, + dev_info(&h->pdev->dev, "%s device c%db%dt%dl%d added.\n", + scsi_device_type(sd->devtype), hostno, sd->bus, sd->target, sd->lun); return 0; } static void -cciss_scsi_remove_entry(int ctlr, int hostno, int entry, +cciss_scsi_remove_entry(ctlr_info_t *h, int hostno, int entry, struct scsi2map *removed, int *nremoved) { - /* assumes hba[ctlr]->scsi_ctlr->lock is held */ + /* assumes h->ctlr]->scsi_ctlr->lock is held */ int i; struct cciss_scsi_dev_t sd; if (entry < 0 || entry >= CCISS_MAX_SCSI_DEVS_PER_HBA) return; - sd = ccissscsi[ctlr].dev[entry]; + sd = ccissscsi[h->ctlr].dev[entry]; removed[*nremoved].bus = sd.bus; removed[*nremoved].target = sd.target; removed[*nremoved].lun = sd.lun; (*nremoved)++; - for (i=entry;i<ccissscsi[ctlr].ndevices-1;i++) - ccissscsi[ctlr].dev[i] = ccissscsi[ctlr].dev[i+1]; - ccissscsi[ctlr].ndevices--; - printk("cciss%d: %s device c%db%dt%dl%d removed.\n", - ctlr, scsi_device_type(sd.devtype), hostno, + for (i = entry; i < ccissscsi[h->ctlr].ndevices-1; i++) + ccissscsi[h->ctlr].dev[i] = ccissscsi[h->ctlr].dev[i+1]; + ccissscsi[h->ctlr].ndevices--; + dev_info(&h->pdev->dev, "%s device c%db%dt%dl%d removed.\n", + scsi_device_type(sd.devtype), hostno, sd.bus, sd.target, sd.lun); } @@ -476,24 +478,24 @@ cciss_scsi_remove_entry(int ctlr, int hostno, int entry, (a)[1] == (b)[1] && \ (a)[0] == (b)[0]) -static void fixup_botched_add(int ctlr, char *scsi3addr) +static void fixup_botched_add(ctlr_info_t *h, char *scsi3addr) { /* called when scsi_add_device fails in order to re-adjust */ /* ccissscsi[] to match the mid layer's view. */ unsigned long flags; int i, j; - CPQ_TAPE_LOCK(ctlr, flags); - for (i = 0; i < ccissscsi[ctlr].ndevices; i++) { + CPQ_TAPE_LOCK(h, flags); + for (i = 0; i < ccissscsi[h->ctlr].ndevices; i++) { if (memcmp(scsi3addr, - ccissscsi[ctlr].dev[i].scsi3addr, 8) == 0) { - for (j = i; j < ccissscsi[ctlr].ndevices-1; j++) - ccissscsi[ctlr].dev[j] = - ccissscsi[ctlr].dev[j+1]; - ccissscsi[ctlr].ndevices--; + ccissscsi[h->ctlr].dev[i].scsi3addr, 8) == 0) { + for (j = i; j < ccissscsi[h->ctlr].ndevices-1; j++) + ccissscsi[h->ctlr].dev[j] = + ccissscsi[h->ctlr].dev[j+1]; + ccissscsi[h->ctlr].ndevices--; break; } } - CPQ_TAPE_UNLOCK(ctlr, flags); + CPQ_TAPE_UNLOCK(h, flags); } static int device_is_the_same(struct cciss_scsi_dev_t *dev1, @@ -513,7 +515,7 @@ static int device_is_the_same(struct cciss_scsi_dev_t *dev1, } static int -adjust_cciss_scsi_table(int ctlr, int hostno, +adjust_cciss_scsi_table(ctlr_info_t *h, int hostno, struct cciss_scsi_dev_t sd[], int nsds) { /* sd contains scsi3 addresses and devtypes, but @@ -534,15 +536,15 @@ adjust_cciss_scsi_table(int ctlr, int hostno, GFP_KERNEL); if (!added || !removed) { - printk(KERN_WARNING "cciss%d: Out of memory in " - "adjust_cciss_scsi_table\n", ctlr); + dev_warn(&h->pdev->dev, + "Out of memory in adjust_cciss_scsi_table\n"); goto free_and_out; } - CPQ_TAPE_LOCK(ctlr, flags); + CPQ_TAPE_LOCK(h, flags); if (hostno != -1) /* if it's not the first time... */ - sh = hba[ctlr]->scsi_ctlr->scsi_host; + sh = h->scsi_ctlr->scsi_host; /* find any devices in ccissscsi[] that are not in sd[] and remove them from ccissscsi[] */ @@ -550,8 +552,8 @@ adjust_cciss_scsi_table(int ctlr, int hostno, i = 0; nremoved = 0; nadded = 0; - while(i<ccissscsi[ctlr].ndevices) { - csd = &ccissscsi[ctlr].dev[i]; + while (i < ccissscsi[h->ctlr].ndevices) { + csd = &ccissscsi[h->ctlr].dev[i]; found=0; for (j=0;j<nsds;j++) { if (SCSI3ADDR_EQ(sd[j].scsi3addr, @@ -566,20 +568,18 @@ adjust_cciss_scsi_table(int ctlr, int hostno, if (found == 0) { /* device no longer present. */ changes++; - /* printk("cciss%d: %s device c%db%dt%dl%d removed.\n", - ctlr, scsi_device_type(csd->devtype), hostno, - csd->bus, csd->target, csd->lun); */ - cciss_scsi_remove_entry(ctlr, hostno, i, + cciss_scsi_remove_entry(h, hostno, i, removed, &nremoved); /* remove ^^^, hence i not incremented */ } else if (found == 1) { /* device is different in some way */ changes++; - printk("cciss%d: device c%db%dt%dl%d has changed.\n", - ctlr, hostno, csd->bus, csd->target, csd->lun); - cciss_scsi_remove_entry(ctlr, hostno, i, + dev_info(&h->pdev->dev, + "device c%db%dt%dl%d has changed.\n", + hostno, csd->bus, csd->target, csd->lun); + cciss_scsi_remove_entry(h, hostno, i, removed, &nremoved); /* remove ^^^, hence i not incremented */ - if (cciss_scsi_add_entry(ctlr, hostno, &sd[j], + if (cciss_scsi_add_entry(h, hostno, &sd[j], added, &nadded) != 0) /* we just removed one, so add can't fail. */ BUG(); @@ -601,8 +601,8 @@ adjust_cciss_scsi_table(int ctlr, int hostno, for (i=0;i<nsds;i++) { found=0; - for (j=0;j<ccissscsi[ctlr].ndevices;j++) { - csd = &ccissscsi[ctlr].dev[j]; + for (j = 0; j < ccissscsi[h->ctlr].ndevices; j++) { + csd = &ccissscsi[h->ctlr].dev[j]; if (SCSI3ADDR_EQ(sd[i].scsi3addr, csd->scsi3addr)) { if (device_is_the_same(&sd[i], csd)) @@ -614,18 +614,18 @@ adjust_cciss_scsi_table(int ctlr, int hostno, } if (!found) { changes++; - if (cciss_scsi_add_entry(ctlr, hostno, &sd[i], + if (cciss_scsi_add_entry(h, hostno, &sd[i], added, &nadded) != 0) break; } else if (found == 1) { /* should never happen... */ changes++; - printk(KERN_WARNING "cciss%d: device " - "unexpectedly changed\n", ctlr); + dev_warn(&h->pdev->dev, + "device unexpectedly changed\n"); /* but if it does happen, we just ignore that device */ } } - CPQ_TAPE_UNLOCK(ctlr, flags); + CPQ_TAPE_UNLOCK(h, flags); /* Don't notify scsi mid layer of any changes the first time through */ /* (or if there are no changes) scsi_scan_host will do it later the */ @@ -645,9 +645,9 @@ adjust_cciss_scsi_table(int ctlr, int hostno, /* We don't expect to get here. */ /* future cmds to this device will get selection */ /* timeout as if the device was gone. */ - printk(KERN_WARNING "cciss%d: didn't find " + dev_warn(&h->pdev->dev, "didn't find " "c%db%dt%dl%d\n for removal.", - ctlr, hostno, removed[i].bus, + hostno, removed[i].bus, removed[i].target, removed[i].lun); } } @@ -659,13 +659,12 @@ adjust_cciss_scsi_table(int ctlr, int hostno, added[i].target, added[i].lun); if (rc == 0) continue; - printk(KERN_WARNING "cciss%d: scsi_add_device " + dev_warn(&h->pdev->dev, "scsi_add_device " "c%db%dt%dl%d failed, device not added.\n", - ctlr, hostno, - added[i].bus, added[i].target, added[i].lun); + hostno, added[i].bus, added[i].target, added[i].lun); /* now we have to remove it from ccissscsi, */ /* since it didn't get added to scsi mid layer */ - fixup_botched_add(ctlr, added[i].scsi3addr); + fixup_botched_add(h, added[i].scsi3addr); } free_and_out: @@ -675,33 +674,33 @@ free_and_out: } static int -lookup_scsi3addr(int ctlr, int bus, int target, int lun, char *scsi3addr) +lookup_scsi3addr(ctlr_info_t *h, int bus, int target, int lun, char *scsi3addr) { int i; struct cciss_scsi_dev_t *sd; unsigned long flags; - CPQ_TAPE_LOCK(ctlr, flags); - for (i=0;i<ccissscsi[ctlr].ndevices;i++) { - sd = &ccissscsi[ctlr].dev[i]; + CPQ_TAPE_LOCK(h, flags); + for (i = 0; i < ccissscsi[h->ctlr].ndevices; i++) { + sd = &ccissscsi[h->ctlr].dev[i]; if (sd->bus == bus && sd->target == target && sd->lun == lun) { memcpy(scsi3addr, &sd->scsi3addr[0], 8); - CPQ_TAPE_UNLOCK(ctlr, flags); + CPQ_TAPE_UNLOCK(h, flags); return 0; } } - CPQ_TAPE_UNLOCK(ctlr, flags); + CPQ_TAPE_UNLOCK(h, flags); return -1; } static void -cciss_scsi_setup(int cntl_num) +cciss_scsi_setup(ctlr_info_t *h) { struct cciss_scsi_adapter_data_t * shba; - ccissscsi[cntl_num].ndevices = 0; + ccissscsi[h->ctlr].ndevices = 0; shba = (struct cciss_scsi_adapter_data_t *) kmalloc(sizeof(*shba), GFP_KERNEL); if (shba == NULL) @@ -709,35 +708,35 @@ cciss_scsi_setup(int cntl_num) shba->scsi_host = NULL; spin_lock_init(&shba->lock); shba->registered = 0; - if (scsi_cmd_stack_setup(cntl_num, shba) != 0) { + if (scsi_cmd_stack_setup(h, shba) != 0) { kfree(shba); shba = NULL; } - hba[cntl_num]->scsi_ctlr = shba; + h->scsi_ctlr = shba; return; } -static void -complete_scsi_command( CommandList_struct *cp, int timeout, __u32 tag) +static void complete_scsi_command(CommandList_struct *c, int timeout, + __u32 tag) { struct scsi_cmnd *cmd; - ctlr_info_t *ctlr; + ctlr_info_t *h; ErrorInfo_struct *ei; - ei = cp->err_info; + ei = c->err_info; /* First, see if it was a message rather than a command */ - if (cp->Request.Type.Type == TYPE_MSG) { - cp->cmd_type = CMD_MSG_DONE; + if (c->Request.Type.Type == TYPE_MSG) { + c->cmd_type = CMD_MSG_DONE; return; } - cmd = (struct scsi_cmnd *) cp->scsi_cmd; - ctlr = hba[cp->ctlr]; + cmd = (struct scsi_cmnd *) c->scsi_cmd; + h = hba[c->ctlr]; scsi_dma_unmap(cmd); - if (cp->Header.SGTotal > ctlr->max_cmd_sgentries) - cciss_unmap_sg_chain_block(ctlr, cp); + if (c->Header.SGTotal > h->max_cmd_sgentries) + cciss_unmap_sg_chain_block(h, c); cmd->result = (DID_OK << 16); /* host byte */ cmd->result |= (COMMAND_COMPLETE << 8); /* msg byte */ @@ -764,9 +763,8 @@ complete_scsi_command( CommandList_struct *cp, int timeout, __u32 tag) { #if 0 printk(KERN_WARNING "cciss: cmd %p " - "has SCSI Status = %x\n", - cp, - ei->ScsiStatus); + "has SCSI Status = %x\n", + c, ei->ScsiStatus); #endif cmd->result |= (ei->ScsiStatus << 1); } @@ -786,13 +784,13 @@ complete_scsi_command( CommandList_struct *cp, int timeout, __u32 tag) case CMD_DATA_UNDERRUN: /* let mid layer handle it. */ break; case CMD_DATA_OVERRUN: - printk(KERN_WARNING "cciss: cp %p has" + dev_warn(&h->pdev->dev, "%p has" " completed with data overrun " - "reported\n", cp); + "reported\n", c); break; case CMD_INVALID: { - /* print_bytes(cp, sizeof(*cp), 1, 0); - print_cmd(cp); */ + /* print_bytes(c, sizeof(*c), 1, 0); + print_cmd(c); */ /* We get CMD_INVALID if you address a non-existent tape drive instead of a selection timeout (no response). You will see this if you yank out a tape drive, then try to access it. This is kind of a shame @@ -802,54 +800,50 @@ complete_scsi_command( CommandList_struct *cp, int timeout, __u32 tag) } break; case CMD_PROTOCOL_ERR: - printk(KERN_WARNING "cciss: cp %p has " - "protocol error \n", cp); + dev_warn(&h->pdev->dev, + "%p has protocol error\n", c); break; case CMD_HARDWARE_ERR: cmd->result = DID_ERROR << 16; - printk(KERN_WARNING "cciss: cp %p had " - " hardware error\n", cp); + dev_warn(&h->pdev->dev, + "%p had hardware error\n", c); break; case CMD_CONNECTION_LOST: cmd->result = DID_ERROR << 16; - printk(KERN_WARNING "cciss: cp %p had " - "connection lost\n", cp); + dev_warn(&h->pdev->dev, + "%p had connection lost\n", c); break; case CMD_ABORTED: cmd->result = DID_ABORT << 16; - printk(KERN_WARNING "cciss: cp %p was " - "aborted\n", cp); + dev_warn(&h->pdev->dev, "%p was aborted\n", c); break; case CMD_ABORT_FAILED: cmd->result = DID_ERROR << 16; - printk(KERN_WARNING "cciss: cp %p reports " - "abort failed\n", cp); + dev_warn(&h->pdev->dev, + "%p reports abort failed\n", c); break; case CMD_UNSOLICITED_ABORT: cmd->result = DID_ABORT << 16; - printk(KERN_WARNING "cciss: cp %p aborted " - "do to an unsolicited abort\n", cp); + dev_warn(&h->pdev->dev, "%p aborted do to an " + "unsolicited abort\n", c); break; case CMD_TIMEOUT: cmd->result = DID_TIME_OUT << 16; - printk(KERN_WARNING "cciss: cp %p timedout\n", - cp); + dev_warn(&h->pdev->dev, "%p timedout\n", c); break; default: cmd->result = DID_ERROR << 16; - printk(KERN_WARNING "cciss: cp %p returned " - "unknown status %x\n", cp, + dev_warn(&h->pdev->dev, + "%p returned unknown status %x\n", c, ei->CommandStatus); } } - // printk("c:%p:c%db%dt%dl%d ", cmd, ctlr->ctlr, cmd->channel, - // cmd->target, cmd->lun); cmd->scsi_done(cmd); - scsi_cmd_free(ctlr, cp); + scsi_cmd_free(h, c); } static int -cciss_scsi_detect(int ctlr) +cciss_scsi_detect(ctlr_info_t *h) { struct Scsi_Host *sh; int error; @@ -860,15 +854,15 @@ cciss_scsi_detect(int ctlr) sh->io_port = 0; // good enough? FIXME, sh->n_io_port = 0; // I don't think we use these two... sh->this_id = SELF_SCSI_ID; - sh->sg_tablesize = hba[ctlr]->maxsgentries; + sh->sg_tablesize = h->maxsgentries; sh->max_cmd_len = MAX_COMMAND_SIZE; ((struct cciss_scsi_adapter_data_t *) - hba[ctlr]->scsi_ctlr)->scsi_host = sh; - sh->hostdata[0] = (unsigned long) hba[ctlr]; - sh->irq = hba[ctlr]->intr[SIMPLE_MODE_INT]; + h->scsi_ctlr)->scsi_host = sh; + sh->hostdata[0] = (unsigned long) h; + sh->irq = h->intr[SIMPLE_MODE_INT]; sh->unique_id = sh->irq; - error = scsi_add_host(sh, &hba[ctlr]->pdev->dev); + error = scsi_add_host(sh, &h->pdev->dev); if (error) goto fail_host_put; scsi_scan_host(sh); @@ -882,20 +876,20 @@ cciss_scsi_detect(int ctlr) static void cciss_unmap_one(struct pci_dev *pdev, - CommandList_struct *cp, + CommandList_struct *c, size_t buflen, int data_direction) { u64bit addr64; - addr64.val32.lower = cp->SG[0].Addr.lower; - addr64.val32.upper = cp->SG[0].Addr.upper; + addr64.val32.lower = c->SG[0].Addr.lower; + addr64.val32.upper = c->SG[0].Addr.upper; pci_unmap_single(pdev, (dma_addr_t) addr64.val, buflen, data_direction); } static void cciss_map_one(struct pci_dev *pdev, - CommandList_struct *cp, + CommandList_struct *c, unsigned char *buf, size_t buflen, int data_direction) @@ -903,164 +897,149 @@ cciss_map_one(struct pci_dev *pdev, __u64 addr64; addr64 = (__u64) pci_map_single(pdev, buf, buflen, data_direction); - cp->SG[0].Addr.lower = + c->SG[0].Addr.lower = (__u32) (addr64 & (__u64) 0x00000000FFFFFFFF); - cp->SG[0].Addr.upper = + c->SG[0].Addr.upper = (__u32) ((addr64 >> 32) & (__u64) 0x00000000FFFFFFFF); - cp->SG[0].Len = buflen; - cp->Header.SGList = (__u8) 1; /* no. SGs contig in this cmd */ - cp->Header.SGTotal = (__u16) 1; /* total sgs in this cmd list */ + c->SG[0].Len = buflen; + c->Header.SGList = (__u8) 1; /* no. SGs contig in this cmd */ + c->Header.SGTotal = (__u16) 1; /* total sgs in this cmd list */ } static int -cciss_scsi_do_simple_cmd(ctlr_info_t *c, - CommandList_struct *cp, +cciss_scsi_do_simple_cmd(ctlr_info_t *h, + CommandList_struct *c, unsigned char *scsi3addr, unsigned char *cdb, unsigned char cdblen, unsigned char *buf, int bufsize, int direction) { - unsigned long flags; DECLARE_COMPLETION_ONSTACK(wait); - cp->cmd_type = CMD_IOCTL_PEND; // treat this like an ioctl - cp->scsi_cmd = NULL; - cp->Header.ReplyQueue = 0; // unused in simple mode - memcpy(&cp->Header.LUN, scsi3addr, sizeof(cp->Header.LUN)); - cp->Header.Tag.lower = cp->busaddr; // Use k. address of cmd as tag + c->cmd_type = CMD_IOCTL_PEND; /* treat this like an ioctl */ + c->scsi_cmd = NULL; + c->Header.ReplyQueue = 0; /* unused in simple mode */ + memcpy(&c->Header.LUN, scsi3addr, sizeof(c->Header.LUN)); + c->Header.Tag.lower = c->busaddr; /* Use k. address of cmd as tag */ // Fill in the request block... /* printk("Using scsi3addr 0x%02x%0x2%0x2%0x2%0x2%0x2%0x2%0x2\n", scsi3addr[0], scsi3addr[1], scsi3addr[2], scsi3addr[3], scsi3addr[4], scsi3addr[5], scsi3addr[6], scsi3addr[7]); */ - memset(cp->Request.CDB, 0, sizeof(cp->Request.CDB)); - memcpy(cp->Request.CDB, cdb, cdblen); - cp->Request.Timeout = 0; - cp->Request.CDBLen = cdblen; - cp->Request.Type.Type = TYPE_CMD; - cp->Request.Type.Attribute = ATTR_SIMPLE; - cp->Request.Type.Direction = direction; + memset(c->Request.CDB, 0, sizeof(c->Request.CDB)); + memcpy(c->Request.CDB, cdb, cdblen); + c->Request.Timeout = 0; + c->Request.CDBLen = cdblen; + c->Request.Type.Type = TYPE_CMD; + c->Request.Type.Attribute = ATTR_SIMPLE; + c->Request.Type.Direction = direction; /* Fill in the SG list and do dma mapping */ - cciss_map_one(c->pdev, cp, (unsigned char *) buf, + cciss_map_one(h->pdev, c, (unsigned char *) buf, bufsize, DMA_FROM_DEVICE); - cp->waiting = &wait; - - /* Put the request on the tail of the request queue */ - spin_lock_irqsave(CCISS_LOCK(c->ctlr), flags); - addQ(&c->reqQ, cp); - c->Qdepth++; - start_io(c); - spin_unlock_irqrestore(CCISS_LOCK(c->ctlr), flags); - + c->waiting = &wait; + enqueue_cmd_and_start_io(h, c); wait_for_completion(&wait); /* undo the dma mapping */ - cciss_unmap_one(c->pdev, cp, bufsize, DMA_FROM_DEVICE); + cciss_unmap_one(h->pdev, c, bufsize, DMA_FROM_DEVICE); return(0); } static void -cciss_scsi_interpret_error(CommandList_struct *cp) +cciss_scsi_interpret_error(ctlr_info_t *h, CommandList_struct *c) { ErrorInfo_struct *ei; - ei = cp->err_info; + ei = c->err_info; switch(ei->CommandStatus) { case CMD_TARGET_STATUS: - printk(KERN_WARNING "cciss: cmd %p has " - "completed with errors\n", cp); - printk(KERN_WARNING "cciss: cmd %p " - "has SCSI Status = %x\n", - cp, - ei->ScsiStatus); + dev_warn(&h->pdev->dev, + "cmd %p has completed with errors\n", c); + dev_warn(&h->pdev->dev, + "cmd %p has SCSI Status = %x\n", + c, ei->ScsiStatus); if (ei->ScsiStatus == 0) - printk(KERN_WARNING - "cciss:SCSI status is abnormally zero. " + dev_warn(&h->pdev->dev, + "SCSI status is abnormally zero. " "(probably indicates selection timeout " "reported incorrectly due to a known " "firmware bug, circa July, 2001.)\n"); break; case CMD_DATA_UNDERRUN: /* let mid layer handle it. */ - printk("UNDERRUN\n"); + dev_info(&h->pdev->dev, "UNDERRUN\n"); break; case CMD_DATA_OVERRUN: - printk(KERN_WARNING "cciss: cp %p has" + dev_warn(&h->pdev->dev, "%p has" " completed with data overrun " - "reported\n", cp); + "reported\n", c); break; case CMD_INVALID: { /* controller unfortunately reports SCSI passthru's */ /* to non-existent targets as invalid commands. */ - printk(KERN_WARNING "cciss: cp %p is " - "reported invalid (probably means " - "target device no longer present)\n", - cp); - /* print_bytes((unsigned char *) cp, sizeof(*cp), 1, 0); - print_cmd(cp); */ + dev_warn(&h->pdev->dev, + "%p is reported invalid (probably means " + "target device no longer present)\n", c); + /* print_bytes((unsigned char *) c, sizeof(*c), 1, 0); + print_cmd(c); */ } break; case CMD_PROTOCOL_ERR: - printk(KERN_WARNING "cciss: cp %p has " - "protocol error \n", cp); + dev_warn(&h->pdev->dev, "%p has protocol error\n", c); break; case CMD_HARDWARE_ERR: /* cmd->result = DID_ERROR << 16; */ - printk(KERN_WARNING "cciss: cp %p had " - " hardware error\n", cp); + dev_warn(&h->pdev->dev, "%p had hardware error\n", c); break; case CMD_CONNECTION_LOST: - printk(KERN_WARNING "cciss: cp %p had " - "connection lost\n", cp); + dev_warn(&h->pdev->dev, "%p had connection lost\n", c); break; case CMD_ABORTED: - printk(KERN_WARNING "cciss: cp %p was " - "aborted\n", cp); + dev_warn(&h->pdev->dev, "%p was aborted\n", c); break; case CMD_ABORT_FAILED: - printk(KERN_WARNING "cciss: cp %p reports " - "abort failed\n", cp); + dev_warn(&h->pdev->dev, + "%p reports abort failed\n", c); break; case CMD_UNSOLICITED_ABORT: - printk(KERN_WARNING "cciss: cp %p aborted " - "do to an unsolicited abort\n", cp); + dev_warn(&h->pdev->dev, + "%p aborted do to an unsolicited abort\n", c); break; case CMD_TIMEOUT: - printk(KERN_WARNING "cciss: cp %p timedout\n", - cp); + dev_warn(&h->pdev->dev, "%p timedout\n", c); break; default: - printk(KERN_WARNING "cciss: cp %p returned " - "unknown status %x\n", cp, - ei->CommandStatus); + dev_warn(&h->pdev->dev, + "%p returned unknown status %x\n", + c, ei->CommandStatus); } } static int -cciss_scsi_do_inquiry(ctlr_info_t *c, unsigned char *scsi3addr, +cciss_scsi_do_inquiry(ctlr_info_t *h, unsigned char *scsi3addr, unsigned char page, unsigned char *buf, unsigned char bufsize) { int rc; - CommandList_struct *cp; + CommandList_struct *c; char cdb[6]; ErrorInfo_struct *ei; unsigned long flags; - spin_lock_irqsave(CCISS_LOCK(c->ctlr), flags); - cp = scsi_cmd_alloc(c); - spin_unlock_irqrestore(CCISS_LOCK(c->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); + c = scsi_cmd_alloc(h); + spin_unlock_irqrestore(&h->lock, flags); - if (cp == NULL) { /* trouble... */ + if (c == NULL) { /* trouble... */ printk("cmd_alloc returned NULL!\n"); return -1; } - ei = cp->err_info; + ei = c->err_info; cdb[0] = CISS_INQUIRY; cdb[1] = (page != 0); @@ -1068,24 +1047,24 @@ cciss_scsi_do_inquiry(ctlr_info_t *c, unsigned char *scsi3addr, cdb[3] = 0; cdb[4] = bufsize; cdb[5] = 0; - rc = cciss_scsi_do_simple_cmd(c, cp, scsi3addr, cdb, + rc = cciss_scsi_do_simple_cmd(h, c, scsi3addr, cdb, 6, buf, bufsize, XFER_READ); if (rc != 0) return rc; /* something went wrong */ if (ei->CommandStatus != 0 && ei->CommandStatus != CMD_DATA_UNDERRUN) { - cciss_scsi_interpret_error(cp); + cciss_scsi_interpret_error(h, c); rc = -1; } - spin_lock_irqsave(CCISS_LOCK(c->ctlr), flags); - scsi_cmd_free(c, cp); - spin_unlock_irqrestore(CCISS_LOCK(c->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); + scsi_cmd_free(h, c); + spin_unlock_irqrestore(&h->lock, flags); return rc; } /* Get the device id from inquiry page 0x83 */ -static int cciss_scsi_get_device_id(ctlr_info_t *c, unsigned char *scsi3addr, +static int cciss_scsi_get_device_id(ctlr_info_t *h, unsigned char *scsi3addr, unsigned char *device_id, int buflen) { int rc; @@ -1096,7 +1075,7 @@ static int cciss_scsi_get_device_id(ctlr_info_t *c, unsigned char *scsi3addr, buf = kzalloc(64, GFP_KERNEL); if (!buf) return -1; - rc = cciss_scsi_do_inquiry(c, scsi3addr, 0x83, buf, 64); + rc = cciss_scsi_do_inquiry(h, scsi3addr, 0x83, buf, 64); if (rc == 0) memcpy(device_id, &buf[8], buflen); kfree(buf); @@ -1104,20 +1083,20 @@ static int cciss_scsi_get_device_id(ctlr_info_t *c, unsigned char *scsi3addr, } static int -cciss_scsi_do_report_phys_luns(ctlr_info_t *c, +cciss_scsi_do_report_phys_luns(ctlr_info_t *h, ReportLunData_struct *buf, int bufsize) { int rc; - CommandList_struct *cp; + CommandList_struct *c; unsigned char cdb[12]; unsigned char scsi3addr[8]; ErrorInfo_struct *ei; unsigned long flags; - spin_lock_irqsave(CCISS_LOCK(c->ctlr), flags); - cp = scsi_cmd_alloc(c); - spin_unlock_irqrestore(CCISS_LOCK(c->ctlr), flags); - if (cp == NULL) { /* trouble... */ + spin_lock_irqsave(&h->lock, flags); + c = scsi_cmd_alloc(h); + spin_unlock_irqrestore(&h->lock, flags); + if (c == NULL) { /* trouble... */ printk("cmd_alloc returned NULL!\n"); return -1; } @@ -1136,27 +1115,27 @@ cciss_scsi_do_report_phys_luns(ctlr_info_t *c, cdb[10] = 0; cdb[11] = 0; - rc = cciss_scsi_do_simple_cmd(c, cp, scsi3addr, + rc = cciss_scsi_do_simple_cmd(h, c, scsi3addr, cdb, 12, (unsigned char *) buf, bufsize, XFER_READ); if (rc != 0) return rc; /* something went wrong */ - ei = cp->err_info; + ei = c->err_info; if (ei->CommandStatus != 0 && ei->CommandStatus != CMD_DATA_UNDERRUN) { - cciss_scsi_interpret_error(cp); + cciss_scsi_interpret_error(h, c); rc = -1; } - spin_lock_irqsave(CCISS_LOCK(c->ctlr), flags); - scsi_cmd_free(c, cp); - spin_unlock_irqrestore(CCISS_LOCK(c->ctlr), flags); + spin_lock_irqsave(&h->lock, flags); + scsi_cmd_free(h, c); + spin_unlock_irqrestore(&h->lock, flags); return rc; } static void -cciss_update_non_disk_devices(int cntl_num, int hostno) +cciss_update_non_disk_devices(ctlr_info_t *h, int hostno) { /* the idea here is we could get notified from /proc that some devices have changed, so we do a report @@ -1189,7 +1168,6 @@ cciss_update_non_disk_devices(int cntl_num, int hostno) ReportLunData_struct *ld_buff; unsigned char *inq_buff; unsigned char scsi3addr[8]; - ctlr_info_t *c; __u32 num_luns=0; unsigned char *ch; struct cciss_scsi_dev_t *currentsd, *this_device; @@ -1197,7 +1175,6 @@ cciss_update_non_disk_devices(int cntl_num, int hostno) int reportlunsize = sizeof(*ld_buff) + CISS_MAX_PHYS_LUN * 8; int i; - c = (ctlr_info_t *) hba[cntl_num]; ld_buff = kzalloc(reportlunsize, GFP_KERNEL); inq_buff = kmalloc(OBDR_TAPE_INQ_SIZE, GFP_KERNEL); currentsd = kzalloc(sizeof(*currentsd) * @@ -1207,7 +1184,7 @@ cciss_update_non_disk_devices(int cntl_num, int hostno) goto out; } this_device = ¤tsd[CCISS_MAX_SCSI_DEVS_PER_HBA]; - if (cciss_scsi_do_report_phys_luns(c, ld_buff, reportlunsize) == 0) { + if (cciss_scsi_do_report_phys_luns(h, ld_buff, reportlunsize) == 0) { ch = &ld_buff->LUNListLength[0]; num_luns = ((ch[0]<<24) | (ch[1]<<16) | (ch[2]<<8) | ch[3]) / 8; if (num_luns > CISS_MAX_PHYS_LUN) { @@ -1231,7 +1208,7 @@ cciss_update_non_disk_devices(int cntl_num, int hostno) memset(inq_buff, 0, OBDR_TAPE_INQ_SIZE); memcpy(&scsi3addr[0], &ld_buff->LUN[i][0], 8); - if (cciss_scsi_do_inquiry(hba[cntl_num], scsi3addr, 0, inq_buff, + if (cciss_scsi_do_inquiry(h, scsi3addr, 0, inq_buff, (unsigned char) OBDR_TAPE_INQ_SIZE) != 0) /* Inquiry failed (msg printed already) */ continue; /* so we will skip this device. */ @@ -1249,7 +1226,7 @@ cciss_update_non_disk_devices(int cntl_num, int hostno) sizeof(this_device->revision)); memset(this_device->device_id, 0, sizeof(this_device->device_id)); - cciss_scsi_get_device_id(hba[cntl_num], scsi3addr, + cciss_scsi_get_device_id(h, scsi3addr, this_device->device_id, sizeof(this_device->device_id)); switch (this_device->devtype) @@ -1276,7 +1253,7 @@ cciss_update_non_disk_devices(int cntl_num, int hostno) case 0x08: /* medium changer */ if (ncurrent >= CCISS_MAX_SCSI_DEVS_PER_HBA) { printk(KERN_INFO "cciss%d: %s ignored, " - "too many devices.\n", cntl_num, + "too many devices.\n", h->ctlr, scsi_device_type(this_device->devtype)); break; } @@ -1288,7 +1265,7 @@ cciss_update_non_disk_devices(int cntl_num, int hostno) } } - adjust_cciss_scsi_table(cntl_num, hostno, currentsd, ncurrent); + adjust_cciss_scsi_table(h, hostno, currentsd, ncurrent); out: kfree(inq_buff); kfree(ld_buff); @@ -1307,12 +1284,12 @@ is_keyword(char *ptr, int len, char *verb) // Thanks to ncr53c8xx.c } static int -cciss_scsi_user_command(int ctlr, int hostno, char *buffer, int length) +cciss_scsi_user_command(ctlr_info_t *h, int hostno, char *buffer, int length) { int arg_len; if ((arg_len = is_keyword(buffer, length, "rescan")) != 0) - cciss_update_non_disk_devices(ctlr, hostno); + cciss_update_non_disk_devices(h, hostno); else return -EINVAL; return length; @@ -1329,20 +1306,16 @@ cciss_scsi_proc_info(struct Scsi_Host *sh, { int buflen, datalen; - ctlr_info_t *ci; + ctlr_info_t *h; int i; - int cntl_num; - - ci = (ctlr_info_t *) sh->hostdata[0]; - if (ci == NULL) /* This really shouldn't ever happen. */ + h = (ctlr_info_t *) sh->hostdata[0]; + if (h == NULL) /* This really shouldn't ever happen. */ return -EINVAL; - cntl_num = ci->ctlr; /* Get our index into the hba[] array */ - if (func == 0) { /* User is reading from /proc/scsi/ciss*?/?* */ buflen = sprintf(buffer, "cciss%d: SCSI host: %d\n", - cntl_num, sh->host_no); + h->ctlr, sh->host_no); /* this information is needed by apps to know which cciss device corresponds to which scsi host number without @@ -1352,8 +1325,9 @@ cciss_scsi_proc_info(struct Scsi_Host *sh, this info is for an app to be able to use to know how to get them back in sync. */ - for (i=0;i<ccissscsi[cntl_num].ndevices;i++) { - struct cciss_scsi_dev_t *sd = &ccissscsi[cntl_num].dev[i]; + for (i = 0; i < ccissscsi[h->ctlr].ndevices; i++) { + struct cciss_scsi_dev_t *sd = + &ccissscsi[h->ctlr].dev[i]; buflen += sprintf(&buffer[buflen], "c%db%dt%dl%d %02d " "0x%02x%02x%02x%02x%02x%02x%02x%02x\n", sh->host_no, sd->bus, sd->target, sd->lun, @@ -1371,15 +1345,15 @@ cciss_scsi_proc_info(struct Scsi_Host *sh, *start = buffer + offset; return(datalen); } else /* User is writing to /proc/scsi/cciss*?/?* ... */ - return cciss_scsi_user_command(cntl_num, sh->host_no, + return cciss_scsi_user_command(h, sh->host_no, buffer, length); } /* cciss_scatter_gather takes a struct scsi_cmnd, (cmd), and does the pci dma mapping and fills in the scatter gather entries of the - cciss command, cp. */ + cciss command, c. */ -static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *cp, +static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *c, struct scsi_cmnd *cmd) { unsigned int len; @@ -1393,7 +1367,7 @@ static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *cp, chained = 0; sg_index = 0; - curr_sg = cp->SG; + curr_sg = c->SG; request_nsgs = scsi_dma_map(cmd); if (request_nsgs) { scsi_for_each_sg(cmd, sg, request_nsgs, i) { @@ -1401,7 +1375,7 @@ static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *cp, !chained && request_nsgs - i > 1) { chained = 1; sg_index = 0; - curr_sg = sa->cmd_sg_list[cp->cmdindex]; + curr_sg = sa->cmd_sg_list[c->cmdindex]; } addr64 = (__u64) sg_dma_address(sg); len = sg_dma_len(sg); @@ -1414,19 +1388,19 @@ static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *cp, ++sg_index; } if (chained) - cciss_map_sg_chain_block(h, cp, - sa->cmd_sg_list[cp->cmdindex], + cciss_map_sg_chain_block(h, c, + sa->cmd_sg_list[c->cmdindex], (request_nsgs - (h->max_cmd_sgentries - 1)) * sizeof(SGDescriptor_struct)); } /* track how many SG entries we are using */ if (request_nsgs > h->maxSG) h->maxSG = request_nsgs; - cp->Header.SGTotal = (__u8) request_nsgs + chained; + c->Header.SGTotal = (__u8) request_nsgs + chained; if (request_nsgs > h->max_cmd_sgentries) - cp->Header.SGList = h->max_cmd_sgentries; + c->Header.SGList = h->max_cmd_sgentries; else - cp->Header.SGList = cp->Header.SGTotal; + c->Header.SGList = c->Header.SGTotal; return; } @@ -1434,18 +1408,17 @@ static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *cp, static int cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd *)) { - ctlr_info_t *c; - int ctlr, rc; + ctlr_info_t *h; + int rc; unsigned char scsi3addr[8]; - CommandList_struct *cp; + CommandList_struct *c; unsigned long flags; // Get the ptr to our adapter structure (hba[i]) out of cmd->host. // We violate cmd->host privacy here. (Is there another way?) - c = (ctlr_info_t *) cmd->device->host->hostdata[0]; - ctlr = c->ctlr; + h = (ctlr_info_t *) cmd->device->host->hostdata[0]; - rc = lookup_scsi3addr(ctlr, cmd->device->channel, cmd->device->id, + rc = lookup_scsi3addr(h, cmd->device->channel, cmd->device->id, cmd->device->lun, scsi3addr); if (rc != 0) { /* the scsi nexus does not match any that we presented... */ @@ -1457,19 +1430,14 @@ cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd return 0; } - /* printk("cciss_queue_command, p=%p, cmd=0x%02x, c%db%dt%dl%d\n", - cmd, cmd->cmnd[0], ctlr, cmd->channel, cmd->target, cmd->lun);*/ - // printk("q:%p:c%db%dt%dl%d ", cmd, ctlr, cmd->channel, - // cmd->target, cmd->lun); - /* Ok, we have a reasonable scsi nexus, so send the cmd down, and see what the device thinks of it. */ - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - cp = scsi_cmd_alloc(c); - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); - if (cp == NULL) { /* trouble... */ - printk("scsi_cmd_alloc returned NULL!\n"); + spin_lock_irqsave(&h->lock, flags); + c = scsi_cmd_alloc(h); + spin_unlock_irqrestore(&h->lock, flags); + if (c == NULL) { /* trouble... */ + dev_warn(&h->pdev->dev, "scsi_cmd_alloc returned NULL!\n"); /* FIXME: next 3 lines are -> BAD! <- */ cmd->result = DID_NO_CONNECT << 16; done(cmd); @@ -1480,35 +1448,41 @@ cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd cmd->scsi_done = done; // save this for use by completion code - // save cp in case we have to abort it - cmd->host_scribble = (unsigned char *) cp; + /* save c in case we have to abort it */ + cmd->host_scribble = (unsigned char *) c; - cp->cmd_type = CMD_SCSI; - cp->scsi_cmd = cmd; - cp->Header.ReplyQueue = 0; // unused in simple mode - memcpy(&cp->Header.LUN.LunAddrBytes[0], &scsi3addr[0], 8); - cp->Header.Tag.lower = cp->busaddr; // Use k. address of cmd as tag + c->cmd_type = CMD_SCSI; + c->scsi_cmd = cmd; + c->Header.ReplyQueue = 0; /* unused in simple mode */ + memcpy(&c->Header.LUN.LunAddrBytes[0], &scsi3addr[0], 8); + c->Header.Tag.lower = c->busaddr; /* Use k. address of cmd as tag */ // Fill in the request block... - cp->Request.Timeout = 0; - memset(cp->Request.CDB, 0, sizeof(cp->Request.CDB)); - BUG_ON(cmd->cmd_len > sizeof(cp->Request.CDB)); - cp->Request.CDBLen = cmd->cmd_len; - memcpy(cp->Request.CDB, cmd->cmnd, cmd->cmd_len); - cp->Request.Type.Type = TYPE_CMD; - cp->Request.Type.Attribute = ATTR_SIMPLE; + c->Request.Timeout = 0; + memset(c->Request.CDB, 0, sizeof(c->Request.CDB)); + BUG_ON(cmd->cmd_len > sizeof(c->Request.CDB)); + c->Request.CDBLen = cmd->cmd_len; + memcpy(c->Request.CDB, cmd->cmnd, cmd->cmd_len); + c->Request.Type.Type = TYPE_CMD; + c->Request.Type.Attribute = ATTR_SIMPLE; switch(cmd->sc_data_direction) { - case DMA_TO_DEVICE: cp->Request.Type.Direction = XFER_WRITE; break; - case DMA_FROM_DEVICE: cp->Request.Type.Direction = XFER_READ; break; - case DMA_NONE: cp->Request.Type.Direction = XFER_NONE; break; + case DMA_TO_DEVICE: + c->Request.Type.Direction = XFER_WRITE; + break; + case DMA_FROM_DEVICE: + c->Request.Type.Direction = XFER_READ; + break; + case DMA_NONE: + c->Request.Type.Direction = XFER_NONE; + break; case DMA_BIDIRECTIONAL: // This can happen if a buggy application does a scsi passthru // and sets both inlen and outlen to non-zero. ( see // ../scsi/scsi_ioctl.c:scsi_ioctl_send_command() ) - cp->Request.Type.Direction = XFER_RSVD; + c->Request.Type.Direction = XFER_RSVD; // This is technically wrong, and cciss controllers should // reject it with CMD_INVALID, which is the most correct // response, but non-fibre backends appear to let it @@ -1519,27 +1493,18 @@ cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd break; default: - printk("cciss: unknown data direction: %d\n", + dev_warn(&h->pdev->dev, "unknown data direction: %d\n", cmd->sc_data_direction); BUG(); break; } - cciss_scatter_gather(c, cp, cmd); - - /* Put the request on the tail of the request queue */ - - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - addQ(&c->reqQ, cp); - c->Qdepth++; - start_io(c); - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); - + cciss_scatter_gather(h, c, cmd); + enqueue_cmd_and_start_io(h, c); /* the cmd'll come back via intr handler in complete_scsi_command() */ return 0; } -static void -cciss_unregister_scsi(int ctlr) +static void cciss_unregister_scsi(ctlr_info_t *h) { struct cciss_scsi_adapter_data_t *sa; struct cciss_scsi_cmd_stack_t *stk; @@ -1547,59 +1512,58 @@ cciss_unregister_scsi(int ctlr) /* we are being forcibly unloaded, and may not refuse. */ - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - sa = hba[ctlr]->scsi_ctlr; + spin_lock_irqsave(&h->lock, flags); + sa = h->scsi_ctlr; stk = &sa->cmd_stack; /* if we weren't ever actually registered, don't unregister */ if (sa->registered) { - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); + spin_unlock_irqrestore(&h->lock, flags); scsi_remove_host(sa->scsi_host); scsi_host_put(sa->scsi_host); - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); + spin_lock_irqsave(&h->lock, flags); } /* set scsi_host to NULL so our detect routine will find us on register */ sa->scsi_host = NULL; - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); - scsi_cmd_stack_free(ctlr); + spin_unlock_irqrestore(&h->lock, flags); + scsi_cmd_stack_free(h); kfree(sa); } -static int -cciss_engage_scsi(int ctlr) +static int cciss_engage_scsi(ctlr_info_t *h) { struct cciss_scsi_adapter_data_t *sa; struct cciss_scsi_cmd_stack_t *stk; unsigned long flags; - spin_lock_irqsave(CCISS_LOCK(ctlr), flags); - sa = hba[ctlr]->scsi_ctlr; + spin_lock_irqsave(&h->lock, flags); + sa = h->scsi_ctlr; stk = &sa->cmd_stack; if (sa->registered) { - printk("cciss%d: SCSI subsystem already engaged.\n", ctlr); - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); + dev_info(&h->pdev->dev, "SCSI subsystem already engaged.\n"); + spin_unlock_irqrestore(&h->lock, flags); return -ENXIO; } sa->registered = 1; - spin_unlock_irqrestore(CCISS_LOCK(ctlr), flags); - cciss_update_non_disk_devices(ctlr, -1); - cciss_scsi_detect(ctlr); + spin_unlock_irqrestore(&h->lock, flags); + cciss_update_non_disk_devices(h, -1); + cciss_scsi_detect(h); return 0; } static void -cciss_seq_tape_report(struct seq_file *seq, int ctlr) +cciss_seq_tape_report(struct seq_file *seq, ctlr_info_t *h) { unsigned long flags; - CPQ_TAPE_LOCK(ctlr, flags); + CPQ_TAPE_LOCK(h, flags); seq_printf(seq, "Sequential access devices: %d\n\n", - ccissscsi[ctlr].ndevices); - CPQ_TAPE_UNLOCK(ctlr, flags); + ccissscsi[h->ctlr].ndevices); + CPQ_TAPE_UNLOCK(h, flags); } static int wait_for_device_to_become_ready(ctlr_info_t *h, @@ -1610,10 +1574,10 @@ static int wait_for_device_to_become_ready(ctlr_info_t *h, int waittime = HZ; CommandList_struct *c; - c = cmd_alloc(h, 1); + c = cmd_alloc(h); if (!c) { - printk(KERN_WARNING "cciss%d: out of memory in " - "wait_for_device_to_become_ready.\n", h->ctlr); + dev_warn(&h->pdev->dev, "out of memory in " + "wait_for_device_to_become_ready.\n"); return IO_ERROR; } @@ -1631,7 +1595,7 @@ static int wait_for_device_to_become_ready(ctlr_info_t *h, waittime = waittime * 2; /* Send the Test Unit Ready */ - rc = fill_cmd(c, TEST_UNIT_READY, h->ctlr, NULL, 0, 0, + rc = fill_cmd(h, c, TEST_UNIT_READY, NULL, 0, 0, lunaddr, TYPE_CMD); if (rc == 0) rc = sendcmd_withirq_core(h, c, 0); @@ -1657,18 +1621,18 @@ static int wait_for_device_to_become_ready(ctlr_info_t *h, } } retry_tur: - printk(KERN_WARNING "cciss%d: Waiting %d secs " + dev_warn(&h->pdev->dev, "Waiting %d secs " "for device to become ready.\n", - h->ctlr, waittime / HZ); + waittime / HZ); rc = 1; /* device not ready. */ } if (rc) - printk("cciss%d: giving up on device.\n", h->ctlr); + dev_warn(&h->pdev->dev, "giving up on device.\n"); else - printk(KERN_WARNING "cciss%d: device is ready.\n", h->ctlr); + dev_warn(&h->pdev->dev, "device is ready.\n"); - cmd_free(h, c, 1); + cmd_free(h, c); return rc; } @@ -1688,26 +1652,24 @@ static int cciss_eh_device_reset_handler(struct scsi_cmnd *scsicmd) int rc; CommandList_struct *cmd_in_trouble; unsigned char lunaddr[8]; - ctlr_info_t *c; - int ctlr; + ctlr_info_t *h; /* find the controller to which the command to be aborted was sent */ - c = (ctlr_info_t *) scsicmd->device->host->hostdata[0]; - if (c == NULL) /* paranoia */ + h = (ctlr_info_t *) scsicmd->device->host->hostdata[0]; + if (h == NULL) /* paranoia */ return FAILED; - ctlr = c->ctlr; - printk(KERN_WARNING "cciss%d: resetting tape drive or medium changer.\n", ctlr); + dev_warn(&h->pdev->dev, "resetting tape drive or medium changer.\n"); /* find the command that's giving us trouble */ cmd_in_trouble = (CommandList_struct *) scsicmd->host_scribble; if (cmd_in_trouble == NULL) /* paranoia */ return FAILED; memcpy(lunaddr, &cmd_in_trouble->Header.LUN.LunAddrBytes[0], 8); /* send a reset to the SCSI LUN which the command was sent to */ - rc = sendcmd_withirq(CCISS_RESET_MSG, ctlr, NULL, 0, 0, lunaddr, + rc = sendcmd_withirq(h, CCISS_RESET_MSG, NULL, 0, 0, lunaddr, TYPE_MSG); - if (rc == 0 && wait_for_device_to_become_ready(c, lunaddr) == 0) + if (rc == 0 && wait_for_device_to_become_ready(h, lunaddr) == 0) return SUCCESS; - printk(KERN_WARNING "cciss%d: resetting device failed.\n", ctlr); + dev_warn(&h->pdev->dev, "resetting device failed.\n"); return FAILED; } @@ -1716,22 +1678,20 @@ static int cciss_eh_abort_handler(struct scsi_cmnd *scsicmd) int rc; CommandList_struct *cmd_to_abort; unsigned char lunaddr[8]; - ctlr_info_t *c; - int ctlr; + ctlr_info_t *h; /* find the controller to which the command to be aborted was sent */ - c = (ctlr_info_t *) scsicmd->device->host->hostdata[0]; - if (c == NULL) /* paranoia */ + h = (ctlr_info_t *) scsicmd->device->host->hostdata[0]; + if (h == NULL) /* paranoia */ return FAILED; - ctlr = c->ctlr; - printk(KERN_WARNING "cciss%d: aborting tardy SCSI cmd\n", ctlr); + dev_warn(&h->pdev->dev, "aborting tardy SCSI cmd\n"); /* find the command to be aborted */ cmd_to_abort = (CommandList_struct *) scsicmd->host_scribble; if (cmd_to_abort == NULL) /* paranoia */ return FAILED; memcpy(lunaddr, &cmd_to_abort->Header.LUN.LunAddrBytes[0], 8); - rc = sendcmd_withirq(CCISS_ABORT_MSG, ctlr, &cmd_to_abort->Header.Tag, + rc = sendcmd_withirq(h, CCISS_ABORT_MSG, &cmd_to_abort->Header.Tag, 0, 0, lunaddr, TYPE_MSG); if (rc == 0) return SUCCESS; diff --git a/drivers/block/cpqarray.c b/drivers/block/cpqarray.c index abb4ec6..d53b029 100644 --- a/drivers/block/cpqarray.c +++ b/drivers/block/cpqarray.c @@ -35,6 +35,7 @@ #include <linux/seq_file.h> #include <linux/init.h> #include <linux/hdreg.h> +#include <linux/smp_lock.h> #include <linux/spinlock.h> #include <linux/blkdev.h> #include <linux/genhd.h> @@ -157,7 +158,7 @@ static int sendcmd( unsigned int blkcnt, unsigned int log_unit ); -static int ida_open(struct block_device *bdev, fmode_t mode); +static int ida_unlocked_open(struct block_device *bdev, fmode_t mode); static int ida_release(struct gendisk *disk, fmode_t mode); static int ida_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg); static int ida_getgeo(struct block_device *bdev, struct hd_geometry *geo); @@ -195,9 +196,9 @@ static inline ctlr_info_t *get_host(struct gendisk *disk) static const struct block_device_operations ida_fops = { .owner = THIS_MODULE, - .open = ida_open, + .open = ida_unlocked_open, .release = ida_release, - .locked_ioctl = ida_ioctl, + .ioctl = ida_ioctl, .getgeo = ida_getgeo, .revalidate_disk= ida_revalidate, }; @@ -840,13 +841,29 @@ static int ida_open(struct block_device *bdev, fmode_t mode) return 0; } +static int ida_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = ida_open(bdev, mode); + unlock_kernel(); + + return ret; +} + /* * Close. Sync first. */ static int ida_release(struct gendisk *disk, fmode_t mode) { - ctlr_info_t *host = get_host(disk); + ctlr_info_t *host; + + lock_kernel(); + host = get_host(disk); host->usage_count--; + unlock_kernel(); + return 0; } @@ -1128,7 +1145,7 @@ static int ida_getgeo(struct block_device *bdev, struct hd_geometry *geo) * ida_ioctl does some miscellaneous stuff like reporting drive geometry, * setting readahead and submitting commands from userspace to the controller. */ -static int ida_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) +static int ida_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { drv_info_t *drv = get_drv(bdev->bd_disk); ctlr_info_t *host = get_host(bdev->bd_disk); @@ -1162,7 +1179,8 @@ out_passthru: return error; case IDAGETCTLRSIG: if (!arg) return -EINVAL; - put_user(host->ctlr_sig, (int __user *)arg); + if (put_user(host->ctlr_sig, (int __user *)arg)) + return -EFAULT; return 0; case IDAREVALIDATEVOLS: if (MINOR(bdev->bd_dev) != 0) @@ -1170,7 +1188,8 @@ out_passthru: return revalidate_allvol(host); case IDADRIVERVERSION: if (!arg) return -EINVAL; - put_user(DRIVER_VERSION, (unsigned long __user *)arg); + if (put_user(DRIVER_VERSION, (unsigned long __user *)arg)) + return -EFAULT; return 0; case IDAGETPCIINFO: { @@ -1192,6 +1211,19 @@ out_passthru: } } + +static int ida_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long param) +{ + int ret; + + lock_kernel(); + ret = ida_locked_ioctl(bdev, mode, cmd, param); + unlock_kernel(); + + return ret; +} + /* * ida_ctlr_ioctl is for passing commands to the controller from userspace. * The command block (io) has already been copied to kernel space for us, @@ -1225,17 +1257,11 @@ static int ida_ctlr_ioctl(ctlr_info_t *h, int dsk, ida_ioctl_t *io) /* Pre submit processing */ switch(io->cmd) { case PASSTHRU_A: - p = kmalloc(io->sg[0].size, GFP_KERNEL); - if (!p) - { - error = -ENOMEM; - cmd_free(h, c, 0); - return(error); - } - if (copy_from_user(p, io->sg[0].addr, io->sg[0].size)) { - kfree(p); - cmd_free(h, c, 0); - return -EFAULT; + p = memdup_user(io->sg[0].addr, io->sg[0].size); + if (IS_ERR(p)) { + error = PTR_ERR(p); + cmd_free(h, c, 0); + return error; } c->req.hdr.blk = pci_map_single(h->pci_dev, &(io->c), sizeof(ida_ioctl_t), @@ -1266,18 +1292,12 @@ static int ida_ctlr_ioctl(ctlr_info_t *h, int dsk, ida_ioctl_t *io) case DIAG_PASS_THRU: case COLLECT_BUFFER: case WRITE_FLASH_ROM: - p = kmalloc(io->sg[0].size, GFP_KERNEL); - if (!p) - { - error = -ENOMEM; - cmd_free(h, c, 0); - return(error); + p = memdup_user(io->sg[0].addr, io->sg[0].size); + if (IS_ERR(p)) { + error = PTR_ERR(p); + cmd_free(h, c, 0); + return error; } - if (copy_from_user(p, io->sg[0].addr, io->sg[0].size)) { - kfree(p); - cmd_free(h, c, 0); - return -EFAULT; - } c->req.sg[0].size = io->sg[0].size; c->req.sg[0].addr = pci_map_single(h->pci_dev, p, c->req.sg[0].size, PCI_DMA_BIDIRECTIONAL); diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c index df01899..9400845 100644 --- a/drivers/block/drbd/drbd_actlog.c +++ b/drivers/block/drbd/drbd_actlog.c @@ -79,8 +79,8 @@ static int _drbd_md_sync_page_io(struct drbd_conf *mdev, md_io.error = 0; if ((rw & WRITE) && !test_bit(MD_NO_BARRIER, &mdev->flags)) - rw |= (1 << BIO_RW_BARRIER); - rw |= ((1<<BIO_RW_UNPLUG) | (1<<BIO_RW_SYNCIO)); + rw |= REQ_HARDBARRIER; + rw |= REQ_UNPLUG | REQ_SYNC; retry: bio = bio_alloc(GFP_NOIO, 1); @@ -103,11 +103,11 @@ static int _drbd_md_sync_page_io(struct drbd_conf *mdev, /* check for unsupported barrier op. * would rather check on EOPNOTSUPP, but that is not reliable. * don't try again for ANY return value != 0 */ - if (unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER) && !ok)) { + if (unlikely((bio->bi_rw & REQ_HARDBARRIER) && !ok)) { /* Try again with no barrier */ dev_warn(DEV, "Barriers not supported on meta data device - disabling\n"); set_bit(MD_NO_BARRIER, &mdev->flags); - rw &= ~(1 << BIO_RW_BARRIER); + rw &= ~REQ_HARDBARRIER; bio_put(bio); goto retry; } diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 485ed8c..352441b 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -550,12 +550,6 @@ struct p_delay_probe { u32 offset; /* usecs the probe got sent after the reference time point */ } __packed; -struct delay_probe { - struct list_head list; - unsigned int seq_num; - struct timeval time; -}; - /* DCBP: Drbd Compressed Bitmap Packet ... */ static inline enum drbd_bitmap_code DCBP_get_code(struct p_compressed_bm *p) @@ -942,11 +936,9 @@ struct drbd_conf { unsigned int ko_count; struct drbd_work resync_work, unplug_work, - md_sync_work, - delay_probe_work; + md_sync_work; struct timer_list resync_timer; struct timer_list md_sync_timer; - struct timer_list delay_probe_timer; /* Used after attach while negotiating new disk state. */ union drbd_state new_state_tmp; @@ -1062,12 +1054,6 @@ struct drbd_conf { u64 ed_uuid; /* UUID of the exposed data */ struct mutex state_mutex; char congestion_reason; /* Why we where congested... */ - struct list_head delay_probes; /* protected by peer_seq_lock */ - int data_delay; /* Delay of packets on the data-sock behind meta-sock */ - unsigned int delay_seq; /* To generate sequence numbers of delay probes */ - struct timeval dps_time; /* delay-probes-start-time */ - unsigned int dp_volume_last; /* send_cnt of last delay probe */ - int c_sync_rate; /* current resync rate after delay_probe magic */ }; static inline struct drbd_conf *minor_to_mdev(unsigned int minor) diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c index 7258c95..fa650dd 100644 --- a/drivers/block/drbd/drbd_main.c +++ b/drivers/block/drbd/drbd_main.c @@ -2184,43 +2184,6 @@ int drbd_send_ov_request(struct drbd_conf *mdev, sector_t sector, int size) return ok; } -static int drbd_send_delay_probe(struct drbd_conf *mdev, struct drbd_socket *ds) -{ - struct p_delay_probe dp; - int offset, ok = 0; - struct timeval now; - - mutex_lock(&ds->mutex); - if (likely(ds->socket)) { - do_gettimeofday(&now); - offset = now.tv_usec - mdev->dps_time.tv_usec + - (now.tv_sec - mdev->dps_time.tv_sec) * 1000000; - dp.seq_num = cpu_to_be32(mdev->delay_seq); - dp.offset = cpu_to_be32(offset); - - ok = _drbd_send_cmd(mdev, ds->socket, P_DELAY_PROBE, - (struct p_header *)&dp, sizeof(dp), 0); - } - mutex_unlock(&ds->mutex); - - return ok; -} - -static int drbd_send_delay_probes(struct drbd_conf *mdev) -{ - int ok; - - mdev->delay_seq++; - do_gettimeofday(&mdev->dps_time); - ok = drbd_send_delay_probe(mdev, &mdev->meta); - ok = ok && drbd_send_delay_probe(mdev, &mdev->data); - - mdev->dp_volume_last = mdev->send_cnt; - mod_timer(&mdev->delay_probe_timer, jiffies + mdev->sync_conf.dp_interval * HZ / 10); - - return ok; -} - /* called on sndtimeo * returns FALSE if we should retry, * TRUE if we think connection is dead @@ -2369,31 +2332,6 @@ static int _drbd_send_zc_ee(struct drbd_conf *mdev, struct drbd_epoch_entry *e) return 1; } -static void consider_delay_probes(struct drbd_conf *mdev) -{ - if (mdev->state.conn != C_SYNC_SOURCE || mdev->agreed_pro_version < 93) - return; - - if (mdev->dp_volume_last + mdev->sync_conf.dp_volume * 2 < mdev->send_cnt) - drbd_send_delay_probes(mdev); -} - -static int w_delay_probes(struct drbd_conf *mdev, struct drbd_work *w, int cancel) -{ - if (!cancel && mdev->state.conn == C_SYNC_SOURCE) - drbd_send_delay_probes(mdev); - - return 1; -} - -static void delay_probe_timer_fn(unsigned long data) -{ - struct drbd_conf *mdev = (struct drbd_conf *) data; - - if (list_empty(&mdev->delay_probe_work.list)) - drbd_queue_work(&mdev->data.work, &mdev->delay_probe_work); -} - /* Used to send write requests * R_PRIMARY -> Peer (P_DATA) */ @@ -2425,15 +2363,15 @@ int drbd_send_dblock(struct drbd_conf *mdev, struct drbd_request *req) /* NOTE: no need to check if barriers supported here as we would * not pass the test in make_request_common in that case */ - if (bio_rw_flagged(req->master_bio, BIO_RW_BARRIER)) { + if (req->master_bio->bi_rw & REQ_HARDBARRIER) { dev_err(DEV, "ASSERT FAILED would have set DP_HARDBARRIER\n"); /* dp_flags |= DP_HARDBARRIER; */ } - if (bio_rw_flagged(req->master_bio, BIO_RW_SYNCIO)) + if (req->master_bio->bi_rw & REQ_SYNC) dp_flags |= DP_RW_SYNC; /* for now handle SYNCIO and UNPLUG * as if they still were one and the same flag */ - if (bio_rw_flagged(req->master_bio, BIO_RW_UNPLUG)) + if (req->master_bio->bi_rw & REQ_UNPLUG) dp_flags |= DP_RW_SYNC; if (mdev->state.conn >= C_SYNC_SOURCE && mdev->state.conn <= C_PAUSED_SYNC_T) @@ -2457,9 +2395,6 @@ int drbd_send_dblock(struct drbd_conf *mdev, struct drbd_request *req) drbd_put_data_sock(mdev); - if (ok) - consider_delay_probes(mdev); - return ok; } @@ -2506,9 +2441,6 @@ int drbd_send_block(struct drbd_conf *mdev, enum drbd_packets cmd, drbd_put_data_sock(mdev); - if (ok) - consider_delay_probes(mdev); - return ok; } @@ -2604,6 +2536,7 @@ static int drbd_open(struct block_device *bdev, fmode_t mode) unsigned long flags; int rv = 0; + lock_kernel(); spin_lock_irqsave(&mdev->req_lock, flags); /* to have a stable mdev->state.role * and no race with updating open_cnt */ @@ -2618,6 +2551,7 @@ static int drbd_open(struct block_device *bdev, fmode_t mode) if (!rv) mdev->open_cnt++; spin_unlock_irqrestore(&mdev->req_lock, flags); + unlock_kernel(); return rv; } @@ -2625,7 +2559,9 @@ static int drbd_open(struct block_device *bdev, fmode_t mode) static int drbd_release(struct gendisk *gd, fmode_t mode) { struct drbd_conf *mdev = gd->private_data; + lock_kernel(); mdev->open_cnt--; + unlock_kernel(); return 0; } @@ -2660,9 +2596,20 @@ static void drbd_unplug_fn(struct request_queue *q) static void drbd_set_defaults(struct drbd_conf *mdev) { - mdev->sync_conf.after = DRBD_AFTER_DEF; - mdev->sync_conf.rate = DRBD_RATE_DEF; - mdev->sync_conf.al_extents = DRBD_AL_EXTENTS_DEF; + /* This way we get a compile error when sync_conf grows, + and we forgot to initialize it here */ + mdev->sync_conf = (struct syncer_conf) { + /* .rate = */ DRBD_RATE_DEF, + /* .after = */ DRBD_AFTER_DEF, + /* .al_extents = */ DRBD_AL_EXTENTS_DEF, + /* .verify_alg = */ {}, 0, + /* .cpu_mask = */ {}, 0, + /* .csums_alg = */ {}, 0, + /* .use_rle = */ 0 + }; + + /* Have to use that way, because the layout differs between + big endian and little endian */ mdev->state = (union drbd_state) { { .role = R_SECONDARY, .peer = R_UNKNOWN, @@ -2721,24 +2668,17 @@ void drbd_init_set_defaults(struct drbd_conf *mdev) INIT_LIST_HEAD(&mdev->unplug_work.list); INIT_LIST_HEAD(&mdev->md_sync_work.list); INIT_LIST_HEAD(&mdev->bm_io_work.w.list); - INIT_LIST_HEAD(&mdev->delay_probes); - INIT_LIST_HEAD(&mdev->delay_probe_work.list); mdev->resync_work.cb = w_resync_inactive; mdev->unplug_work.cb = w_send_write_hint; mdev->md_sync_work.cb = w_md_sync; mdev->bm_io_work.w.cb = w_bitmap_io; - mdev->delay_probe_work.cb = w_delay_probes; init_timer(&mdev->resync_timer); init_timer(&mdev->md_sync_timer); - init_timer(&mdev->delay_probe_timer); mdev->resync_timer.function = resync_timer_fn; mdev->resync_timer.data = (unsigned long) mdev; mdev->md_sync_timer.function = md_sync_timer_fn; mdev->md_sync_timer.data = (unsigned long) mdev; - mdev->delay_probe_timer.function = delay_probe_timer_fn; - mdev->delay_probe_timer.data = (unsigned long) mdev; - init_waitqueue_head(&mdev->misc_wait); init_waitqueue_head(&mdev->state_wait); diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c index 2151f18..73131c5 100644 --- a/drivers/block/drbd/drbd_nl.c +++ b/drivers/block/drbd/drbd_nl.c @@ -1557,10 +1557,6 @@ static int drbd_nl_syncer_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *n sc.rate = DRBD_RATE_DEF; sc.after = DRBD_AFTER_DEF; sc.al_extents = DRBD_AL_EXTENTS_DEF; - sc.dp_volume = DRBD_DP_VOLUME_DEF; - sc.dp_interval = DRBD_DP_INTERVAL_DEF; - sc.throttle_th = DRBD_RS_THROTTLE_TH_DEF; - sc.hold_off_th = DRBD_RS_HOLD_OFF_TH_DEF; } else memcpy(&sc, &mdev->sync_conf, sizeof(struct syncer_conf)); diff --git a/drivers/block/drbd/drbd_proc.c b/drivers/block/drbd/drbd_proc.c index d0f1767..be3374b 100644 --- a/drivers/block/drbd/drbd_proc.c +++ b/drivers/block/drbd/drbd_proc.c @@ -73,21 +73,14 @@ static void drbd_syncer_progress(struct drbd_conf *mdev, struct seq_file *seq) seq_printf(seq, "sync'ed:%3u.%u%% ", res / 10, res % 10); /* if more than 1 GB display in MB */ if (mdev->rs_total > 0x100000L) - seq_printf(seq, "(%lu/%lu)M", + seq_printf(seq, "(%lu/%lu)M\n\t", (unsigned long) Bit2KB(rs_left >> 10), (unsigned long) Bit2KB(mdev->rs_total >> 10)); else - seq_printf(seq, "(%lu/%lu)K", + seq_printf(seq, "(%lu/%lu)K\n\t", (unsigned long) Bit2KB(rs_left), (unsigned long) Bit2KB(mdev->rs_total)); - if (mdev->state.conn == C_SYNC_TARGET) - seq_printf(seq, " queue_delay: %d.%d ms\n\t", - mdev->data_delay / 1000, - (mdev->data_delay % 1000) / 100); - else if (mdev->state.conn == C_SYNC_SOURCE) - seq_printf(seq, " delay_probe: %u\n\t", mdev->delay_seq); - /* see drivers/md/md.c * We do not want to overflow, so the order of operands and * the * 100 / 100 trick are important. We do a +1 to be @@ -135,14 +128,6 @@ static void drbd_syncer_progress(struct drbd_conf *mdev, struct seq_file *seq) else seq_printf(seq, " (%ld)", dbdt); - if (mdev->state.conn == C_SYNC_TARGET) { - if (mdev->c_sync_rate > 1000) - seq_printf(seq, " want: %d,%03d", - mdev->c_sync_rate / 1000, mdev->c_sync_rate % 1000); - else - seq_printf(seq, " want: %d", mdev->c_sync_rate); - } - seq_printf(seq, " K/sec\n"); } diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index ec1711f..081522d 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -1180,7 +1180,7 @@ next_bio: bio->bi_sector = sector; bio->bi_bdev = mdev->ldev->backing_bdev; /* we special case some flags in the multi-bio case, see below - * (BIO_RW_UNPLUG, BIO_RW_BARRIER) */ + * (REQ_UNPLUG, REQ_HARDBARRIER) */ bio->bi_rw = rw; bio->bi_private = e; bio->bi_end_io = drbd_endio_sec; @@ -1209,16 +1209,16 @@ next_bio: bios = bios->bi_next; bio->bi_next = NULL; - /* strip off BIO_RW_UNPLUG unless it is the last bio */ + /* strip off REQ_UNPLUG unless it is the last bio */ if (bios) - bio->bi_rw &= ~(1<<BIO_RW_UNPLUG); + bio->bi_rw &= ~REQ_UNPLUG; drbd_generic_make_request(mdev, fault_type, bio); - /* strip off BIO_RW_BARRIER, + /* strip off REQ_HARDBARRIER, * unless it is the first or last bio */ if (bios && bios->bi_next) - bios->bi_rw &= ~(1<<BIO_RW_BARRIER); + bios->bi_rw &= ~REQ_HARDBARRIER; } while (bios); maybe_kick_lo(mdev); return 0; @@ -1233,7 +1233,7 @@ fail: } /** - * w_e_reissue() - Worker callback; Resubmit a bio, without BIO_RW_BARRIER set + * w_e_reissue() - Worker callback; Resubmit a bio, without REQ_HARDBARRIER set * @mdev: DRBD device. * @w: work object. * @cancel: The connection will be closed anyways (unused in this callback) @@ -1245,7 +1245,7 @@ int w_e_reissue(struct drbd_conf *mdev, struct drbd_work *w, int cancel) __relea (and DE_BARRIER_IN_NEXT_EPOCH_ISSUED in the previous Epoch) so that we can finish that epoch in drbd_may_finish_epoch(). That is necessary if we already have a long chain of Epochs, before - we realize that BIO_RW_BARRIER is actually not supported */ + we realize that REQ_HARDBARRIER is actually not supported */ /* As long as the -ENOTSUPP on the barrier is reported immediately that will never trigger. If it is reported late, we will just @@ -1824,14 +1824,14 @@ static int receive_Data(struct drbd_conf *mdev, struct p_header *h) epoch = list_entry(e->epoch->list.prev, struct drbd_epoch, list); if (epoch == e->epoch) { set_bit(DE_CONTAINS_A_BARRIER, &e->epoch->flags); - rw |= (1<<BIO_RW_BARRIER); + rw |= REQ_HARDBARRIER; e->flags |= EE_IS_BARRIER; } else { if (atomic_read(&epoch->epoch_size) > 1 || !test_bit(DE_CONTAINS_A_BARRIER, &epoch->flags)) { set_bit(DE_BARRIER_IN_NEXT_EPOCH_ISSUED, &epoch->flags); set_bit(DE_CONTAINS_A_BARRIER, &e->epoch->flags); - rw |= (1<<BIO_RW_BARRIER); + rw |= REQ_HARDBARRIER; e->flags |= EE_IS_BARRIER; } } @@ -1841,10 +1841,10 @@ static int receive_Data(struct drbd_conf *mdev, struct p_header *h) dp_flags = be32_to_cpu(p->dp_flags); if (dp_flags & DP_HARDBARRIER) { dev_err(DEV, "ASSERT FAILED would have submitted barrier request\n"); - /* rw |= (1<<BIO_RW_BARRIER); */ + /* rw |= REQ_HARDBARRIER; */ } if (dp_flags & DP_RW_SYNC) - rw |= (1<<BIO_RW_SYNCIO) | (1<<BIO_RW_UNPLUG); + rw |= REQ_SYNC | REQ_UNPLUG; if (dp_flags & DP_MAY_SET_IN_SYNC) e->flags |= EE_MAY_SET_IN_SYNC; @@ -3555,14 +3555,15 @@ static int receive_bitmap(struct drbd_conf *mdev, struct p_header *h) return ok; } -static int receive_skip(struct drbd_conf *mdev, struct p_header *h) +static int receive_skip_(struct drbd_conf *mdev, struct p_header *h, int silent) { /* TODO zero copy sink :) */ static char sink[128]; int size, want, r; - dev_warn(DEV, "skipping unknown optional packet type %d, l: %d!\n", - h->command, h->length); + if (!silent) + dev_warn(DEV, "skipping unknown optional packet type %d, l: %d!\n", + h->command, h->length); size = h->length; while (size > 0) { @@ -3574,101 +3575,25 @@ static int receive_skip(struct drbd_conf *mdev, struct p_header *h) return size == 0; } -static int receive_UnplugRemote(struct drbd_conf *mdev, struct p_header *h) -{ - if (mdev->state.disk >= D_INCONSISTENT) - drbd_kick_lo(mdev); - - /* Make sure we've acked all the TCP data associated - * with the data requests being unplugged */ - drbd_tcp_quickack(mdev->data.socket); - - return TRUE; -} - -static void timeval_sub_us(struct timeval* tv, unsigned int us) +static int receive_skip(struct drbd_conf *mdev, struct p_header *h) { - tv->tv_sec -= us / 1000000; - us = us % 1000000; - if (tv->tv_usec > us) { - tv->tv_usec += 1000000; - tv->tv_sec--; - } - tv->tv_usec -= us; + return receive_skip_(mdev, h, 0); } -static void got_delay_probe(struct drbd_conf *mdev, int from, struct p_delay_probe *p) +static int receive_skip_silent(struct drbd_conf *mdev, struct p_header *h) { - struct delay_probe *dp; - struct list_head *le; - struct timeval now; - int seq_num; - int offset; - int data_delay; - - seq_num = be32_to_cpu(p->seq_num); - offset = be32_to_cpu(p->offset); - - spin_lock(&mdev->peer_seq_lock); - if (!list_empty(&mdev->delay_probes)) { - if (from == USE_DATA_SOCKET) - le = mdev->delay_probes.next; - else - le = mdev->delay_probes.prev; - - dp = list_entry(le, struct delay_probe, list); - - if (dp->seq_num == seq_num) { - list_del(le); - spin_unlock(&mdev->peer_seq_lock); - do_gettimeofday(&now); - timeval_sub_us(&now, offset); - data_delay = - now.tv_usec - dp->time.tv_usec + - (now.tv_sec - dp->time.tv_sec) * 1000000; - - if (data_delay > 0) - mdev->data_delay = data_delay; - - kfree(dp); - return; - } - - if (dp->seq_num > seq_num) { - spin_unlock(&mdev->peer_seq_lock); - dev_warn(DEV, "Previous allocation failure of struct delay_probe?\n"); - return; /* Do not alloca a struct delay_probe.... */ - } - } - spin_unlock(&mdev->peer_seq_lock); - - dp = kmalloc(sizeof(struct delay_probe), GFP_NOIO); - if (!dp) { - dev_warn(DEV, "Failed to allocate a struct delay_probe, do not worry.\n"); - return; - } - - dp->seq_num = seq_num; - do_gettimeofday(&dp->time); - timeval_sub_us(&dp->time, offset); - - spin_lock(&mdev->peer_seq_lock); - if (from == USE_DATA_SOCKET) - list_add(&dp->list, &mdev->delay_probes); - else - list_add_tail(&dp->list, &mdev->delay_probes); - spin_unlock(&mdev->peer_seq_lock); + return receive_skip_(mdev, h, 1); } -static int receive_delay_probe(struct drbd_conf *mdev, struct p_header *h) +static int receive_UnplugRemote(struct drbd_conf *mdev, struct p_header *h) { - struct p_delay_probe *p = (struct p_delay_probe *)h; + if (mdev->state.disk >= D_INCONSISTENT) + drbd_kick_lo(mdev); - ERR_IF(h->length != (sizeof(*p)-sizeof(*h))) return FALSE; - if (drbd_recv(mdev, h->payload, h->length) != h->length) - return FALSE; + /* Make sure we've acked all the TCP data associated + * with the data requests being unplugged */ + drbd_tcp_quickack(mdev->data.socket); - got_delay_probe(mdev, USE_DATA_SOCKET, p); return TRUE; } @@ -3695,7 +3620,7 @@ static drbd_cmd_handler_f drbd_default_handler[] = { [P_OV_REQUEST] = receive_DataRequest, [P_OV_REPLY] = receive_DataRequest, [P_CSUM_RS_REQUEST] = receive_DataRequest, - [P_DELAY_PROBE] = receive_delay_probe, + [P_DELAY_PROBE] = receive_skip_silent, /* anything missing from this table is in * the asender_tbl, see get_asender_cmd */ [P_MAX_CMD] = NULL, @@ -4472,11 +4397,9 @@ static int got_OVResult(struct drbd_conf *mdev, struct p_header *h) return TRUE; } -static int got_delay_probe_m(struct drbd_conf *mdev, struct p_header *h) +static int got_something_to_ignore_m(struct drbd_conf *mdev, struct p_header *h) { - struct p_delay_probe *p = (struct p_delay_probe *)h; - - got_delay_probe(mdev, USE_META_SOCKET, p); + /* IGNORE */ return TRUE; } @@ -4504,7 +4427,7 @@ static struct asender_cmd *get_asender_cmd(int cmd) [P_BARRIER_ACK] = { sizeof(struct p_barrier_ack), got_BarrierAck }, [P_STATE_CHG_REPLY] = { sizeof(struct p_req_state_reply), got_RqSReply }, [P_RS_IS_IN_SYNC] = { sizeof(struct p_block_ack), got_IsInSync }, - [P_DELAY_PROBE] = { sizeof(struct p_delay_probe), got_delay_probe_m }, + [P_DELAY_PROBE] = { sizeof(struct p_delay_probe), got_something_to_ignore_m }, [P_MAX_CMD] = { 0, NULL }, }; if (cmd > P_MAX_CMD || asender_tbl[cmd].process == NULL) diff --git a/drivers/block/drbd/drbd_req.c b/drivers/block/drbd/drbd_req.c index 654f1ef..f761d98 100644 --- a/drivers/block/drbd/drbd_req.c +++ b/drivers/block/drbd/drbd_req.c @@ -997,7 +997,7 @@ int drbd_make_request_26(struct request_queue *q, struct bio *bio) * because of those XXX, this is not yet enabled, * i.e. in drbd_init_set_defaults we set the NO_BARRIER_SUPP bit. */ - if (unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER) && test_bit(NO_BARRIER_SUPP, &mdev->flags))) { + if (unlikely(bio->bi_rw & REQ_HARDBARRIER) && test_bit(NO_BARRIER_SUPP, &mdev->flags)) { /* dev_warn(DEV, "Rejecting barrier request as underlying device does not support\n"); */ bio_endio(bio, -EOPNOTSUPP); return 0; diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c index b623cee..ca4a16c 100644 --- a/drivers/block/drbd/drbd_worker.c +++ b/drivers/block/drbd/drbd_worker.c @@ -424,18 +424,6 @@ void resync_timer_fn(unsigned long data) drbd_queue_work(&mdev->data.work, &mdev->resync_work); } -static int calc_resync_rate(struct drbd_conf *mdev) -{ - int d = mdev->data_delay / 1000; /* us -> ms */ - int td = mdev->sync_conf.throttle_th * 100; /* 0.1s -> ms */ - int hd = mdev->sync_conf.hold_off_th * 100; /* 0.1s -> ms */ - int cr = mdev->sync_conf.rate; - - return d <= td ? cr : - d >= hd ? 0 : - cr + (cr * (td - d) / (hd - td)); -} - int w_make_resync_request(struct drbd_conf *mdev, struct drbd_work *w, int cancel) { @@ -473,8 +461,7 @@ int w_make_resync_request(struct drbd_conf *mdev, max_segment_size = mdev->agreed_pro_version < 94 ? queue_max_segment_size(mdev->rq_queue) : DRBD_MAX_SEGMENT_SIZE; - mdev->c_sync_rate = calc_resync_rate(mdev); - number = SLEEP_TIME * mdev->c_sync_rate / ((BM_BLOCK_SIZE / 1024) * HZ); + number = SLEEP_TIME * mdev->sync_conf.rate / ((BM_BLOCK_SIZE / 1024) * HZ); pe = atomic_read(&mdev->rs_pending_cnt); mutex_lock(&mdev->data.mutex); diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c index 90c4038..cf04c1b 100644 --- a/drivers/block/floppy.c +++ b/drivers/block/floppy.c @@ -178,6 +178,7 @@ static int print_unex = 1; #include <linux/slab.h> #include <linux/mm.h> #include <linux/bio.h> +#include <linux/smp_lock.h> #include <linux/string.h> #include <linux/jiffies.h> #include <linux/fcntl.h> @@ -514,8 +515,6 @@ static unsigned long fdc_busy; static DECLARE_WAIT_QUEUE_HEAD(fdc_wait); static DECLARE_WAIT_QUEUE_HEAD(command_done); -#define NO_SIGNAL (!interruptible || !signal_pending(current)) - /* Errors during formatting are counted here. */ static int format_errors; @@ -539,7 +538,7 @@ static int max_buffer_sectors; static int *errors; typedef void (*done_f)(int); -static struct cont_t { +static const struct cont_t { void (*interrupt)(void); /* this is called after the interrupt of the * main command */ @@ -578,7 +577,7 @@ static void reset_fdc(void); #define NEED_1_RECAL -2 #define NEED_2_RECAL -3 -static int usage_count; +static atomic_t usage_count = ATOMIC_INIT(0); /* buffer related variables */ static int buffer_track = -1; @@ -858,36 +857,15 @@ static void set_fdc(int drive) } /* locks the driver */ -static int _lock_fdc(int drive, bool interruptible, int line) +static int lock_fdc(int drive, bool interruptible) { - if (!usage_count) { - pr_err("Trying to lock fdc while usage count=0 at line %d\n", - line); + if (WARN(atomic_read(&usage_count) == 0, + "Trying to lock fdc while usage count=0\n")) return -1; - } - - if (test_and_set_bit(0, &fdc_busy)) { - DECLARE_WAITQUEUE(wait, current); - add_wait_queue(&fdc_wait, &wait); - - for (;;) { - set_current_state(TASK_INTERRUPTIBLE); - - if (!test_and_set_bit(0, &fdc_busy)) - break; - schedule(); - - if (!NO_SIGNAL) { - remove_wait_queue(&fdc_wait, &wait); - return -EINTR; - } - } + if (wait_event_interruptible(fdc_wait, !test_and_set_bit(0, &fdc_busy))) + return -EINTR; - set_current_state(TASK_RUNNING); - remove_wait_queue(&fdc_wait, &wait); - flush_scheduled_work(); - } command_status = FD_COMMAND_NONE; __reschedule_timeout(drive, "lock fdc"); @@ -895,11 +873,8 @@ static int _lock_fdc(int drive, bool interruptible, int line) return 0; } -#define lock_fdc(drive, interruptible) \ - _lock_fdc(drive, interruptible, __LINE__) - /* unlocks the driver */ -static inline void unlock_fdc(void) +static void unlock_fdc(void) { unsigned long flags; @@ -1224,7 +1199,7 @@ static int need_more_output(void) /* Set perpendicular mode as required, based on data rate, if supported. * 82077 Now tested. 1Mbps data rate only possible with 82077-1. */ -static inline void perpendicular_mode(void) +static void perpendicular_mode(void) { unsigned char perp_mode; @@ -1995,14 +1970,14 @@ static void do_wakeup(void) wake_up(&command_done); } -static struct cont_t wakeup_cont = { +static const struct cont_t wakeup_cont = { .interrupt = empty, .redo = do_wakeup, .error = empty, .done = (done_f)empty }; -static struct cont_t intr_cont = { +static const struct cont_t intr_cont = { .interrupt = empty, .redo = process_fd_request, .error = empty, @@ -2015,25 +1990,10 @@ static int wait_til_done(void (*handler)(void), bool interruptible) schedule_bh(handler); - if (command_status < 2 && NO_SIGNAL) { - DECLARE_WAITQUEUE(wait, current); - - add_wait_queue(&command_done, &wait); - for (;;) { - set_current_state(interruptible ? - TASK_INTERRUPTIBLE : - TASK_UNINTERRUPTIBLE); - - if (command_status >= 2 || !NO_SIGNAL) - break; - - is_alive(__func__, ""); - schedule(); - } - - set_current_state(TASK_RUNNING); - remove_wait_queue(&command_done, &wait); - } + if (interruptible) + wait_event_interruptible(command_done, command_status >= 2); + else + wait_event(command_done, command_status >= 2); if (command_status < 2) { cancel_activity(); @@ -2223,7 +2183,7 @@ static void redo_format(void) debugt(__func__, "queue format request"); } -static struct cont_t format_cont = { +static const struct cont_t format_cont = { .interrupt = format_interrupt, .redo = redo_format, .error = bad_flp_intr, @@ -2583,10 +2543,8 @@ static int make_raw_rw_request(void) int tracksize; int ssize; - if (max_buffer_sectors == 0) { - pr_info("VFS: Block I/O scheduled on unopened device\n"); + if (WARN(max_buffer_sectors == 0, "VFS: Block I/O scheduled on unopened device\n")) return 0; - } set_fdc((long)current_req->rq_disk->private_data); @@ -2921,7 +2879,7 @@ do_request: return; } -static struct cont_t rw_cont = { +static const struct cont_t rw_cont = { .interrupt = rw_interrupt, .redo = redo_fd_request, .error = bad_flp_intr, @@ -2936,19 +2894,16 @@ static void process_fd_request(void) static void do_fd_request(struct request_queue *q) { - if (max_buffer_sectors == 0) { - pr_info("VFS: %s called on non-open device\n", __func__); + if (WARN(max_buffer_sectors == 0, + "VFS: %s called on non-open device\n", __func__)) return; - } - if (usage_count == 0) { - pr_info("warning: usage count=0, current_req=%p exiting\n", - current_req); - pr_info("sect=%ld type=%x flags=%x\n", - (long)blk_rq_pos(current_req), current_req->cmd_type, - current_req->cmd_flags); + if (WARN(atomic_read(&usage_count) == 0, + "warning: usage count=0, current_req=%p sect=%ld type=%x flags=%x\n", + current_req, (long)blk_rq_pos(current_req), current_req->cmd_type, + current_req->cmd_flags)) return; - } + if (test_bit(0, &fdc_busy)) { /* fdc busy, this new request will be treated when the current one is done */ @@ -2960,7 +2915,7 @@ static void do_fd_request(struct request_queue *q) is_alive(__func__, ""); } -static struct cont_t poll_cont = { +static const struct cont_t poll_cont = { .interrupt = success_and_wakeup, .redo = floppy_ready, .error = generic_failure, @@ -2991,7 +2946,7 @@ static void reset_intr(void) pr_info("weird, reset interrupt called\n"); } -static struct cont_t reset_cont = { +static const struct cont_t reset_cont = { .interrupt = reset_intr, .redo = success_and_wakeup, .error = generic_failure, @@ -3033,7 +2988,7 @@ static inline int fd_copyin(void __user *param, void *address, return copy_from_user(address, param, size) ? -EFAULT : 0; } -static inline const char *drive_name(int type, int drive) +static const char *drive_name(int type, int drive) { struct floppy_struct *floppy; @@ -3096,14 +3051,14 @@ static void raw_cmd_done(int flag) generic_done(flag); } -static struct cont_t raw_cmd_cont = { +static const struct cont_t raw_cmd_cont = { .interrupt = success_and_wakeup, .redo = floppy_start, .error = generic_failure, .done = raw_cmd_done }; -static inline int raw_cmd_copyout(int cmd, void __user *param, +static int raw_cmd_copyout(int cmd, void __user *param, struct floppy_raw_cmd *ptr) { int ret; @@ -3148,7 +3103,7 @@ static void raw_cmd_free(struct floppy_raw_cmd **ptr) } } -static inline int raw_cmd_copyin(int cmd, void __user *param, +static int raw_cmd_copyin(int cmd, void __user *param, struct floppy_raw_cmd **rcmd) { struct floppy_raw_cmd *ptr; @@ -3266,7 +3221,7 @@ static int invalidate_drive(struct block_device *bdev) return 0; } -static inline int set_geometry(unsigned int cmd, struct floppy_struct *g, +static int set_geometry(unsigned int cmd, struct floppy_struct *g, int drive, int type, struct block_device *bdev) { int cnt; @@ -3337,7 +3292,7 @@ static inline int set_geometry(unsigned int cmd, struct floppy_struct *g, } /* handle obsolete ioctl's */ -static int ioctl_table[] = { +static unsigned int ioctl_table[] = { FDCLRPRM, FDSETPRM, FDDEFPRM, @@ -3365,7 +3320,7 @@ static int ioctl_table[] = { FDTWADDLE }; -static inline int normalize_ioctl(int *cmd, int *size) +static int normalize_ioctl(unsigned int *cmd, int *size) { int i; @@ -3417,7 +3372,7 @@ static int fd_getgeo(struct block_device *bdev, struct hd_geometry *geo) return 0; } -static int fd_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, +static int fd_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long param) { int drive = (long)bdev->bd_disk->private_data; @@ -3593,6 +3548,18 @@ static int fd_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, return 0; } +static int fd_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long param) +{ + int ret; + + lock_kernel(); + ret = fd_locked_ioctl(bdev, mode, cmd, param); + unlock_kernel(); + + return ret; +} + static void __init config_types(void) { bool has_drive = false; @@ -3649,6 +3616,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode) { int drive = (long)disk->private_data; + lock_kernel(); mutex_lock(&open_lock); if (UDRS->fd_ref < 0) UDRS->fd_ref = 0; @@ -3659,6 +3627,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode) if (!UDRS->fd_ref) opened_bdev[drive] = NULL; mutex_unlock(&open_lock); + unlock_kernel(); return 0; } @@ -3676,6 +3645,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) int res = -EBUSY; char *tmp; + lock_kernel(); mutex_lock(&open_lock); old_dev = UDRS->fd_device; if (opened_bdev[drive] && opened_bdev[drive] != bdev) @@ -3752,6 +3722,7 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) goto out; } mutex_unlock(&open_lock); + unlock_kernel(); return 0; out: if (UDRS->fd_ref < 0) @@ -3762,6 +3733,7 @@ out: opened_bdev[drive] = NULL; out2: mutex_unlock(&open_lock); + unlock_kernel(); return res; } @@ -3829,6 +3801,7 @@ static int __floppy_read_block_0(struct block_device *bdev) bio.bi_size = size; bio.bi_bdev = bdev; bio.bi_sector = 0; + bio.bi_flags = BIO_QUIET; init_completion(&complete); bio.bi_private = &complete; bio.bi_end_io = floppy_rb0_complete; @@ -3857,10 +3830,10 @@ static int floppy_revalidate(struct gendisk *disk) if (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags) || test_bit(FD_VERIFY_BIT, &UDRS->flags) || test_bit(drive, &fake_change) || NO_GEOM) { - if (usage_count == 0) { - pr_info("VFS: revalidate called on non-open device.\n"); + if (WARN(atomic_read(&usage_count) == 0, + "VFS: revalidate called on non-open device.\n")) return -EFAULT; - } + lock_fdc(drive, false); cf = (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags) || test_bit(FD_VERIFY_BIT, &UDRS->flags)); @@ -3893,7 +3866,7 @@ static const struct block_device_operations floppy_fops = { .owner = THIS_MODULE, .open = floppy_open, .release = floppy_release, - .locked_ioctl = fd_ioctl, + .ioctl = fd_ioctl, .getgeo = fd_getgeo, .media_changed = check_floppy_change, .revalidate_disk = floppy_revalidate, @@ -4126,7 +4099,7 @@ static ssize_t floppy_cmos_show(struct device *dev, return sprintf(buf, "%X\n", UDP->cmos); } -DEVICE_ATTR(cmos, S_IRUGO, floppy_cmos_show, NULL); +static DEVICE_ATTR(cmos, S_IRUGO, floppy_cmos_show, NULL); static void floppy_device_release(struct device *dev) { @@ -4175,6 +4148,9 @@ static int __init floppy_init(void) int i, unit, drive; int err, dr; + set_debugt(); + interruptjiffies = resultjiffies = jiffies; + #if defined(CONFIG_PPC) if (check_legacy_ioport(FDC1)) return -ENODEV; @@ -4353,7 +4329,7 @@ out_unreg_platform_dev: platform_device_unregister(&floppy_device[drive]); out_flush_work: flush_scheduled_work(); - if (usage_count) + if (atomic_read(&usage_count)) floppy_release_irq_and_dma(); out_unreg_region: blk_unregister_region(MKDEV(FLOPPY_MAJOR, 0), 256); @@ -4370,8 +4346,6 @@ out_put_disk: return err; } -static DEFINE_SPINLOCK(floppy_usage_lock); - static const struct io_region { int offset; int size; @@ -4417,14 +4391,8 @@ static void floppy_release_regions(int fdc) static int floppy_grab_irq_and_dma(void) { - unsigned long flags; - - spin_lock_irqsave(&floppy_usage_lock, flags); - if (usage_count++) { - spin_unlock_irqrestore(&floppy_usage_lock, flags); + if (atomic_inc_return(&usage_count) > 1) return 0; - } - spin_unlock_irqrestore(&floppy_usage_lock, flags); /* * We might have scheduled a free_irq(), wait it to @@ -4435,9 +4403,7 @@ static int floppy_grab_irq_and_dma(void) if (fd_request_irq()) { DPRINT("Unable to grab IRQ%d for the floppy driver\n", FLOPPY_IRQ); - spin_lock_irqsave(&floppy_usage_lock, flags); - usage_count--; - spin_unlock_irqrestore(&floppy_usage_lock, flags); + atomic_dec(&usage_count); return -1; } if (fd_request_dma()) { @@ -4447,9 +4413,7 @@ static int floppy_grab_irq_and_dma(void) use_virtual_dma = can_use_virtual_dma = 1; if (!(can_use_virtual_dma & 1)) { fd_free_irq(); - spin_lock_irqsave(&floppy_usage_lock, flags); - usage_count--; - spin_unlock_irqrestore(&floppy_usage_lock, flags); + atomic_dec(&usage_count); return -1; } } @@ -4484,9 +4448,7 @@ cleanup: fd_free_dma(); while (--fdc >= 0) floppy_release_regions(fdc); - spin_lock_irqsave(&floppy_usage_lock, flags); - usage_count--; - spin_unlock_irqrestore(&floppy_usage_lock, flags); + atomic_dec(&usage_count); return -1; } @@ -4498,14 +4460,10 @@ static void floppy_release_irq_and_dma(void) #endif long tmpsize; unsigned long tmpaddr; - unsigned long flags; - spin_lock_irqsave(&floppy_usage_lock, flags); - if (--usage_count) { - spin_unlock_irqrestore(&floppy_usage_lock, flags); + if (!atomic_dec_and_test(&usage_count)) return; - } - spin_unlock_irqrestore(&floppy_usage_lock, flags); + if (irqdma_allocated) { fd_disable_dma(); fd_free_dma(); @@ -4598,7 +4556,7 @@ static void __exit floppy_module_exit(void) del_timer_sync(&fd_timer); blk_cleanup_queue(floppy_queue); - if (usage_count) + if (atomic_read(&usage_count)) floppy_release_irq_and_dma(); /* eject disk, if any */ diff --git a/drivers/block/hd.c b/drivers/block/hd.c index 81c78b3..30ec6b3 100644 --- a/drivers/block/hd.c +++ b/drivers/block/hd.c @@ -627,7 +627,7 @@ repeat: req_data_dir(req) == READ ? "read" : "writ", cyl, head, sec, nsect, req->buffer); #endif - if (blk_fs_request(req)) { + if (req->cmd_type == REQ_TYPE_FS) { switch (rq_data_dir(req)) { case READ: hd_out(disk, nsect, sec, head, cyl, ATA_CMD_PIO_READ, diff --git a/drivers/block/loop.c b/drivers/block/loop.c index 6120922..f3c636d 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -67,6 +67,7 @@ #include <linux/compat.h> #include <linux/suspend.h> #include <linux/freezer.h> +#include <linux/smp_lock.h> #include <linux/writeback.h> #include <linux/buffer_head.h> /* for invalidate_bdev() */ #include <linux/completion.h> @@ -476,7 +477,7 @@ static int do_bio_filebacked(struct loop_device *lo, struct bio *bio) pos = ((loff_t) bio->bi_sector << 9) + lo->lo_offset; if (bio_rw(bio) == WRITE) { - bool barrier = bio_rw_flagged(bio, BIO_RW_BARRIER); + bool barrier = (bio->bi_rw & REQ_HARDBARRIER); struct file *file = lo->lo_backing_file; if (barrier) { @@ -831,7 +832,7 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, lo->lo_queue->unplug_fn = loop_unplug; if (!(lo_flags & LO_FLAGS_READ_ONLY) && file->f_op->fsync) - blk_queue_ordered(lo->lo_queue, QUEUE_ORDERED_DRAIN, NULL); + blk_queue_ordered(lo->lo_queue, QUEUE_ORDERED_DRAIN); set_capacity(lo->lo_disk, size); bd_set_size(bdev, size << 9); @@ -1408,9 +1409,11 @@ static int lo_open(struct block_device *bdev, fmode_t mode) { struct loop_device *lo = bdev->bd_disk->private_data; + lock_kernel(); mutex_lock(&lo->lo_ctl_mutex); lo->lo_refcnt++; mutex_unlock(&lo->lo_ctl_mutex); + unlock_kernel(); return 0; } @@ -1420,6 +1423,7 @@ static int lo_release(struct gendisk *disk, fmode_t mode) struct loop_device *lo = disk->private_data; int err; + lock_kernel(); mutex_lock(&lo->lo_ctl_mutex); if (--lo->lo_refcnt) @@ -1444,6 +1448,7 @@ static int lo_release(struct gendisk *disk, fmode_t mode) out: mutex_unlock(&lo->lo_ctl_mutex); out_unlocked: + lock_kernel(); return 0; } diff --git a/drivers/block/mg_disk.c b/drivers/block/mg_disk.c index 28db925..b82c5ce 100644 --- a/drivers/block/mg_disk.c +++ b/drivers/block/mg_disk.c @@ -670,7 +670,7 @@ static void mg_request_poll(struct request_queue *q) break; } - if (unlikely(!blk_fs_request(host->req))) { + if (unlikely(host->req->cmd_type != REQ_TYPE_FS)) { mg_end_request_cur(host, -EIO); continue; } @@ -756,7 +756,7 @@ static void mg_request(struct request_queue *q) continue; } - if (unlikely(!blk_fs_request(req))) { + if (unlikely(req->cmd_type != REQ_TYPE_FS)) { mg_end_request_cur(host, -EIO); continue; } diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index 16c3c86..0daa422 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -24,6 +24,7 @@ #include <linux/errno.h> #include <linux/file.h> #include <linux/ioctl.h> +#include <linux/smp_lock.h> #include <linux/compiler.h> #include <linux/err.h> #include <linux/kernel.h> @@ -448,7 +449,7 @@ static void nbd_clear_que(struct nbd_device *lo) static void nbd_handle_req(struct nbd_device *lo, struct request *req) { - if (!blk_fs_request(req)) + if (req->cmd_type != REQ_TYPE_FS) goto error_out; nbd_cmd(req) = NBD_CMD_READ; @@ -716,9 +717,11 @@ static int nbd_ioctl(struct block_device *bdev, fmode_t mode, dprintk(DBG_IOCTL, "%s: nbd_ioctl cmd=%s(0x%x) arg=%lu\n", lo->disk->disk_name, ioctl_cmd_to_ascii(cmd), cmd, arg); + lock_kernel(); mutex_lock(&lo->tx_lock); error = __nbd_ioctl(bdev, lo, cmd, arg); mutex_unlock(&lo->tx_lock); + unlock_kernel(); return error; } @@ -726,7 +729,7 @@ static int nbd_ioctl(struct block_device *bdev, fmode_t mode, static const struct block_device_operations nbd_fops = { .owner = THIS_MODULE, - .locked_ioctl = nbd_ioctl, + .ioctl = nbd_ioctl, }; /* diff --git a/drivers/block/osdblk.c b/drivers/block/osdblk.c index 6cd8b70..2284b4f 100644 --- a/drivers/block/osdblk.c +++ b/drivers/block/osdblk.c @@ -310,7 +310,8 @@ static void osdblk_rq_fn(struct request_queue *q) break; /* filter out block requests we don't understand */ - if (!blk_fs_request(rq) && !blk_barrier_rq(rq)) { + if (rq->cmd_type != REQ_TYPE_FS && + !(rq->cmd_flags & REQ_HARDBARRIER)) { blk_end_request_all(rq, 0); continue; } @@ -322,7 +323,7 @@ static void osdblk_rq_fn(struct request_queue *q) * driver-specific, etc. */ - do_flush = (rq->special == (void *) 0xdeadbeefUL); + do_flush = rq->cmd_flags & REQ_FLUSH; do_write = (rq_data_dir(rq) == WRITE); if (!do_flush) { /* osd_flush does not use a bio */ @@ -379,14 +380,6 @@ static void osdblk_rq_fn(struct request_queue *q) } } -static void osdblk_prepare_flush(struct request_queue *q, struct request *rq) -{ - /* add driver-specific marker, to indicate that this request - * is a flush command - */ - rq->special = (void *) 0xdeadbeefUL; -} - static void osdblk_free_disk(struct osdblk_device *osdev) { struct gendisk *disk = osdev->disk; @@ -446,7 +439,7 @@ static int osdblk_init_disk(struct osdblk_device *osdev) blk_queue_stack_limits(q, osd_request_queue(osdev->osd)); blk_queue_prep_rq(q, blk_queue_start_tag); - blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH, osdblk_prepare_flush); + blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH); disk->queue = q; diff --git a/drivers/block/paride/pcd.c b/drivers/block/paride/pcd.c index 71acf4e..76f8565 100644 --- a/drivers/block/paride/pcd.c +++ b/drivers/block/paride/pcd.c @@ -138,6 +138,7 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_DLY}; #include <linux/cdrom.h> #include <linux/spinlock.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <asm/uaccess.h> static DEFINE_SPINLOCK(pcd_lock); @@ -224,13 +225,21 @@ static char *pcd_buf; /* buffer for request in progress */ static int pcd_block_open(struct block_device *bdev, fmode_t mode) { struct pcd_unit *cd = bdev->bd_disk->private_data; - return cdrom_open(&cd->info, bdev, mode); + int ret; + + lock_kernel(); + ret = cdrom_open(&cd->info, bdev, mode); + unlock_kernel(); + + return ret; } static int pcd_block_release(struct gendisk *disk, fmode_t mode) { struct pcd_unit *cd = disk->private_data; + lock_kernel(); cdrom_release(&cd->info, mode); + unlock_kernel(); return 0; } @@ -238,7 +247,13 @@ static int pcd_block_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg) { struct pcd_unit *cd = bdev->bd_disk->private_data; - return cdrom_ioctl(&cd->info, bdev, mode, cmd, arg); + int ret; + + lock_kernel(); + ret = cdrom_ioctl(&cd->info, bdev, mode, cmd, arg); + unlock_kernel(); + + return ret; } static int pcd_block_media_changed(struct gendisk *disk) @@ -251,7 +266,7 @@ static const struct block_device_operations pcd_bdops = { .owner = THIS_MODULE, .open = pcd_block_open, .release = pcd_block_release, - .locked_ioctl = pcd_block_ioctl, + .ioctl = pcd_block_ioctl, .media_changed = pcd_block_media_changed, }; diff --git a/drivers/block/paride/pd.c b/drivers/block/paride/pd.c index c1e5cd0..985f0d4 100644 --- a/drivers/block/paride/pd.c +++ b/drivers/block/paride/pd.c @@ -153,6 +153,7 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_GEO, D_SBY, D_DLY, D_SLV}; #include <linux/blkdev.h> #include <linux/blkpg.h> #include <linux/kernel.h> +#include <linux/smp_lock.h> #include <asm/uaccess.h> #include <linux/workqueue.h> @@ -439,7 +440,7 @@ static char *pd_buf; /* buffer for request in progress */ static enum action do_pd_io_start(void) { - if (blk_special_request(pd_req)) { + if (pd_req->cmd_type == REQ_TYPE_SPECIAL) { phase = pd_special; return pd_special(); } @@ -735,12 +736,14 @@ static int pd_open(struct block_device *bdev, fmode_t mode) { struct pd_unit *disk = bdev->bd_disk->private_data; + lock_kernel(); disk->access++; if (disk->removable) { pd_special_command(disk, pd_media_check); pd_special_command(disk, pd_door_lock); } + unlock_kernel(); return 0; } @@ -768,8 +771,10 @@ static int pd_ioctl(struct block_device *bdev, fmode_t mode, switch (cmd) { case CDROMEJECT: + lock_kernel(); if (disk->access == 1) pd_special_command(disk, pd_eject); + unlock_kernel(); return 0; default: return -EINVAL; @@ -780,8 +785,10 @@ static int pd_release(struct gendisk *p, fmode_t mode) { struct pd_unit *disk = p->private_data; + lock_kernel(); if (!--disk->access && disk->removable) pd_special_command(disk, pd_door_unlock); + unlock_kernel(); return 0; } @@ -812,7 +819,7 @@ static const struct block_device_operations pd_fops = { .owner = THIS_MODULE, .open = pd_open, .release = pd_release, - .locked_ioctl = pd_ioctl, + .ioctl = pd_ioctl, .getgeo = pd_getgeo, .media_changed = pd_check_media, .revalidate_disk= pd_revalidate diff --git a/drivers/block/paride/pf.c b/drivers/block/paride/pf.c index c059aab..4457b49 100644 --- a/drivers/block/paride/pf.c +++ b/drivers/block/paride/pf.c @@ -152,6 +152,7 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_LUN, D_DLY}; #include <linux/spinlock.h> #include <linux/blkdev.h> #include <linux/blkpg.h> +#include <linux/smp_lock.h> #include <asm/uaccess.h> static DEFINE_SPINLOCK(pf_spin_lock); @@ -266,7 +267,7 @@ static const struct block_device_operations pf_fops = { .owner = THIS_MODULE, .open = pf_open, .release = pf_release, - .locked_ioctl = pf_ioctl, + .ioctl = pf_ioctl, .getgeo = pf_getgeo, .media_changed = pf_check_media, }; @@ -299,20 +300,26 @@ static void __init pf_init_units(void) static int pf_open(struct block_device *bdev, fmode_t mode) { struct pf_unit *pf = bdev->bd_disk->private_data; + int ret; + lock_kernel(); pf_identify(pf); + ret = -ENODEV; if (pf->media_status == PF_NM) - return -ENODEV; + goto out; + ret = -EROFS; if ((pf->media_status == PF_RO) && (mode & FMODE_WRITE)) - return -EROFS; + goto out; + ret = 0; pf->access++; if (pf->removable) pf_lock(pf, 1); - - return 0; +out: + unlock_kernel(); + return ret; } static int pf_getgeo(struct block_device *bdev, struct hd_geometry *geo) @@ -342,7 +349,10 @@ static int pf_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, u if (pf->access != 1) return -EBUSY; + lock_kernel(); pf_eject(pf); + unlock_kernel(); + return 0; } @@ -350,14 +360,18 @@ static int pf_release(struct gendisk *disk, fmode_t mode) { struct pf_unit *pf = disk->private_data; - if (pf->access <= 0) + lock_kernel(); + if (pf->access <= 0) { + unlock_kernel(); return -EINVAL; + } pf->access--; if (!pf->access && pf->removable) pf_lock(pf, 0); + unlock_kernel(); return 0; } diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index 8a549db..b1cbeb5 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -57,6 +57,7 @@ #include <linux/seq_file.h> #include <linux/miscdevice.h> #include <linux/freezer.h> +#include <linux/smp_lock.h> #include <linux/mutex.h> #include <linux/slab.h> #include <scsi/scsi_cmnd.h> @@ -1221,7 +1222,7 @@ static int pkt_start_recovery(struct packet_data *pkt) pkt->bio->bi_flags = 1 << BIO_UPTODATE; pkt->bio->bi_idx = 0; - BUG_ON(pkt->bio->bi_rw != (1 << BIO_RW)); + BUG_ON(pkt->bio->bi_rw != REQ_WRITE); BUG_ON(pkt->bio->bi_vcnt != pkt->frames); BUG_ON(pkt->bio->bi_size != pkt->frames * CD_FRAMESIZE); BUG_ON(pkt->bio->bi_end_io != pkt_end_io_packet_write); @@ -2382,6 +2383,7 @@ static int pkt_open(struct block_device *bdev, fmode_t mode) VPRINTK(DRIVER_NAME": entering open\n"); + lock_kernel(); mutex_lock(&ctl_mutex); pd = pkt_find_dev_from_minor(MINOR(bdev->bd_dev)); if (!pd) { @@ -2409,6 +2411,7 @@ static int pkt_open(struct block_device *bdev, fmode_t mode) } mutex_unlock(&ctl_mutex); + unlock_kernel(); return 0; out_dec: @@ -2416,6 +2419,7 @@ out_dec: out: VPRINTK(DRIVER_NAME": failed open (%d)\n", ret); mutex_unlock(&ctl_mutex); + unlock_kernel(); return ret; } @@ -2424,6 +2428,7 @@ static int pkt_close(struct gendisk *disk, fmode_t mode) struct pktcdvd_device *pd = disk->private_data; int ret = 0; + lock_kernel(); mutex_lock(&ctl_mutex); pd->refcnt--; BUG_ON(pd->refcnt < 0); @@ -2432,6 +2437,7 @@ static int pkt_close(struct gendisk *disk, fmode_t mode) pkt_release_dev(pd, flush); } mutex_unlock(&ctl_mutex); + unlock_kernel(); return ret; } @@ -2762,10 +2768,12 @@ out_mem: static int pkt_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { struct pktcdvd_device *pd = bdev->bd_disk->private_data; + int ret; VPRINTK("pkt_ioctl: cmd %x, dev %d:%d\n", cmd, MAJOR(bdev->bd_dev), MINOR(bdev->bd_dev)); + lock_kernel(); switch (cmd) { case CDROMEJECT: /* @@ -2783,14 +2791,16 @@ static int pkt_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, case CDROM_LAST_WRITTEN: case CDROM_SEND_PACKET: case SCSI_IOCTL_SEND_COMMAND: - return __blkdev_driver_ioctl(pd->bdev, mode, cmd, arg); + ret = __blkdev_driver_ioctl(pd->bdev, mode, cmd, arg); + break; default: VPRINTK(DRIVER_NAME": Unknown ioctl for %s (%x)\n", pd->name, cmd); - return -ENOTTY; + ret = -ENOTTY; } + unlock_kernel(); - return 0; + return ret; } static int pkt_media_changed(struct gendisk *disk) @@ -2812,7 +2822,7 @@ static const struct block_device_operations pktcdvd_ops = { .owner = THIS_MODULE, .open = pkt_open, .release = pkt_close, - .locked_ioctl = pkt_ioctl, + .ioctl = pkt_ioctl, .media_changed = pkt_media_changed, }; diff --git a/drivers/block/ps3disk.c b/drivers/block/ps3disk.c index 3b419e3..e9da874 100644 --- a/drivers/block/ps3disk.c +++ b/drivers/block/ps3disk.c @@ -196,13 +196,12 @@ static void ps3disk_do_request(struct ps3_storage_device *dev, dev_dbg(&dev->sbd.core, "%s:%u\n", __func__, __LINE__); while ((req = blk_fetch_request(q))) { - if (blk_fs_request(req)) { - if (ps3disk_submit_request_sg(dev, req)) - break; - } else if (req->cmd_type == REQ_TYPE_LINUX_BLOCK && - req->cmd[0] == REQ_LB_OP_FLUSH) { + if (req->cmd_flags & REQ_FLUSH) { if (ps3disk_submit_flush_request(dev, req)) break; + } else if (req->cmd_type == REQ_TYPE_FS) { + if (ps3disk_submit_request_sg(dev, req)) + break; } else { blk_dump_rq_flags(req, DEVICE_NAME " bad request"); __blk_end_request_all(req, -EIO); @@ -257,8 +256,7 @@ static irqreturn_t ps3disk_interrupt(int irq, void *data) return IRQ_HANDLED; } - if (req->cmd_type == REQ_TYPE_LINUX_BLOCK && - req->cmd[0] == REQ_LB_OP_FLUSH) { + if (req->cmd_flags & REQ_FLUSH) { read = 0; op = "flush"; } else { @@ -398,16 +396,6 @@ static int ps3disk_identify(struct ps3_storage_device *dev) return 0; } -static void ps3disk_prepare_flush(struct request_queue *q, struct request *req) -{ - struct ps3_storage_device *dev = q->queuedata; - - dev_dbg(&dev->sbd.core, "%s:%u\n", __func__, __LINE__); - - req->cmd_type = REQ_TYPE_LINUX_BLOCK; - req->cmd[0] = REQ_LB_OP_FLUSH; -} - static unsigned long ps3disk_mask; static DEFINE_MUTEX(ps3disk_mask_mutex); @@ -480,8 +468,7 @@ static int __devinit ps3disk_probe(struct ps3_system_bus_device *_dev) blk_queue_dma_alignment(queue, dev->blk_size-1); blk_queue_logical_block_size(queue, dev->blk_size); - blk_queue_ordered(queue, QUEUE_ORDERED_DRAIN_FLUSH, - ps3disk_prepare_flush); + blk_queue_ordered(queue, QUEUE_ORDERED_DRAIN_FLUSH); blk_queue_max_segments(queue, -1); blk_queue_max_segment_size(queue, dev->bounce_size); diff --git a/drivers/block/swim.c b/drivers/block/swim.c index e463657..2e46815 100644 --- a/drivers/block/swim.c +++ b/drivers/block/swim.c @@ -20,6 +20,7 @@ #include <linux/fd.h> #include <linux/slab.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <linux/hdreg.h> #include <linux/kernel.h> #include <linux/delay.h> @@ -661,11 +662,23 @@ out: return err; } +static int floppy_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = floppy_open(bdev, mode); + unlock_kernel(); + + return ret; +} + static int floppy_release(struct gendisk *disk, fmode_t mode) { struct floppy_state *fs = disk->private_data; struct swim __iomem *base = fs->swd->base; + lock_kernel(); if (fs->ref_count < 0) fs->ref_count = 0; else if (fs->ref_count > 0) @@ -673,6 +686,7 @@ static int floppy_release(struct gendisk *disk, fmode_t mode) if (fs->ref_count == 0) swim_motor(base, OFF); + unlock_kernel(); return 0; } @@ -690,7 +704,9 @@ static int floppy_ioctl(struct block_device *bdev, fmode_t mode, case FDEJECT: if (fs->ref_count != 1) return -EBUSY; + lock_kernel(); err = floppy_eject(fs); + unlock_kernel(); return err; case FDGETPRM: @@ -751,9 +767,9 @@ static int floppy_revalidate(struct gendisk *disk) static const struct block_device_operations floppy_fops = { .owner = THIS_MODULE, - .open = floppy_open, + .open = floppy_unlocked_open, .release = floppy_release, - .locked_ioctl = floppy_ioctl, + .ioctl = floppy_ioctl, .getgeo = floppy_getgeo, .media_changed = floppy_check_change, .revalidate_disk = floppy_revalidate, diff --git a/drivers/block/swim3.c b/drivers/block/swim3.c index ed6fb91..cc6a386 100644 --- a/drivers/block/swim3.c +++ b/drivers/block/swim3.c @@ -25,6 +25,7 @@ #include <linux/ioctl.h> #include <linux/blkdev.h> #include <linux/interrupt.h> +#include <linux/smp_lock.h> #include <linux/module.h> #include <linux/spinlock.h> #include <asm/io.h> @@ -839,7 +840,7 @@ static int fd_eject(struct floppy_state *fs) static struct floppy_struct floppy_type = { 2880,18,2,80,0,0x1B,0x00,0xCF,0x6C,NULL }; /* 7 1.44MB 3.5" */ -static int floppy_ioctl(struct block_device *bdev, fmode_t mode, +static int floppy_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long param) { struct floppy_state *fs = bdev->bd_disk->private_data; @@ -867,6 +868,18 @@ static int floppy_ioctl(struct block_device *bdev, fmode_t mode, return -ENOTTY; } +static int floppy_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long param) +{ + int ret; + + lock_kernel(); + ret = floppy_locked_ioctl(bdev, mode, cmd, param); + unlock_kernel(); + + return ret; +} + static int floppy_open(struct block_device *bdev, fmode_t mode) { struct floppy_state *fs = bdev->bd_disk->private_data; @@ -936,15 +949,28 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) return 0; } +static int floppy_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = floppy_open(bdev, mode); + unlock_kernel(); + + return ret; +} + static int floppy_release(struct gendisk *disk, fmode_t mode) { struct floppy_state *fs = disk->private_data; struct swim3 __iomem *sw = fs->swim3; + lock_kernel(); if (fs->ref_count > 0 && --fs->ref_count == 0) { swim3_action(fs, MOTOR_OFF); out_8(&sw->control_bic, 0xff); swim3_select(fs, RELAX); } + unlock_kernel(); return 0; } @@ -995,9 +1021,9 @@ static int floppy_revalidate(struct gendisk *disk) } static const struct block_device_operations floppy_fops = { - .open = floppy_open, + .open = floppy_unlocked_open, .release = floppy_release, - .locked_ioctl = floppy_ioctl, + .ioctl = floppy_ioctl, .media_changed = floppy_check_change, .revalidate_disk= floppy_revalidate, }; diff --git a/drivers/block/ub.c b/drivers/block/ub.c index 0536b5b..c48e148 100644 --- a/drivers/block/ub.c +++ b/drivers/block/ub.c @@ -28,6 +28,7 @@ #include <linux/timer.h> #include <linux/scatterlist.h> #include <linux/slab.h> +#include <linux/smp_lock.h> #include <scsi/scsi.h> #define DRV_NAME "ub" @@ -648,7 +649,7 @@ static int ub_request_fn_1(struct ub_lun *lun, struct request *rq) return 0; } - if (lun->changed && !blk_pc_request(rq)) { + if (lun->changed && rq->cmd_type != REQ_TYPE_BLOCK_PC) { blk_start_request(rq); ub_end_rq(rq, SAM_STAT_CHECK_CONDITION); return 0; @@ -684,7 +685,7 @@ static int ub_request_fn_1(struct ub_lun *lun, struct request *rq) } urq->nsg = n_elem; - if (blk_pc_request(rq)) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { ub_cmd_build_packet(sc, lun, cmd, urq); } else { ub_cmd_build_block(sc, lun, cmd, urq); @@ -781,7 +782,7 @@ static void ub_rw_cmd_done(struct ub_dev *sc, struct ub_scsi_cmd *cmd) rq = urq->rq; if (cmd->error == 0) { - if (blk_pc_request(rq)) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { if (cmd->act_len >= rq->resid_len) rq->resid_len = 0; else @@ -795,7 +796,7 @@ static void ub_rw_cmd_done(struct ub_dev *sc, struct ub_scsi_cmd *cmd) } } } else { - if (blk_pc_request(rq)) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { /* UB_SENSE_SIZE is smaller than SCSI_SENSE_BUFFERSIZE */ memcpy(rq->sense, sc->top_sense, UB_SENSE_SIZE); rq->sense_len = UB_SENSE_SIZE; @@ -1710,6 +1711,18 @@ err_open: return rc; } +static int ub_bd_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = ub_bd_open(bdev, mode); + unlock_kernel(); + + return ret; +} + + /* */ static int ub_bd_release(struct gendisk *disk, fmode_t mode) @@ -1717,7 +1730,10 @@ static int ub_bd_release(struct gendisk *disk, fmode_t mode) struct ub_lun *lun = disk->private_data; struct ub_dev *sc = lun->udev; + lock_kernel(); ub_put(sc); + unlock_kernel(); + return 0; } @@ -1729,8 +1745,13 @@ static int ub_bd_ioctl(struct block_device *bdev, fmode_t mode, { struct gendisk *disk = bdev->bd_disk; void __user *usermem = (void __user *) arg; + int ret; + + lock_kernel(); + ret = scsi_cmd_ioctl(disk->queue, disk, mode, cmd, usermem); + unlock_kernel(); - return scsi_cmd_ioctl(disk->queue, disk, mode, cmd, usermem); + return ret; } /* @@ -1792,9 +1813,9 @@ static int ub_bd_media_changed(struct gendisk *disk) static const struct block_device_operations ub_bd_fops = { .owner = THIS_MODULE, - .open = ub_bd_open, + .open = ub_bd_unlocked_open, .release = ub_bd_release, - .locked_ioctl = ub_bd_ioctl, + .ioctl = ub_bd_ioctl, .media_changed = ub_bd_media_changed, .revalidate_disk = ub_bd_revalidate, }; diff --git a/drivers/block/umem.c b/drivers/block/umem.c index 2f9470f..8be5715 100644 --- a/drivers/block/umem.c +++ b/drivers/block/umem.c @@ -478,7 +478,7 @@ static void process_page(unsigned long data) le32_to_cpu(desc->local_addr)>>9, le32_to_cpu(desc->transfer_size)); dump_dmastat(card, control); - } else if (test_bit(BIO_RW, &bio->bi_rw) && + } else if ((bio->bi_rw & REQ_WRITE) && le32_to_cpu(desc->local_addr) >> 9 == card->init_size) { card->init_size += le32_to_cpu(desc->transfer_size) >> 9; diff --git a/drivers/block/viodasd.c b/drivers/block/viodasd.c index 788d938..f651e51 100644 --- a/drivers/block/viodasd.c +++ b/drivers/block/viodasd.c @@ -41,6 +41,7 @@ #include <linux/errno.h> #include <linux/init.h> #include <linux/string.h> +#include <linux/smp_lock.h> #include <linux/dma-mapping.h> #include <linux/completion.h> #include <linux/device.h> @@ -175,6 +176,18 @@ static int viodasd_open(struct block_device *bdev, fmode_t mode) return 0; } +static int viodasd_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = viodasd_open(bdev, mode); + unlock_kernel(); + + return ret; +} + + /* * External release entry point. */ @@ -183,6 +196,7 @@ static int viodasd_release(struct gendisk *disk, fmode_t mode) struct viodasd_device *d = disk->private_data; HvLpEvent_Rc hvrc; + lock_kernel(); /* Send the event to OS/400. We DON'T expect a response */ hvrc = HvCallEvent_signalLpEventFast(viopath_hostLp, HvLpEvent_Type_VirtualIo, @@ -195,6 +209,9 @@ static int viodasd_release(struct gendisk *disk, fmode_t mode) 0, 0, 0); if (hvrc != 0) pr_warning("HV close call failed %d\n", (int)hvrc); + + unlock_kernel(); + return 0; } @@ -219,7 +236,7 @@ static int viodasd_getgeo(struct block_device *bdev, struct hd_geometry *geo) */ static const struct block_device_operations viodasd_fops = { .owner = THIS_MODULE, - .open = viodasd_open, + .open = viodasd_unlocked_open, .release = viodasd_release, .getgeo = viodasd_getgeo, }; @@ -361,7 +378,7 @@ static void do_viodasd_request(struct request_queue *q) if (req == NULL) return; /* check that request contains a valid command */ - if (!blk_fs_request(req)) { + if (req->cmd_type != REQ_TYPE_FS) { viodasd_end_request(req, -EIO, blk_rq_sectors(req)); continue; } diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 23b7c48..2aafafc 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -2,6 +2,7 @@ #include <linux/spinlock.h> #include <linux/slab.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <linux/hdreg.h> #include <linux/virtio.h> #include <linux/virtio_blk.h> @@ -65,13 +66,18 @@ static void blk_done(struct virtqueue *vq) break; } - if (blk_pc_request(vbr->req)) { + switch (vbr->req->cmd_type) { + case REQ_TYPE_BLOCK_PC: vbr->req->resid_len = vbr->in_hdr.residual; vbr->req->sense_len = vbr->in_hdr.sense_len; vbr->req->errors = vbr->in_hdr.errors; - } - if (blk_special_request(vbr->req)) + break; + case REQ_TYPE_SPECIAL: vbr->req->errors = (error != 0); + break; + default: + break; + } __blk_end_request_all(vbr->req, error); list_del(&vbr->list); @@ -94,36 +100,35 @@ static bool do_req(struct request_queue *q, struct virtio_blk *vblk, return false; vbr->req = req; - switch (req->cmd_type) { - case REQ_TYPE_FS: - vbr->out_hdr.type = 0; - vbr->out_hdr.sector = blk_rq_pos(vbr->req); - vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); - break; - case REQ_TYPE_BLOCK_PC: - vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD; - vbr->out_hdr.sector = 0; - vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); - break; - case REQ_TYPE_SPECIAL: - vbr->out_hdr.type = VIRTIO_BLK_T_GET_ID; + + if (req->cmd_flags & REQ_FLUSH) { + vbr->out_hdr.type = VIRTIO_BLK_T_FLUSH; vbr->out_hdr.sector = 0; vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); - break; - case REQ_TYPE_LINUX_BLOCK: - if (req->cmd[0] == REQ_LB_OP_FLUSH) { - vbr->out_hdr.type = VIRTIO_BLK_T_FLUSH; + } else { + switch (req->cmd_type) { + case REQ_TYPE_FS: + vbr->out_hdr.type = 0; + vbr->out_hdr.sector = blk_rq_pos(vbr->req); + vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); + break; + case REQ_TYPE_BLOCK_PC: + vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD; vbr->out_hdr.sector = 0; vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); break; + case REQ_TYPE_SPECIAL: + vbr->out_hdr.type = VIRTIO_BLK_T_GET_ID; + vbr->out_hdr.sector = 0; + vbr->out_hdr.ioprio = req_get_ioprio(vbr->req); + break; + default: + /* We don't put anything else in the queue. */ + BUG(); } - /*FALLTHRU*/ - default: - /* We don't put anything else in the queue. */ - BUG(); } - if (blk_barrier_rq(vbr->req)) + if (vbr->req->cmd_flags & REQ_HARDBARRIER) vbr->out_hdr.type |= VIRTIO_BLK_T_BARRIER; sg_set_buf(&vblk->sg[out++], &vbr->out_hdr, sizeof(vbr->out_hdr)); @@ -134,12 +139,12 @@ static bool do_req(struct request_queue *q, struct virtio_blk *vblk, * block, and before the normal inhdr we put the sense data and the * inhdr with additional status information before the normal inhdr. */ - if (blk_pc_request(vbr->req)) + if (vbr->req->cmd_type == REQ_TYPE_BLOCK_PC) sg_set_buf(&vblk->sg[out++], vbr->req->cmd, vbr->req->cmd_len); num = blk_rq_map_sg(q, vbr->req, vblk->sg + out); - if (blk_pc_request(vbr->req)) { + if (vbr->req->cmd_type == REQ_TYPE_BLOCK_PC) { sg_set_buf(&vblk->sg[num + out + in++], vbr->req->sense, 96); sg_set_buf(&vblk->sg[num + out + in++], &vbr->in_hdr, sizeof(vbr->in_hdr)); @@ -190,12 +195,6 @@ static void do_virtblk_request(struct request_queue *q) virtqueue_kick(vblk->vq); } -static void virtblk_prepare_flush(struct request_queue *q, struct request *req) -{ - req->cmd_type = REQ_TYPE_LINUX_BLOCK; - req->cmd[0] = REQ_LB_OP_FLUSH; -} - /* return id (s/n) string for *disk to *id_str */ static int virtblk_get_id(struct gendisk *disk, char *id_str) @@ -219,7 +218,7 @@ static int virtblk_get_id(struct gendisk *disk, char *id_str) return blk_execute_rq(vblk->disk->queue, vblk->disk, req, false); } -static int virtblk_ioctl(struct block_device *bdev, fmode_t mode, +static int virtblk_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long data) { struct gendisk *disk = bdev->bd_disk; @@ -235,6 +234,18 @@ static int virtblk_ioctl(struct block_device *bdev, fmode_t mode, (void __user *)data); } +static int virtblk_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long param) +{ + int ret; + + lock_kernel(); + ret = virtblk_locked_ioctl(bdev, mode, cmd, param); + unlock_kernel(); + + return ret; +} + /* We provide getgeo only to please some old bootloader/partitioning tools */ static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo) { @@ -261,7 +272,7 @@ static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo) } static const struct block_device_operations virtblk_fops = { - .locked_ioctl = virtblk_ioctl, + .ioctl = virtblk_ioctl, .owner = THIS_MODULE, .getgeo = virtblk_getgeo, }; @@ -383,8 +394,7 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) * flushing a volatile write cache on the host. Use that * to implement write barrier support. */ - blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH, - virtblk_prepare_flush); + blk_queue_ordered(q, QUEUE_ORDERED_DRAIN_FLUSH); } else if (virtio_has_feature(vdev, VIRTIO_BLK_F_BARRIER)) { /* * If the BARRIER feature is supported the host expects us @@ -393,7 +403,7 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) * never re-orders outstanding I/O. This feature is not * useful for real life scenarious and deprecated. */ - blk_queue_ordered(q, QUEUE_ORDERED_TAG, NULL); + blk_queue_ordered(q, QUEUE_ORDERED_TAG); } else { /* * If the FLUSH feature is not supported we must assume that @@ -401,7 +411,7 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) * caching. We still need to drain the queue to provider * proper barrier semantics. */ - blk_queue_ordered(q, QUEUE_ORDERED_DRAIN, NULL); + blk_queue_ordered(q, QUEUE_ORDERED_DRAIN); } /* If disk is read-only in the host, the guest should obey */ diff --git a/drivers/block/xd.c b/drivers/block/xd.c index 18a80ff..d5a3cd7 100644 --- a/drivers/block/xd.c +++ b/drivers/block/xd.c @@ -46,6 +46,7 @@ #include <linux/init.h> #include <linux/wait.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <linux/blkpg.h> #include <linux/delay.h> #include <linux/io.h> @@ -133,7 +134,7 @@ static int xd_getgeo(struct block_device *bdev, struct hd_geometry *geo); static const struct block_device_operations xd_fops = { .owner = THIS_MODULE, - .locked_ioctl = xd_ioctl, + .ioctl = xd_ioctl, .getgeo = xd_getgeo, }; static DECLARE_WAIT_QUEUE_HEAD(xd_wait_int); @@ -322,7 +323,7 @@ static void do_xd_request (struct request_queue * q) int res = -EIO; int retry; - if (!blk_fs_request(req)) + if (req->cmd_type != REQ_TYPE_FS) goto done; if (block + count > get_capacity(req->rq_disk)) goto done; @@ -347,7 +348,7 @@ static int xd_getgeo(struct block_device *bdev, struct hd_geometry *geo) } /* xd_ioctl: handle device ioctl's */ -static int xd_ioctl(struct block_device *bdev, fmode_t mode, u_int cmd, u_long arg) +static int xd_locked_ioctl(struct block_device *bdev, fmode_t mode, u_int cmd, u_long arg) { switch (cmd) { case HDIO_SET_DMA: @@ -375,6 +376,18 @@ static int xd_ioctl(struct block_device *bdev, fmode_t mode, u_int cmd, u_long a } } +static int xd_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long param) +{ + int ret; + + lock_kernel(); + ret = xd_locked_ioctl(bdev, mode, cmd, param); + unlock_kernel(); + + return ret; +} + /* xd_readwrite: handle a read/write request */ static int xd_readwrite (u_char operation,XD_INFO *p,char *buffer,u_int block,u_int count) { diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c index f63ac3d..ac1b682 100644 --- a/drivers/block/xen-blkfront.c +++ b/drivers/block/xen-blkfront.c @@ -41,6 +41,7 @@ #include <linux/cdrom.h> #include <linux/module.h> #include <linux/slab.h> +#include <linux/smp_lock.h> #include <linux/scatterlist.h> #include <xen/xen.h> @@ -79,6 +80,7 @@ static const struct block_device_operations xlvbd_block_fops; */ struct blkfront_info { + struct mutex mutex; struct xenbus_device *xbdev; struct gendisk *gd; int vdevice; @@ -95,16 +97,14 @@ struct blkfront_info unsigned long shadow_free; int feature_barrier; int is_ready; - - /** - * The number of people holding this device open. We won't allow a - * hot-unplug unless this is 0. - */ - int users; }; static DEFINE_SPINLOCK(blkif_io_lock); +static unsigned int nr_minors; +static unsigned long *minors; +static DEFINE_SPINLOCK(minor_lock); + #define MAXIMUM_OUTSTANDING_BLOCK_REQS \ (BLKIF_MAX_SEGMENTS_PER_REQUEST * BLK_RING_SIZE) #define GRANT_INVALID_REF 0 @@ -139,6 +139,55 @@ static void add_id_to_freelist(struct blkfront_info *info, info->shadow_free = id; } +static int xlbd_reserve_minors(unsigned int minor, unsigned int nr) +{ + unsigned int end = minor + nr; + int rc; + + if (end > nr_minors) { + unsigned long *bitmap, *old; + + bitmap = kzalloc(BITS_TO_LONGS(end) * sizeof(*bitmap), + GFP_KERNEL); + if (bitmap == NULL) + return -ENOMEM; + + spin_lock(&minor_lock); + if (end > nr_minors) { + old = minors; + memcpy(bitmap, minors, + BITS_TO_LONGS(nr_minors) * sizeof(*bitmap)); + minors = bitmap; + nr_minors = BITS_TO_LONGS(end) * BITS_PER_LONG; + } else + old = bitmap; + spin_unlock(&minor_lock); + kfree(old); + } + + spin_lock(&minor_lock); + if (find_next_bit(minors, end, minor) >= end) { + for (; minor < end; ++minor) + __set_bit(minor, minors); + rc = 0; + } else + rc = -EBUSY; + spin_unlock(&minor_lock); + + return rc; +} + +static void xlbd_release_minors(unsigned int minor, unsigned int nr) +{ + unsigned int end = minor + nr; + + BUG_ON(end > nr_minors); + spin_lock(&minor_lock); + for (; minor < end; ++minor) + __clear_bit(minor, minors); + spin_unlock(&minor_lock); +} + static void blkif_restart_queue_callback(void *arg) { struct blkfront_info *info = (struct blkfront_info *)arg; @@ -239,7 +288,7 @@ static int blkif_queue_request(struct request *req) ring_req->operation = rq_data_dir(req) ? BLKIF_OP_WRITE : BLKIF_OP_READ; - if (blk_barrier_rq(req)) + if (req->cmd_flags & REQ_HARDBARRIER) ring_req->operation = BLKIF_OP_WRITE_BARRIER; ring_req->nr_segments = blk_rq_map_sg(req->q, req, info->sg); @@ -310,7 +359,7 @@ static void do_blkif_request(struct request_queue *rq) blk_start_request(req); - if (!blk_fs_request(req)) { + if (req->cmd_type != REQ_TYPE_FS) { __blk_end_request_all(req, -EIO); continue; } @@ -372,17 +421,22 @@ static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size) static int xlvbd_barrier(struct blkfront_info *info) { int err; + const char *barrier; - err = blk_queue_ordered(info->rq, - info->feature_barrier ? QUEUE_ORDERED_DRAIN : QUEUE_ORDERED_NONE, - NULL); + switch (info->feature_barrier) { + case QUEUE_ORDERED_DRAIN: barrier = "enabled (drain)"; break; + case QUEUE_ORDERED_TAG: barrier = "enabled (tag)"; break; + case QUEUE_ORDERED_NONE: barrier = "disabled"; break; + default: return -EINVAL; + } + + err = blk_queue_ordered(info->rq, info->feature_barrier); if (err) return err; printk(KERN_INFO "blkfront: %s: barriers %s\n", - info->gd->disk_name, - info->feature_barrier ? "enabled" : "disabled"); + info->gd->disk_name, barrier); return 0; } @@ -418,9 +472,14 @@ static int xlvbd_alloc_gendisk(blkif_sector_t capacity, if ((minor % nr_parts) == 0) nr_minors = nr_parts; + err = xlbd_reserve_minors(minor, nr_minors); + if (err) + goto out; + err = -ENODEV; + gd = alloc_disk(nr_minors); if (gd == NULL) - goto out; + goto release; offset = minor / nr_parts; @@ -451,14 +510,13 @@ static int xlvbd_alloc_gendisk(blkif_sector_t capacity, if (xlvbd_init_blk_queue(gd, sector_size)) { del_gendisk(gd); - goto out; + goto release; } info->rq = gd->queue; info->gd = gd; - if (info->feature_barrier) - xlvbd_barrier(info); + xlvbd_barrier(info); if (vdisk_info & VDISK_READONLY) set_disk_ro(gd, 1); @@ -471,10 +529,45 @@ static int xlvbd_alloc_gendisk(blkif_sector_t capacity, return 0; + release: + xlbd_release_minors(minor, nr_minors); out: return err; } +static void xlvbd_release_gendisk(struct blkfront_info *info) +{ + unsigned int minor, nr_minors; + unsigned long flags; + + if (info->rq == NULL) + return; + + spin_lock_irqsave(&blkif_io_lock, flags); + + /* No more blkif_request(). */ + blk_stop_queue(info->rq); + + /* No more gnttab callback work. */ + gnttab_cancel_free_callback(&info->callback); + spin_unlock_irqrestore(&blkif_io_lock, flags); + + /* Flush gnttab callback work. Must be done with no locks held. */ + flush_scheduled_work(); + + del_gendisk(info->gd); + + minor = info->gd->first_minor; + nr_minors = info->gd->minors; + xlbd_release_minors(minor, nr_minors); + + blk_cleanup_queue(info->rq); + info->rq = NULL; + + put_disk(info->gd); + info->gd = NULL; +} + static void kick_pending_request_queues(struct blkfront_info *info) { if (!RING_FULL(&info->ring)) { @@ -569,7 +662,7 @@ static irqreturn_t blkif_interrupt(int irq, void *dev_id) printk(KERN_WARNING "blkfront: %s: write barrier op failed\n", info->gd->disk_name); error = -EOPNOTSUPP; - info->feature_barrier = 0; + info->feature_barrier = QUEUE_ORDERED_NONE; xlvbd_barrier(info); } /* fall through */ @@ -652,7 +745,7 @@ fail: /* Common code used when first setting up, and when resuming. */ -static int talk_to_backend(struct xenbus_device *dev, +static int talk_to_blkback(struct xenbus_device *dev, struct blkfront_info *info) { const char *message = NULL; @@ -712,7 +805,6 @@ again: return err; } - /** * Entry point to this code when a new device is created. Allocate the basic * structures and the ring buffer for communication with the backend, and @@ -773,6 +865,7 @@ static int blkfront_probe(struct xenbus_device *dev, return -ENOMEM; } + mutex_init(&info->mutex); info->xbdev = dev; info->vdevice = vdevice; info->connected = BLKIF_STATE_DISCONNECTED; @@ -786,7 +879,7 @@ static int blkfront_probe(struct xenbus_device *dev, info->handle = simple_strtoul(strrchr(dev->nodename, '/')+1, NULL, 0); dev_set_drvdata(&dev->dev, info); - err = talk_to_backend(dev, info); + err = talk_to_blkback(dev, info); if (err) { kfree(info); dev_set_drvdata(&dev->dev, NULL); @@ -881,13 +974,50 @@ static int blkfront_resume(struct xenbus_device *dev) blkif_free(info, info->connected == BLKIF_STATE_CONNECTED); - err = talk_to_backend(dev, info); + err = talk_to_blkback(dev, info); if (info->connected == BLKIF_STATE_SUSPENDED && !err) err = blkif_recover(info); return err; } +static void +blkfront_closing(struct blkfront_info *info) +{ + struct xenbus_device *xbdev = info->xbdev; + struct block_device *bdev = NULL; + + mutex_lock(&info->mutex); + + if (xbdev->state == XenbusStateClosing) { + mutex_unlock(&info->mutex); + return; + } + + if (info->gd) + bdev = bdget_disk(info->gd, 0); + + mutex_unlock(&info->mutex); + + if (!bdev) { + xenbus_frontend_closed(xbdev); + return; + } + + mutex_lock(&bdev->bd_mutex); + + if (bdev->bd_openers) { + xenbus_dev_error(xbdev, -EBUSY, + "Device in use; refusing to close"); + xenbus_switch_state(xbdev, XenbusStateClosing); + } else { + xlvbd_release_gendisk(info); + xenbus_frontend_closed(xbdev); + } + + mutex_unlock(&bdev->bd_mutex); + bdput(bdev); +} /* * Invoked when the backend is finally 'ready' (and has told produced @@ -899,11 +1029,31 @@ static void blkfront_connect(struct blkfront_info *info) unsigned long sector_size; unsigned int binfo; int err; - - if ((info->connected == BLKIF_STATE_CONNECTED) || - (info->connected == BLKIF_STATE_SUSPENDED) ) + int barrier; + + switch (info->connected) { + case BLKIF_STATE_CONNECTED: + /* + * Potentially, the back-end may be signalling + * a capacity change; update the capacity. + */ + err = xenbus_scanf(XBT_NIL, info->xbdev->otherend, + "sectors", "%Lu", §ors); + if (XENBUS_EXIST_ERR(err)) + return; + printk(KERN_INFO "Setting capacity to %Lu\n", + sectors); + set_capacity(info->gd, sectors); + revalidate_disk(info->gd); + + /* fall through */ + case BLKIF_STATE_SUSPENDED: return; + default: + break; + } + dev_dbg(&info->xbdev->dev, "%s:%s.\n", __func__, info->xbdev->otherend); @@ -920,10 +1070,26 @@ static void blkfront_connect(struct blkfront_info *info) } err = xenbus_gather(XBT_NIL, info->xbdev->otherend, - "feature-barrier", "%lu", &info->feature_barrier, + "feature-barrier", "%lu", &barrier, NULL); + + /* + * If there's no "feature-barrier" defined, then it means + * we're dealing with a very old backend which writes + * synchronously; draining will do what needs to get done. + * + * If there are barriers, then we can do full queued writes + * with tagged barriers. + * + * If barriers are not supported, then there's no much we can + * do, so just set ordering to NONE. + */ if (err) - info->feature_barrier = 0; + info->feature_barrier = QUEUE_ORDERED_DRAIN; + else if (barrier) + info->feature_barrier = QUEUE_ORDERED_TAG; + else + info->feature_barrier = QUEUE_ORDERED_NONE; err = xlvbd_alloc_gendisk(sectors, info, binfo, sector_size); if (err) { @@ -946,52 +1112,14 @@ static void blkfront_connect(struct blkfront_info *info) } /** - * Handle the change of state of the backend to Closing. We must delete our - * device-layer structures now, to ensure that writes are flushed through to - * the backend. Once is this done, we can switch to Closed in - * acknowledgement. - */ -static void blkfront_closing(struct xenbus_device *dev) -{ - struct blkfront_info *info = dev_get_drvdata(&dev->dev); - unsigned long flags; - - dev_dbg(&dev->dev, "blkfront_closing: %s removed\n", dev->nodename); - - if (info->rq == NULL) - goto out; - - spin_lock_irqsave(&blkif_io_lock, flags); - - /* No more blkif_request(). */ - blk_stop_queue(info->rq); - - /* No more gnttab callback work. */ - gnttab_cancel_free_callback(&info->callback); - spin_unlock_irqrestore(&blkif_io_lock, flags); - - /* Flush gnttab callback work. Must be done with no locks held. */ - flush_scheduled_work(); - - blk_cleanup_queue(info->rq); - info->rq = NULL; - - del_gendisk(info->gd); - - out: - xenbus_frontend_closed(dev); -} - -/** * Callback received when the backend's state changes. */ -static void backend_changed(struct xenbus_device *dev, +static void blkback_changed(struct xenbus_device *dev, enum xenbus_state backend_state) { struct blkfront_info *info = dev_get_drvdata(&dev->dev); - struct block_device *bd; - dev_dbg(&dev->dev, "blkfront:backend_changed.\n"); + dev_dbg(&dev->dev, "blkfront:blkback_changed to state %d.\n", backend_state); switch (backend_state) { case XenbusStateInitialising: @@ -1006,35 +1134,56 @@ static void backend_changed(struct xenbus_device *dev, break; case XenbusStateClosing: - if (info->gd == NULL) { - xenbus_frontend_closed(dev); - break; - } - bd = bdget_disk(info->gd, 0); - if (bd == NULL) - xenbus_dev_fatal(dev, -ENODEV, "bdget failed"); - - mutex_lock(&bd->bd_mutex); - if (info->users > 0) - xenbus_dev_error(dev, -EBUSY, - "Device in use; refusing to close"); - else - blkfront_closing(dev); - mutex_unlock(&bd->bd_mutex); - bdput(bd); + blkfront_closing(info); break; } } -static int blkfront_remove(struct xenbus_device *dev) +static int blkfront_remove(struct xenbus_device *xbdev) { - struct blkfront_info *info = dev_get_drvdata(&dev->dev); + struct blkfront_info *info = dev_get_drvdata(&xbdev->dev); + struct block_device *bdev = NULL; + struct gendisk *disk; - dev_dbg(&dev->dev, "blkfront_remove: %s removed\n", dev->nodename); + dev_dbg(&xbdev->dev, "%s removed", xbdev->nodename); blkif_free(info, 0); - kfree(info); + mutex_lock(&info->mutex); + + disk = info->gd; + if (disk) + bdev = bdget_disk(disk, 0); + + info->xbdev = NULL; + mutex_unlock(&info->mutex); + + if (!bdev) { + kfree(info); + return 0; + } + + /* + * The xbdev was removed before we reached the Closed + * state. See if it's safe to remove the disk. If the bdev + * isn't closed yet, we let release take care of it. + */ + + mutex_lock(&bdev->bd_mutex); + info = disk->private_data; + + dev_warn(disk_to_dev(disk), + "%s was hot-unplugged, %d stale handles\n", + xbdev->nodename, bdev->bd_openers); + + if (info && !bdev->bd_openers) { + xlvbd_release_gendisk(info); + disk->private_data = NULL; + kfree(info); + } + + mutex_unlock(&bdev->bd_mutex); + bdput(bdev); return 0; } @@ -1043,30 +1192,78 @@ static int blkfront_is_ready(struct xenbus_device *dev) { struct blkfront_info *info = dev_get_drvdata(&dev->dev); - return info->is_ready; + return info->is_ready && info->xbdev; } static int blkif_open(struct block_device *bdev, fmode_t mode) { - struct blkfront_info *info = bdev->bd_disk->private_data; - info->users++; - return 0; + struct gendisk *disk = bdev->bd_disk; + struct blkfront_info *info; + int err = 0; + + lock_kernel(); + + info = disk->private_data; + if (!info) { + /* xbdev gone */ + err = -ERESTARTSYS; + goto out; + } + + mutex_lock(&info->mutex); + + if (!info->gd) + /* xbdev is closed */ + err = -ERESTARTSYS; + + mutex_unlock(&info->mutex); + +out: + unlock_kernel(); + return err; } static int blkif_release(struct gendisk *disk, fmode_t mode) { struct blkfront_info *info = disk->private_data; - info->users--; - if (info->users == 0) { - /* Check whether we have been instructed to close. We will - have ignored this request initially, as the device was - still mounted. */ - struct xenbus_device *dev = info->xbdev; - enum xenbus_state state = xenbus_read_driver_state(dev->otherend); - - if (state == XenbusStateClosing && info->is_ready) - blkfront_closing(dev); + struct block_device *bdev; + struct xenbus_device *xbdev; + + lock_kernel(); + + bdev = bdget_disk(disk, 0); + bdput(bdev); + + if (bdev->bd_openers) + goto out; + + /* + * Check if we have been instructed to close. We will have + * deferred this request, because the bdev was still open. + */ + + mutex_lock(&info->mutex); + xbdev = info->xbdev; + + if (xbdev && xbdev->state == XenbusStateClosing) { + /* pending switch to state closed */ + dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n"); + xlvbd_release_gendisk(info); + xenbus_frontend_closed(info->xbdev); + } + + mutex_unlock(&info->mutex); + + if (!xbdev) { + /* sudden device removal */ + dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n"); + xlvbd_release_gendisk(info); + disk->private_data = NULL; + kfree(info); } + +out: + unlock_kernel(); return 0; } @@ -1076,7 +1273,7 @@ static const struct block_device_operations xlvbd_block_fops = .open = blkif_open, .release = blkif_release, .getgeo = blkif_getgeo, - .locked_ioctl = blkif_ioctl, + .ioctl = blkif_ioctl, }; @@ -1092,7 +1289,7 @@ static struct xenbus_driver blkfront = { .probe = blkfront_probe, .remove = blkfront_remove, .resume = blkfront_resume, - .otherend_changed = backend_changed, + .otherend_changed = blkback_changed, .is_ready = blkfront_is_ready, }; diff --git a/drivers/block/xsysace.c b/drivers/block/xsysace.c index a7b83c0..b71888b 100644 --- a/drivers/block/xsysace.c +++ b/drivers/block/xsysace.c @@ -89,6 +89,7 @@ #include <linux/delay.h> #include <linux/slab.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <linux/ata.h> #include <linux/hdreg.h> #include <linux/platform_device.h> @@ -465,7 +466,7 @@ struct request *ace_get_next_request(struct request_queue * q) struct request *req; while ((req = blk_peek_request(q)) != NULL) { - if (blk_fs_request(req)) + if (req->cmd_type == REQ_TYPE_FS) break; blk_start_request(req); __blk_end_request_all(req, -EIO); @@ -901,11 +902,14 @@ static int ace_open(struct block_device *bdev, fmode_t mode) dev_dbg(ace->dev, "ace_open() users=%i\n", ace->users + 1); + lock_kernel(); spin_lock_irqsave(&ace->lock, flags); ace->users++; spin_unlock_irqrestore(&ace->lock, flags); check_disk_change(bdev); + unlock_kernel(); + return 0; } @@ -917,6 +921,7 @@ static int ace_release(struct gendisk *disk, fmode_t mode) dev_dbg(ace->dev, "ace_release() users=%i\n", ace->users - 1); + lock_kernel(); spin_lock_irqsave(&ace->lock, flags); ace->users--; if (ace->users == 0) { @@ -924,6 +929,7 @@ static int ace_release(struct gendisk *disk, fmode_t mode) ace_out(ace, ACE_CTRL, val & ~ACE_CTRL_LOCKREQ); } spin_unlock_irqrestore(&ace->lock, flags); + unlock_kernel(); return 0; } diff --git a/drivers/block/z2ram.c b/drivers/block/z2ram.c index 9114654..d75b2bb 100644 --- a/drivers/block/z2ram.c +++ b/drivers/block/z2ram.c @@ -33,6 +33,7 @@ #include <linux/module.h> #include <linux/blkdev.h> #include <linux/bitops.h> +#include <linux/smp_lock.h> #include <linux/slab.h> #include <asm/setup.h> @@ -153,6 +154,7 @@ static int z2_open(struct block_device *bdev, fmode_t mode) device = MINOR(bdev->bd_dev); + lock_kernel(); if ( current_device != -1 && current_device != device ) { rc = -EBUSY; @@ -294,20 +296,25 @@ static int z2_open(struct block_device *bdev, fmode_t mode) set_capacity(z2ram_gendisk, z2ram_size >> 9); } + unlock_kernel(); return 0; err_out_kfree: kfree(z2ram_map); err_out: + unlock_kernel(); return rc; } static int z2_release(struct gendisk *disk, fmode_t mode) { - if ( current_device == -1 ) - return 0; - + lock_kernel(); + if ( current_device == -1 ) { + unlock_kernel(); + return 0; + } + unlock_kernel(); /* * FIXME: unmap memory */ diff --git a/drivers/cdrom/cdrom.c b/drivers/cdrom/cdrom.c index e3749d0..af13c62 100644 --- a/drivers/cdrom/cdrom.c +++ b/drivers/cdrom/cdrom.c @@ -242,6 +242,8 @@ -------------------------------------------------------------------------*/ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #define REVISION "Revision: 3.20" #define VERSION "Id: cdrom.c 3.20 2003/12/17" @@ -314,11 +316,17 @@ static const char *mrw_format_status[] = { static const char *mrw_address_space[] = { "DMA", "GAA" }; #if (ERRLOGMASK!=CD_NOTHING) -#define cdinfo(type, fmt, args...) \ - if ((ERRLOGMASK & type) || debug==1 ) \ - printk(KERN_INFO "cdrom: " fmt, ## args) +#define cdinfo(type, fmt, args...) \ +do { \ + if ((ERRLOGMASK & type) || debug == 1) \ + pr_info(fmt, ##args); \ +} while (0) #else -#define cdinfo(type, fmt, args...) +#define cdinfo(type, fmt, args...) \ +do { \ + if (0 && (ERRLOGMASK & type) || debug == 1) \ + pr_info(fmt, ##args); \ +} while (0) #endif /* These are used to simplify getting data in from and back to user land */ @@ -395,7 +403,7 @@ int register_cdrom(struct cdrom_device_info *cdi) if (cdo->open == NULL || cdo->release == NULL) return -EINVAL; if (!banner_printed) { - printk(KERN_INFO "Uniform CD-ROM driver " REVISION "\n"); + pr_info("Uniform CD-ROM driver " REVISION "\n"); banner_printed = 1; cdrom_sysctl_register(); } @@ -546,7 +554,7 @@ static int cdrom_mrw_bgformat(struct cdrom_device_info *cdi, int cont) unsigned char buffer[12]; int ret; - printk(KERN_INFO "cdrom: %sstarting format\n", cont ? "Re" : ""); + pr_info("%sstarting format\n", cont ? "Re" : ""); /* * FmtData bit set (bit 4), format type is 1 @@ -576,7 +584,7 @@ static int cdrom_mrw_bgformat(struct cdrom_device_info *cdi, int cont) ret = cdi->ops->generic_packet(cdi, &cgc); if (ret) - printk(KERN_INFO "cdrom: bgformat failed\n"); + pr_info("bgformat failed\n"); return ret; } @@ -622,8 +630,7 @@ static int cdrom_mrw_exit(struct cdrom_device_info *cdi) ret = 0; if (di.mrw_status == CDM_MRW_BGFORMAT_ACTIVE) { - printk(KERN_INFO "cdrom: issuing MRW back ground " - "format suspend\n"); + pr_info("issuing MRW background format suspend\n"); ret = cdrom_mrw_bgformat_susp(cdi, 0); } @@ -658,7 +665,8 @@ static int cdrom_mrw_set_lba_space(struct cdrom_device_info *cdi, int space) if ((ret = cdrom_mode_select(cdi, &cgc))) return ret; - printk(KERN_INFO "cdrom: %s: mrw address space %s selected\n", cdi->name, mrw_address_space[space]); + pr_info("%s: mrw address space %s selected\n", + cdi->name, mrw_address_space[space]); return 0; } @@ -762,7 +770,7 @@ static int cdrom_mrw_open_write(struct cdrom_device_info *cdi) * always reset to DMA lba space on open */ if (cdrom_mrw_set_lba_space(cdi, MRW_LBA_DMA)) { - printk(KERN_ERR "cdrom: failed setting lba address space\n"); + pr_err("failed setting lba address space\n"); return 1; } @@ -781,8 +789,7 @@ static int cdrom_mrw_open_write(struct cdrom_device_info *cdi) * 3 - MRW formatting complete */ ret = 0; - printk(KERN_INFO "cdrom open: mrw_status '%s'\n", - mrw_format_status[di.mrw_status]); + pr_info("open: mrw_status '%s'\n", mrw_format_status[di.mrw_status]); if (!di.mrw_status) ret = 1; else if (di.mrw_status == CDM_MRW_BGFORMAT_INACTIVE && @@ -932,8 +939,7 @@ static void cdrom_dvd_rw_close_write(struct cdrom_device_info *cdi) return; } - printk(KERN_INFO "cdrom: %s: dirty DVD+RW media, \"finalizing\"\n", - cdi->name); + pr_info("%s: dirty DVD+RW media, \"finalizing\"\n", cdi->name); init_cdrom_command(&cgc, NULL, 0, CGC_DATA_NONE); cgc.cmd[0] = GPCMD_FLUSH_CACHE; @@ -2176,7 +2182,7 @@ retry: * frame dma, so drop to single frame dma if we need to */ if (cdi->cdda_method == CDDA_BPC_FULL && nframes > 1) { - printk("cdrom: dropping to single frame dma\n"); + pr_info("dropping to single frame dma\n"); cdi->cdda_method = CDDA_BPC_SINGLE; goto retry; } @@ -2189,7 +2195,7 @@ retry: if (cdi->last_sense != 0x04 && cdi->last_sense != 0x0b) return ret; - printk("cdrom: dropping to old style cdda (sense=%x)\n", cdi->last_sense); + pr_info("dropping to old style cdda (sense=%x)\n", cdi->last_sense); cdi->cdda_method = CDDA_OLD; return cdrom_read_cdda_old(cdi, ubuf, lba, nframes); } @@ -3401,7 +3407,7 @@ static int cdrom_print_info(const char *header, int val, char *info, "\t%d", CDROM_CAN(val) != 0); break; default: - printk(KERN_INFO "cdrom: invalid option%d\n", option); + pr_info("invalid option%d\n", option); return 1; } if (!ret) @@ -3491,7 +3497,7 @@ doit: mutex_unlock(&cdrom_mutex); return proc_dostring(ctl, write, buffer, lenp, ppos); done: - printk(KERN_INFO "cdrom: info buffer too small\n"); + pr_info("info buffer too small\n"); goto doit; } @@ -3665,7 +3671,7 @@ static int __init cdrom_init(void) static void __exit cdrom_exit(void) { - printk(KERN_INFO "Uniform CD-ROM driver unloaded\n"); + pr_info("Uniform CD-ROM driver unloaded\n"); cdrom_sysctl_unregister(); } diff --git a/drivers/cdrom/gdrom.c b/drivers/cdrom/gdrom.c index 03c71f7..261107d 100644 --- a/drivers/cdrom/gdrom.c +++ b/drivers/cdrom/gdrom.c @@ -19,6 +19,8 @@ * */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/init.h> #include <linux/module.h> #include <linux/fs.h> @@ -32,6 +34,7 @@ #include <linux/blkdev.h> #include <linux/interrupt.h> #include <linux/device.h> +#include <linux/smp_lock.h> #include <linux/wait.h> #include <linux/workqueue.h> #include <linux/platform_device.h> @@ -339,8 +342,7 @@ static int gdrom_get_last_session(struct cdrom_device_info *cd_info, tocuse = 0; err = gdrom_readtoc_cmd(gd.toc, 0); if (err) { - printk(KERN_INFO "GDROM: Could not get CD " - "table of contents\n"); + pr_info("Could not get CD table of contents\n"); return -ENXIO; } } @@ -357,8 +359,7 @@ static int gdrom_get_last_session(struct cdrom_device_info *cd_info, } while (track >= fentry); if ((track > 100) || (track < get_entry_track(gd.toc->first))) { - printk(KERN_INFO "GDROM: No data on the last " - "session of the CD\n"); + pr_info("No data on the last session of the CD\n"); gdrom_getsense(NULL); return -ENXIO; } @@ -451,14 +452,14 @@ static int gdrom_getsense(short *bufstring) goto cleanup_sense; insw(GDROM_DATA_REG, &sense, sense_command->buflen/2); if (sense[1] & 40) { - printk(KERN_INFO "GDROM: Drive not ready - command aborted\n"); + pr_info("Drive not ready - command aborted\n"); goto cleanup_sense; } sense_key = sense[1] & 0x0F; if (sense_key < ARRAY_SIZE(sense_texts)) - printk(KERN_INFO "GDROM: %s\n", sense_texts[sense_key].text); + pr_info("%s\n", sense_texts[sense_key].text); else - printk(KERN_ERR "GDROM: Unknown sense key: %d\n", sense_key); + pr_err("Unknown sense key: %d\n", sense_key); if (bufstring) /* return addional sense data */ memcpy(bufstring, &sense[4], 2); if (sense_key < 2) @@ -492,12 +493,18 @@ static struct cdrom_device_ops gdrom_ops = { static int gdrom_bdops_open(struct block_device *bdev, fmode_t mode) { - return cdrom_open(gd.cd_info, bdev, mode); + int ret; + lock_kernel(); + ret = cdrom_open(gd.cd_info, bdev, mode); + unlock_kernel(); + return ret; } static int gdrom_bdops_release(struct gendisk *disk, fmode_t mode) { + lock_kernel(); cdrom_release(gd.cd_info, mode); + unlock_kernel(); return 0; } @@ -509,7 +516,13 @@ static int gdrom_bdops_mediachanged(struct gendisk *disk) static int gdrom_bdops_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg) { - return cdrom_ioctl(gd.cd_info, bdev, mode, cmd, arg); + int ret; + + lock_kernel(); + ret = cdrom_ioctl(gd.cd_info, bdev, mode, cmd, arg); + unlock_kernel(); + + return ret; } static const struct block_device_operations gdrom_bdops = { @@ -517,7 +530,7 @@ static const struct block_device_operations gdrom_bdops = { .open = gdrom_bdops_open, .release = gdrom_bdops_release, .media_changed = gdrom_bdops_mediachanged, - .locked_ioctl = gdrom_bdops_ioctl, + .ioctl = gdrom_bdops_ioctl, }; static irqreturn_t gdrom_command_interrupt(int irq, void *dev_id) @@ -643,14 +656,13 @@ static void gdrom_request(struct request_queue *rq) struct request *req; while ((req = blk_fetch_request(rq)) != NULL) { - if (!blk_fs_request(req)) { - printk(KERN_DEBUG "GDROM: Non-fs request ignored\n"); + if (req->cmd_type != REQ_TYPE_FS) { + printk(KERN_DEBUG "gdrom: Non-fs request ignored\n"); __blk_end_request_all(req, -EIO); continue; } if (rq_data_dir(req) != READ) { - printk(KERN_NOTICE "GDROM: Read only device -"); - printk(" write request ignored\n"); + pr_notice("Read only device - write request ignored\n"); __blk_end_request_all(req, -EIO); continue; } @@ -685,7 +697,7 @@ static int __devinit gdrom_outputversion(void) firmw_ver = kstrndup(id->firmver, 16, GFP_KERNEL); if (!firmw_ver) goto free_manuf_name; - printk(KERN_INFO "GDROM: %s from %s with firmware %s\n", + pr_info("%s from %s with firmware %s\n", model_name, manuf_name, firmw_ver); err = 0; kfree(firmw_ver); @@ -757,7 +769,7 @@ static int __devinit probe_gdrom(struct platform_device *devptr) int err; /* Start the device */ if (gdrom_execute_diagnostic() != 1) { - printk(KERN_WARNING "GDROM: ATA Probe for GDROM failed.\n"); + pr_warning("ATA Probe for GDROM failed\n"); return -ENODEV; } /* Print out firmware ID */ @@ -767,7 +779,7 @@ static int __devinit probe_gdrom(struct platform_device *devptr) gdrom_major = register_blkdev(0, GDROM_DEV_NAME); if (gdrom_major <= 0) return gdrom_major; - printk(KERN_INFO "GDROM: Registered with major number %d\n", + pr_info("Registered with major number %d\n", gdrom_major); /* Specify basic properties of drive */ gd.cd_info = kzalloc(sizeof(struct cdrom_device_info), GFP_KERNEL); @@ -818,7 +830,7 @@ probe_fail_no_disk: unregister_blkdev(gdrom_major, GDROM_DEV_NAME); gdrom_major = 0; probe_fail_no_mem: - printk(KERN_WARNING "GDROM: Probe failed - error is 0x%X\n", err); + pr_warning("Probe failed - error is 0x%X\n", err); return err; } diff --git a/drivers/cdrom/viocd.c b/drivers/cdrom/viocd.c index 451cd70..56bf9f4 100644 --- a/drivers/cdrom/viocd.c +++ b/drivers/cdrom/viocd.c @@ -31,6 +31,8 @@ * the OS/400 partition. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/major.h> #include <linux/blkdev.h> #include <linux/cdrom.h> @@ -40,6 +42,7 @@ #include <linux/module.h> #include <linux/completion.h> #include <linux/proc_fs.h> +#include <linux/smp_lock.h> #include <linux/seq_file.h> #include <linux/scatterlist.h> @@ -53,9 +56,6 @@ #define VIOCD_VERS "1.06" -#define VIOCD_KERN_WARNING KERN_WARNING "viocd: " -#define VIOCD_KERN_INFO KERN_INFO "viocd: " - /* * Should probably make this a module parameter....sigh */ @@ -154,13 +154,21 @@ static const struct file_operations proc_viocd_operations = { static int viocd_blk_open(struct block_device *bdev, fmode_t mode) { struct disk_info *di = bdev->bd_disk->private_data; - return cdrom_open(&di->viocd_info, bdev, mode); + int ret; + + lock_kernel(); + ret = cdrom_open(&di->viocd_info, bdev, mode); + unlock_kernel(); + + return ret; } static int viocd_blk_release(struct gendisk *disk, fmode_t mode) { struct disk_info *di = disk->private_data; + lock_kernel(); cdrom_release(&di->viocd_info, mode); + unlock_kernel(); return 0; } @@ -168,7 +176,13 @@ static int viocd_blk_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, unsigned long arg) { struct disk_info *di = bdev->bd_disk->private_data; - return cdrom_ioctl(&di->viocd_info, bdev, mode, cmd, arg); + int ret; + + lock_kernel(); + ret = cdrom_ioctl(&di->viocd_info, bdev, mode, cmd, arg); + unlock_kernel(); + + return ret; } static int viocd_blk_media_changed(struct gendisk *disk) @@ -181,7 +195,7 @@ static const struct block_device_operations viocd_fops = { .owner = THIS_MODULE, .open = viocd_blk_open, .release = viocd_blk_release, - .locked_ioctl = viocd_blk_ioctl, + .ioctl = viocd_blk_ioctl, .media_changed = viocd_blk_media_changed, }; @@ -202,9 +216,8 @@ static int viocd_open(struct cdrom_device_info *cdi, int purpose) (u64)&we, VIOVERSION << 16, ((u64)device_no << 48), 0, 0, 0); if (hvrc != 0) { - printk(VIOCD_KERN_WARNING - "bad rc on HvCallEvent_signalLpEventFast %d\n", - (int)hvrc); + pr_warning("bad rc on HvCallEvent_signalLpEventFast %d\n", + (int)hvrc); return -EIO; } @@ -213,8 +226,8 @@ static int viocd_open(struct cdrom_device_info *cdi, int purpose) if (we.rc) { const struct vio_error_entry *err = vio_lookup_rc(viocd_err_table, we.sub_result); - printk(VIOCD_KERN_WARNING "bad rc %d:0x%04X on open: %s\n", - we.rc, we.sub_result, err->msg); + pr_warning("bad rc %d:0x%04X on open: %s\n", + we.rc, we.sub_result, err->msg); return -err->errno; } @@ -234,9 +247,8 @@ static void viocd_release(struct cdrom_device_info *cdi) viopath_targetinst(viopath_hostLp), 0, VIOVERSION << 16, ((u64)device_no << 48), 0, 0, 0); if (hvrc != 0) - printk(VIOCD_KERN_WARNING - "bad rc on HvCallEvent_signalLpEventFast %d\n", - (int)hvrc); + pr_warning("bad rc on HvCallEvent_signalLpEventFast %d\n", + (int)hvrc); } /* Send a read or write request to OS/400 */ @@ -262,13 +274,12 @@ static int send_request(struct request *req) sg_init_table(&sg, 1); if (blk_rq_map_sg(req->q, req, &sg) == 0) { - printk(VIOCD_KERN_WARNING - "error setting up scatter/gather list\n"); + pr_warning("error setting up scatter/gather list\n"); return -1; } if (dma_map_sg(diskinfo->dev, &sg, 1, direction) == 0) { - printk(VIOCD_KERN_WARNING "error allocating sg tce\n"); + pr_warning("error allocating sg tce\n"); return -1; } dmaaddr = sg_dma_address(&sg); @@ -284,7 +295,7 @@ static int send_request(struct request *req) ((u64)DEVICE_NR(diskinfo) << 48) | dmaaddr, (u64)blk_rq_pos(req) * 512, len, 0); if (hvrc != HvLpEvent_Rc_Good) { - printk(VIOCD_KERN_WARNING "hv error on op %d\n", (int)hvrc); + pr_warning("hv error on op %d\n", (int)hvrc); return -1; } @@ -298,11 +309,10 @@ static void do_viocd_request(struct request_queue *q) struct request *req; while ((rwreq == 0) && ((req = blk_fetch_request(q)) != NULL)) { - if (!blk_fs_request(req)) + if (req->cmd_type != REQ_TYPE_FS) __blk_end_request_all(req, -EIO); else if (send_request(req) < 0) { - printk(VIOCD_KERN_WARNING - "unable to send message to OS/400!"); + pr_warning("unable to send message to OS/400!\n"); __blk_end_request_all(req, -EIO); } else rwreq++; @@ -327,8 +337,8 @@ static int viocd_media_changed(struct cdrom_device_info *cdi, int disc_nr) (u64)&we, VIOVERSION << 16, ((u64)device_no << 48), 0, 0, 0); if (hvrc != 0) { - printk(VIOCD_KERN_WARNING "bad rc on HvCallEvent_signalLpEventFast %d\n", - (int)hvrc); + pr_warning("bad rc on HvCallEvent_signalLpEventFast %d\n", + (int)hvrc); return -EIO; } @@ -338,9 +348,8 @@ static int viocd_media_changed(struct cdrom_device_info *cdi, int disc_nr) if (we.rc) { const struct vio_error_entry *err = vio_lookup_rc(viocd_err_table, we.sub_result); - printk(VIOCD_KERN_WARNING - "bad rc %d:0x%04X on check_change: %s; Assuming no change\n", - we.rc, we.sub_result, err->msg); + pr_warning("bad rc %d:0x%04X on check_change: %s; Assuming no change\n", + we.rc, we.sub_result, err->msg); return 0; } @@ -367,8 +376,8 @@ static int viocd_lock_door(struct cdrom_device_info *cdi, int locking) (u64)&we, VIOVERSION << 16, (device_no << 48) | (flags << 32), 0, 0, 0); if (hvrc != 0) { - printk(VIOCD_KERN_WARNING "bad rc on HvCallEvent_signalLpEventFast %d\n", - (int)hvrc); + pr_warning("bad rc on HvCallEvent_signalLpEventFast %d\n", + (int)hvrc); return -EIO; } @@ -455,8 +464,7 @@ static void vio_handle_cd_event(struct HvLpEvent *event) return; /* First, we should NEVER get an int here...only acks */ if (hvlpevent_is_int(event)) { - printk(VIOCD_KERN_WARNING - "Yikes! got an int in viocd event handler!\n"); + pr_warning("Yikes! got an int in viocd event handler!\n"); if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); @@ -510,10 +518,9 @@ return_complete: const struct vio_error_entry *err = vio_lookup_rc(viocd_err_table, bevent->sub_result); - printk(VIOCD_KERN_WARNING "request %p failed " - "with rc %d:0x%04X: %s\n", - req, event->xRc, - bevent->sub_result, err->msg); + pr_warning("request %p failed with rc %d:0x%04X: %s\n", + req, event->xRc, + bevent->sub_result, err->msg); __blk_end_request_all(req, -EIO); } else __blk_end_request_all(req, 0); @@ -524,9 +531,8 @@ return_complete: break; default: - printk(VIOCD_KERN_WARNING - "message with invalid subtype %0x04X!\n", - event->xSubtype & VIOMINOR_SUBTYPE_MASK); + pr_warning("message with invalid subtype %0x04X!\n", + event->xSubtype & VIOMINOR_SUBTYPE_MASK); if (hvlpevent_need_ack(event)) { event->xRc = HvLpEvent_Rc_InvalidSubtype; HvCallEvent_ackLpEvent(event); @@ -593,23 +599,19 @@ static int viocd_probe(struct vio_dev *vdev, const struct vio_device_id *id) sprintf(c->name, VIOCD_DEVICE "%c", 'a' + deviceno); if (register_cdrom(c) != 0) { - printk(VIOCD_KERN_WARNING "Cannot register viocd CD-ROM %s!\n", - c->name); + pr_warning("Cannot register viocd CD-ROM %s!\n", c->name); goto out; } - printk(VIOCD_KERN_INFO "cd %s is iSeries resource %10.10s " - "type %4.4s, model %3.3s\n", - c->name, d->rsrcname, d->type, d->model); + pr_info("cd %s is iSeries resource %10.10s type %4.4s, model %3.3s\n", + c->name, d->rsrcname, d->type, d->model); q = blk_init_queue(do_viocd_request, &viocd_reqlock); if (q == NULL) { - printk(VIOCD_KERN_WARNING "Cannot allocate queue for %s!\n", - c->name); + pr_warning("Cannot allocate queue for %s!\n", c->name); goto out_unregister_cdrom; } gendisk = alloc_disk(1); if (gendisk == NULL) { - printk(VIOCD_KERN_WARNING "Cannot create gendisk for %s!\n", - c->name); + pr_warning("Cannot create gendisk for %s!\n", c->name); goto out_cleanup_queue; } gendisk->major = VIOCD_MAJOR; @@ -682,21 +684,19 @@ static int __init viocd_init(void) return -ENODEV; } - printk(VIOCD_KERN_INFO "vers " VIOCD_VERS ", hosting partition %d\n", - viopath_hostLp); + pr_info("vers " VIOCD_VERS ", hosting partition %d\n", viopath_hostLp); if (register_blkdev(VIOCD_MAJOR, VIOCD_DEVICE) != 0) { - printk(VIOCD_KERN_WARNING "Unable to get major %d for %s\n", - VIOCD_MAJOR, VIOCD_DEVICE); + pr_warning("Unable to get major %d for %s\n", + VIOCD_MAJOR, VIOCD_DEVICE); return -EIO; } ret = viopath_open(viopath_hostLp, viomajorsubtype_cdio, MAX_CD_REQ + 2); if (ret) { - printk(VIOCD_KERN_WARNING - "error opening path to host partition %d\n", - viopath_hostLp); + pr_warning("error opening path to host partition %d\n", + viopath_hostLp); goto out_unregister; } diff --git a/drivers/ide/ide-atapi.c b/drivers/ide/ide-atapi.c index f9daffd..e88a2cf 100644 --- a/drivers/ide/ide-atapi.c +++ b/drivers/ide/ide-atapi.c @@ -190,7 +190,7 @@ void ide_prep_sense(ide_drive_t *drive, struct request *rq) BUG_ON(sense_len > sizeof(*sense)); - if (blk_sense_request(rq) || drive->sense_rq_armed) + if (rq->cmd_type == REQ_TYPE_SENSE || drive->sense_rq_armed) return; memset(sense, 0, sizeof(*sense)); @@ -307,13 +307,16 @@ EXPORT_SYMBOL_GPL(ide_cd_expiry); int ide_cd_get_xferlen(struct request *rq) { - if (blk_fs_request(rq)) + switch (rq->cmd_type) { + case REQ_TYPE_FS: return 32768; - else if (blk_sense_request(rq) || blk_pc_request(rq) || - rq->cmd_type == REQ_TYPE_ATA_PC) + case REQ_TYPE_SENSE: + case REQ_TYPE_BLOCK_PC: + case REQ_TYPE_ATA_PC: return blk_rq_bytes(rq); - else + default: return 0; + } } EXPORT_SYMBOL_GPL(ide_cd_get_xferlen); @@ -474,12 +477,12 @@ static ide_startstop_t ide_pc_intr(ide_drive_t *drive) if (uptodate == 0) drive->failed_pc = NULL; - if (blk_special_request(rq)) { + if (rq->cmd_type == REQ_TYPE_SPECIAL) { rq->errors = 0; error = 0; } else { - if (blk_fs_request(rq) == 0 && uptodate <= 0) { + if (rq->cmd_type != REQ_TYPE_FS && uptodate <= 0) { if (rq->errors == 0) rq->errors = -EIO; } diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c index 2de76cc..31fc769 100644 --- a/drivers/ide/ide-cd.c +++ b/drivers/ide/ide-cd.c @@ -31,6 +31,7 @@ #include <linux/delay.h> #include <linux/timer.h> #include <linux/seq_file.h> +#include <linux/smp_lock.h> #include <linux/slab.h> #include <linux/interrupt.h> #include <linux/errno.h> @@ -176,7 +177,7 @@ static void cdrom_analyze_sense_data(ide_drive_t *drive, if (!sense->valid) break; if (failed_command == NULL || - !blk_fs_request(failed_command)) + failed_command->cmd_type != REQ_TYPE_FS) break; sector = (sense->information[0] << 24) | (sense->information[1] << 16) | @@ -292,7 +293,7 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat) "stat 0x%x", rq->cmd[0], rq->cmd_type, err, stat); - if (blk_sense_request(rq)) { + if (rq->cmd_type == REQ_TYPE_SENSE) { /* * We got an error trying to get sense info from the drive * (probably while trying to recover from a former error). @@ -303,7 +304,7 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat) } /* if we have an error, pass CHECK_CONDITION as the SCSI status byte */ - if (blk_pc_request(rq) && !rq->errors) + if (rq->cmd_type == REQ_TYPE_BLOCK_PC && !rq->errors) rq->errors = SAM_STAT_CHECK_CONDITION; if (blk_noretry_request(rq)) @@ -311,13 +312,14 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat) switch (sense_key) { case NOT_READY: - if (blk_fs_request(rq) && rq_data_dir(rq) == WRITE) { + if (rq->cmd_type == REQ_TYPE_FS && rq_data_dir(rq) == WRITE) { if (ide_cd_breathe(drive, rq)) return 1; } else { cdrom_saw_media_change(drive); - if (blk_fs_request(rq) && !blk_rq_quiet(rq)) + if (rq->cmd_type == REQ_TYPE_FS && + !(rq->cmd_flags & REQ_QUIET)) printk(KERN_ERR PFX "%s: tray open\n", drive->name); } @@ -326,7 +328,7 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat) case UNIT_ATTENTION: cdrom_saw_media_change(drive); - if (blk_fs_request(rq) == 0) + if (rq->cmd_type != REQ_TYPE_FS) return 0; /* @@ -352,7 +354,7 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat) * No point in retrying after an illegal request or data * protect error. */ - if (!blk_rq_quiet(rq)) + if (!(rq->cmd_flags & REQ_QUIET)) ide_dump_status(drive, "command error", stat); do_end_request = 1; break; @@ -361,20 +363,20 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat) * No point in re-trying a zillion times on a bad sector. * If we got here the error is not correctable. */ - if (!blk_rq_quiet(rq)) + if (!(rq->cmd_flags & REQ_QUIET)) ide_dump_status(drive, "media error " "(bad sector)", stat); do_end_request = 1; break; case BLANK_CHECK: /* disk appears blank? */ - if (!blk_rq_quiet(rq)) + if (!(rq->cmd_flags & REQ_QUIET)) ide_dump_status(drive, "media error (blank)", stat); do_end_request = 1; break; default: - if (blk_fs_request(rq) == 0) + if (rq->cmd_type != REQ_TYPE_FS) break; if (err & ~ATA_ABORTED) { /* go to the default handler for other errors */ @@ -385,7 +387,7 @@ static int cdrom_decode_status(ide_drive_t *drive, u8 stat) do_end_request = 1; } - if (blk_fs_request(rq) == 0) { + if (rq->cmd_type != REQ_TYPE_FS) { rq->cmd_flags |= REQ_FAILED; do_end_request = 1; } @@ -532,7 +534,7 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive) ide_expiry_t *expiry = NULL; int dma_error = 0, dma, thislen, uptodate = 0; int write = (rq_data_dir(rq) == WRITE) ? 1 : 0, rc = 0; - int sense = blk_sense_request(rq); + int sense = (rq->cmd_type == REQ_TYPE_SENSE); unsigned int timeout; u16 len; u8 ireason, stat; @@ -575,7 +577,7 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive) ide_read_bcount_and_ireason(drive, &len, &ireason); - thislen = blk_fs_request(rq) ? len : cmd->nleft; + thislen = (rq->cmd_type == REQ_TYPE_FS) ? len : cmd->nleft; if (thislen > len) thislen = len; @@ -584,7 +586,7 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive) /* If DRQ is clear, the command has completed. */ if ((stat & ATA_DRQ) == 0) { - if (blk_fs_request(rq)) { + if (rq->cmd_type == REQ_TYPE_FS) { /* * If we're not done reading/writing, complain. * Otherwise, complete the command normally. @@ -598,7 +600,7 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive) rq->cmd_flags |= REQ_FAILED; uptodate = 0; } - } else if (!blk_pc_request(rq)) { + } else if (rq->cmd_type != REQ_TYPE_BLOCK_PC) { ide_cd_request_sense_fixup(drive, cmd); uptodate = cmd->nleft ? 0 : 1; @@ -647,7 +649,7 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive) /* pad, if necessary */ if (len > 0) { - if (blk_fs_request(rq) == 0 || write == 0) + if (rq->cmd_type != REQ_TYPE_FS || write == 0) ide_pad_transfer(drive, write, len); else { printk(KERN_ERR PFX "%s: confused, missing data\n", @@ -656,11 +658,11 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive) } } - if (blk_pc_request(rq)) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { timeout = rq->timeout; } else { timeout = ATAPI_WAIT_PC; - if (!blk_fs_request(rq)) + if (rq->cmd_type != REQ_TYPE_FS) expiry = ide_cd_expiry; } @@ -669,7 +671,7 @@ static ide_startstop_t cdrom_newpc_intr(ide_drive_t *drive) return ide_started; out_end: - if (blk_pc_request(rq) && rc == 0) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC && rc == 0) { rq->resid_len = 0; blk_end_request_all(rq, 0); hwif->rq = NULL; @@ -677,7 +679,7 @@ out_end: if (sense && uptodate) ide_cd_complete_failed_rq(drive, rq); - if (blk_fs_request(rq)) { + if (rq->cmd_type == REQ_TYPE_FS) { if (cmd->nleft == 0) uptodate = 1; } else { @@ -690,7 +692,7 @@ out_end: return ide_stopped; /* make sure it's fully ended */ - if (blk_fs_request(rq) == 0) { + if (rq->cmd_type != REQ_TYPE_FS) { rq->resid_len -= cmd->nbytes - cmd->nleft; if (uptodate == 0 && (cmd->tf_flags & IDE_TFLAG_WRITE)) rq->resid_len += cmd->last_xfer_len; @@ -750,7 +752,7 @@ static void cdrom_do_block_pc(ide_drive_t *drive, struct request *rq) ide_debug_log(IDE_DBG_PC, "rq->cmd[0]: 0x%x, rq->cmd_type: 0x%x", rq->cmd[0], rq->cmd_type); - if (blk_pc_request(rq)) + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) rq->cmd_flags |= REQ_QUIET; else rq->cmd_flags &= ~REQ_FAILED; @@ -791,21 +793,26 @@ static ide_startstop_t ide_cd_do_request(ide_drive_t *drive, struct request *rq, if (drive->debug_mask & IDE_DBG_RQ) blk_dump_rq_flags(rq, "ide_cd_do_request"); - if (blk_fs_request(rq)) { + switch (rq->cmd_type) { + case REQ_TYPE_FS: if (cdrom_start_rw(drive, rq) == ide_stopped) goto out_end; - } else if (blk_sense_request(rq) || blk_pc_request(rq) || - rq->cmd_type == REQ_TYPE_ATA_PC) { + break; + case REQ_TYPE_SENSE: + case REQ_TYPE_BLOCK_PC: + case REQ_TYPE_ATA_PC: if (!rq->timeout) rq->timeout = ATAPI_WAIT_PC; cdrom_do_block_pc(drive, rq); - } else if (blk_special_request(rq)) { + break; + case REQ_TYPE_SPECIAL: /* right now this can only be a reset... */ uptodate = 1; goto out_end; - } else + default: BUG(); + } /* prepare sense request for this command */ ide_prep_sense(drive, rq); @@ -817,7 +824,7 @@ static ide_startstop_t ide_cd_do_request(ide_drive_t *drive, struct request *rq, cmd.rq = rq; - if (blk_fs_request(rq) || blk_rq_bytes(rq)) { + if (rq->cmd_type == REQ_TYPE_FS || blk_rq_bytes(rq)) { ide_init_sg_cmd(&cmd, blk_rq_bytes(rq)); ide_map_sg(drive, &cmd); } @@ -1373,9 +1380,9 @@ static int ide_cdrom_prep_pc(struct request *rq) static int ide_cdrom_prep_fn(struct request_queue *q, struct request *rq) { - if (blk_fs_request(rq)) + if (rq->cmd_type == REQ_TYPE_FS) return ide_cdrom_prep_fs(q, rq); - else if (blk_pc_request(rq)) + else if (rq->cmd_type == REQ_TYPE_BLOCK_PC) return ide_cdrom_prep_pc(rq); return 0; @@ -1592,17 +1599,19 @@ static struct ide_driver ide_cdrom_driver = { static int idecd_open(struct block_device *bdev, fmode_t mode) { - struct cdrom_info *info = ide_cd_get(bdev->bd_disk); - int rc = -ENOMEM; + struct cdrom_info *info; + int rc = -ENXIO; + lock_kernel(); + info = ide_cd_get(bdev->bd_disk); if (!info) - return -ENXIO; + goto out; rc = cdrom_open(&info->devinfo, bdev, mode); - if (rc < 0) ide_cd_put(info); - +out: + unlock_kernel(); return rc; } @@ -1610,9 +1619,11 @@ static int idecd_release(struct gendisk *disk, fmode_t mode) { struct cdrom_info *info = ide_drv_g(disk, cdrom_info); + lock_kernel(); cdrom_release(&info->devinfo, mode); ide_cd_put(info); + unlock_kernel(); return 0; } @@ -1656,7 +1667,7 @@ static int idecd_get_spindown(struct cdrom_device_info *cdi, unsigned long arg) return 0; } -static int idecd_ioctl(struct block_device *bdev, fmode_t mode, +static int idecd_locked_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, unsigned long arg) { struct cdrom_info *info = ide_drv_g(bdev->bd_disk, cdrom_info); @@ -1678,6 +1689,19 @@ static int idecd_ioctl(struct block_device *bdev, fmode_t mode, return err; } +static int idecd_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long arg) +{ + int ret; + + lock_kernel(); + ret = idecd_locked_ioctl(bdev, mode, cmd, arg); + unlock_kernel(); + + return ret; +} + + static int idecd_media_changed(struct gendisk *disk) { struct cdrom_info *info = ide_drv_g(disk, cdrom_info); @@ -1698,7 +1722,7 @@ static const struct block_device_operations idecd_ops = { .owner = THIS_MODULE, .open = idecd_open, .release = idecd_release, - .locked_ioctl = idecd_ioctl, + .ioctl = idecd_ioctl, .media_changed = idecd_media_changed, .revalidate_disk = idecd_revalidate_disk }; diff --git a/drivers/ide/ide-cd_ioctl.c b/drivers/ide/ide-cd_ioctl.c index 02712bf..766b3de 100644 --- a/drivers/ide/ide-cd_ioctl.c +++ b/drivers/ide/ide-cd_ioctl.c @@ -454,7 +454,7 @@ int ide_cdrom_packet(struct cdrom_device_info *cdi, touch it at all. */ if (cgc->data_direction == CGC_DATA_WRITE) - flags |= REQ_RW; + flags |= REQ_WRITE; if (cgc->sense) memset(cgc->sense, 0, sizeof(struct request_sense)); diff --git a/drivers/ide/ide-disk.c b/drivers/ide/ide-disk.c index 33d6503..7433e07 100644 --- a/drivers/ide/ide-disk.c +++ b/drivers/ide/ide-disk.c @@ -184,7 +184,7 @@ static ide_startstop_t ide_do_rw_disk(ide_drive_t *drive, struct request *rq, ide_hwif_t *hwif = drive->hwif; BUG_ON(drive->dev_flags & IDE_DFLAG_BLOCKED); - BUG_ON(!blk_fs_request(rq)); + BUG_ON(rq->cmd_type != REQ_TYPE_FS); ledtrig_ide_activity(); @@ -427,10 +427,15 @@ static void ide_disk_unlock_native_capacity(ide_drive_t *drive) drive->dev_flags |= IDE_DFLAG_NOHPA; /* disable HPA on resume */ } -static void idedisk_prepare_flush(struct request_queue *q, struct request *rq) +static int idedisk_prep_fn(struct request_queue *q, struct request *rq) { ide_drive_t *drive = q->queuedata; - struct ide_cmd *cmd = kmalloc(sizeof(*cmd), GFP_ATOMIC); + struct ide_cmd *cmd; + + if (!(rq->cmd_flags & REQ_FLUSH)) + return BLKPREP_OK; + + cmd = kmalloc(sizeof(*cmd), GFP_ATOMIC); /* FIXME: map struct ide_taskfile on rq->cmd[] */ BUG_ON(cmd == NULL); @@ -448,6 +453,8 @@ static void idedisk_prepare_flush(struct request_queue *q, struct request *rq) rq->cmd_type = REQ_TYPE_ATA_TASKFILE; rq->special = cmd; cmd->rq = rq; + + return BLKPREP_OK; } ide_devset_get(multcount, mult_count); @@ -513,7 +520,6 @@ static void update_ordered(ide_drive_t *drive) { u16 *id = drive->id; unsigned ordered = QUEUE_ORDERED_NONE; - prepare_flush_fn *prep_fn = NULL; if (drive->dev_flags & IDE_DFLAG_WCACHE) { unsigned long long capacity; @@ -538,12 +544,12 @@ static void update_ordered(ide_drive_t *drive) if (barrier) { ordered = QUEUE_ORDERED_DRAIN_FLUSH; - prep_fn = idedisk_prepare_flush; + blk_queue_prep_rq(drive->queue, idedisk_prep_fn); } } else ordered = QUEUE_ORDERED_DRAIN; - blk_queue_ordered(drive->queue, ordered, prep_fn); + blk_queue_ordered(drive->queue, ordered); } ide_devset_get_flag(wcache, IDE_DFLAG_WCACHE); diff --git a/drivers/ide/ide-disk_ioctl.c b/drivers/ide/ide-disk_ioctl.c index 7b783dd..ec94c66 100644 --- a/drivers/ide/ide-disk_ioctl.c +++ b/drivers/ide/ide-disk_ioctl.c @@ -1,6 +1,7 @@ #include <linux/kernel.h> #include <linux/ide.h> #include <linux/hdreg.h> +#include <linux/smp_lock.h> #include "ide-disk.h" @@ -18,9 +19,13 @@ int ide_disk_ioctl(ide_drive_t *drive, struct block_device *bdev, fmode_t mode, { int err; + lock_kernel(); err = ide_setting_ioctl(drive, bdev, cmd, arg, ide_disk_ioctl_settings); if (err != -EOPNOTSUPP) - return err; + goto out; - return generic_ide_ioctl(drive, bdev, cmd, arg); + err = generic_ide_ioctl(drive, bdev, cmd, arg); +out: + unlock_kernel(); + return err; } diff --git a/drivers/ide/ide-eh.c b/drivers/ide/ide-eh.c index e9abf2c..c0aa93f 100644 --- a/drivers/ide/ide-eh.c +++ b/drivers/ide/ide-eh.c @@ -122,7 +122,7 @@ ide_startstop_t ide_error(ide_drive_t *drive, const char *msg, u8 stat) return ide_stopped; /* retry only "normal" I/O: */ - if (!blk_fs_request(rq)) { + if (rq->cmd_type != REQ_TYPE_FS) { if (rq->cmd_type == REQ_TYPE_ATA_TASKFILE) { struct ide_cmd *cmd = rq->special; @@ -146,7 +146,8 @@ static inline void ide_complete_drive_reset(ide_drive_t *drive, int err) { struct request *rq = drive->hwif->rq; - if (rq && blk_special_request(rq) && rq->cmd[0] == REQ_DRIVE_RESET) { + if (rq && rq->cmd_type == REQ_TYPE_SPECIAL && + rq->cmd[0] == REQ_DRIVE_RESET) { if (err <= 0 && rq->errors == 0) rq->errors = -EIO; ide_complete_rq(drive, err ? err : 0, blk_rq_bytes(rq)); diff --git a/drivers/ide/ide-floppy.c b/drivers/ide/ide-floppy.c index 4713bdc..5406b6e 100644 --- a/drivers/ide/ide-floppy.c +++ b/drivers/ide/ide-floppy.c @@ -73,7 +73,7 @@ static int ide_floppy_callback(ide_drive_t *drive, int dsc) drive->failed_pc = NULL; if (pc->c[0] == GPCMD_READ_10 || pc->c[0] == GPCMD_WRITE_10 || - (rq && blk_pc_request(rq))) + (rq && rq->cmd_type == REQ_TYPE_BLOCK_PC)) uptodate = 1; /* FIXME */ else if (pc->c[0] == GPCMD_REQUEST_SENSE) { @@ -98,7 +98,7 @@ static int ide_floppy_callback(ide_drive_t *drive, int dsc) "Aborting request!\n"); } - if (blk_special_request(rq)) + if (rq->cmd_type == REQ_TYPE_SPECIAL) rq->errors = uptodate ? 0 : IDE_DRV_ERROR_GENERAL; return uptodate; @@ -207,7 +207,7 @@ static void idefloppy_create_rw_cmd(ide_drive_t *drive, memcpy(rq->cmd, pc->c, 12); pc->rq = rq; - if (rq->cmd_flags & REQ_RW) + if (rq->cmd_flags & REQ_WRITE) pc->flags |= PC_FLAG_WRITING; pc->flags |= PC_FLAG_DMA_OK; @@ -247,14 +247,16 @@ static ide_startstop_t ide_floppy_do_request(ide_drive_t *drive, } else printk(KERN_ERR PFX "%s: I/O error\n", drive->name); - if (blk_special_request(rq)) { + if (rq->cmd_type == REQ_TYPE_SPECIAL) { rq->errors = 0; ide_complete_rq(drive, 0, blk_rq_bytes(rq)); return ide_stopped; } else goto out_end; } - if (blk_fs_request(rq)) { + + switch (rq->cmd_type) { + case REQ_TYPE_FS: if (((long)blk_rq_pos(rq) % floppy->bs_factor) || (blk_rq_sectors(rq) % floppy->bs_factor)) { printk(KERN_ERR PFX "%s: unsupported r/w rq size\n", @@ -263,13 +265,18 @@ static ide_startstop_t ide_floppy_do_request(ide_drive_t *drive, } pc = &floppy->queued_pc; idefloppy_create_rw_cmd(drive, pc, rq, (unsigned long)block); - } else if (blk_special_request(rq) || blk_sense_request(rq)) { + break; + case REQ_TYPE_SPECIAL: + case REQ_TYPE_SENSE: pc = (struct ide_atapi_pc *)rq->special; - } else if (blk_pc_request(rq)) { + break; + case REQ_TYPE_BLOCK_PC: pc = &floppy->queued_pc; idefloppy_blockpc_cmd(floppy, pc, rq); - } else + break; + default: BUG(); + } ide_prep_sense(drive, rq); @@ -280,7 +287,7 @@ static ide_startstop_t ide_floppy_do_request(ide_drive_t *drive, cmd.rq = rq; - if (blk_fs_request(rq) || blk_rq_bytes(rq)) { + if (rq->cmd_type == REQ_TYPE_FS || blk_rq_bytes(rq)) { ide_init_sg_cmd(&cmd, blk_rq_bytes(rq)); ide_map_sg(drive, &cmd); } @@ -290,7 +297,7 @@ static ide_startstop_t ide_floppy_do_request(ide_drive_t *drive, return ide_floppy_issue_pc(drive, &cmd, pc); out_end: drive->failed_pc = NULL; - if (blk_fs_request(rq) == 0 && rq->errors == 0) + if (rq->cmd_type != REQ_TYPE_FS && rq->errors == 0) rq->errors = -EIO; ide_complete_rq(drive, -EIO, blk_rq_bytes(rq)); return ide_stopped; diff --git a/drivers/ide/ide-floppy_ioctl.c b/drivers/ide/ide-floppy_ioctl.c index 9c22882..fd3d05a 100644 --- a/drivers/ide/ide-floppy_ioctl.c +++ b/drivers/ide/ide-floppy_ioctl.c @@ -5,6 +5,7 @@ #include <linux/kernel.h> #include <linux/ide.h> #include <linux/cdrom.h> +#include <linux/smp_lock.h> #include <asm/unaligned.h> @@ -275,12 +276,15 @@ int ide_floppy_ioctl(ide_drive_t *drive, struct block_device *bdev, void __user *argp = (void __user *)arg; int err; - if (cmd == CDROMEJECT || cmd == CDROM_LOCKDOOR) - return ide_floppy_lockdoor(drive, &pc, arg, cmd); + lock_kernel(); + if (cmd == CDROMEJECT || cmd == CDROM_LOCKDOOR) { + err = ide_floppy_lockdoor(drive, &pc, arg, cmd); + goto out; + } err = ide_floppy_format_ioctl(drive, &pc, mode, cmd, argp); if (err != -ENOTTY) - return err; + goto out; /* * skip SCSI_IOCTL_SEND_COMMAND (deprecated) @@ -293,5 +297,7 @@ int ide_floppy_ioctl(ide_drive_t *drive, struct block_device *bdev, if (err == -ENOTTY) err = generic_ide_ioctl(drive, bdev, cmd, arg); +out: + unlock_kernel(); return err; } diff --git a/drivers/ide/ide-gd.c b/drivers/ide/ide-gd.c index 7939953..70aeeb1 100644 --- a/drivers/ide/ide-gd.c +++ b/drivers/ide/ide-gd.c @@ -1,3 +1,4 @@ +#include <linux/smp_lock.h> #include <linux/module.h> #include <linux/types.h> #include <linux/string.h> @@ -237,6 +238,18 @@ out_put_idkp: return ret; } +static int ide_gd_unlocked_open(struct block_device *bdev, fmode_t mode) +{ + int ret; + + lock_kernel(); + ret = ide_gd_open(bdev, mode); + unlock_kernel(); + + return ret; +} + + static int ide_gd_release(struct gendisk *disk, fmode_t mode) { struct ide_disk_obj *idkp = ide_drv_g(disk, ide_disk_obj); @@ -244,6 +257,7 @@ static int ide_gd_release(struct gendisk *disk, fmode_t mode) ide_debug_log(IDE_DBG_FUNC, "enter"); + lock_kernel(); if (idkp->openers == 1) drive->disk_ops->flush(drive); @@ -255,6 +269,7 @@ static int ide_gd_release(struct gendisk *disk, fmode_t mode) idkp->openers--; ide_disk_put(idkp); + unlock_kernel(); return 0; } @@ -321,9 +336,9 @@ static int ide_gd_ioctl(struct block_device *bdev, fmode_t mode, static const struct block_device_operations ide_gd_ops = { .owner = THIS_MODULE, - .open = ide_gd_open, + .open = ide_gd_unlocked_open, .release = ide_gd_release, - .locked_ioctl = ide_gd_ioctl, + .ioctl = ide_gd_ioctl, .getgeo = ide_gd_getgeo, .media_changed = ide_gd_media_changed, .unlock_native_capacity = ide_gd_unlock_native_capacity, diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c index 172ac92..a381be8 100644 --- a/drivers/ide/ide-io.c +++ b/drivers/ide/ide-io.c @@ -135,7 +135,7 @@ EXPORT_SYMBOL(ide_complete_rq); void ide_kill_rq(ide_drive_t *drive, struct request *rq) { - u8 drv_req = blk_special_request(rq) && rq->rq_disk; + u8 drv_req = (rq->cmd_type == REQ_TYPE_SPECIAL) && rq->rq_disk; u8 media = drive->media; drive->failed_pc = NULL; @@ -145,7 +145,7 @@ void ide_kill_rq(ide_drive_t *drive, struct request *rq) } else { if (media == ide_tape) rq->errors = IDE_DRV_ERROR_GENERAL; - else if (blk_fs_request(rq) == 0 && rq->errors == 0) + else if (rq->cmd_type != REQ_TYPE_FS && rq->errors == 0) rq->errors = -EIO; } @@ -307,7 +307,7 @@ static ide_startstop_t start_request (ide_drive_t *drive, struct request *rq) { ide_startstop_t startstop; - BUG_ON(!blk_rq_started(rq)); + BUG_ON(!(rq->cmd_flags & REQ_STARTED)); #ifdef DEBUG printk("%s: start_request: current=0x%08lx\n", @@ -353,7 +353,7 @@ static ide_startstop_t start_request (ide_drive_t *drive, struct request *rq) pm->pm_step == IDE_PM_COMPLETED) ide_complete_pm_rq(drive, rq); return startstop; - } else if (!rq->rq_disk && blk_special_request(rq)) + } else if (!rq->rq_disk && rq->cmd_type == REQ_TYPE_SPECIAL) /* * TODO: Once all ULDs have been modified to * check for specific op codes rather than diff --git a/drivers/ide/ide-pm.c b/drivers/ide/ide-pm.c index 1c08311..9240609 100644 --- a/drivers/ide/ide-pm.c +++ b/drivers/ide/ide-pm.c @@ -191,10 +191,10 @@ void ide_complete_pm_rq(ide_drive_t *drive, struct request *rq) #ifdef DEBUG_PM printk("%s: completing PM request, %s\n", drive->name, - blk_pm_suspend_request(rq) ? "suspend" : "resume"); + (rq->cmd_type == REQ_TYPE_PM_SUSPEND) ? "suspend" : "resume"); #endif spin_lock_irqsave(q->queue_lock, flags); - if (blk_pm_suspend_request(rq)) + if (rq->cmd_type == REQ_TYPE_PM_SUSPEND) blk_stop_queue(q); else drive->dev_flags &= ~IDE_DFLAG_BLOCKED; @@ -210,11 +210,11 @@ void ide_check_pm_state(ide_drive_t *drive, struct request *rq) { struct request_pm_state *pm = rq->special; - if (blk_pm_suspend_request(rq) && + if (rq->cmd_type == REQ_TYPE_PM_SUSPEND && pm->pm_step == IDE_PM_START_SUSPEND) /* Mark drive blocked when starting the suspend sequence. */ drive->dev_flags |= IDE_DFLAG_BLOCKED; - else if (blk_pm_resume_request(rq) && + else if (rq->cmd_type == REQ_TYPE_PM_RESUME && pm->pm_step == IDE_PM_START_RESUME) { /* * The first thing we do on wakeup is to wait for BSY bit to diff --git a/drivers/ide/ide-tape.c b/drivers/ide/ide-tape.c index b072328..6d622cb 100644 --- a/drivers/ide/ide-tape.c +++ b/drivers/ide/ide-tape.c @@ -32,6 +32,7 @@ #include <linux/errno.h> #include <linux/genhd.h> #include <linux/seq_file.h> +#include <linux/smp_lock.h> #include <linux/slab.h> #include <linux/pci.h> #include <linux/ide.h> @@ -577,7 +578,8 @@ static ide_startstop_t idetape_do_request(ide_drive_t *drive, rq->cmd[0], (unsigned long long)blk_rq_pos(rq), blk_rq_sectors(rq)); - BUG_ON(!(blk_special_request(rq) || blk_sense_request(rq))); + BUG_ON(!(rq->cmd_type == REQ_TYPE_SPECIAL || + rq->cmd_type == REQ_TYPE_SENSE)); /* Retry a failed packet command */ if (drive->failed_pc && drive->pc->c[0] == REQUEST_SENSE) { @@ -1905,7 +1907,11 @@ static const struct file_operations idetape_fops = { static int idetape_open(struct block_device *bdev, fmode_t mode) { - struct ide_tape_obj *tape = ide_tape_get(bdev->bd_disk, false, 0); + struct ide_tape_obj *tape; + + lock_kernel(); + tape = ide_tape_get(bdev->bd_disk, false, 0); + unlock_kernel(); if (!tape) return -ENXIO; @@ -1917,7 +1923,10 @@ static int idetape_release(struct gendisk *disk, fmode_t mode) { struct ide_tape_obj *tape = ide_drv_g(disk, ide_tape_obj); + lock_kernel(); ide_tape_put(tape); + unlock_kernel(); + return 0; } @@ -1926,9 +1935,14 @@ static int idetape_ioctl(struct block_device *bdev, fmode_t mode, { struct ide_tape_obj *tape = ide_drv_g(bdev->bd_disk, ide_tape_obj); ide_drive_t *drive = tape->drive; - int err = generic_ide_ioctl(drive, bdev, cmd, arg); + int err; + + lock_kernel(); + err = generic_ide_ioctl(drive, bdev, cmd, arg); if (err == -EINVAL) err = idetape_blkdev_ioctl(drive, cmd, arg); + unlock_kernel(); + return err; } @@ -1936,7 +1950,7 @@ static const struct block_device_operations idetape_block_ops = { .owner = THIS_MODULE, .open = idetape_open, .release = idetape_release, - .locked_ioctl = idetape_ioctl, + .ioctl = idetape_ioctl, }; static int ide_tape_probe(ide_drive_t *drive) diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c index 10f457c..0590c75 100644 --- a/drivers/md/dm-io.c +++ b/drivers/md/dm-io.c @@ -356,7 +356,7 @@ static void dispatch_io(int rw, unsigned int num_regions, BUG_ON(num_regions > DM_IO_MAX_REGIONS); if (sync) - rw |= (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG); + rw |= REQ_SYNC | REQ_UNPLUG; /* * For multiple regions we need to be careful to rewind @@ -364,7 +364,7 @@ static void dispatch_io(int rw, unsigned int num_regions, */ for (i = 0; i < num_regions; i++) { *dp = old_pages; - if (where[i].count || (rw & (1 << BIO_RW_BARRIER))) + if (where[i].count || (rw & REQ_HARDBARRIER)) do_region(rw, i, where + i, dp, io); } @@ -412,8 +412,8 @@ retry: } set_current_state(TASK_RUNNING); - if (io->eopnotsupp_bits && (rw & (1 << BIO_RW_BARRIER))) { - rw &= ~(1 << BIO_RW_BARRIER); + if (io->eopnotsupp_bits && (rw & REQ_HARDBARRIER)) { + rw &= ~REQ_HARDBARRIER; goto retry; } @@ -479,8 +479,8 @@ static int dp_init(struct dm_io_request *io_req, struct dpages *dp) * New collapsed (a)synchronous interface. * * If the IO is asynchronous (i.e. it has notify.fn), you must either unplug - * the queue with blk_unplug() some time later or set the BIO_RW_SYNC bit in - * io_req->bi_rw. If you fail to do one of these, the IO will be submitted to + * the queue with blk_unplug() some time later or set REQ_SYNC in +io_req->bi_rw. If you fail to do one of these, the IO will be submitted to * the disk after q->unplug_delay, which defaults to 3ms in blk-settings.c. */ int dm_io(struct dm_io_request *io_req, unsigned num_regions, diff --git a/drivers/md/dm-kcopyd.c b/drivers/md/dm-kcopyd.c index addf834..d8587ba 100644 --- a/drivers/md/dm-kcopyd.c +++ b/drivers/md/dm-kcopyd.c @@ -345,7 +345,7 @@ static int run_io_job(struct kcopyd_job *job) { int r; struct dm_io_request io_req = { - .bi_rw = job->rw | (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG), + .bi_rw = job->rw | REQ_SYNC | REQ_UNPLUG, .mem.type = DM_IO_PAGE_LIST, .mem.ptr.pl = job->pages, .mem.offset = job->offset, diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c index ddda531..7413626 100644 --- a/drivers/md/dm-raid1.c +++ b/drivers/md/dm-raid1.c @@ -1211,7 +1211,7 @@ static int mirror_end_io(struct dm_target *ti, struct bio *bio, if (error == -EOPNOTSUPP) goto out; - if ((error == -EWOULDBLOCK) && bio_rw_flagged(bio, BIO_RW_AHEAD)) + if ((error == -EWOULDBLOCK) && (bio->bi_rw & REQ_RAHEAD)) goto out; if (unlikely(error)) { diff --git a/drivers/md/dm-stripe.c b/drivers/md/dm-stripe.c index e610725..d6e28d7 100644 --- a/drivers/md/dm-stripe.c +++ b/drivers/md/dm-stripe.c @@ -284,7 +284,7 @@ static int stripe_end_io(struct dm_target *ti, struct bio *bio, if (!error) return 0; /* I/O complete */ - if ((error == -EWOULDBLOCK) && bio_rw_flagged(bio, BIO_RW_AHEAD)) + if ((error == -EWOULDBLOCK) && (bio->bi_rw & REQ_RAHEAD)) return error; if (error == -EOPNOTSUPP) diff --git a/drivers/md/dm.c b/drivers/md/dm.c index d21e128..a3f21dc 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -15,6 +15,7 @@ #include <linux/blkpg.h> #include <linux/bio.h> #include <linux/buffer_head.h> +#include <linux/smp_lock.h> #include <linux/mempool.h> #include <linux/slab.h> #include <linux/idr.h> @@ -338,6 +339,7 @@ static int dm_blk_open(struct block_device *bdev, fmode_t mode) { struct mapped_device *md; + lock_kernel(); spin_lock(&_minor_lock); md = bdev->bd_disk->private_data; @@ -355,6 +357,7 @@ static int dm_blk_open(struct block_device *bdev, fmode_t mode) out: spin_unlock(&_minor_lock); + unlock_kernel(); return md ? 0 : -ENXIO; } @@ -362,8 +365,12 @@ out: static int dm_blk_close(struct gendisk *disk, fmode_t mode) { struct mapped_device *md = disk->private_data; + + lock_kernel(); atomic_dec(&md->open_count); dm_put(md); + unlock_kernel(); + return 0; } @@ -614,7 +621,7 @@ static void dec_pending(struct dm_io *io, int error) */ spin_lock_irqsave(&md->deferred_lock, flags); if (__noflush_suspending(md)) { - if (!bio_rw_flagged(io->bio, BIO_RW_BARRIER)) + if (!(io->bio->bi_rw & REQ_HARDBARRIER)) bio_list_add_head(&md->deferred, io->bio); } else @@ -626,7 +633,7 @@ static void dec_pending(struct dm_io *io, int error) io_error = io->error; bio = io->bio; - if (bio_rw_flagged(bio, BIO_RW_BARRIER)) { + if (bio->bi_rw & REQ_HARDBARRIER) { /* * There can be just one barrier request so we use * a per-device variable for error reporting. @@ -792,12 +799,12 @@ static void dm_end_request(struct request *clone, int error) { int rw = rq_data_dir(clone); int run_queue = 1; - bool is_barrier = blk_barrier_rq(clone); + bool is_barrier = clone->cmd_flags & REQ_HARDBARRIER; struct dm_rq_target_io *tio = clone->end_io_data; struct mapped_device *md = tio->md; struct request *rq = tio->orig; - if (blk_pc_request(rq) && !is_barrier) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC && !is_barrier) { rq->errors = clone->errors; rq->resid_len = clone->resid_len; @@ -844,7 +851,7 @@ void dm_requeue_unmapped_request(struct request *clone) struct request_queue *q = rq->q; unsigned long flags; - if (unlikely(blk_barrier_rq(clone))) { + if (unlikely(clone->cmd_flags & REQ_HARDBARRIER)) { /* * Barrier clones share an original request. * Leave it to dm_end_request(), which handles this special @@ -943,7 +950,7 @@ static void dm_complete_request(struct request *clone, int error) struct dm_rq_target_io *tio = clone->end_io_data; struct request *rq = tio->orig; - if (unlikely(blk_barrier_rq(clone))) { + if (unlikely(clone->cmd_flags & REQ_HARDBARRIER)) { /* * Barrier clones share an original request. So can't use * softirq_done with the original. @@ -972,7 +979,7 @@ void dm_kill_unmapped_request(struct request *clone, int error) struct dm_rq_target_io *tio = clone->end_io_data; struct request *rq = tio->orig; - if (unlikely(blk_barrier_rq(clone))) { + if (unlikely(clone->cmd_flags & REQ_HARDBARRIER)) { /* * Barrier clones share an original request. * Leave it to dm_end_request(), which handles this special @@ -1106,7 +1113,7 @@ static struct bio *split_bvec(struct bio *bio, sector_t sector, clone->bi_sector = sector; clone->bi_bdev = bio->bi_bdev; - clone->bi_rw = bio->bi_rw & ~(1 << BIO_RW_BARRIER); + clone->bi_rw = bio->bi_rw & ~REQ_HARDBARRIER; clone->bi_vcnt = 1; clone->bi_size = to_bytes(len); clone->bi_io_vec->bv_offset = offset; @@ -1133,7 +1140,7 @@ static struct bio *clone_bio(struct bio *bio, sector_t sector, clone = bio_alloc_bioset(GFP_NOIO, bio->bi_max_vecs, bs); __bio_clone(clone, bio); - clone->bi_rw &= ~(1 << BIO_RW_BARRIER); + clone->bi_rw &= ~REQ_HARDBARRIER; clone->bi_destructor = dm_bio_destructor; clone->bi_sector = sector; clone->bi_idx = idx; @@ -1301,7 +1308,7 @@ static void __split_and_process_bio(struct mapped_device *md, struct bio *bio) ci.map = dm_get_live_table(md); if (unlikely(!ci.map)) { - if (!bio_rw_flagged(bio, BIO_RW_BARRIER)) + if (!(bio->bi_rw & REQ_HARDBARRIER)) bio_io_error(bio); else if (!md->barrier_error) @@ -1414,7 +1421,7 @@ static int _dm_request(struct request_queue *q, struct bio *bio) * we have to queue this io for later. */ if (unlikely(test_bit(DMF_QUEUE_IO_TO_THREAD, &md->flags)) || - unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER))) { + unlikely(bio->bi_rw & REQ_HARDBARRIER)) { up_read(&md->io_lock); if (unlikely(test_bit(DMF_BLOCK_IO_FOR_SUSPEND, &md->flags)) && @@ -1455,20 +1462,9 @@ static int dm_request(struct request_queue *q, struct bio *bio) return _dm_request(q, bio); } -/* - * Mark this request as flush request, so that dm_request_fn() can - * recognize. - */ -static void dm_rq_prepare_flush(struct request_queue *q, struct request *rq) -{ - rq->cmd_type = REQ_TYPE_LINUX_BLOCK; - rq->cmd[0] = REQ_LB_OP_FLUSH; -} - static bool dm_rq_is_flush_request(struct request *rq) { - if (rq->cmd_type == REQ_TYPE_LINUX_BLOCK && - rq->cmd[0] == REQ_LB_OP_FLUSH) + if (rq->cmd_flags & REQ_FLUSH) return true; else return false; @@ -1912,8 +1908,7 @@ static struct mapped_device *alloc_dev(int minor) blk_queue_softirq_done(md->queue, dm_softirq_done); blk_queue_prep_rq(md->queue, dm_prep_fn); blk_queue_lld_busy(md->queue, dm_lld_busy); - blk_queue_ordered(md->queue, QUEUE_ORDERED_DRAIN_FLUSH, - dm_rq_prepare_flush); + blk_queue_ordered(md->queue, QUEUE_ORDERED_DRAIN_FLUSH); md->disk = alloc_disk(1); if (!md->disk) @@ -2296,7 +2291,7 @@ static void dm_wq_work(struct work_struct *work) if (dm_request_based(md)) generic_make_request(c); else { - if (bio_rw_flagged(c, BIO_RW_BARRIER)) + if (c->bi_rw & REQ_HARDBARRIER) process_barrier(md, c); else __split_and_process_bio(md, c); diff --git a/drivers/md/linear.c b/drivers/md/linear.c index 7e0e057..ba19060 100644 --- a/drivers/md/linear.c +++ b/drivers/md/linear.c @@ -294,7 +294,7 @@ static int linear_make_request (mddev_t *mddev, struct bio *bio) dev_info_t *tmp_dev; sector_t start_sector; - if (unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER))) { + if (unlikely(bio->bi_rw & REQ_HARDBARRIER)) { md_barrier_request(mddev, bio); return 0; } diff --git a/drivers/md/md.c b/drivers/md/md.c index cb20d0b..700c96e 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -36,6 +36,7 @@ #include <linux/blkdev.h> #include <linux/sysctl.h> #include <linux/seq_file.h> +#include <linux/smp_lock.h> #include <linux/buffer_head.h> /* for invalidate_bdev */ #include <linux/poll.h> #include <linux/ctype.h> @@ -353,7 +354,7 @@ static void md_submit_barrier(struct work_struct *ws) /* an empty barrier - all done */ bio_endio(bio, 0); else { - bio->bi_rw &= ~(1<<BIO_RW_BARRIER); + bio->bi_rw &= ~REQ_HARDBARRIER; if (mddev->pers->make_request(mddev, bio)) generic_make_request(bio); mddev->barrier = POST_REQUEST_BARRIER; @@ -675,11 +676,11 @@ void md_super_write(mddev_t *mddev, mdk_rdev_t *rdev, * if zero is reached. * If an error occurred, call md_error * - * As we might need to resubmit the request if BIO_RW_BARRIER + * As we might need to resubmit the request if REQ_HARDBARRIER * causes ENOTSUPP, we allocate a spare bio... */ struct bio *bio = bio_alloc(GFP_NOIO, 1); - int rw = (1<<BIO_RW) | (1<<BIO_RW_SYNCIO) | (1<<BIO_RW_UNPLUG); + int rw = REQ_WRITE | REQ_SYNC | REQ_UNPLUG; bio->bi_bdev = rdev->bdev; bio->bi_sector = sector; @@ -691,7 +692,7 @@ void md_super_write(mddev_t *mddev, mdk_rdev_t *rdev, atomic_inc(&mddev->pending_writes); if (!test_bit(BarriersNotsupp, &rdev->flags)) { struct bio *rbio; - rw |= (1<<BIO_RW_BARRIER); + rw |= REQ_HARDBARRIER; rbio = bio_clone(bio, GFP_NOIO); rbio->bi_private = bio; rbio->bi_end_io = super_written_barrier; @@ -736,7 +737,7 @@ int sync_page_io(struct block_device *bdev, sector_t sector, int size, struct completion event; int ret; - rw |= (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG); + rw |= REQ_SYNC | REQ_UNPLUG; bio->bi_bdev = bdev; bio->bi_sector = sector; @@ -5902,6 +5903,7 @@ static int md_open(struct block_device *bdev, fmode_t mode) mddev_t *mddev = mddev_find(bdev->bd_dev); int err; + lock_kernel(); if (mddev->gendisk != bdev->bd_disk) { /* we are racing with mddev_put which is discarding this * bd_disk. @@ -5910,6 +5912,7 @@ static int md_open(struct block_device *bdev, fmode_t mode) /* Wait until bdev->bd_disk is definitely gone */ flush_scheduled_work(); /* Then retry the open from the top */ + unlock_kernel(); return -ERESTARTSYS; } BUG_ON(mddev != bdev->bd_disk->private_data); @@ -5923,6 +5926,7 @@ static int md_open(struct block_device *bdev, fmode_t mode) check_disk_size_change(mddev->gendisk, bdev); out: + unlock_kernel(); return err; } @@ -5931,8 +5935,10 @@ static int md_release(struct gendisk *disk, fmode_t mode) mddev_t *mddev = disk->private_data; BUG_ON(!mddev); + lock_kernel(); atomic_dec(&mddev->openers); mddev_put(mddev); + unlock_kernel(); return 0; } diff --git a/drivers/md/md.h b/drivers/md/md.h index 10597bf..fc56e0f 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -67,7 +67,7 @@ struct mdk_rdev_s #define Faulty 1 /* device is known to have a fault */ #define In_sync 2 /* device is in_sync with rest of array */ #define WriteMostly 4 /* Avoid reading if at all possible */ -#define BarriersNotsupp 5 /* BIO_RW_BARRIER is not supported */ +#define BarriersNotsupp 5 /* REQ_HARDBARRIER is not supported */ #define AllReserved 6 /* If whole device is reserved for * one array */ #define AutoDetected 7 /* added by auto-detect */ @@ -254,7 +254,7 @@ struct mddev_s * fails. Only supported */ struct bio *biolist; /* bios that need to be retried - * because BIO_RW_BARRIER is not supported + * because REQ_HARDBARRIER is not supported */ atomic_t recovery_active; /* blocks scheduled, but not written */ diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c index 410fb60..0307d21 100644 --- a/drivers/md/multipath.c +++ b/drivers/md/multipath.c @@ -91,7 +91,7 @@ static void multipath_end_request(struct bio *bio, int error) if (uptodate) multipath_end_bh_io(mp_bh, 0); - else if (!bio_rw_flagged(bio, BIO_RW_AHEAD)) { + else if (!(bio->bi_rw & REQ_RAHEAD)) { /* * oops, IO error: */ @@ -142,7 +142,7 @@ static int multipath_make_request(mddev_t *mddev, struct bio * bio) struct multipath_bh * mp_bh; struct multipath_info *multipath; - if (unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER))) { + if (unlikely(bio->bi_rw & REQ_HARDBARRIER)) { md_barrier_request(mddev, bio); return 0; } @@ -163,7 +163,7 @@ static int multipath_make_request(mddev_t *mddev, struct bio * bio) mp_bh->bio = *bio; mp_bh->bio.bi_sector += multipath->rdev->data_offset; mp_bh->bio.bi_bdev = multipath->rdev->bdev; - mp_bh->bio.bi_rw |= (1 << BIO_RW_FAILFAST_TRANSPORT); + mp_bh->bio.bi_rw |= REQ_FAILFAST_TRANSPORT; mp_bh->bio.bi_end_io = multipath_end_request; mp_bh->bio.bi_private = mp_bh; generic_make_request(&mp_bh->bio); @@ -398,7 +398,7 @@ static void multipathd (mddev_t *mddev) *bio = *(mp_bh->master_bio); bio->bi_sector += conf->multipaths[mp_bh->path].rdev->data_offset; bio->bi_bdev = conf->multipaths[mp_bh->path].rdev->bdev; - bio->bi_rw |= (1 << BIO_RW_FAILFAST_TRANSPORT); + bio->bi_rw |= REQ_FAILFAST_TRANSPORT; bio->bi_end_io = multipath_end_request; bio->bi_private = mp_bh; generic_make_request(bio); diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c index 563abed..6f7af46 100644 --- a/drivers/md/raid0.c +++ b/drivers/md/raid0.c @@ -483,7 +483,7 @@ static int raid0_make_request(mddev_t *mddev, struct bio *bio) struct strip_zone *zone; mdk_rdev_t *tmp_dev; - if (unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER))) { + if (unlikely(bio->bi_rw & REQ_HARDBARRIER)) { md_barrier_request(mddev, bio); return 0; } diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index a948da8..73cc74f 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -787,7 +787,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) struct bio_list bl; struct page **behind_pages = NULL; const int rw = bio_data_dir(bio); - const bool do_sync = bio_rw_flagged(bio, BIO_RW_SYNCIO); + const bool do_sync = (bio->bi_rw & REQ_SYNC); bool do_barriers; mdk_rdev_t *blocked_rdev; @@ -822,7 +822,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) finish_wait(&conf->wait_barrier, &w); } if (unlikely(!mddev->barriers_work && - bio_rw_flagged(bio, BIO_RW_BARRIER))) { + (bio->bi_rw & REQ_HARDBARRIER))) { if (rw == WRITE) md_write_end(mddev); bio_endio(bio, -EOPNOTSUPP); @@ -877,7 +877,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) read_bio->bi_sector = r1_bio->sector + mirror->rdev->data_offset; read_bio->bi_bdev = mirror->rdev->bdev; read_bio->bi_end_io = raid1_end_read_request; - read_bio->bi_rw = READ | (do_sync << BIO_RW_SYNCIO); + read_bio->bi_rw = READ | do_sync; read_bio->bi_private = r1_bio; generic_make_request(read_bio); @@ -959,7 +959,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) atomic_set(&r1_bio->remaining, 0); atomic_set(&r1_bio->behind_remaining, 0); - do_barriers = bio_rw_flagged(bio, BIO_RW_BARRIER); + do_barriers = bio->bi_rw & REQ_HARDBARRIER; if (do_barriers) set_bit(R1BIO_Barrier, &r1_bio->state); @@ -975,8 +975,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) mbio->bi_sector = r1_bio->sector + conf->mirrors[i].rdev->data_offset; mbio->bi_bdev = conf->mirrors[i].rdev->bdev; mbio->bi_end_io = raid1_end_write_request; - mbio->bi_rw = WRITE | (do_barriers << BIO_RW_BARRIER) | - (do_sync << BIO_RW_SYNCIO); + mbio->bi_rw = WRITE | do_barriers | do_sync; mbio->bi_private = r1_bio; if (behind_pages) { @@ -1633,7 +1632,7 @@ static void raid1d(mddev_t *mddev) sync_request_write(mddev, r1_bio); unplug = 1; } else if (test_bit(R1BIO_BarrierRetry, &r1_bio->state)) { - /* some requests in the r1bio were BIO_RW_BARRIER + /* some requests in the r1bio were REQ_HARDBARRIER * requests which failed with -EOPNOTSUPP. Hohumm.. * Better resubmit without the barrier. * We know which devices to resubmit for, because @@ -1641,7 +1640,7 @@ static void raid1d(mddev_t *mddev) * We already have a nr_pending reference on these rdevs. */ int i; - const bool do_sync = bio_rw_flagged(r1_bio->master_bio, BIO_RW_SYNCIO); + const bool do_sync = (r1_bio->master_bio->bi_rw & REQ_SYNC); clear_bit(R1BIO_BarrierRetry, &r1_bio->state); clear_bit(R1BIO_Barrier, &r1_bio->state); for (i=0; i < conf->raid_disks; i++) @@ -1662,8 +1661,7 @@ static void raid1d(mddev_t *mddev) conf->mirrors[i].rdev->data_offset; bio->bi_bdev = conf->mirrors[i].rdev->bdev; bio->bi_end_io = raid1_end_write_request; - bio->bi_rw = WRITE | - (do_sync << BIO_RW_SYNCIO); + bio->bi_rw = WRITE | do_sync; bio->bi_private = r1_bio; r1_bio->bios[i] = bio; generic_make_request(bio); @@ -1698,7 +1696,7 @@ static void raid1d(mddev_t *mddev) (unsigned long long)r1_bio->sector); raid_end_bio_io(r1_bio); } else { - const bool do_sync = bio_rw_flagged(r1_bio->master_bio, BIO_RW_SYNCIO); + const bool do_sync = r1_bio->master_bio->bi_rw & REQ_SYNC; r1_bio->bios[r1_bio->read_disk] = mddev->ro ? IO_BLOCKED : NULL; r1_bio->read_disk = disk; @@ -1715,7 +1713,7 @@ static void raid1d(mddev_t *mddev) bio->bi_sector = r1_bio->sector + rdev->data_offset; bio->bi_bdev = rdev->bdev; bio->bi_end_io = raid1_end_read_request; - bio->bi_rw = READ | (do_sync << BIO_RW_SYNCIO); + bio->bi_rw = READ | do_sync; bio->bi_private = r1_bio; unplug = 1; generic_make_request(bio); diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 42e64e4..62ecb66 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -799,12 +799,12 @@ static int make_request(mddev_t *mddev, struct bio * bio) int i; int chunk_sects = conf->chunk_mask + 1; const int rw = bio_data_dir(bio); - const bool do_sync = bio_rw_flagged(bio, BIO_RW_SYNCIO); + const bool do_sync = (bio->bi_rw & REQ_SYNC); struct bio_list bl; unsigned long flags; mdk_rdev_t *blocked_rdev; - if (unlikely(bio_rw_flagged(bio, BIO_RW_BARRIER))) { + if (unlikely(bio->bi_rw & REQ_HARDBARRIER)) { md_barrier_request(mddev, bio); return 0; } @@ -879,7 +879,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) mirror->rdev->data_offset; read_bio->bi_bdev = mirror->rdev->bdev; read_bio->bi_end_io = raid10_end_read_request; - read_bio->bi_rw = READ | (do_sync << BIO_RW_SYNCIO); + read_bio->bi_rw = READ | do_sync; read_bio->bi_private = r10_bio; generic_make_request(read_bio); @@ -947,7 +947,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) conf->mirrors[d].rdev->data_offset; mbio->bi_bdev = conf->mirrors[d].rdev->bdev; mbio->bi_end_io = raid10_end_write_request; - mbio->bi_rw = WRITE | (do_sync << BIO_RW_SYNCIO); + mbio->bi_rw = WRITE | do_sync; mbio->bi_private = r10_bio; atomic_inc(&r10_bio->remaining); @@ -1716,7 +1716,7 @@ static void raid10d(mddev_t *mddev) raid_end_bio_io(r10_bio); bio_put(bio); } else { - const bool do_sync = bio_rw_flagged(r10_bio->master_bio, BIO_RW_SYNCIO); + const bool do_sync = (r10_bio->master_bio->bi_rw & REQ_SYNC); bio_put(bio); rdev = conf->mirrors[mirror].rdev; if (printk_ratelimit()) @@ -1730,7 +1730,7 @@ static void raid10d(mddev_t *mddev) bio->bi_sector = r10_bio->devs[r10_bio->read_slot].addr + rdev->data_offset; bio->bi_bdev = rdev->bdev; - bio->bi_rw = READ | (do_sync << BIO_RW_SYNCIO); + bio->bi_rw = READ | do_sync; bio->bi_private = r10_bio; bio->bi_end_io = raid10_end_read_request; unplug = 1; diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 96c6902..20ac2f1 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -3958,7 +3958,7 @@ static int make_request(mddev_t *mddev, struct bio * bi) const int rw = bio_data_dir(bi); int remaining; - if (unlikely(bio_rw_flagged(bi, BIO_RW_BARRIER))) { + if (unlikely(bi->bi_rw & REQ_HARDBARRIER)) { /* Drain all pending writes. We only really need * to ensure they have been submitted, but this is * easier. diff --git a/drivers/memstick/core/mspro_block.c b/drivers/memstick/core/mspro_block.c index 8327e24..eef78a0 100644 --- a/drivers/memstick/core/mspro_block.c +++ b/drivers/memstick/core/mspro_block.c @@ -18,6 +18,7 @@ #include <linux/kthread.h> #include <linux/delay.h> #include <linux/slab.h> +#include <linux/smp_lock.h> #include <linux/memstick.h> #define DRIVER_NAME "mspro_block" @@ -179,6 +180,7 @@ static int mspro_block_bd_open(struct block_device *bdev, fmode_t mode) struct mspro_block_data *msb = disk->private_data; int rc = -ENXIO; + lock_kernel(); mutex_lock(&mspro_block_disk_lock); if (msb && msb->card) { @@ -190,6 +192,7 @@ static int mspro_block_bd_open(struct block_device *bdev, fmode_t mode) } mutex_unlock(&mspro_block_disk_lock); + unlock_kernel(); return rc; } @@ -221,7 +224,11 @@ static int mspro_block_disk_release(struct gendisk *disk) static int mspro_block_bd_release(struct gendisk *disk, fmode_t mode) { - return mspro_block_disk_release(disk); + int ret; + lock_kernel(); + ret = mspro_block_disk_release(disk); + unlock_kernel(); + return ret; } static int mspro_block_bd_getgeo(struct block_device *bdev, @@ -805,7 +812,8 @@ static void mspro_block_start(struct memstick_dev *card) static int mspro_block_prepare_req(struct request_queue *q, struct request *req) { - if (!blk_fs_request(req) && !blk_pc_request(req)) { + if (req->cmd_type != REQ_TYPE_FS && + req->cmd_type != REQ_TYPE_BLOCK_PC) { blk_dump_rq_flags(req, "MSPro unsupported request"); return BLKPREP_KILL; } diff --git a/drivers/message/i2o/i2o_block.c b/drivers/message/i2o/i2o_block.c index fc593fb..e6733bc 100644 --- a/drivers/message/i2o/i2o_block.c +++ b/drivers/message/i2o/i2o_block.c @@ -53,6 +53,7 @@ #include <linux/module.h> #include <linux/slab.h> #include <linux/i2o.h> +#include <linux/smp_lock.h> #include <linux/mempool.h> @@ -577,6 +578,7 @@ static int i2o_block_open(struct block_device *bdev, fmode_t mode) if (!dev->i2o_dev) return -ENODEV; + lock_kernel(); if (dev->power > 0x1f) i2o_block_device_power(dev, 0x02); @@ -585,6 +587,7 @@ static int i2o_block_open(struct block_device *bdev, fmode_t mode) i2o_block_device_lock(dev->i2o_dev, -1); osm_debug("Ready.\n"); + unlock_kernel(); return 0; }; @@ -615,6 +618,7 @@ static int i2o_block_release(struct gendisk *disk, fmode_t mode) if (!dev->i2o_dev) return 0; + lock_kernel(); i2o_block_device_flush(dev->i2o_dev); i2o_block_device_unlock(dev->i2o_dev, -1); @@ -625,6 +629,7 @@ static int i2o_block_release(struct gendisk *disk, fmode_t mode) operation = 0x24; i2o_block_device_power(dev, operation); + unlock_kernel(); return 0; } @@ -652,30 +657,40 @@ static int i2o_block_ioctl(struct block_device *bdev, fmode_t mode, { struct gendisk *disk = bdev->bd_disk; struct i2o_block_device *dev = disk->private_data; + int ret = -ENOTTY; /* Anyone capable of this syscall can do *real bad* things */ if (!capable(CAP_SYS_ADMIN)) return -EPERM; + lock_kernel(); switch (cmd) { case BLKI2OGRSTRAT: - return put_user(dev->rcache, (int __user *)arg); + ret = put_user(dev->rcache, (int __user *)arg); + break; case BLKI2OGWSTRAT: - return put_user(dev->wcache, (int __user *)arg); + ret = put_user(dev->wcache, (int __user *)arg); + break; case BLKI2OSRSTRAT: + ret = -EINVAL; if (arg < 0 || arg > CACHE_SMARTFETCH) - return -EINVAL; + break; dev->rcache = arg; + ret = 0; break; case BLKI2OSWSTRAT: + ret = -EINVAL; if (arg != 0 && (arg < CACHE_WRITETHROUGH || arg > CACHE_SMARTBACK)) - return -EINVAL; + break; dev->wcache = arg; + ret = 0; break; } - return -ENOTTY; + unlock_kernel(); + + return ret; }; /** @@ -883,7 +898,7 @@ static void i2o_block_request_fn(struct request_queue *q) if (!req) break; - if (blk_fs_request(req)) { + if (req->cmd_type == REQ_TYPE_FS) { struct i2o_block_delayed_request *dreq; struct i2o_block_request *ireq = req->special; unsigned int queue_depth; @@ -930,7 +945,8 @@ static const struct block_device_operations i2o_block_fops = { .owner = THIS_MODULE, .open = i2o_block_open, .release = i2o_block_release, - .locked_ioctl = i2o_block_ioctl, + .ioctl = i2o_block_ioctl, + .compat_ioctl = i2o_block_ioctl, .getgeo = i2o_block_getgeo, .media_changed = i2o_block_media_changed }; diff --git a/drivers/mmc/card/block.c b/drivers/mmc/card/block.c index cb9fbc8..8433cde 100644 --- a/drivers/mmc/card/block.c +++ b/drivers/mmc/card/block.c @@ -29,6 +29,7 @@ #include <linux/kdev_t.h> #include <linux/blkdev.h> #include <linux/mutex.h> +#include <linux/smp_lock.h> #include <linux/scatterlist.h> #include <linux/string_helpers.h> @@ -107,6 +108,7 @@ static int mmc_blk_open(struct block_device *bdev, fmode_t mode) struct mmc_blk_data *md = mmc_blk_get(bdev->bd_disk); int ret = -ENXIO; + lock_kernel(); if (md) { if (md->usage == 2) check_disk_change(bdev); @@ -117,6 +119,7 @@ static int mmc_blk_open(struct block_device *bdev, fmode_t mode) ret = -EROFS; } } + unlock_kernel(); return ret; } @@ -125,7 +128,9 @@ static int mmc_blk_release(struct gendisk *disk, fmode_t mode) { struct mmc_blk_data *md = disk->private_data; + lock_kernel(); mmc_blk_put(md); + unlock_kernel(); return 0; } diff --git a/drivers/mmc/card/queue.c b/drivers/mmc/card/queue.c index d6ded24..c77eb49 100644 --- a/drivers/mmc/card/queue.c +++ b/drivers/mmc/card/queue.c @@ -32,7 +32,7 @@ static int mmc_prep_request(struct request_queue *q, struct request *req) /* * We only like normal block requests. */ - if (!blk_fs_request(req)) { + if (req->cmd_type != REQ_TYPE_FS) { blk_dump_rq_flags(req, "MMC bad request"); return BLKPREP_KILL; } @@ -128,7 +128,7 @@ int mmc_init_queue(struct mmc_queue *mq, struct mmc_card *card, spinlock_t *lock mq->req = NULL; blk_queue_prep_rq(mq->queue, mmc_prep_request); - blk_queue_ordered(mq->queue, QUEUE_ORDERED_DRAIN, NULL); + blk_queue_ordered(mq->queue, QUEUE_ORDERED_DRAIN); queue_flag_set_unlocked(QUEUE_FLAG_NONROT, mq->queue); #ifdef CONFIG_MMC_BLOCK_BOUNCE diff --git a/drivers/mtd/mtd_blkdevs.c b/drivers/mtd/mtd_blkdevs.c index 1d2144d..62e6870 100644 --- a/drivers/mtd/mtd_blkdevs.c +++ b/drivers/mtd/mtd_blkdevs.c @@ -29,6 +29,7 @@ #include <linux/blkdev.h> #include <linux/blkpg.h> #include <linux/spinlock.h> +#include <linux/smp_lock.h> #include <linux/hdreg.h> #include <linux/init.h> #include <linux/mutex.h> @@ -87,14 +88,14 @@ static int do_blktrans_request(struct mtd_blktrans_ops *tr, buf = req->buffer; - if (!blk_fs_request(req)) + if (req->cmd_type != REQ_TYPE_FS) return -EIO; if (blk_rq_pos(req) + blk_rq_cur_sectors(req) > get_capacity(req->rq_disk)) return -EIO; - if (blk_discard_rq(req)) + if (req->cmd_flags & REQ_DISCARD) return tr->discard(dev, block, nsect); switch(rq_data_dir(req)) { @@ -178,8 +179,9 @@ static int blktrans_open(struct block_device *bdev, fmode_t mode) int ret; if (!dev) - return -ERESTARTSYS; + return -ERESTARTSYS; /* FIXME: busy loop! -arnd*/ + lock_kernel(); mutex_lock(&dev->lock); if (!dev->mtd) { @@ -196,6 +198,7 @@ static int blktrans_open(struct block_device *bdev, fmode_t mode) unlock: mutex_unlock(&dev->lock); blktrans_dev_put(dev); + unlock_kernel(); return ret; } @@ -207,6 +210,7 @@ static int blktrans_release(struct gendisk *disk, fmode_t mode) if (!dev) return ret; + lock_kernel(); mutex_lock(&dev->lock); /* Release one reference, we sure its not the last one here*/ @@ -219,6 +223,7 @@ static int blktrans_release(struct gendisk *disk, fmode_t mode) unlock: mutex_unlock(&dev->lock); blktrans_dev_put(dev); + unlock_kernel(); return ret; } @@ -251,6 +256,7 @@ static int blktrans_ioctl(struct block_device *bdev, fmode_t mode, if (!dev) return ret; + lock_kernel(); mutex_lock(&dev->lock); if (!dev->mtd) @@ -265,6 +271,7 @@ static int blktrans_ioctl(struct block_device *bdev, fmode_t mode, } unlock: mutex_unlock(&dev->lock); + unlock_kernel(); blktrans_dev_put(dev); return ret; } @@ -273,7 +280,7 @@ static const struct block_device_operations mtd_blktrans_ops = { .owner = THIS_MODULE, .open = blktrans_open, .release = blktrans_release, - .locked_ioctl = blktrans_ioctl, + .ioctl = blktrans_ioctl, .getgeo = blktrans_getgeo, }; diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c index 33975e9..1a84fae 100644 --- a/drivers/s390/block/dasd.c +++ b/drivers/s390/block/dasd.c @@ -21,6 +21,7 @@ #include <linux/hdreg.h> #include <linux/async.h> #include <linux/mutex.h> +#include <linux/smp_lock.h> #include <asm/ccwdev.h> #include <asm/ebcdic.h> @@ -2196,7 +2197,7 @@ static void dasd_setup_queue(struct dasd_block *block) */ blk_queue_max_segment_size(block->request_queue, PAGE_SIZE); blk_queue_segment_boundary(block->request_queue, PAGE_SIZE - 1); - blk_queue_ordered(block->request_queue, QUEUE_ORDERED_DRAIN, NULL); + blk_queue_ordered(block->request_queue, QUEUE_ORDERED_DRAIN); } /* @@ -2235,6 +2236,7 @@ static int dasd_open(struct block_device *bdev, fmode_t mode) if (!block) return -ENODEV; + lock_kernel(); base = block->base; atomic_inc(&block->open_count); if (test_bit(DASD_FLAG_OFFLINE, &base->flags)) { @@ -2269,12 +2271,14 @@ static int dasd_open(struct block_device *bdev, fmode_t mode) goto out; } + unlock_kernel(); return 0; out: module_put(base->discipline->owner); unlock: atomic_dec(&block->open_count); + unlock_kernel(); return rc; } @@ -2282,8 +2286,10 @@ static int dasd_release(struct gendisk *disk, fmode_t mode) { struct dasd_block *block = disk->private_data; + lock_kernel(); atomic_dec(&block->open_count); module_put(block->base->discipline->owner); + unlock_kernel(); return 0; } diff --git a/drivers/s390/block/dcssblk.c b/drivers/s390/block/dcssblk.c index 9b43ae9..2bd72aa 100644 --- a/drivers/s390/block/dcssblk.c +++ b/drivers/s390/block/dcssblk.c @@ -14,6 +14,7 @@ #include <linux/init.h> #include <linux/slab.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <linux/completion.h> #include <linux/interrupt.h> #include <linux/platform_device.h> @@ -775,6 +776,7 @@ dcssblk_open(struct block_device *bdev, fmode_t mode) struct dcssblk_dev_info *dev_info; int rc; + lock_kernel(); dev_info = bdev->bd_disk->private_data; if (NULL == dev_info) { rc = -ENODEV; @@ -784,6 +786,7 @@ dcssblk_open(struct block_device *bdev, fmode_t mode) bdev->bd_block_size = 4096; rc = 0; out: + unlock_kernel(); return rc; } @@ -794,6 +797,7 @@ dcssblk_release(struct gendisk *disk, fmode_t mode) struct segment_info *entry; int rc; + lock_kernel(); if (!dev_info) { rc = -ENODEV; goto out; @@ -811,6 +815,7 @@ dcssblk_release(struct gendisk *disk, fmode_t mode) up_write(&dcssblk_devices_sem); rc = 0; out: + unlock_kernel(); return rc; } diff --git a/drivers/s390/char/tape_block.c b/drivers/s390/char/tape_block.c index 097da8c..b7de025 100644 --- a/drivers/s390/char/tape_block.c +++ b/drivers/s390/char/tape_block.c @@ -16,6 +16,7 @@ #include <linux/fs.h> #include <linux/module.h> #include <linux/blkdev.h> +#include <linux/smp_lock.h> #include <linux/interrupt.h> #include <linux/buffer_head.h> #include <linux/kernel.h> @@ -361,6 +362,7 @@ tapeblock_open(struct block_device *bdev, fmode_t mode) struct tape_device * device; int rc; + lock_kernel(); device = tape_get_device(disk->private_data); if (device->required_tapemarks) { @@ -384,12 +386,14 @@ tapeblock_open(struct block_device *bdev, fmode_t mode) * is called. */ tape_state_set(device, TS_BLKUSE); + unlock_kernel(); return 0; release: tape_release(device); put_device: tape_put_device(device); + unlock_kernel(); return rc; } @@ -403,10 +407,12 @@ static int tapeblock_release(struct gendisk *disk, fmode_t mode) { struct tape_device *device = disk->private_data; - + + lock_kernel(); tape_state_set(device, TS_IN_USE); tape_release(device); tape_put_device(device); + unlock_kernel(); return 0; } diff --git a/drivers/scsi/aha1542.c b/drivers/scsi/aha1542.c index 2a8cf13..4f785f2 100644 --- a/drivers/scsi/aha1542.c +++ b/drivers/scsi/aha1542.c @@ -52,22 +52,6 @@ #define SCSI_BUF_PA(address) isa_virt_to_bus(address) #define SCSI_SG_PA(sgent) (isa_page_to_bus(sg_page((sgent))) + (sgent)->offset) -static void BAD_SG_DMA(Scsi_Cmnd * SCpnt, - struct scatterlist *sgp, - int nseg, - int badseg) -{ - printk(KERN_CRIT "sgpnt[%d:%d] page %p/0x%llx length %u\n", - badseg, nseg, sg_virt(sgp), - (unsigned long long)SCSI_SG_PA(sgp), - sgp->length); - - /* - * Not safe to continue. - */ - panic("Buffer at physical address > 16Mb used for aha1542"); -} - #include<linux/stat.h> #ifdef DEBUG @@ -691,8 +675,6 @@ static int aha1542_queuecommand(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *)) } scsi_for_each_sg(SCpnt, sg, sg_count, i) { any2scsi(cptr[i].dataptr, SCSI_SG_PA(sg)); - if (SCSI_SG_PA(sg) + sg->length - 1 > ISA_DMA_THRESHOLD) - BAD_SG_DMA(SCpnt, scsi_sglist(SCpnt), sg_count, i); any2scsi(cptr[i].datalen, sg->length); }; any2scsi(ccb[mbo].datalen, sg_count * sizeof(struct chain)); @@ -1133,16 +1115,9 @@ static int __init aha1542_detect(struct scsi_host_template * tpnt) release_region(bases[indx], 4); continue; } - /* For now we do this - until kmalloc is more intelligent - we are resigned to stupid hacks like this */ - if (SCSI_BUF_PA(shpnt) >= ISA_DMA_THRESHOLD) { - printk(KERN_ERR "Invalid address for shpnt with 1542.\n"); - goto unregister; - } if (!aha1542_test_port(bases[indx], shpnt)) goto unregister; - base_io = bases[indx]; /* Set the Bus on/off-times as not to ruin floppy performance */ diff --git a/drivers/scsi/osd/osd_initiator.c b/drivers/scsi/osd/osd_initiator.c index ee4b691..fda4de3 100644 --- a/drivers/scsi/osd/osd_initiator.c +++ b/drivers/scsi/osd/osd_initiator.c @@ -716,7 +716,7 @@ static int _osd_req_list_objects(struct osd_request *or, return PTR_ERR(bio); } - bio->bi_rw &= ~(1 << BIO_RW); + bio->bi_rw &= ~REQ_WRITE; or->in.bio = bio; or->in.total_bytes = bio->bi_size; return 0; @@ -814,7 +814,7 @@ void osd_req_write(struct osd_request *or, { _osd_req_encode_common(or, OSD_ACT_WRITE, obj, offset, len); WARN_ON(or->out.bio || or->out.total_bytes); - WARN_ON(0 == bio_rw_flagged(bio, BIO_RW)); + WARN_ON(0 == (bio->bi_rw & REQ_WRITE)); or->out.bio = bio; or->out.total_bytes = len; } @@ -829,7 +829,7 @@ int osd_req_write_kern(struct osd_request *or, if (IS_ERR(bio)) return PTR_ERR(bio); - bio->bi_rw |= (1 << BIO_RW); /* FIXME: bio_set_dir() */ + bio->bi_rw |= REQ_WRITE; /* FIXME: bio_set_dir() */ osd_req_write(or, obj, offset, bio, len); return 0; } @@ -865,7 +865,7 @@ void osd_req_read(struct osd_request *or, { _osd_req_encode_common(or, OSD_ACT_READ, obj, offset, len); WARN_ON(or->in.bio || or->in.total_bytes); - WARN_ON(1 == bio_rw_flagged(bio, BIO_RW)); + WARN_ON(1 == (bio->bi_rw & REQ_WRITE)); or->in.bio = bio; or->in.total_bytes = len; } diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c index 2bf9846..bbbc186 100644 --- a/drivers/scsi/scsi_error.c +++ b/drivers/scsi/scsi_error.c @@ -320,7 +320,7 @@ static int scsi_check_sense(struct scsi_cmnd *scmd) "changed. The Linux SCSI layer does not " "automatically adjust these parameters.\n"); - if (blk_barrier_rq(scmd->request)) + if (scmd->request->cmd_flags & REQ_HARDBARRIER) /* * barrier requests should always retry on UA * otherwise block will get a spurious error @@ -1331,16 +1331,16 @@ int scsi_noretry_cmd(struct scsi_cmnd *scmd) case DID_OK: break; case DID_BUS_BUSY: - return blk_failfast_transport(scmd->request); + return (scmd->request->cmd_flags & REQ_FAILFAST_TRANSPORT); case DID_PARITY: - return blk_failfast_dev(scmd->request); + return (scmd->request->cmd_flags & REQ_FAILFAST_DEV); case DID_ERROR: if (msg_byte(scmd->result) == COMMAND_COMPLETE && status_byte(scmd->result) == RESERVATION_CONFLICT) return 0; /* fall through */ case DID_SOFT_ERROR: - return blk_failfast_driver(scmd->request); + return (scmd->request->cmd_flags & REQ_FAILFAST_DRIVER); } switch (status_byte(scmd->result)) { @@ -1349,7 +1349,9 @@ int scsi_noretry_cmd(struct scsi_cmnd *scmd) * assume caller has checked sense and determinted * the check condition was retryable. */ - return blk_failfast_dev(scmd->request); + if (scmd->request->cmd_flags & REQ_FAILFAST_DEV || + scmd->request->cmd_type == REQ_TYPE_BLOCK_PC) + return 1; } return 0; diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 1646fe7..b8de389 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -85,7 +85,7 @@ static void scsi_unprep_request(struct request *req) { struct scsi_cmnd *cmd = req->special; - req->cmd_flags &= ~REQ_DONTPREP; + blk_unprep_request(req); req->special = NULL; scsi_put_command(cmd); @@ -722,7 +722,7 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes) sense_deferred = scsi_sense_is_deferred(&sshdr); } - if (blk_pc_request(req)) { /* SG_IO ioctl from block level */ + if (req->cmd_type == REQ_TYPE_BLOCK_PC) { /* SG_IO ioctl from block level */ req->errors = result; if (result) { if (sense_valid && req->sense) { @@ -757,7 +757,8 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes) } } - BUG_ON(blk_bidi_rq(req)); /* bidi not support for !blk_pc_request yet */ + /* no bidi support for !REQ_TYPE_BLOCK_PC yet */ + BUG_ON(blk_bidi_rq(req)); /* * Next deal with any sectors which we were able to correctly @@ -1010,11 +1011,8 @@ int scsi_init_io(struct scsi_cmnd *cmd, gfp_t gfp_mask) err_exit: scsi_release_buffers(cmd); - if (error == BLKPREP_KILL) - scsi_put_command(cmd); - else /* BLKPREP_DEFER */ - scsi_unprep_request(cmd->request); - + scsi_put_command(cmd); + cmd->request->special = NULL; return error; } EXPORT_SYMBOL(scsi_init_io); diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index cc8a1d1..8e2e893 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -46,6 +46,7 @@ #include <linux/blkdev.h> #include <linux/blkpg.h> #include <linux/delay.h> +#include <linux/smp_lock.h> #include <linux/mutex.h> #include <linux/string_helpers.h> #include <linux/async.h> @@ -411,54 +412,85 @@ static void sd_prot_op(struct scsi_cmnd *scmd, unsigned int dif) } /** - * sd_prepare_discard - unmap blocks on thinly provisioned device + * scsi_setup_discard_cmnd - unmap blocks on thinly provisioned device + * @sdp: scsi device to operate one * @rq: Request to prepare * * Will issue either UNMAP or WRITE SAME(16) depending on preference * indicated by target device. **/ -static int sd_prepare_discard(struct request *rq) +static int scsi_setup_discard_cmnd(struct scsi_device *sdp, struct request *rq) { struct scsi_disk *sdkp = scsi_disk(rq->rq_disk); struct bio *bio = rq->bio; sector_t sector = bio->bi_sector; - unsigned int num = bio_sectors(bio); + unsigned int nr_sectors = bio_sectors(bio); + unsigned int len; + int ret; + struct page *page; if (sdkp->device->sector_size == 4096) { sector >>= 3; - num >>= 3; + nr_sectors >>= 3; } - rq->cmd_type = REQ_TYPE_BLOCK_PC; rq->timeout = SD_TIMEOUT; memset(rq->cmd, 0, rq->cmd_len); + page = alloc_page(GFP_ATOMIC | __GFP_ZERO); + if (!page) + return BLKPREP_DEFER; + if (sdkp->unmap) { - char *buf = kmap_atomic(bio_page(bio), KM_USER0); + char *buf = page_address(page); + rq->cmd_len = 10; rq->cmd[0] = UNMAP; rq->cmd[8] = 24; - rq->cmd_len = 10; - - /* Ensure that data length matches payload */ - rq->__data_len = bio->bi_size = bio->bi_io_vec->bv_len = 24; put_unaligned_be16(6 + 16, &buf[0]); put_unaligned_be16(16, &buf[2]); put_unaligned_be64(sector, &buf[8]); - put_unaligned_be32(num, &buf[16]); + put_unaligned_be32(nr_sectors, &buf[16]); - kunmap_atomic(buf, KM_USER0); + len = 24; } else { + rq->cmd_len = 16; rq->cmd[0] = WRITE_SAME_16; rq->cmd[1] = 0x8; /* UNMAP */ put_unaligned_be64(sector, &rq->cmd[2]); - put_unaligned_be32(num, &rq->cmd[10]); - rq->cmd_len = 16; + put_unaligned_be32(nr_sectors, &rq->cmd[10]); + + len = sdkp->device->sector_size; } - return BLKPREP_OK; + blk_add_request_payload(rq, page, len); + ret = scsi_setup_blk_pc_cmnd(sdp, rq); + rq->buffer = page_address(page); + if (ret != BLKPREP_OK) { + __free_page(page); + rq->buffer = NULL; + } + return ret; +} + +static int scsi_setup_flush_cmnd(struct scsi_device *sdp, struct request *rq) +{ + rq->timeout = SD_TIMEOUT; + rq->retries = SD_MAX_RETRIES; + rq->cmd[0] = SYNCHRONIZE_CACHE; + rq->cmd_len = 10; + + return scsi_setup_blk_pc_cmnd(sdp, rq); +} + +static void sd_unprep_fn(struct request_queue *q, struct request *rq) +{ + if (rq->cmd_flags & REQ_DISCARD) { + free_page((unsigned long)rq->buffer); + rq->buffer = NULL; + } } /** @@ -485,10 +517,13 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq) * Discard request come in as REQ_TYPE_FS but we turn them into * block PC requests to make life easier. */ - if (blk_discard_rq(rq)) - ret = sd_prepare_discard(rq); - - if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { + if (rq->cmd_flags & REQ_DISCARD) { + ret = scsi_setup_discard_cmnd(sdp, rq); + goto out; + } else if (rq->cmd_flags & REQ_FLUSH) { + ret = scsi_setup_flush_cmnd(sdp, rq); + goto out; + } else if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { ret = scsi_setup_blk_pc_cmnd(sdp, rq); goto out; } else if (rq->cmd_type != REQ_TYPE_FS) { @@ -636,7 +671,7 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq) SCpnt->cmnd[0] = VARIABLE_LENGTH_CMD; SCpnt->cmnd[7] = 0x18; SCpnt->cmnd[9] = (rq_data_dir(rq) == READ) ? READ_32 : WRITE_32; - SCpnt->cmnd[10] = protect | (blk_fua_rq(rq) ? 0x8 : 0); + SCpnt->cmnd[10] = protect | ((rq->cmd_flags & REQ_FUA) ? 0x8 : 0); /* LBA */ SCpnt->cmnd[12] = sizeof(block) > 4 ? (unsigned char) (block >> 56) & 0xff : 0; @@ -661,7 +696,7 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq) SCpnt->cmnd[31] = (unsigned char) this_count & 0xff; } else if (block > 0xffffffff) { SCpnt->cmnd[0] += READ_16 - READ_6; - SCpnt->cmnd[1] = protect | (blk_fua_rq(rq) ? 0x8 : 0); + SCpnt->cmnd[1] = protect | ((rq->cmd_flags & REQ_FUA) ? 0x8 : 0); SCpnt->cmnd[2] = sizeof(block) > 4 ? (unsigned char) (block >> 56) & 0xff : 0; SCpnt->cmnd[3] = sizeof(block) > 4 ? (unsigned char) (block >> 48) & 0xff : 0; SCpnt->cmnd[4] = sizeof(block) > 4 ? (unsigned char) (block >> 40) & 0xff : 0; @@ -682,7 +717,7 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq) this_count = 0xffff; SCpnt->cmnd[0] += READ_10 - READ_6; - SCpnt->cmnd[1] = protect | (blk_fua_rq(rq) ? 0x8 : 0); + SCpnt->cmnd[1] = protect | ((rq->cmd_flags & REQ_FUA) ? 0x8 : 0); SCpnt->cmnd[2] = (unsigned char) (block >> 24) & 0xff; SCpnt->cmnd[3] = (unsigned char) (block >> 16) & 0xff; SCpnt->cmnd[4] = (unsigned char) (block >> 8) & 0xff; @@ -691,7 +726,7 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq) SCpnt->cmnd[7] = (unsigned char) (this_count >> 8) & 0xff; SCpnt->cmnd[8] = (unsigned char) this_count & 0xff; } else { - if (unlikely(blk_fua_rq(rq))) { + if (unlikely(rq->cmd_flags & REQ_FUA)) { /* * This happens only if this drive failed * 10byte rw command with ILLEGAL_REQUEST @@ -745,6 +780,8 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq) * or from within the kernel (e.g. as a result of a mount(1) ). * In the latter case @inode and @filp carry an abridged amount * of information as noted above. + * + * Locking: called with bdev->bd_mutex held. **/ static int sd_open(struct block_device *bdev, fmode_t mode) { @@ -799,7 +836,7 @@ static int sd_open(struct block_device *bdev, fmode_t mode) if (!scsi_device_online(sdev)) goto error_out; - if (!sdkp->openers++ && sdev->removable) { + if ((atomic_inc_return(&sdkp->openers) == 1) && sdev->removable) { if (scsi_block_when_processing_errors(sdev)) scsi_set_medium_removal(sdev, SCSI_REMOVAL_PREVENT); } @@ -823,6 +860,8 @@ error_autopm: * * Note: may block (uninterruptible) if error recovery is underway * on this disk. + * + * Locking: called with bdev->bd_mutex held. **/ static int sd_release(struct gendisk *disk, fmode_t mode) { @@ -831,7 +870,7 @@ static int sd_release(struct gendisk *disk, fmode_t mode) SCSI_LOG_HLQUEUE(3, sd_printk(KERN_INFO, sdkp, "sd_release\n")); - if (!--sdkp->openers && sdev->removable) { + if (atomic_dec_return(&sdkp->openers) && sdev->removable) { if (scsi_block_when_processing_errors(sdev)) scsi_set_medium_removal(sdev, SCSI_REMOVAL_ALLOW); } @@ -904,7 +943,7 @@ static int sd_ioctl(struct block_device *bdev, fmode_t mode, error = scsi_nonblockable_ioctl(sdp, cmd, p, (mode & FMODE_NDELAY) != 0); if (!scsi_block_when_processing_errors(sdp) || !error) - return error; + goto out; /* * Send SCSI addressing ioctls directly to mid level, send other @@ -914,13 +953,17 @@ static int sd_ioctl(struct block_device *bdev, fmode_t mode, switch (cmd) { case SCSI_IOCTL_GET_IDLUN: case SCSI_IOCTL_GET_BUS_NUMBER: - return scsi_ioctl(sdp, cmd, p); + error = scsi_ioctl(sdp, cmd, p); + break; default: error = scsi_cmd_ioctl(disk->queue, disk, mode, cmd, p); if (error != -ENOTTY) - return error; + break; + error = scsi_ioctl(sdp, cmd, p); + break; } - return scsi_ioctl(sdp, cmd, p); +out: + return error; } static void set_media_not_present(struct scsi_disk *sdkp) @@ -1045,15 +1088,6 @@ static int sd_sync_cache(struct scsi_disk *sdkp) return 0; } -static void sd_prepare_flush(struct request_queue *q, struct request *rq) -{ - rq->cmd_type = REQ_TYPE_BLOCK_PC; - rq->timeout = SD_TIMEOUT; - rq->retries = SD_MAX_RETRIES; - rq->cmd[0] = SYNCHRONIZE_CACHE; - rq->cmd_len = 10; -} - static void sd_rescan(struct device *dev) { struct scsi_disk *sdkp = scsi_disk_get_from_dev(dev); @@ -1103,7 +1137,7 @@ static const struct block_device_operations sd_fops = { .owner = THIS_MODULE, .open = sd_open, .release = sd_release, - .locked_ioctl = sd_ioctl, + .ioctl = sd_ioctl, .getgeo = sd_getgeo, #ifdef CONFIG_COMPAT .compat_ioctl = sd_compat_ioctl, @@ -1120,7 +1154,7 @@ static unsigned int sd_completed_bytes(struct scsi_cmnd *scmd) u64 bad_lba; int info_valid; - if (!blk_fs_request(scmd->request)) + if (scmd->request->cmd_type != REQ_TYPE_FS) return 0; info_valid = scsi_get_sense_info_fld(scmd->sense_buffer, @@ -1171,6 +1205,12 @@ static int sd_done(struct scsi_cmnd *SCpnt) int sense_valid = 0; int sense_deferred = 0; + if (SCpnt->request->cmd_flags & REQ_DISCARD) { + if (!result) + scsi_set_resid(SCpnt, 0); + return good_bytes; + } + if (result) { sense_valid = scsi_command_normalize_sense(SCpnt, &sshdr); if (sense_valid) @@ -2121,7 +2161,7 @@ static int sd_revalidate_disk(struct gendisk *disk) else ordered = QUEUE_ORDERED_DRAIN; - blk_queue_ordered(sdkp->disk->queue, ordered, sd_prepare_flush); + blk_queue_ordered(sdkp->disk->queue, ordered); set_capacity(disk, sdkp->capacity); kfree(buffer); @@ -2234,6 +2274,7 @@ static void sd_probe_async(void *data, async_cookie_t cookie) sd_revalidate_disk(gd); blk_queue_prep_rq(sdp->request_queue, sd_prep_fn); + blk_queue_unprep_rq(sdp->request_queue, sd_unprep_fn); gd->driverfs_dev = &sdp->sdev_gendev; gd->flags = GENHD_FL_EXT_DEVT; @@ -2313,7 +2354,7 @@ static int sd_probe(struct device *dev) sdkp->driver = &sd_template; sdkp->disk = gd; sdkp->index = index; - sdkp->openers = 0; + atomic_set(&sdkp->openers, 0); sdkp->previous_state = 1; if (!sdp->request_queue->rq_timeout) { @@ -2372,6 +2413,7 @@ static int sd_remove(struct device *dev) async_synchronize_full(); blk_queue_prep_rq(sdkp->device->request_queue, scsi_prep_fn); + blk_queue_unprep_rq(sdkp->device->request_queue, NULL); device_del(&sdkp->dev); del_gendisk(sdkp->disk); sd_shutdown(dev); diff --git a/drivers/scsi/sd.h b/drivers/scsi/sd.h index 43d3caf..f81a930 100644 --- a/drivers/scsi/sd.h +++ b/drivers/scsi/sd.h @@ -47,7 +47,7 @@ struct scsi_disk { struct scsi_device *device; struct device dev; struct gendisk *disk; - unsigned int openers; /* protected by BKL for now, yuck */ + atomic_t openers; sector_t capacity; /* size in 512-byte sectors */ u32 index; unsigned short hw_sector_size; diff --git a/drivers/scsi/sr.c b/drivers/scsi/sr.c index 0a90abc..ba9c3e0 100644 --- a/drivers/scsi/sr.c +++ b/drivers/scsi/sr.c @@ -44,6 +44,7 @@ #include <linux/init.h> #include <linux/blkdev.h> #include <linux/mutex.h> +#include <linux/smp_lock.h> #include <linux/slab.h> #include <asm/uaccess.h> @@ -466,22 +467,27 @@ static int sr_prep_fn(struct request_queue *q, struct request *rq) static int sr_block_open(struct block_device *bdev, fmode_t mode) { - struct scsi_cd *cd = scsi_cd_get(bdev->bd_disk); + struct scsi_cd *cd; int ret = -ENXIO; + lock_kernel(); + cd = scsi_cd_get(bdev->bd_disk); if (cd) { ret = cdrom_open(&cd->cdi, bdev, mode); if (ret) scsi_cd_put(cd); } + unlock_kernel(); return ret; } static int sr_block_release(struct gendisk *disk, fmode_t mode) { struct scsi_cd *cd = scsi_cd(disk); + lock_kernel(); cdrom_release(&cd->cdi, mode); scsi_cd_put(cd); + unlock_kernel(); return 0; } @@ -493,6 +499,8 @@ static int sr_block_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, void __user *argp = (void __user *)arg; int ret; + lock_kernel(); + /* * Send SCSI addressing ioctls directly to mid level, send other * ioctls to cdrom/block level. @@ -500,12 +508,13 @@ static int sr_block_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, switch (cmd) { case SCSI_IOCTL_GET_IDLUN: case SCSI_IOCTL_GET_BUS_NUMBER: - return scsi_ioctl(sdev, cmd, argp); + ret = scsi_ioctl(sdev, cmd, argp); + goto out; } ret = cdrom_ioctl(&cd->cdi, bdev, mode, cmd, arg); if (ret != -ENOSYS) - return ret; + goto out; /* * ENODEV means that we didn't recognise the ioctl, or that we @@ -516,8 +525,12 @@ static int sr_block_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd, ret = scsi_nonblockable_ioctl(sdev, cmd, argp, (mode & FMODE_NDELAY) != 0); if (ret != -ENODEV) - return ret; - return scsi_ioctl(sdev, cmd, argp); + goto out; + ret = scsi_ioctl(sdev, cmd, argp); + +out: + unlock_kernel(); + return ret; } static int sr_block_media_changed(struct gendisk *disk) @@ -531,7 +544,7 @@ static const struct block_device_operations sr_bdops = .owner = THIS_MODULE, .open = sr_block_open, .release = sr_block_release, - .locked_ioctl = sr_block_ioctl, + .ioctl = sr_block_ioctl, .media_changed = sr_block_media_changed, /* * No compat_ioctl for now because sr_block_ioctl never diff --git a/drivers/scsi/sun3_NCR5380.c b/drivers/scsi/sun3_NCR5380.c index b5838d5..713620e 100644 --- a/drivers/scsi/sun3_NCR5380.c +++ b/drivers/scsi/sun3_NCR5380.c @@ -2022,7 +2022,7 @@ static void NCR5380_information_transfer (struct Scsi_Host *instance) if((count > SUN3_DMA_MINSIZE) && (sun3_dma_setup_done != cmd)) { - if(blk_fs_request(cmd->request)) { + if (cmd->request->cmd_type == REQ_TYPE_FS) { sun3scsi_dma_setup(d, count, rq_data_dir(cmd->request)); sun3_dma_setup_done = cmd; diff --git a/drivers/scsi/sun3_scsi.c b/drivers/scsi/sun3_scsi.c index e606cf0..613f588 100644 --- a/drivers/scsi/sun3_scsi.c +++ b/drivers/scsi/sun3_scsi.c @@ -524,7 +524,7 @@ static inline unsigned long sun3scsi_dma_xfer_len(unsigned long wanted, struct scsi_cmnd *cmd, int write_flag) { - if(blk_fs_request(cmd->request)) + if (cmd->request->cmd_type == REQ_TYPE_FS) return wanted; else return 0; diff --git a/drivers/scsi/sun3_scsi_vme.c b/drivers/scsi/sun3_scsi_vme.c index aaa4fd0..7c526b8 100644 --- a/drivers/scsi/sun3_scsi_vme.c +++ b/drivers/scsi/sun3_scsi_vme.c @@ -458,7 +458,7 @@ static inline unsigned long sun3scsi_dma_xfer_len(unsigned long wanted, struct scsi_cmnd *cmd, int write_flag) { - if(blk_fs_request(cmd->request)) + if (cmd->request->cmd_type == REQ_TYPE_FS) return wanted; else return 0; diff --git a/drivers/staging/hv/blkvsc_drv.c b/drivers/staging/hv/blkvsc_drv.c index f7ea2a3..ff1d247 100644 --- a/drivers/staging/hv/blkvsc_drv.c +++ b/drivers/staging/hv/blkvsc_drv.c @@ -25,6 +25,7 @@ #include <linux/major.h> #include <linux/delay.h> #include <linux/hdreg.h> +#include <linux/smp_lock.h> #include <linux/slab.h> #include <scsi/scsi.h> #include <scsi/scsi_cmnd.h> @@ -805,7 +806,8 @@ static void blkvsc_init_rw(struct blkvsc_request *blkvsc_req) blkvsc_req->cmnd[0] = READ_16; } - blkvsc_req->cmnd[1] |= blk_fua_rq(blkvsc_req->req) ? 0x8 : 0; + blkvsc_req->cmnd[1] |= + (blkvsc_req->req->cmd_flags & REQ_FUA) ? 0x8 : 0; *(unsigned long long *)&blkvsc_req->cmnd[2] = cpu_to_be64(blkvsc_req->sector_start); @@ -821,7 +823,8 @@ static void blkvsc_init_rw(struct blkvsc_request *blkvsc_req) blkvsc_req->cmnd[0] = READ_10; } - blkvsc_req->cmnd[1] |= blk_fua_rq(blkvsc_req->req) ? 0x8 : 0; + blkvsc_req->cmnd[1] |= + (blkvsc_req->req->cmd_flags & REQ_FUA) ? 0x8 : 0; *(unsigned int *)&blkvsc_req->cmnd[2] = cpu_to_be32(blkvsc_req->sector_start); @@ -1268,7 +1271,7 @@ static void blkvsc_request(struct request_queue *queue) DPRINT_DBG(BLKVSC_DRV, "- req %p\n", req); blkdev = req->rq_disk->private_data; - if (blkdev->shutting_down || !blk_fs_request(req) || + if (blkdev->shutting_down || req->cmd_type != REQ_TYPE_FS || blkdev->media_not_present) { __blk_end_request_cur(req, 0); continue; @@ -1306,6 +1309,7 @@ static int blkvsc_open(struct block_device *bdev, fmode_t mode) DPRINT_DBG(BLKVSC_DRV, "- users %d disk %s\n", blkdev->users, blkdev->gd->disk_name); + lock_kernel(); spin_lock(&blkdev->lock); if (!blkdev->users && blkdev->device_type == DVD_TYPE) { @@ -1317,6 +1321,7 @@ static int blkvsc_open(struct block_device *bdev, fmode_t mode) blkdev->users++; spin_unlock(&blkdev->lock); + unlock_kernel(); return 0; } @@ -1327,6 +1332,7 @@ static int blkvsc_release(struct gendisk *disk, fmode_t mode) DPRINT_DBG(BLKVSC_DRV, "- users %d disk %s\n", blkdev->users, blkdev->gd->disk_name); + lock_kernel(); spin_lock(&blkdev->lock); if (blkdev->users == 1) { spin_unlock(&blkdev->lock); @@ -1337,6 +1343,7 @@ static int blkvsc_release(struct gendisk *disk, fmode_t mode) blkdev->users--; spin_unlock(&blkdev->lock); + unlock_kernel(); return 0; } diff --git a/drivers/xen/xenbus/xenbus_client.c b/drivers/xen/xenbus/xenbus_client.c index 7b3e973..7e49527 100644 --- a/drivers/xen/xenbus/xenbus_client.c +++ b/drivers/xen/xenbus/xenbus_client.c @@ -133,17 +133,12 @@ int xenbus_watch_pathfmt(struct xenbus_device *dev, } EXPORT_SYMBOL_GPL(xenbus_watch_pathfmt); +static void xenbus_switch_fatal(struct xenbus_device *, int, int, + const char *, ...); -/** - * xenbus_switch_state - * @dev: xenbus device - * @state: new state - * - * Advertise in the store a change of the given driver to the given new_state. - * Return 0 on success, or -errno on error. On error, the device will switch - * to XenbusStateClosing, and the error will be saved in the store. - */ -int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state state) +static int +__xenbus_switch_state(struct xenbus_device *dev, + enum xenbus_state state, int depth) { /* We check whether the state is currently set to the given value, and if not, then the state is set. We don't want to unconditionally @@ -152,35 +147,65 @@ int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state state) to it, as the device will be tearing down, and we don't want to resurrect that directory. - Note that, because of this cached value of our state, this function - will not work inside a Xenstore transaction (something it was - trying to in the past) because dev->state would not get reset if - the transaction was aborted. - + Note that, because of this cached value of our state, this + function will not take a caller's Xenstore transaction + (something it was trying to in the past) because dev->state + would not get reset if the transaction was aborted. */ + struct xenbus_transaction xbt; int current_state; - int err; + int err, abort; if (state == dev->state) return 0; - err = xenbus_scanf(XBT_NIL, dev->nodename, "state", "%d", - ¤t_state); - if (err != 1) +again: + abort = 1; + + err = xenbus_transaction_start(&xbt); + if (err) { + xenbus_switch_fatal(dev, depth, err, "starting transaction"); return 0; + } + + err = xenbus_scanf(xbt, dev->nodename, "state", "%d", ¤t_state); + if (err != 1) + goto abort; - err = xenbus_printf(XBT_NIL, dev->nodename, "state", "%d", state); + err = xenbus_printf(xbt, dev->nodename, "state", "%d", state); if (err) { - if (state != XenbusStateClosing) /* Avoid looping */ - xenbus_dev_fatal(dev, err, "writing new state"); - return err; + xenbus_switch_fatal(dev, depth, err, "writing new state"); + goto abort; } - dev->state = state; + abort = 0; +abort: + err = xenbus_transaction_end(xbt, abort); + if (err) { + if (err == -EAGAIN && !abort) + goto again; + xenbus_switch_fatal(dev, depth, err, "ending transaction"); + } else + dev->state = state; return 0; } + +/** + * xenbus_switch_state + * @dev: xenbus device + * @state: new state + * + * Advertise in the store a change of the given driver to the given new_state. + * Return 0 on success, or -errno on error. On error, the device will switch + * to XenbusStateClosing, and the error will be saved in the store. + */ +int xenbus_switch_state(struct xenbus_device *dev, enum xenbus_state state) +{ + return __xenbus_switch_state(dev, state, 0); +} + EXPORT_SYMBOL_GPL(xenbus_switch_state); int xenbus_frontend_closed(struct xenbus_device *dev) @@ -284,6 +309,23 @@ void xenbus_dev_fatal(struct xenbus_device *dev, int err, const char *fmt, ...) EXPORT_SYMBOL_GPL(xenbus_dev_fatal); /** + * Equivalent to xenbus_dev_fatal(dev, err, fmt, args), but helps + * avoiding recursion within xenbus_switch_state. + */ +static void xenbus_switch_fatal(struct xenbus_device *dev, int depth, int err, + const char *fmt, ...) +{ + va_list ap; + + va_start(ap, fmt); + xenbus_va_dev_error(dev, err, fmt, ap); + va_end(ap); + + if (!depth) + __xenbus_switch_state(dev, XenbusStateClosing, 1); +} + +/** * xenbus_grant_ring * @dev: xenbus device * @ring_mfn: mfn of ring to grant @@ -843,7 +843,8 @@ struct bio *bio_copy_user_iov(struct request_queue *q, if (!bio) goto out_bmd; - bio->bi_rw |= (!write_to_vm << BIO_RW); + if (!write_to_vm) + bio->bi_rw |= REQ_WRITE; ret = 0; @@ -1024,7 +1025,7 @@ static struct bio *__bio_map_user_iov(struct request_queue *q, * set data direction, and check if mapped pages need bouncing */ if (!write_to_vm) - bio->bi_rw |= (1 << BIO_RW); + bio->bi_rw |= REQ_WRITE; bio->bi_bdev = bdev; bio->bi_flags |= (1 << BIO_USER_MAPPED); diff --git a/fs/block_dev.c b/fs/block_dev.c index 451afbd..6641146 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1346,13 +1346,12 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) return ret; } - lock_kernel(); restart: ret = -ENXIO; disk = get_gendisk(bdev->bd_dev, &partno); if (!disk) - goto out_unlock_kernel; + goto out; mutex_lock_nested(&bdev->bd_mutex, for_part); if (!bdev->bd_openers) { @@ -1432,7 +1431,6 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) if (for_part) bdev->bd_part_count++; mutex_unlock(&bdev->bd_mutex); - unlock_kernel(); return 0; out_clear: @@ -1445,9 +1443,7 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) bdev->bd_contains = NULL; out_unlock_bdev: mutex_unlock(&bdev->bd_mutex); - out_unlock_kernel: - unlock_kernel(); - + out: if (disk) module_put(disk->fops->owner); put_disk(disk); @@ -1516,7 +1512,6 @@ static int __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part) struct block_device *victim = NULL; mutex_lock_nested(&bdev->bd_mutex, for_part); - lock_kernel(); if (for_part) bdev->bd_part_count--; @@ -1541,7 +1536,6 @@ static int __blkdev_put(struct block_device *bdev, fmode_t mode, int for_part) victim = bdev->bd_contains; bdev->bd_contains = NULL; } - unlock_kernel(); mutex_unlock(&bdev->bd_mutex); bdput(bdev); if (victim) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 34f7c37..64f1008 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -480,7 +480,7 @@ static void end_workqueue_bio(struct bio *bio, int err) end_io_wq->work.func = end_workqueue_fn; end_io_wq->work.flags = 0; - if (bio->bi_rw & (1 << BIO_RW)) { + if (bio->bi_rw & REQ_WRITE) { if (end_io_wq->metadata) btrfs_queue_worker(&fs_info->endio_meta_write_workers, &end_io_wq->work); @@ -604,7 +604,7 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode, atomic_inc(&fs_info->nr_async_submits); - if (rw & (1 << BIO_RW_SYNCIO)) + if (rw & REQ_SYNC) btrfs_set_work_high_prio(&async->work); btrfs_queue_worker(&fs_info->workers, &async->work); @@ -668,7 +668,7 @@ static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio, bio, 1); BUG_ON(ret); - if (!(rw & (1 << BIO_RW))) { + if (!(rw & REQ_WRITE)) { /* * called for a read, do the setup so that checksum validation * can happen in the async kernel threads @@ -1427,7 +1427,7 @@ static void end_workqueue_fn(struct btrfs_work *work) * ram and up to date before trying to verify things. For * blocksize <= pagesize, it is basically a noop */ - if (!(bio->bi_rw & (1 << BIO_RW)) && end_io_wq->metadata && + if (!(bio->bi_rw & REQ_WRITE) && end_io_wq->metadata && !bio_ready_for_csum(bio)) { btrfs_queue_worker(&fs_info->endio_meta_workers, &end_io_wq->work); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 8976c33..c038644 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1429,7 +1429,7 @@ static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio, ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0); BUG_ON(ret); - if (!(rw & (1 << BIO_RW))) { + if (!(rw & REQ_WRITE)) { if (bio_flags & EXTENT_BIO_COMPRESSED) { return btrfs_submit_compressed_read(inode, bio, mirror_num, bio_flags); @@ -1841,7 +1841,7 @@ static int btrfs_io_failed_hook(struct bio *failed_bio, bio->bi_size = 0; bio_add_page(bio, page, failrec->len, start - page_offset(page)); - if (failed_bio->bi_rw & (1 << BIO_RW)) + if (failed_bio->bi_rw & REQ_WRITE) rw = WRITE; else rw = READ; @@ -5647,7 +5647,7 @@ static void btrfs_submit_direct(int rw, struct bio *bio, struct inode *inode, struct bio_vec *bvec = bio->bi_io_vec; u64 start; int skip_sum; - int write = rw & (1 << BIO_RW); + int write = rw & REQ_WRITE; int ret = 0; skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index d6e3af8..dd318ff 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -258,7 +258,7 @@ loop_lock: BUG_ON(atomic_read(&cur->bi_cnt) == 0); - if (bio_rw_flagged(cur, BIO_RW_SYNCIO)) + if (cur->bi_rw & REQ_SYNC) num_sync_run++; submit_bio(cur->bi_rw, cur); @@ -2651,7 +2651,7 @@ static int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, int max_errors = 0; struct btrfs_multi_bio *multi = NULL; - if (multi_ret && !(rw & (1 << BIO_RW))) + if (multi_ret && !(rw & REQ_WRITE)) stripes_allocated = 1; again: if (multi_ret) { @@ -2687,7 +2687,7 @@ again: mirror_num = 0; /* if our multi bio struct is too small, back off and try again */ - if (rw & (1 << BIO_RW)) { + if (rw & REQ_WRITE) { if (map->type & (BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_DUP)) { stripes_required = map->num_stripes; @@ -2697,7 +2697,7 @@ again: max_errors = 1; } } - if (multi_ret && (rw & (1 << BIO_RW)) && + if (multi_ret && (rw & REQ_WRITE) && stripes_allocated < stripes_required) { stripes_allocated = map->num_stripes; free_extent_map(em); @@ -2733,7 +2733,7 @@ again: num_stripes = 1; stripe_index = 0; if (map->type & BTRFS_BLOCK_GROUP_RAID1) { - if (unplug_page || (rw & (1 << BIO_RW))) + if (unplug_page || (rw & REQ_WRITE)) num_stripes = map->num_stripes; else if (mirror_num) stripe_index = mirror_num - 1; @@ -2744,7 +2744,7 @@ again: } } else if (map->type & BTRFS_BLOCK_GROUP_DUP) { - if (rw & (1 << BIO_RW)) + if (rw & REQ_WRITE) num_stripes = map->num_stripes; else if (mirror_num) stripe_index = mirror_num - 1; @@ -2755,7 +2755,7 @@ again: stripe_index = do_div(stripe_nr, factor); stripe_index *= map->sub_stripes; - if (unplug_page || (rw & (1 << BIO_RW))) + if (unplug_page || (rw & REQ_WRITE)) num_stripes = map->sub_stripes; else if (mirror_num) stripe_index += mirror_num - 1; @@ -2945,7 +2945,7 @@ static noinline int schedule_bio(struct btrfs_root *root, struct btrfs_pending_bios *pending_bios; /* don't bother with additional async steps for reads, right now */ - if (!(rw & (1 << BIO_RW))) { + if (!(rw & REQ_WRITE)) { bio_get(bio); submit_bio(rw, bio); bio_put(bio); @@ -2964,7 +2964,7 @@ static noinline int schedule_bio(struct btrfs_root *root, bio->bi_rw |= rw; spin_lock(&device->io_lock); - if (bio_rw_flagged(bio, BIO_RW_SYNCIO)) + if (bio->bi_rw & REQ_SYNC) pending_bios = &device->pending_sync_bios; else pending_bios = &device->pending_bios; diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c index 66b9cf7..de89645 100644 --- a/fs/coda/psdev.c +++ b/fs/coda/psdev.c @@ -177,7 +177,7 @@ static ssize_t coda_psdev_write(struct file *file, const char __user *buf, nbytes = req->uc_outSize; /* don't have more space! */ } if (copy_from_user(req->uc_data, buf, nbytes)) { - req->uc_flags |= REQ_ABORT; + req->uc_flags |= CODA_REQ_ABORT; wake_up(&req->uc_sleep); retval = -EFAULT; goto out; @@ -254,8 +254,8 @@ static ssize_t coda_psdev_read(struct file * file, char __user * buf, retval = -EFAULT; /* If request was not a signal, enqueue and don't free */ - if (!(req->uc_flags & REQ_ASYNC)) { - req->uc_flags |= REQ_READ; + if (!(req->uc_flags & CODA_REQ_ASYNC)) { + req->uc_flags |= CODA_REQ_READ; list_add_tail(&(req->uc_chain), &vcp->vc_processing); goto out; } @@ -315,19 +315,19 @@ static int coda_psdev_release(struct inode * inode, struct file * file) list_del(&req->uc_chain); /* Async requests need to be freed here */ - if (req->uc_flags & REQ_ASYNC) { + if (req->uc_flags & CODA_REQ_ASYNC) { CODA_FREE(req->uc_data, sizeof(struct coda_in_hdr)); kfree(req); continue; } - req->uc_flags |= REQ_ABORT; + req->uc_flags |= CODA_REQ_ABORT; wake_up(&req->uc_sleep); } list_for_each_entry_safe(req, tmp, &vcp->vc_processing, uc_chain) { list_del(&req->uc_chain); - req->uc_flags |= REQ_ABORT; + req->uc_flags |= CODA_REQ_ABORT; wake_up(&req->uc_sleep); } diff --git a/fs/coda/upcall.c b/fs/coda/upcall.c index f09c5ed..b8893ab 100644 --- a/fs/coda/upcall.c +++ b/fs/coda/upcall.c @@ -604,7 +604,7 @@ static void coda_unblock_signals(sigset_t *old) (((r)->uc_opcode != CODA_CLOSE && \ (r)->uc_opcode != CODA_STORE && \ (r)->uc_opcode != CODA_RELEASE) || \ - (r)->uc_flags & REQ_READ)) + (r)->uc_flags & CODA_REQ_READ)) static inline void coda_waitfor_upcall(struct upc_req *req) { @@ -624,7 +624,7 @@ static inline void coda_waitfor_upcall(struct upc_req *req) set_current_state(TASK_UNINTERRUPTIBLE); /* got a reply */ - if (req->uc_flags & (REQ_WRITE | REQ_ABORT)) + if (req->uc_flags & (CODA_REQ_WRITE | CODA_REQ_ABORT)) break; if (blocked && time_after(jiffies, timeout) && @@ -708,7 +708,7 @@ static int coda_upcall(struct venus_comm *vcp, coda_waitfor_upcall(req); /* Op went through, interrupt or not... */ - if (req->uc_flags & REQ_WRITE) { + if (req->uc_flags & CODA_REQ_WRITE) { out = (union outputArgs *)req->uc_data; /* here we map positive Venus errors to kernel errors */ error = -out->oh.result; @@ -717,13 +717,13 @@ static int coda_upcall(struct venus_comm *vcp, } error = -EINTR; - if ((req->uc_flags & REQ_ABORT) || !signal_pending(current)) { + if ((req->uc_flags & CODA_REQ_ABORT) || !signal_pending(current)) { printk(KERN_WARNING "coda: Unexpected interruption.\n"); goto exit; } /* Interrupted before venus read it. */ - if (!(req->uc_flags & REQ_READ)) + if (!(req->uc_flags & CODA_REQ_READ)) goto exit; /* Venus saw the upcall, make sure we can send interrupt signal */ @@ -747,7 +747,7 @@ static int coda_upcall(struct venus_comm *vcp, sig_inputArgs->ih.opcode = CODA_SIGNAL; sig_inputArgs->ih.unique = req->uc_unique; - sig_req->uc_flags = REQ_ASYNC; + sig_req->uc_flags = CODA_REQ_ASYNC; sig_req->uc_opcode = sig_inputArgs->ih.opcode; sig_req->uc_unique = sig_inputArgs->ih.unique; sig_req->uc_inSize = sizeof(struct coda_in_hdr); diff --git a/fs/exofs/ios.c b/fs/exofs/ios.c index 4337cad..e273220 100644 --- a/fs/exofs/ios.c +++ b/fs/exofs/ios.c @@ -599,7 +599,7 @@ static int _sbi_write_mirror(struct exofs_io_state *ios, int cur_comp) } else { bio = master_dev->bio; /* FIXME: bio_set_dir() */ - bio->bi_rw |= (1 << BIO_RW); + bio->bi_rw |= REQ_WRITE; } osd_req_write(or, &ios->obj, per_dev->offset, bio, diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index b7c7586..2f76c4a 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -26,15 +26,9 @@ #include <linux/blkdev.h> #include <linux/backing-dev.h> #include <linux/buffer_head.h> +#include <linux/tracepoint.h> #include "internal.h" -#define inode_to_bdi(inode) ((inode)->i_mapping->backing_dev_info) - -/* - * We don't actually have pdflush, but this one is exported though /proc... - */ -int nr_pdflush_threads; - /* * Passed into wb_writeback(), essentially a subset of writeback_control */ @@ -50,6 +44,21 @@ struct wb_writeback_work { struct completion *done; /* set if the caller waits */ }; +/* + * Include the creation of the trace points after defining the + * wb_writeback_work structure so that the definition remains local to this + * file. + */ +#define CREATE_TRACE_POINTS +#include <trace/events/writeback.h> + +#define inode_to_bdi(inode) ((inode)->i_mapping->backing_dev_info) + +/* + * We don't actually have pdflush, but this one is exported though /proc... + */ +int nr_pdflush_threads; + /** * writeback_in_progress - determine whether there is writeback in progress * @bdi: the device's backing_dev_info structure. @@ -65,22 +74,21 @@ int writeback_in_progress(struct backing_dev_info *bdi) static void bdi_queue_work(struct backing_dev_info *bdi, struct wb_writeback_work *work) { - spin_lock(&bdi->wb_lock); - list_add_tail(&work->list, &bdi->work_list); - spin_unlock(&bdi->wb_lock); + trace_writeback_queue(bdi, work); - /* - * If the default thread isn't there, make sure we add it. When - * it gets created and wakes up, we'll run this work. - */ - if (unlikely(list_empty_careful(&bdi->wb_list))) + spin_lock_bh(&bdi->wb_lock); + list_add_tail(&work->list, &bdi->work_list); + if (bdi->wb.task) { + wake_up_process(bdi->wb.task); + } else { + /* + * The bdi thread isn't there, wake up the forker thread which + * will create and run it. + */ + trace_writeback_nothread(bdi, work); wake_up_process(default_backing_dev_info.wb.task); - else { - struct bdi_writeback *wb = &bdi->wb; - - if (wb->task) - wake_up_process(wb->task); } + spin_unlock_bh(&bdi->wb_lock); } static void @@ -95,8 +103,10 @@ __bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages, */ work = kzalloc(sizeof(*work), GFP_ATOMIC); if (!work) { - if (bdi->wb.task) + if (bdi->wb.task) { + trace_writeback_nowork(bdi); wake_up_process(bdi->wb.task); + } return; } @@ -643,10 +653,14 @@ static long wb_writeback(struct bdi_writeback *wb, wbc.more_io = 0; wbc.nr_to_write = MAX_WRITEBACK_PAGES; wbc.pages_skipped = 0; + + trace_wbc_writeback_start(&wbc, wb->bdi); if (work->sb) __writeback_inodes_sb(work->sb, wb, &wbc); else writeback_inodes_wb(wb, &wbc); + trace_wbc_writeback_written(&wbc, wb->bdi); + work->nr_pages -= MAX_WRITEBACK_PAGES - wbc.nr_to_write; wrote += MAX_WRITEBACK_PAGES - wbc.nr_to_write; @@ -674,6 +688,7 @@ static long wb_writeback(struct bdi_writeback *wb, if (!list_empty(&wb->b_more_io)) { inode = list_entry(wb->b_more_io.prev, struct inode, i_list); + trace_wbc_writeback_wait(&wbc, wb->bdi); inode_wait_for_writeback(inode); } spin_unlock(&inode_lock); @@ -686,17 +701,17 @@ static long wb_writeback(struct bdi_writeback *wb, * Return the next wb_writeback_work struct that hasn't been processed yet. */ static struct wb_writeback_work * -get_next_work_item(struct backing_dev_info *bdi, struct bdi_writeback *wb) +get_next_work_item(struct backing_dev_info *bdi) { struct wb_writeback_work *work = NULL; - spin_lock(&bdi->wb_lock); + spin_lock_bh(&bdi->wb_lock); if (!list_empty(&bdi->work_list)) { work = list_entry(bdi->work_list.next, struct wb_writeback_work, list); list_del_init(&work->list); } - spin_unlock(&bdi->wb_lock); + spin_unlock_bh(&bdi->wb_lock); return work; } @@ -744,7 +759,7 @@ long wb_do_writeback(struct bdi_writeback *wb, int force_wait) struct wb_writeback_work *work; long wrote = 0; - while ((work = get_next_work_item(bdi, wb)) != NULL) { + while ((work = get_next_work_item(bdi)) != NULL) { /* * Override sync mode, in case we must wait for completion * because this thread is exiting now. @@ -752,6 +767,8 @@ long wb_do_writeback(struct bdi_writeback *wb, int force_wait) if (force_wait) work->sync_mode = WB_SYNC_ALL; + trace_writeback_exec(bdi, work); + wrote += wb_writeback(wb, work); /* @@ -776,47 +793,66 @@ long wb_do_writeback(struct bdi_writeback *wb, int force_wait) * Handle writeback of dirty data for the device backed by this bdi. Also * wakes up periodically and does kupdated style flushing. */ -int bdi_writeback_task(struct bdi_writeback *wb) +int bdi_writeback_thread(void *data) { - unsigned long last_active = jiffies; - unsigned long wait_jiffies = -1UL; + struct bdi_writeback *wb = data; + struct backing_dev_info *bdi = wb->bdi; long pages_written; + current->flags |= PF_FLUSHER | PF_SWAPWRITE; + set_freezable(); + wb->last_active = jiffies; + + /* + * Our parent may run at a different priority, just set us to normal + */ + set_user_nice(current, 0); + + trace_writeback_thread_start(bdi); + while (!kthread_should_stop()) { + /* + * Remove own delayed wake-up timer, since we are already awake + * and we'll take care of the preriodic write-back. + */ + del_timer(&wb->wakeup_timer); + pages_written = wb_do_writeback(wb, 0); + trace_writeback_pages_written(pages_written); + if (pages_written) - last_active = jiffies; - else if (wait_jiffies != -1UL) { - unsigned long max_idle; + wb->last_active = jiffies; - /* - * Longest period of inactivity that we tolerate. If we - * see dirty data again later, the task will get - * recreated automatically. - */ - max_idle = max(5UL * 60 * HZ, wait_jiffies); - if (time_after(jiffies, max_idle + last_active)) - break; + set_current_state(TASK_INTERRUPTIBLE); + if (!list_empty(&bdi->work_list)) { + __set_current_state(TASK_RUNNING); + continue; } - if (dirty_writeback_interval) { - wait_jiffies = msecs_to_jiffies(dirty_writeback_interval * 10); - schedule_timeout_interruptible(wait_jiffies); - } else { - set_current_state(TASK_INTERRUPTIBLE); - if (list_empty_careful(&wb->bdi->work_list) && - !kthread_should_stop()) - schedule(); - __set_current_state(TASK_RUNNING); + if (wb_has_dirty_io(wb) && dirty_writeback_interval) + schedule_timeout(msecs_to_jiffies(dirty_writeback_interval * 10)); + else { + /* + * We have nothing to do, so can go sleep without any + * timeout and save power. When a work is queued or + * something is made dirty - we will be woken up. + */ + schedule(); } try_to_freeze(); } + /* Flush any work that raced with us exiting */ + if (!list_empty(&bdi->work_list)) + wb_do_writeback(wb, 1); + + trace_writeback_thread_stop(bdi); return 0; } + /* * Start writeback of `nr_pages' pages. If `nr_pages' is zero, write back * the whole world. @@ -891,6 +927,8 @@ static noinline void block_dump___mark_inode_dirty(struct inode *inode) void __mark_inode_dirty(struct inode *inode, int flags) { struct super_block *sb = inode->i_sb; + struct backing_dev_info *bdi = NULL; + bool wakeup_bdi = false; /* * Don't do this for I_DIRTY_PAGES - that doesn't actually @@ -944,22 +982,31 @@ void __mark_inode_dirty(struct inode *inode, int flags) * reposition it (that would break b_dirty time-ordering). */ if (!was_dirty) { - struct bdi_writeback *wb = &inode_to_bdi(inode)->wb; - struct backing_dev_info *bdi = wb->bdi; - - if (bdi_cap_writeback_dirty(bdi) && - !test_bit(BDI_registered, &bdi->state)) { - WARN_ON(1); - printk(KERN_ERR "bdi-%s not registered\n", - bdi->name); + bdi = inode_to_bdi(inode); + + if (bdi_cap_writeback_dirty(bdi)) { + WARN(!test_bit(BDI_registered, &bdi->state), + "bdi-%s not registered\n", bdi->name); + + /* + * If this is the first dirty inode for this + * bdi, we have to wake-up the corresponding + * bdi thread to make sure background + * write-back happens later. + */ + if (!wb_has_dirty_io(&bdi->wb)) + wakeup_bdi = true; } inode->dirtied_when = jiffies; - list_move(&inode->i_list, &wb->b_dirty); + list_move(&inode->i_list, &bdi->wb.b_dirty); } } out: spin_unlock(&inode_lock); + + if (wakeup_bdi) + bdi_wakeup_thread_delayed(bdi); } EXPORT_SYMBOL(__mark_inode_dirty); diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index 6a857e2..cde1248 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -595,7 +595,7 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull) if (test_bit(SDF_NOBARRIERS, &sdp->sd_flags)) goto skip_barrier; get_bh(bh); - submit_bh(WRITE_SYNC | (1 << BIO_RW_BARRIER) | (1 << BIO_RW_META), bh); + submit_bh(WRITE_BARRIER | REQ_META, bh); wait_on_buffer(bh); if (buffer_eopnotsupp(bh)) { clear_buffer_eopnotsupp(bh); @@ -605,7 +605,7 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull) lock_buffer(bh); skip_barrier: get_bh(bh); - submit_bh(WRITE_SYNC | (1 << BIO_RW_META), bh); + submit_bh(WRITE_SYNC | REQ_META, bh); wait_on_buffer(bh); } if (!buffer_uptodate(bh)) diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c index 18176d0..f3b071f 100644 --- a/fs/gfs2/meta_io.c +++ b/fs/gfs2/meta_io.c @@ -36,8 +36,8 @@ static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wb { struct buffer_head *bh, *head; int nr_underway = 0; - int write_op = (1 << BIO_RW_META) | ((wbc->sync_mode == WB_SYNC_ALL ? - WRITE_SYNC_PLUG : WRITE)); + int write_op = REQ_META | + (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC_PLUG : WRITE); BUG_ON(!PageLocked(page)); BUG_ON(!page_has_buffers(page)); @@ -225,7 +225,7 @@ int gfs2_meta_read(struct gfs2_glock *gl, u64 blkno, int flags, } bh->b_end_io = end_buffer_read_sync; get_bh(bh); - submit_bh(READ_SYNC | (1 << BIO_RW_META), bh); + submit_bh(READ_SYNC | REQ_META, bh); if (!(flags & DIO_WAIT)) return 0; @@ -432,7 +432,7 @@ struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen) if (buffer_uptodate(first_bh)) goto out; if (!buffer_locked(first_bh)) - ll_rw_block(READ_SYNC | (1 << BIO_RW_META), 1, &first_bh); + ll_rw_block(READ_SYNC | REQ_META, 1, &first_bh); dblock++; extlen--; diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 4f44bde..4d4b1e8 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -274,7 +274,7 @@ static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector) bio->bi_end_io = end_bio_io_page; bio->bi_private = page; - submit_bio(READ_SYNC | (1 << BIO_RW_META), bio); + submit_bio(READ_SYNC | REQ_META, bio); wait_on_page_locked(page); bio_put(bio); if (!PageUptodate(page)) { diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c index 2e6a272..4588fb9 100644 --- a/fs/nilfs2/segbuf.c +++ b/fs/nilfs2/segbuf.c @@ -508,7 +508,7 @@ static int nilfs_segbuf_write(struct nilfs_segment_buffer *segbuf, * Last BIO is always sent through the following * submission. */ - rw |= (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG); + rw |= REQ_SYNC | REQ_UNPLUG; res = nilfs_segbuf_submit_bio(segbuf, &wi, rw); } diff --git a/fs/splice.c b/fs/splice.c index efdbfec..8f1dfae 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -399,17 +399,7 @@ __generic_file_splice_read(struct file *in, loff_t *ppos, * If the page isn't uptodate, we may need to start io on it */ if (!PageUptodate(page)) { - /* - * If in nonblock mode then dont block on waiting - * for an in-flight io page - */ - if (flags & SPLICE_F_NONBLOCK) { - if (!trylock_page(page)) { - error = -EAGAIN; - break; - } - } else - lock_page(page); + lock_page(page); /* * Page was truncated, or invalidated by the @@ -597,7 +587,6 @@ ssize_t default_file_splice_read(struct file *in, loff_t *ppos, struct page *pages[PIPE_DEF_BUFFERS]; struct partial_page partial[PIPE_DEF_BUFFERS]; struct iovec *vec, __vec[PIPE_DEF_BUFFERS]; - pgoff_t index; ssize_t res; size_t this_len; int error; @@ -621,7 +610,6 @@ ssize_t default_file_splice_read(struct file *in, loff_t *ppos, goto shrink_ret; } - index = *ppos >> PAGE_CACHE_SHIFT; offset = *ppos & ~PAGE_CACHE_MASK; nr_pages = (len + offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; diff --git a/include/linux/Kbuild b/include/linux/Kbuild index 2547daf..9d65d4d 100644 --- a/include/linux/Kbuild +++ b/include/linux/Kbuild @@ -39,6 +39,7 @@ header-y += ax25.h header-y += b1lli.h header-y += baycom.h header-y += bfs_fs.h +header-y += blk_types.h header-y += blkpg.h header-y += bpqether.h header-y += bsg.h diff --git a/include/linux/audit.h b/include/linux/audit.h index f391d45..e24afab 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -544,7 +544,7 @@ extern int audit_signals; #define audit_putname(n) do { ; } while (0) #define __audit_inode(n,d) do { ; } while (0) #define __audit_inode_child(i,p) do { ; } while (0) -#define audit_inode(n,d) do { ; } while (0) +#define audit_inode(n,d) do { (void)(d); } while (0) #define audit_inode_child(i,p) do { ; } while (0) #define audit_core_dumps(i) do { ; } while (0) #define auditsc_get_stamp(c,t,s) (0) diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h index e9aec0d..7628219 100644 --- a/include/linux/backing-dev.h +++ b/include/linux/backing-dev.h @@ -45,22 +45,21 @@ enum bdi_stat_item { #define BDI_STAT_BATCH (8*(1+ilog2(nr_cpu_ids))) struct bdi_writeback { - struct list_head list; /* hangs off the bdi */ - - struct backing_dev_info *bdi; /* our parent bdi */ + struct backing_dev_info *bdi; /* our parent bdi */ unsigned int nr; - unsigned long last_old_flush; /* last old data flush */ + unsigned long last_old_flush; /* last old data flush */ + unsigned long last_active; /* last time bdi thread was active */ - struct task_struct *task; /* writeback task */ - struct list_head b_dirty; /* dirty inodes */ - struct list_head b_io; /* parked for writeback */ - struct list_head b_more_io; /* parked for more writeback */ + struct task_struct *task; /* writeback thread */ + struct timer_list wakeup_timer; /* used for delayed bdi thread wakeup */ + struct list_head b_dirty; /* dirty inodes */ + struct list_head b_io; /* parked for writeback */ + struct list_head b_more_io; /* parked for more writeback */ }; struct backing_dev_info { struct list_head bdi_list; - struct rcu_head rcu_head; unsigned long ra_pages; /* max readahead in PAGE_CACHE_SIZE units */ unsigned long state; /* Always use atomic bitops on this */ unsigned int capabilities; /* Device capabilities */ @@ -80,8 +79,7 @@ struct backing_dev_info { unsigned int max_ratio, max_prop_frac; struct bdi_writeback wb; /* default writeback info for this bdi */ - spinlock_t wb_lock; /* protects update side of wb_list */ - struct list_head wb_list; /* the flusher threads hanging off this bdi */ + spinlock_t wb_lock; /* protects work_list */ struct list_head work_list; @@ -105,9 +103,10 @@ void bdi_unregister(struct backing_dev_info *bdi); int bdi_setup_and_register(struct backing_dev_info *, char *, unsigned int); void bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages); void bdi_start_background_writeback(struct backing_dev_info *bdi); -int bdi_writeback_task(struct bdi_writeback *wb); +int bdi_writeback_thread(void *data); int bdi_has_dirty_io(struct backing_dev_info *bdi); void bdi_arm_supers_timer(void); +void bdi_wakeup_thread_delayed(struct backing_dev_info *bdi); extern spinlock_t bdi_lock; extern struct list_head bdi_list; diff --git a/include/linux/bio.h b/include/linux/bio.h index 7fc5606..5274103 100644 --- a/include/linux/bio.h +++ b/include/linux/bio.h @@ -9,7 +9,7 @@ * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of - + * * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * @@ -28,6 +28,9 @@ #include <asm/io.h> +/* struct bio, bio_vec and BIO_* flags are defined in blk_types.h */ +#include <linux/blk_types.h> + #define BIO_DEBUG #ifdef BIO_DEBUG @@ -41,154 +44,6 @@ #define BIO_MAX_SECTORS (BIO_MAX_SIZE >> 9) /* - * was unsigned short, but we might as well be ready for > 64kB I/O pages - */ -struct bio_vec { - struct page *bv_page; - unsigned int bv_len; - unsigned int bv_offset; -}; - -struct bio_set; -struct bio; -struct bio_integrity_payload; -typedef void (bio_end_io_t) (struct bio *, int); -typedef void (bio_destructor_t) (struct bio *); - -/* - * main unit of I/O for the block layer and lower layers (ie drivers and - * stacking drivers) - */ -struct bio { - sector_t bi_sector; /* device address in 512 byte - sectors */ - struct bio *bi_next; /* request queue link */ - struct block_device *bi_bdev; - unsigned long bi_flags; /* status, command, etc */ - unsigned long bi_rw; /* bottom bits READ/WRITE, - * top bits priority - */ - - unsigned short bi_vcnt; /* how many bio_vec's */ - unsigned short bi_idx; /* current index into bvl_vec */ - - /* Number of segments in this BIO after - * physical address coalescing is performed. - */ - unsigned int bi_phys_segments; - - unsigned int bi_size; /* residual I/O count */ - - /* - * To keep track of the max segment size, we account for the - * sizes of the first and last mergeable segments in this bio. - */ - unsigned int bi_seg_front_size; - unsigned int bi_seg_back_size; - - unsigned int bi_max_vecs; /* max bvl_vecs we can hold */ - - unsigned int bi_comp_cpu; /* completion CPU */ - - atomic_t bi_cnt; /* pin count */ - - struct bio_vec *bi_io_vec; /* the actual vec list */ - - bio_end_io_t *bi_end_io; - - void *bi_private; -#if defined(CONFIG_BLK_DEV_INTEGRITY) - struct bio_integrity_payload *bi_integrity; /* data integrity */ -#endif - - bio_destructor_t *bi_destructor; /* destructor */ - - /* - * We can inline a number of vecs at the end of the bio, to avoid - * double allocations for a small number of bio_vecs. This member - * MUST obviously be kept at the very end of the bio. - */ - struct bio_vec bi_inline_vecs[0]; -}; - -/* - * bio flags - */ -#define BIO_UPTODATE 0 /* ok after I/O completion */ -#define BIO_RW_BLOCK 1 /* RW_AHEAD set, and read/write would block */ -#define BIO_EOF 2 /* out-out-bounds error */ -#define BIO_SEG_VALID 3 /* bi_phys_segments valid */ -#define BIO_CLONED 4 /* doesn't own data */ -#define BIO_BOUNCED 5 /* bio is a bounce bio */ -#define BIO_USER_MAPPED 6 /* contains user pages */ -#define BIO_EOPNOTSUPP 7 /* not supported */ -#define BIO_CPU_AFFINE 8 /* complete bio on same CPU as submitted */ -#define BIO_NULL_MAPPED 9 /* contains invalid user pages */ -#define BIO_FS_INTEGRITY 10 /* fs owns integrity data, not block layer */ -#define BIO_QUIET 11 /* Make BIO Quiet */ -#define bio_flagged(bio, flag) ((bio)->bi_flags & (1 << (flag))) - -/* - * top 4 bits of bio flags indicate the pool this bio came from - */ -#define BIO_POOL_BITS (4) -#define BIO_POOL_NONE ((1UL << BIO_POOL_BITS) - 1) -#define BIO_POOL_OFFSET (BITS_PER_LONG - BIO_POOL_BITS) -#define BIO_POOL_MASK (1UL << BIO_POOL_OFFSET) -#define BIO_POOL_IDX(bio) ((bio)->bi_flags >> BIO_POOL_OFFSET) - -/* - * bio bi_rw flags - * - * bit 0 -- data direction - * If not set, bio is a read from device. If set, it's a write to device. - * bit 1 -- fail fast device errors - * bit 2 -- fail fast transport errors - * bit 3 -- fail fast driver errors - * bit 4 -- rw-ahead when set - * bit 5 -- barrier - * Insert a serialization point in the IO queue, forcing previously - * submitted IO to be completed before this one is issued. - * bit 6 -- synchronous I/O hint. - * bit 7 -- Unplug the device immediately after submitting this bio. - * bit 8 -- metadata request - * Used for tracing to differentiate metadata and data IO. May also - * get some preferential treatment in the IO scheduler - * bit 9 -- discard sectors - * Informs the lower level device that this range of sectors is no longer - * used by the file system and may thus be freed by the device. Used - * for flash based storage. - * Don't want driver retries for any fast fail whatever the reason. - * bit 10 -- Tell the IO scheduler not to wait for more requests after this - one has been submitted, even if it is a SYNC request. - */ -enum bio_rw_flags { - BIO_RW, - BIO_RW_FAILFAST_DEV, - BIO_RW_FAILFAST_TRANSPORT, - BIO_RW_FAILFAST_DRIVER, - /* above flags must match REQ_* */ - BIO_RW_AHEAD, - BIO_RW_BARRIER, - BIO_RW_SYNCIO, - BIO_RW_UNPLUG, - BIO_RW_META, - BIO_RW_DISCARD, - BIO_RW_NOIDLE, -}; - -/* - * First four bits must match between bio->bi_rw and rq->cmd_flags, make - * that explicit here. - */ -#define BIO_RW_RQ_MASK 0xf - -static inline bool bio_rw_flagged(struct bio *bio, enum bio_rw_flags flag) -{ - return (bio->bi_rw & (1 << flag)) != 0; -} - -/* * upper 16 bits of bi_rw define the io priority of this bio */ #define BIO_PRIO_SHIFT (8 * sizeof(unsigned long) - IOPRIO_BITS) @@ -211,7 +66,10 @@ static inline bool bio_rw_flagged(struct bio *bio, enum bio_rw_flags flag) #define bio_offset(bio) bio_iovec((bio))->bv_offset #define bio_segments(bio) ((bio)->bi_vcnt - (bio)->bi_idx) #define bio_sectors(bio) ((bio)->bi_size >> 9) -#define bio_empty_barrier(bio) (bio_rw_flagged(bio, BIO_RW_BARRIER) && !bio_has_data(bio) && !bio_rw_flagged(bio, BIO_RW_DISCARD)) +#define bio_empty_barrier(bio) \ + ((bio->bi_rw & REQ_HARDBARRIER) && \ + !bio_has_data(bio) && \ + !(bio->bi_rw & REQ_DISCARD)) static inline unsigned int bio_cur_bytes(struct bio *bio) { diff --git a/include/linux/blk_types.h b/include/linux/blk_types.h new file mode 100644 index 0000000..5369177 --- /dev/null +++ b/include/linux/blk_types.h @@ -0,0 +1,194 @@ +/* + * Block data types and constants. Directly include this file only to + * break include dependency loop. + */ +#ifndef __LINUX_BLK_TYPES_H +#define __LINUX_BLK_TYPES_H + +#ifdef CONFIG_BLOCK + +#include <linux/types.h> + +struct bio_set; +struct bio; +struct bio_integrity_payload; +struct page; +struct block_device; +typedef void (bio_end_io_t) (struct bio *, int); +typedef void (bio_destructor_t) (struct bio *); + +/* + * was unsigned short, but we might as well be ready for > 64kB I/O pages + */ +struct bio_vec { + struct page *bv_page; + unsigned int bv_len; + unsigned int bv_offset; +}; + +/* + * main unit of I/O for the block layer and lower layers (ie drivers and + * stacking drivers) + */ +struct bio { + sector_t bi_sector; /* device address in 512 byte + sectors */ + struct bio *bi_next; /* request queue link */ + struct block_device *bi_bdev; + unsigned long bi_flags; /* status, command, etc */ + unsigned long bi_rw; /* bottom bits READ/WRITE, + * top bits priority + */ + + unsigned short bi_vcnt; /* how many bio_vec's */ + unsigned short bi_idx; /* current index into bvl_vec */ + + /* Number of segments in this BIO after + * physical address coalescing is performed. + */ + unsigned int bi_phys_segments; + + unsigned int bi_size; /* residual I/O count */ + + /* + * To keep track of the max segment size, we account for the + * sizes of the first and last mergeable segments in this bio. + */ + unsigned int bi_seg_front_size; + unsigned int bi_seg_back_size; + + unsigned int bi_max_vecs; /* max bvl_vecs we can hold */ + + unsigned int bi_comp_cpu; /* completion CPU */ + + atomic_t bi_cnt; /* pin count */ + + struct bio_vec *bi_io_vec; /* the actual vec list */ + + bio_end_io_t *bi_end_io; + + void *bi_private; +#if defined(CONFIG_BLK_DEV_INTEGRITY) + struct bio_integrity_payload *bi_integrity; /* data integrity */ +#endif + + bio_destructor_t *bi_destructor; /* destructor */ + + /* + * We can inline a number of vecs at the end of the bio, to avoid + * double allocations for a small number of bio_vecs. This member + * MUST obviously be kept at the very end of the bio. + */ + struct bio_vec bi_inline_vecs[0]; +}; + +/* + * bio flags + */ +#define BIO_UPTODATE 0 /* ok after I/O completion */ +#define BIO_RW_BLOCK 1 /* RW_AHEAD set, and read/write would block */ +#define BIO_EOF 2 /* out-out-bounds error */ +#define BIO_SEG_VALID 3 /* bi_phys_segments valid */ +#define BIO_CLONED 4 /* doesn't own data */ +#define BIO_BOUNCED 5 /* bio is a bounce bio */ +#define BIO_USER_MAPPED 6 /* contains user pages */ +#define BIO_EOPNOTSUPP 7 /* not supported */ +#define BIO_CPU_AFFINE 8 /* complete bio on same CPU as submitted */ +#define BIO_NULL_MAPPED 9 /* contains invalid user pages */ +#define BIO_FS_INTEGRITY 10 /* fs owns integrity data, not block layer */ +#define BIO_QUIET 11 /* Make BIO Quiet */ +#define bio_flagged(bio, flag) ((bio)->bi_flags & (1 << (flag))) + +/* + * top 4 bits of bio flags indicate the pool this bio came from + */ +#define BIO_POOL_BITS (4) +#define BIO_POOL_NONE ((1UL << BIO_POOL_BITS) - 1) +#define BIO_POOL_OFFSET (BITS_PER_LONG - BIO_POOL_BITS) +#define BIO_POOL_MASK (1UL << BIO_POOL_OFFSET) +#define BIO_POOL_IDX(bio) ((bio)->bi_flags >> BIO_POOL_OFFSET) + +#endif /* CONFIG_BLOCK */ + +/* + * Request flags. For use in the cmd_flags field of struct request, and in + * bi_rw of struct bio. Note that some flags are only valid in either one. + */ +enum rq_flag_bits { + /* common flags */ + __REQ_WRITE, /* not set, read. set, write */ + __REQ_FAILFAST_DEV, /* no driver retries of device errors */ + __REQ_FAILFAST_TRANSPORT, /* no driver retries of transport errors */ + __REQ_FAILFAST_DRIVER, /* no driver retries of driver errors */ + + __REQ_HARDBARRIER, /* may not be passed by drive either */ + __REQ_SYNC, /* request is sync (sync write or read) */ + __REQ_META, /* metadata io request */ + __REQ_DISCARD, /* request to discard sectors */ + __REQ_NOIDLE, /* don't anticipate more IO after this one */ + + /* bio only flags */ + __REQ_UNPLUG, /* unplug the immediately after submission */ + __REQ_RAHEAD, /* read ahead, can fail anytime */ + + /* request only flags */ + __REQ_SORTED, /* elevator knows about this request */ + __REQ_SOFTBARRIER, /* may not be passed by ioscheduler */ + __REQ_FUA, /* forced unit access */ + __REQ_NOMERGE, /* don't touch this for merging */ + __REQ_STARTED, /* drive already may have started this one */ + __REQ_DONTPREP, /* don't call prep for this one */ + __REQ_QUEUED, /* uses queueing */ + __REQ_ELVPRIV, /* elevator private data attached */ + __REQ_FAILED, /* set if the request failed */ + __REQ_QUIET, /* don't worry about errors */ + __REQ_PREEMPT, /* set for "ide_preempt" requests */ + __REQ_ORDERED_COLOR, /* is before or after barrier */ + __REQ_ALLOCED, /* request came from our alloc pool */ + __REQ_COPY_USER, /* contains copies of user pages */ + __REQ_INTEGRITY, /* integrity metadata has been remapped */ + __REQ_FLUSH, /* request for cache flush */ + __REQ_IO_STAT, /* account I/O stat */ + __REQ_MIXED_MERGE, /* merge of different types, fail separately */ + __REQ_NR_BITS, /* stops here */ +}; + +#define REQ_WRITE (1 << __REQ_WRITE) +#define REQ_FAILFAST_DEV (1 << __REQ_FAILFAST_DEV) +#define REQ_FAILFAST_TRANSPORT (1 << __REQ_FAILFAST_TRANSPORT) +#define REQ_FAILFAST_DRIVER (1 << __REQ_FAILFAST_DRIVER) +#define REQ_HARDBARRIER (1 << __REQ_HARDBARRIER) +#define REQ_SYNC (1 << __REQ_SYNC) +#define REQ_META (1 << __REQ_META) +#define REQ_DISCARD (1 << __REQ_DISCARD) +#define REQ_NOIDLE (1 << __REQ_NOIDLE) + +#define REQ_FAILFAST_MASK \ + (REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT | REQ_FAILFAST_DRIVER) +#define REQ_COMMON_MASK \ + (REQ_WRITE | REQ_FAILFAST_MASK | REQ_HARDBARRIER | REQ_SYNC | \ + REQ_META| REQ_DISCARD | REQ_NOIDLE) + +#define REQ_UNPLUG (1 << __REQ_UNPLUG) +#define REQ_RAHEAD (1 << __REQ_RAHEAD) + +#define REQ_SORTED (1 << __REQ_SORTED) +#define REQ_SOFTBARRIER (1 << __REQ_SOFTBARRIER) +#define REQ_FUA (1 << __REQ_FUA) +#define REQ_NOMERGE (1 << __REQ_NOMERGE) +#define REQ_STARTED (1 << __REQ_STARTED) +#define REQ_DONTPREP (1 << __REQ_DONTPREP) +#define REQ_QUEUED (1 << __REQ_QUEUED) +#define REQ_ELVPRIV (1 << __REQ_ELVPRIV) +#define REQ_FAILED (1 << __REQ_FAILED) +#define REQ_QUIET (1 << __REQ_QUIET) +#define REQ_PREEMPT (1 << __REQ_PREEMPT) +#define REQ_ORDERED_COLOR (1 << __REQ_ORDERED_COLOR) +#define REQ_ALLOCED (1 << __REQ_ALLOCED) +#define REQ_COPY_USER (1 << __REQ_COPY_USER) +#define REQ_INTEGRITY (1 << __REQ_INTEGRITY) +#define REQ_FLUSH (1 << __REQ_FLUSH) +#define REQ_IO_STAT (1 << __REQ_IO_STAT) +#define REQ_MIXED_MERGE (1 << __REQ_MIXED_MERGE) + +#endif /* __LINUX_BLK_TYPES_H */ diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 09a8402..89c855c 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -60,7 +60,6 @@ enum rq_cmd_type_bits { REQ_TYPE_PM_RESUME, /* resume request */ REQ_TYPE_PM_SHUTDOWN, /* shutdown request */ REQ_TYPE_SPECIAL, /* driver defined type */ - REQ_TYPE_LINUX_BLOCK, /* generic block layer message */ /* * for ATA/ATAPI devices. this really doesn't belong here, ide should * use REQ_TYPE_SPECIAL and use rq->cmd[0] with the range of driver @@ -70,84 +69,6 @@ enum rq_cmd_type_bits { REQ_TYPE_ATA_PC, }; -/* - * For request of type REQ_TYPE_LINUX_BLOCK, rq->cmd[0] is the opcode being - * sent down (similar to how REQ_TYPE_BLOCK_PC means that ->cmd[] holds a - * SCSI cdb. - * - * 0x00 -> 0x3f are driver private, to be used for whatever purpose they need, - * typically to differentiate REQ_TYPE_SPECIAL requests. - * - */ -enum { - REQ_LB_OP_EJECT = 0x40, /* eject request */ - REQ_LB_OP_FLUSH = 0x41, /* flush request */ -}; - -/* - * request type modified bits. first four bits match BIO_RW* bits, important - */ -enum rq_flag_bits { - __REQ_RW, /* not set, read. set, write */ - __REQ_FAILFAST_DEV, /* no driver retries of device errors */ - __REQ_FAILFAST_TRANSPORT, /* no driver retries of transport errors */ - __REQ_FAILFAST_DRIVER, /* no driver retries of driver errors */ - /* above flags must match BIO_RW_* */ - __REQ_DISCARD, /* request to discard sectors */ - __REQ_SORTED, /* elevator knows about this request */ - __REQ_SOFTBARRIER, /* may not be passed by ioscheduler */ - __REQ_HARDBARRIER, /* may not be passed by drive either */ - __REQ_FUA, /* forced unit access */ - __REQ_NOMERGE, /* don't touch this for merging */ - __REQ_STARTED, /* drive already may have started this one */ - __REQ_DONTPREP, /* don't call prep for this one */ - __REQ_QUEUED, /* uses queueing */ - __REQ_ELVPRIV, /* elevator private data attached */ - __REQ_FAILED, /* set if the request failed */ - __REQ_QUIET, /* don't worry about errors */ - __REQ_PREEMPT, /* set for "ide_preempt" requests */ - __REQ_ORDERED_COLOR, /* is before or after barrier */ - __REQ_RW_SYNC, /* request is sync (sync write or read) */ - __REQ_ALLOCED, /* request came from our alloc pool */ - __REQ_RW_META, /* metadata io request */ - __REQ_COPY_USER, /* contains copies of user pages */ - __REQ_INTEGRITY, /* integrity metadata has been remapped */ - __REQ_NOIDLE, /* Don't anticipate more IO after this one */ - __REQ_IO_STAT, /* account I/O stat */ - __REQ_MIXED_MERGE, /* merge of different types, fail separately */ - __REQ_NR_BITS, /* stops here */ -}; - -#define REQ_RW (1 << __REQ_RW) -#define REQ_FAILFAST_DEV (1 << __REQ_FAILFAST_DEV) -#define REQ_FAILFAST_TRANSPORT (1 << __REQ_FAILFAST_TRANSPORT) -#define REQ_FAILFAST_DRIVER (1 << __REQ_FAILFAST_DRIVER) -#define REQ_DISCARD (1 << __REQ_DISCARD) -#define REQ_SORTED (1 << __REQ_SORTED) -#define REQ_SOFTBARRIER (1 << __REQ_SOFTBARRIER) -#define REQ_HARDBARRIER (1 << __REQ_HARDBARRIER) -#define REQ_FUA (1 << __REQ_FUA) -#define REQ_NOMERGE (1 << __REQ_NOMERGE) -#define REQ_STARTED (1 << __REQ_STARTED) -#define REQ_DONTPREP (1 << __REQ_DONTPREP) -#define REQ_QUEUED (1 << __REQ_QUEUED) -#define REQ_ELVPRIV (1 << __REQ_ELVPRIV) -#define REQ_FAILED (1 << __REQ_FAILED) -#define REQ_QUIET (1 << __REQ_QUIET) -#define REQ_PREEMPT (1 << __REQ_PREEMPT) -#define REQ_ORDERED_COLOR (1 << __REQ_ORDERED_COLOR) -#define REQ_RW_SYNC (1 << __REQ_RW_SYNC) -#define REQ_ALLOCED (1 << __REQ_ALLOCED) -#define REQ_RW_META (1 << __REQ_RW_META) -#define REQ_COPY_USER (1 << __REQ_COPY_USER) -#define REQ_INTEGRITY (1 << __REQ_INTEGRITY) -#define REQ_NOIDLE (1 << __REQ_NOIDLE) -#define REQ_IO_STAT (1 << __REQ_IO_STAT) -#define REQ_MIXED_MERGE (1 << __REQ_MIXED_MERGE) - -#define REQ_FAILFAST_MASK (REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT | \ - REQ_FAILFAST_DRIVER) - #define BLK_MAX_CDB 16 /* @@ -264,6 +185,7 @@ struct request_pm_state typedef void (request_fn_proc) (struct request_queue *q); typedef int (make_request_fn) (struct request_queue *q, struct bio *bio); typedef int (prep_rq_fn) (struct request_queue *, struct request *); +typedef void (unprep_rq_fn) (struct request_queue *, struct request *); typedef void (unplug_fn) (struct request_queue *); struct bio_vec; @@ -275,7 +197,6 @@ struct bvec_merge_data { }; typedef int (merge_bvec_fn) (struct request_queue *, struct bvec_merge_data *, struct bio_vec *); -typedef void (prepare_flush_fn) (struct request_queue *, struct request *); typedef void (softirq_done_fn)(struct request *); typedef int (dma_drain_needed_fn)(struct request *); typedef int (lld_busy_fn) (struct request_queue *q); @@ -346,9 +267,9 @@ struct request_queue request_fn_proc *request_fn; make_request_fn *make_request_fn; prep_rq_fn *prep_rq_fn; + unprep_rq_fn *unprep_rq_fn; unplug_fn *unplug_fn; merge_bvec_fn *merge_bvec_fn; - prepare_flush_fn *prepare_flush_fn; softirq_done_fn *softirq_done_fn; rq_timed_out_fn *rq_timed_out_fn; dma_drain_needed_fn *dma_drain_needed; @@ -467,11 +388,13 @@ struct request_queue #define QUEUE_FLAG_IO_STAT 15 /* do IO stats */ #define QUEUE_FLAG_DISCARD 16 /* supports DISCARD */ #define QUEUE_FLAG_NOXMERGES 17 /* No extended merges */ +#define QUEUE_FLAG_ADD_RANDOM 18 /* Contributes to random pool */ #define QUEUE_FLAG_DEFAULT ((1 << QUEUE_FLAG_IO_STAT) | \ (1 << QUEUE_FLAG_CLUSTER) | \ (1 << QUEUE_FLAG_STACKABLE) | \ - (1 << QUEUE_FLAG_SAME_COMP)) + (1 << QUEUE_FLAG_SAME_COMP) | \ + (1 << QUEUE_FLAG_ADD_RANDOM)) static inline int queue_is_locked(struct request_queue *q) { @@ -596,38 +519,26 @@ enum { test_bit(QUEUE_FLAG_NOXMERGES, &(q)->queue_flags) #define blk_queue_nonrot(q) test_bit(QUEUE_FLAG_NONROT, &(q)->queue_flags) #define blk_queue_io_stat(q) test_bit(QUEUE_FLAG_IO_STAT, &(q)->queue_flags) +#define blk_queue_add_random(q) test_bit(QUEUE_FLAG_ADD_RANDOM, &(q)->queue_flags) #define blk_queue_flushing(q) ((q)->ordseq) #define blk_queue_stackable(q) \ test_bit(QUEUE_FLAG_STACKABLE, &(q)->queue_flags) #define blk_queue_discard(q) test_bit(QUEUE_FLAG_DISCARD, &(q)->queue_flags) -#define blk_fs_request(rq) ((rq)->cmd_type == REQ_TYPE_FS) -#define blk_pc_request(rq) ((rq)->cmd_type == REQ_TYPE_BLOCK_PC) -#define blk_special_request(rq) ((rq)->cmd_type == REQ_TYPE_SPECIAL) -#define blk_sense_request(rq) ((rq)->cmd_type == REQ_TYPE_SENSE) - -#define blk_failfast_dev(rq) ((rq)->cmd_flags & REQ_FAILFAST_DEV) -#define blk_failfast_transport(rq) ((rq)->cmd_flags & REQ_FAILFAST_TRANSPORT) -#define blk_failfast_driver(rq) ((rq)->cmd_flags & REQ_FAILFAST_DRIVER) -#define blk_noretry_request(rq) (blk_failfast_dev(rq) || \ - blk_failfast_transport(rq) || \ - blk_failfast_driver(rq)) -#define blk_rq_started(rq) ((rq)->cmd_flags & REQ_STARTED) -#define blk_rq_io_stat(rq) ((rq)->cmd_flags & REQ_IO_STAT) -#define blk_rq_quiet(rq) ((rq)->cmd_flags & REQ_QUIET) - -#define blk_account_rq(rq) (blk_rq_started(rq) && (blk_fs_request(rq) || blk_discard_rq(rq))) - -#define blk_pm_suspend_request(rq) ((rq)->cmd_type == REQ_TYPE_PM_SUSPEND) -#define blk_pm_resume_request(rq) ((rq)->cmd_type == REQ_TYPE_PM_RESUME) +#define blk_noretry_request(rq) \ + ((rq)->cmd_flags & (REQ_FAILFAST_DEV|REQ_FAILFAST_TRANSPORT| \ + REQ_FAILFAST_DRIVER)) + +#define blk_account_rq(rq) \ + (((rq)->cmd_flags & REQ_STARTED) && \ + ((rq)->cmd_type == REQ_TYPE_FS || \ + ((rq)->cmd_flags & REQ_DISCARD))) + #define blk_pm_request(rq) \ - (blk_pm_suspend_request(rq) || blk_pm_resume_request(rq)) + ((rq)->cmd_type == REQ_TYPE_PM_SUSPEND || \ + (rq)->cmd_type == REQ_TYPE_PM_RESUME) #define blk_rq_cpu_valid(rq) ((rq)->cpu != -1) -#define blk_sorted_rq(rq) ((rq)->cmd_flags & REQ_SORTED) -#define blk_barrier_rq(rq) ((rq)->cmd_flags & REQ_HARDBARRIER) -#define blk_fua_rq(rq) ((rq)->cmd_flags & REQ_FUA) -#define blk_discard_rq(rq) ((rq)->cmd_flags & REQ_DISCARD) #define blk_bidi_rq(rq) ((rq)->next_rq != NULL) /* rq->queuelist of dequeued request must be list_empty() */ #define blk_queued_rq(rq) (!list_empty(&(rq)->queuelist)) @@ -641,7 +552,7 @@ enum { */ static inline bool rw_is_sync(unsigned int rw_flags) { - return !(rw_flags & REQ_RW) || (rw_flags & REQ_RW_SYNC); + return !(rw_flags & REQ_WRITE) || (rw_flags & REQ_SYNC); } static inline bool rq_is_sync(struct request *rq) @@ -649,9 +560,6 @@ static inline bool rq_is_sync(struct request *rq) return rw_is_sync(rq->cmd_flags); } -#define rq_is_meta(rq) ((rq)->cmd_flags & REQ_RW_META) -#define rq_noidle(rq) ((rq)->cmd_flags & REQ_NOIDLE) - static inline int blk_queue_full(struct request_queue *q, int sync) { if (sync) @@ -684,7 +592,8 @@ static inline void blk_clear_queue_full(struct request_queue *q, int sync) (REQ_NOMERGE | REQ_STARTED | REQ_HARDBARRIER | REQ_SOFTBARRIER) #define rq_mergeable(rq) \ (!((rq)->cmd_flags & RQ_NOMERGE_FLAGS) && \ - (blk_discard_rq(rq) || blk_fs_request((rq)))) + (((rq)->cmd_flags & REQ_DISCARD) || \ + (rq)->cmd_type == REQ_TYPE_FS)) /* * q->prep_rq_fn return values @@ -709,7 +618,7 @@ extern unsigned long blk_max_low_pfn, blk_max_pfn; #define BLK_BOUNCE_HIGH -1ULL #endif #define BLK_BOUNCE_ANY (-1ULL) -#define BLK_BOUNCE_ISA (ISA_DMA_THRESHOLD) +#define BLK_BOUNCE_ISA (DMA_BIT_MASK(24)) /* * default timeout for SG_IO if none specified @@ -781,6 +690,8 @@ extern struct request *blk_make_request(struct request_queue *, struct bio *, gfp_t); extern void blk_insert_request(struct request_queue *, struct request *, int, void *); extern void blk_requeue_request(struct request_queue *, struct request *); +extern void blk_add_request_payload(struct request *rq, struct page *page, + unsigned int len); extern int blk_rq_check_limits(struct request_queue *q, struct request *rq); extern int blk_lld_busy(struct request_queue *q); extern int blk_rq_prep_clone(struct request *rq, struct request *rq_src, @@ -915,6 +826,7 @@ extern void blk_complete_request(struct request *); extern void __blk_complete_request(struct request *); extern void blk_abort_request(struct request *); extern void blk_abort_queue(struct request_queue *); +extern void blk_unprep_request(struct request *); /* * Access functions for manipulating queue properties @@ -959,6 +871,7 @@ extern int blk_queue_dma_drain(struct request_queue *q, extern void blk_queue_lld_busy(struct request_queue *q, lld_busy_fn *fn); extern void blk_queue_segment_boundary(struct request_queue *, unsigned long); extern void blk_queue_prep_rq(struct request_queue *, prep_rq_fn *pfn); +extern void blk_queue_unprep_rq(struct request_queue *, unprep_rq_fn *ufn); extern void blk_queue_merge_bvec(struct request_queue *, merge_bvec_fn *); extern void blk_queue_dma_alignment(struct request_queue *, int); extern void blk_queue_update_dma_alignment(struct request_queue *, int); @@ -966,7 +879,7 @@ extern void blk_queue_softirq_done(struct request_queue *, softirq_done_fn *); extern void blk_queue_rq_timed_out(struct request_queue *, rq_timed_out_fn *); extern void blk_queue_rq_timeout(struct request_queue *, unsigned int); extern struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev); -extern int blk_queue_ordered(struct request_queue *, unsigned, prepare_flush_fn *); +extern int blk_queue_ordered(struct request_queue *, unsigned); extern bool blk_do_ordered(struct request_queue *, struct request **); extern unsigned blk_ordered_cur_seq(struct request_queue *); extern unsigned blk_ordered_req_seq(struct request *); @@ -1020,7 +933,7 @@ static inline int sb_issue_discard(struct super_block *sb, { block <<= (sb->s_blocksize_bits - 9); nr_blocks <<= (sb->s_blocksize_bits - 9); - return blkdev_issue_discard(sb->s_bdev, block, nr_blocks, GFP_KERNEL, + return blkdev_issue_discard(sb->s_bdev, block, nr_blocks, GFP_NOFS, BLKDEV_IFL_WAIT | BLKDEV_IFL_BARRIER); } @@ -1333,7 +1246,6 @@ static inline int blk_integrity_rq(struct request *rq) struct block_device_operations { int (*open) (struct block_device *, fmode_t); int (*release) (struct gendisk *, fmode_t); - int (*locked_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long); int (*ioctl) (struct block_device *, fmode_t, unsigned, unsigned long); int (*compat_ioctl) (struct block_device *, fmode_t, unsigned, unsigned long); int (*direct_access) (struct block_device *, sector_t, diff --git a/include/linux/blktrace_api.h b/include/linux/blktrace_api.h index 416bf62..3395cf7 100644 --- a/include/linux/blktrace_api.h +++ b/include/linux/blktrace_api.h @@ -5,6 +5,7 @@ #ifdef __KERNEL__ #include <linux/blkdev.h> #include <linux/relay.h> +#include <linux/compat.h> #endif /* @@ -220,11 +221,26 @@ static inline int blk_trace_init_sysfs(struct device *dev) #endif /* CONFIG_BLK_DEV_IO_TRACE */ +#ifdef CONFIG_COMPAT + +struct compat_blk_user_trace_setup { + char name[32]; + u16 act_mask; + u32 buf_size; + u32 buf_nr; + compat_u64 start_lba; + compat_u64 end_lba; + u32 pid; +}; +#define BLKTRACESETUP32 _IOWR(0x12, 115, struct compat_blk_user_trace_setup) + +#endif + #if defined(CONFIG_EVENT_TRACING) && defined(CONFIG_BLOCK) static inline int blk_cmd_buf_len(struct request *rq) { - return blk_pc_request(rq) ? rq->cmd_len * 3 : 1; + return (rq->cmd_type == REQ_TYPE_BLOCK_PC) ? rq->cmd_len * 3 : 1; } extern void blk_dump_cmd(char *buf, struct request *rq); diff --git a/include/linux/coda_psdev.h b/include/linux/coda_psdev.h index 8859e2e..284b520 100644 --- a/include/linux/coda_psdev.h +++ b/include/linux/coda_psdev.h @@ -86,9 +86,9 @@ struct upc_req { wait_queue_head_t uc_sleep; /* process' wait queue */ }; -#define REQ_ASYNC 0x1 -#define REQ_READ 0x2 -#define REQ_WRITE 0x4 -#define REQ_ABORT 0x8 +#define CODA_REQ_ASYNC 0x1 +#define CODA_REQ_READ 0x2 +#define CODA_REQ_WRITE 0x4 +#define CODA_REQ_ABORT 0x8 #endif diff --git a/include/linux/drbd.h b/include/linux/drbd.h index b8d2516..479ee3a 100644 --- a/include/linux/drbd.h +++ b/include/linux/drbd.h @@ -53,7 +53,7 @@ extern const char *drbd_buildtag(void); -#define REL_VERSION "8.3.8" +#define REL_VERSION "8.3.8.1" #define API_VERSION 88 #define PRO_VERSION_MIN 86 #define PRO_VERSION_MAX 94 diff --git a/include/linux/drbd_nl.h b/include/linux/drbd_nl.h index ce77a74..5f04281 100644 --- a/include/linux/drbd_nl.h +++ b/include/linux/drbd_nl.h @@ -78,10 +78,11 @@ NL_PACKET(syncer_conf, 8, NL_INTEGER( 30, T_MAY_IGNORE, rate) NL_INTEGER( 31, T_MAY_IGNORE, after) NL_INTEGER( 32, T_MAY_IGNORE, al_extents) - NL_INTEGER( 71, T_MAY_IGNORE, dp_volume) - NL_INTEGER( 72, T_MAY_IGNORE, dp_interval) - NL_INTEGER( 73, T_MAY_IGNORE, throttle_th) - NL_INTEGER( 74, T_MAY_IGNORE, hold_off_th) +/* NL_INTEGER( 71, T_MAY_IGNORE, dp_volume) + * NL_INTEGER( 72, T_MAY_IGNORE, dp_interval) + * NL_INTEGER( 73, T_MAY_IGNORE, throttle_th) + * NL_INTEGER( 74, T_MAY_IGNORE, hold_off_th) + * feature will be reimplemented differently with 8.3.9 */ NL_STRING( 52, T_MAY_IGNORE, verify_alg, SHARED_SECRET_MAX) NL_STRING( 51, T_MAY_IGNORE, cpu_mask, 32) NL_STRING( 64, T_MAY_IGNORE, csums_alg, SHARED_SECRET_MAX) diff --git a/include/linux/fs.h b/include/linux/fs.h index a8ccf85..1542e0e 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -8,6 +8,7 @@ #include <linux/limits.h> #include <linux/ioctl.h> +#include <linux/blk_types.h> /* * It's silly to have NR_OPEN bigger than NR_FILE, but you can change @@ -121,7 +122,7 @@ struct inodes_stat_t { * immediately wait on this read without caring about * unplugging. * READA Used for read-ahead operations. Lower priority, and the - * block layer could (in theory) choose to ignore this + * block layer could (in theory) choose to ignore this * request if it runs into resource problems. * WRITE A normal async write. Device will be plugged. * SWRITE Like WRITE, but a special case for ll_rw_block() that @@ -140,7 +141,7 @@ struct inodes_stat_t { * SWRITE_SYNC * SWRITE_SYNC_PLUG Like WRITE_SYNC/WRITE_SYNC_PLUG, but locks the buffer. * See SWRITE. - * WRITE_BARRIER Like WRITE, but tells the block layer that all + * WRITE_BARRIER Like WRITE_SYNC, but tells the block layer that all * previously submitted writes must be safely on storage * before this one is started. Also guarantees that when * this write is complete, it itself is also safely on @@ -148,29 +149,31 @@ struct inodes_stat_t { * of this IO. * */ -#define RW_MASK 1 -#define RWA_MASK 2 -#define READ 0 -#define WRITE 1 -#define READA 2 /* read-ahead - don't block if no resources */ -#define SWRITE 3 /* for ll_rw_block() - wait for buffer lock */ -#define READ_SYNC (READ | (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG)) -#define READ_META (READ | (1 << BIO_RW_META)) -#define WRITE_SYNC_PLUG (WRITE | (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_NOIDLE)) -#define WRITE_SYNC (WRITE_SYNC_PLUG | (1 << BIO_RW_UNPLUG)) -#define WRITE_ODIRECT_PLUG (WRITE | (1 << BIO_RW_SYNCIO)) -#define WRITE_META (WRITE | (1 << BIO_RW_META)) -#define SWRITE_SYNC_PLUG \ - (SWRITE | (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_NOIDLE)) -#define SWRITE_SYNC (SWRITE_SYNC_PLUG | (1 << BIO_RW_UNPLUG)) -#define WRITE_BARRIER (WRITE | (1 << BIO_RW_BARRIER)) +#define RW_MASK REQ_WRITE +#define RWA_MASK REQ_RAHEAD + +#define READ 0 +#define WRITE RW_MASK +#define READA RWA_MASK +#define SWRITE (WRITE | READA) + +#define READ_SYNC (READ | REQ_SYNC | REQ_UNPLUG) +#define READ_META (READ | REQ_META) +#define WRITE_SYNC_PLUG (WRITE | REQ_SYNC | REQ_NOIDLE) +#define WRITE_SYNC (WRITE | REQ_SYNC | REQ_NOIDLE | REQ_UNPLUG) +#define WRITE_ODIRECT_PLUG (WRITE | REQ_SYNC) +#define WRITE_META (WRITE | REQ_META) +#define WRITE_BARRIER (WRITE | REQ_SYNC | REQ_NOIDLE | REQ_UNPLUG | \ + REQ_HARDBARRIER) +#define SWRITE_SYNC_PLUG (SWRITE | REQ_SYNC | REQ_NOIDLE) +#define SWRITE_SYNC (SWRITE | REQ_SYNC | REQ_NOIDLE | REQ_UNPLUG) /* * These aren't really reads or writes, they pass down information about * parts of device that are now unused by the file system. */ -#define DISCARD_NOBARRIER (WRITE | (1 << BIO_RW_DISCARD)) -#define DISCARD_BARRIER (DISCARD_NOBARRIER | (1 << BIO_RW_BARRIER)) +#define DISCARD_NOBARRIER (WRITE | REQ_DISCARD) +#define DISCARD_BARRIER (WRITE | REQ_DISCARD | REQ_HARDBARRIER) #define SEL_IN 1 #define SEL_OUT 2 @@ -2196,7 +2199,6 @@ static inline void insert_inode_hash(struct inode *inode) { extern void file_move(struct file *f, struct list_head *list); extern void file_kill(struct file *f); #ifdef CONFIG_BLOCK -struct bio; extern void submit_bio(int, struct bio *); extern int bdev_read_only(struct block_device *); #endif @@ -2263,7 +2265,6 @@ static inline int xip_truncate_page(struct address_space *mapping, loff_t from) #endif #ifdef CONFIG_BLOCK -struct bio; typedef void (dio_submit_t)(int rw, struct bio *bio, struct inode *inode, loff_t file_offset); diff --git a/include/trace/events/block.h b/include/trace/events/block.h index d870a91..d8ce278 100644 --- a/include/trace/events/block.h +++ b/include/trace/events/block.h @@ -25,8 +25,10 @@ DECLARE_EVENT_CLASS(block_rq_with_error, TP_fast_assign( __entry->dev = rq->rq_disk ? disk_devt(rq->rq_disk) : 0; - __entry->sector = blk_pc_request(rq) ? 0 : blk_rq_pos(rq); - __entry->nr_sector = blk_pc_request(rq) ? 0 : blk_rq_sectors(rq); + __entry->sector = (rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + 0 : blk_rq_pos(rq); + __entry->nr_sector = (rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + 0 : blk_rq_sectors(rq); __entry->errors = rq->errors; blk_fill_rwbs_rq(__entry->rwbs, rq); @@ -109,9 +111,12 @@ DECLARE_EVENT_CLASS(block_rq, TP_fast_assign( __entry->dev = rq->rq_disk ? disk_devt(rq->rq_disk) : 0; - __entry->sector = blk_pc_request(rq) ? 0 : blk_rq_pos(rq); - __entry->nr_sector = blk_pc_request(rq) ? 0 : blk_rq_sectors(rq); - __entry->bytes = blk_pc_request(rq) ? blk_rq_bytes(rq) : 0; + __entry->sector = (rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + 0 : blk_rq_pos(rq); + __entry->nr_sector = (rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + 0 : blk_rq_sectors(rq); + __entry->bytes = (rq->cmd_type == REQ_TYPE_BLOCK_PC) ? + blk_rq_bytes(rq) : 0; blk_fill_rwbs_rq(__entry->rwbs, rq); blk_dump_cmd(__get_str(cmd), rq); diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h new file mode 100644 index 0000000..f345f66 --- /dev/null +++ b/include/trace/events/writeback.h @@ -0,0 +1,159 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM writeback + +#if !defined(_TRACE_WRITEBACK_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_WRITEBACK_H + +#include <linux/backing-dev.h> +#include <linux/device.h> +#include <linux/writeback.h> + +struct wb_writeback_work; + +DECLARE_EVENT_CLASS(writeback_work_class, + TP_PROTO(struct backing_dev_info *bdi, struct wb_writeback_work *work), + TP_ARGS(bdi, work), + TP_STRUCT__entry( + __array(char, name, 32) + __field(long, nr_pages) + __field(dev_t, sb_dev) + __field(int, sync_mode) + __field(int, for_kupdate) + __field(int, range_cyclic) + __field(int, for_background) + ), + TP_fast_assign( + strncpy(__entry->name, dev_name(bdi->dev), 32); + __entry->nr_pages = work->nr_pages; + __entry->sb_dev = work->sb ? work->sb->s_dev : 0; + __entry->sync_mode = work->sync_mode; + __entry->for_kupdate = work->for_kupdate; + __entry->range_cyclic = work->range_cyclic; + __entry->for_background = work->for_background; + ), + TP_printk("bdi %s: sb_dev %d:%d nr_pages=%ld sync_mode=%d " + "kupdate=%d range_cyclic=%d background=%d", + __entry->name, + MAJOR(__entry->sb_dev), MINOR(__entry->sb_dev), + __entry->nr_pages, + __entry->sync_mode, + __entry->for_kupdate, + __entry->range_cyclic, + __entry->for_background + ) +); +#define DEFINE_WRITEBACK_WORK_EVENT(name) \ +DEFINE_EVENT(writeback_work_class, name, \ + TP_PROTO(struct backing_dev_info *bdi, struct wb_writeback_work *work), \ + TP_ARGS(bdi, work)) +DEFINE_WRITEBACK_WORK_EVENT(writeback_nothread); +DEFINE_WRITEBACK_WORK_EVENT(writeback_queue); +DEFINE_WRITEBACK_WORK_EVENT(writeback_exec); + +TRACE_EVENT(writeback_pages_written, + TP_PROTO(long pages_written), + TP_ARGS(pages_written), + TP_STRUCT__entry( + __field(long, pages) + ), + TP_fast_assign( + __entry->pages = pages_written; + ), + TP_printk("%ld", __entry->pages) +); + +DECLARE_EVENT_CLASS(writeback_class, + TP_PROTO(struct backing_dev_info *bdi), + TP_ARGS(bdi), + TP_STRUCT__entry( + __array(char, name, 32) + ), + TP_fast_assign( + strncpy(__entry->name, dev_name(bdi->dev), 32); + ), + TP_printk("bdi %s", + __entry->name + ) +); +#define DEFINE_WRITEBACK_EVENT(name) \ +DEFINE_EVENT(writeback_class, name, \ + TP_PROTO(struct backing_dev_info *bdi), \ + TP_ARGS(bdi)) + +DEFINE_WRITEBACK_EVENT(writeback_nowork); +DEFINE_WRITEBACK_EVENT(writeback_wake_thread); +DEFINE_WRITEBACK_EVENT(writeback_wake_forker_thread); +DEFINE_WRITEBACK_EVENT(writeback_bdi_register); +DEFINE_WRITEBACK_EVENT(writeback_bdi_unregister); +DEFINE_WRITEBACK_EVENT(writeback_thread_start); +DEFINE_WRITEBACK_EVENT(writeback_thread_stop); + +DECLARE_EVENT_CLASS(wbc_class, + TP_PROTO(struct writeback_control *wbc, struct backing_dev_info *bdi), + TP_ARGS(wbc, bdi), + TP_STRUCT__entry( + __array(char, name, 32) + __field(long, nr_to_write) + __field(long, pages_skipped) + __field(int, sync_mode) + __field(int, nonblocking) + __field(int, encountered_congestion) + __field(int, for_kupdate) + __field(int, for_background) + __field(int, for_reclaim) + __field(int, range_cyclic) + __field(int, more_io) + __field(unsigned long, older_than_this) + __field(long, range_start) + __field(long, range_end) + ), + + TP_fast_assign( + strncpy(__entry->name, dev_name(bdi->dev), 32); + __entry->nr_to_write = wbc->nr_to_write; + __entry->pages_skipped = wbc->pages_skipped; + __entry->sync_mode = wbc->sync_mode; + __entry->for_kupdate = wbc->for_kupdate; + __entry->for_background = wbc->for_background; + __entry->for_reclaim = wbc->for_reclaim; + __entry->range_cyclic = wbc->range_cyclic; + __entry->more_io = wbc->more_io; + __entry->older_than_this = wbc->older_than_this ? + *wbc->older_than_this : 0; + __entry->range_start = (long)wbc->range_start; + __entry->range_end = (long)wbc->range_end; + ), + + TP_printk("bdi %s: towrt=%ld skip=%ld mode=%d kupd=%d " + "bgrd=%d reclm=%d cyclic=%d more=%d older=0x%lx " + "start=0x%lx end=0x%lx", + __entry->name, + __entry->nr_to_write, + __entry->pages_skipped, + __entry->sync_mode, + __entry->for_kupdate, + __entry->for_background, + __entry->for_reclaim, + __entry->range_cyclic, + __entry->more_io, + __entry->older_than_this, + __entry->range_start, + __entry->range_end) +) + +#define DEFINE_WBC_EVENT(name) \ +DEFINE_EVENT(wbc_class, name, \ + TP_PROTO(struct writeback_control *wbc, struct backing_dev_info *bdi), \ + TP_ARGS(wbc, bdi)) +DEFINE_WBC_EVENT(wbc_writeback_start); +DEFINE_WBC_EVENT(wbc_writeback_written); +DEFINE_WBC_EVENT(wbc_writeback_wait); +DEFINE_WBC_EVENT(wbc_balance_dirty_start); +DEFINE_WBC_EVENT(wbc_balance_dirty_written); +DEFINE_WBC_EVENT(wbc_balance_dirty_wait); +DEFINE_WBC_EVENT(wbc_writepage); + +#endif /* _TRACE_WRITEBACK_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/kernel/power/block_io.c b/kernel/power/block_io.c index 97024fd..83bbc7c 100644 --- a/kernel/power/block_io.c +++ b/kernel/power/block_io.c @@ -28,7 +28,7 @@ static int submit(int rw, struct block_device *bdev, sector_t sector, struct page *page, struct bio **bio_chain) { - const int bio_rw = rw | (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG); + const int bio_rw = rw | REQ_SYNC | REQ_UNPLUG; struct bio *bio; bio = bio_alloc(__GFP_WAIT | __GFP_HIGH, 1); diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 638711c..82499a5 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -169,9 +169,12 @@ static int act_log_check(struct blk_trace *bt, u32 what, sector_t sector, static const u32 ddir_act[2] = { BLK_TC_ACT(BLK_TC_READ), BLK_TC_ACT(BLK_TC_WRITE) }; +#define BLK_TC_HARDBARRIER BLK_TC_BARRIER +#define BLK_TC_RAHEAD BLK_TC_AHEAD + /* The ilog2() calls fall out because they're constant */ -#define MASK_TC_BIT(rw, __name) ((rw & (1 << BIO_RW_ ## __name)) << \ - (ilog2(BLK_TC_ ## __name) + BLK_TC_SHIFT - BIO_RW_ ## __name)) +#define MASK_TC_BIT(rw, __name) ((rw & REQ_ ## __name) << \ + (ilog2(BLK_TC_ ## __name) + BLK_TC_SHIFT - __REQ_ ## __name)) /* * The worker for the various blk_add_trace*() types. Fills out a @@ -194,9 +197,9 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes, return; what |= ddir_act[rw & WRITE]; - what |= MASK_TC_BIT(rw, BARRIER); - what |= MASK_TC_BIT(rw, SYNCIO); - what |= MASK_TC_BIT(rw, AHEAD); + what |= MASK_TC_BIT(rw, HARDBARRIER); + what |= MASK_TC_BIT(rw, SYNC); + what |= MASK_TC_BIT(rw, RAHEAD); what |= MASK_TC_BIT(rw, META); what |= MASK_TC_BIT(rw, DISCARD); @@ -549,6 +552,41 @@ int blk_trace_setup(struct request_queue *q, char *name, dev_t dev, } EXPORT_SYMBOL_GPL(blk_trace_setup); +#if defined(CONFIG_COMPAT) && defined(CONFIG_X86_64) +static int compat_blk_trace_setup(struct request_queue *q, char *name, + dev_t dev, struct block_device *bdev, + char __user *arg) +{ + struct blk_user_trace_setup buts; + struct compat_blk_user_trace_setup cbuts; + int ret; + + if (copy_from_user(&cbuts, arg, sizeof(cbuts))) + return -EFAULT; + + buts = (struct blk_user_trace_setup) { + .act_mask = cbuts.act_mask, + .buf_size = cbuts.buf_size, + .buf_nr = cbuts.buf_nr, + .start_lba = cbuts.start_lba, + .end_lba = cbuts.end_lba, + .pid = cbuts.pid, + }; + memcpy(&buts.name, &cbuts.name, 32); + + ret = do_blk_trace_setup(q, name, dev, bdev, &buts); + if (ret) + return ret; + + if (copy_to_user(arg, &buts.name, 32)) { + blk_trace_remove(q); + return -EFAULT; + } + + return 0; +} +#endif + int blk_trace_startstop(struct request_queue *q, int start) { int ret; @@ -601,6 +639,7 @@ int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg) if (!q) return -ENXIO; + lock_kernel(); mutex_lock(&bdev->bd_mutex); switch (cmd) { @@ -608,6 +647,12 @@ int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg) bdevname(bdev, b); ret = blk_trace_setup(q, b, bdev->bd_dev, bdev, arg); break; +#if defined(CONFIG_COMPAT) && defined(CONFIG_X86_64) + case BLKTRACESETUP32: + bdevname(bdev, b); + ret = compat_blk_trace_setup(q, b, bdev->bd_dev, bdev, arg); + break; +#endif case BLKTRACESTART: start = 1; case BLKTRACESTOP: @@ -622,6 +667,7 @@ int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg) } mutex_unlock(&bdev->bd_mutex); + unlock_kernel(); return ret; } @@ -661,10 +707,10 @@ static void blk_add_trace_rq(struct request_queue *q, struct request *rq, if (likely(!bt)) return; - if (blk_discard_rq(rq)) - rw |= (1 << BIO_RW_DISCARD); + if (rq->cmd_flags & REQ_DISCARD) + rw |= REQ_DISCARD; - if (blk_pc_request(rq)) { + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) { what |= BLK_TC_ACT(BLK_TC_PC); __blk_add_trace(bt, 0, blk_rq_bytes(rq), rw, what, rq->errors, rq->cmd_len, rq->cmd); @@ -925,7 +971,7 @@ void blk_add_driver_data(struct request_queue *q, if (likely(!bt)) return; - if (blk_pc_request(rq)) + if (rq->cmd_type == REQ_TYPE_BLOCK_PC) __blk_add_trace(bt, 0, blk_rq_bytes(rq), 0, BLK_TA_DRV_DATA, rq->errors, len, data); else @@ -1730,7 +1776,7 @@ void blk_dump_cmd(char *buf, struct request *rq) int len = rq->cmd_len; unsigned char *cmd = rq->cmd; - if (!blk_pc_request(rq)) { + if (rq->cmd_type != REQ_TYPE_BLOCK_PC) { buf[0] = '\0'; return; } @@ -1755,20 +1801,20 @@ void blk_fill_rwbs(char *rwbs, u32 rw, int bytes) if (rw & WRITE) rwbs[i++] = 'W'; - else if (rw & 1 << BIO_RW_DISCARD) + else if (rw & REQ_DISCARD) rwbs[i++] = 'D'; else if (bytes) rwbs[i++] = 'R'; else rwbs[i++] = 'N'; - if (rw & 1 << BIO_RW_AHEAD) + if (rw & REQ_RAHEAD) rwbs[i++] = 'A'; - if (rw & 1 << BIO_RW_BARRIER) + if (rw & REQ_HARDBARRIER) rwbs[i++] = 'B'; - if (rw & 1 << BIO_RW_SYNCIO) + if (rw & REQ_SYNC) rwbs[i++] = 'S'; - if (rw & 1 << BIO_RW_META) + if (rw & REQ_META) rwbs[i++] = 'M'; rwbs[i] = '\0'; @@ -1779,8 +1825,8 @@ void blk_fill_rwbs_rq(char *rwbs, struct request *rq) int rw = rq->cmd_flags & 0x03; int bytes; - if (blk_discard_rq(rq)) - rw |= (1 << BIO_RW_DISCARD); + if (rq->cmd_flags & REQ_DISCARD) + rw |= REQ_DISCARD; bytes = blk_rq_bytes(rq); diff --git a/mm/backing-dev.c b/mm/backing-dev.c index f9fd3dd..08d3575 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -10,6 +10,7 @@ #include <linux/module.h> #include <linux/writeback.h> #include <linux/device.h> +#include <trace/events/writeback.h> static atomic_long_t bdi_seq = ATOMIC_LONG_INIT(0); @@ -49,8 +50,6 @@ static struct timer_list sync_supers_timer; static int bdi_sync_supers(void *); static void sync_supers_timer_fn(unsigned long); -static void bdi_add_default_flusher_task(struct backing_dev_info *bdi); - #ifdef CONFIG_DEBUG_FS #include <linux/debugfs.h> #include <linux/seq_file.h> @@ -65,28 +64,21 @@ static void bdi_debug_init(void) static int bdi_debug_stats_show(struct seq_file *m, void *v) { struct backing_dev_info *bdi = m->private; - struct bdi_writeback *wb; + struct bdi_writeback *wb = &bdi->wb; unsigned long background_thresh; unsigned long dirty_thresh; unsigned long bdi_thresh; unsigned long nr_dirty, nr_io, nr_more_io, nr_wb; struct inode *inode; - /* - * inode lock is enough here, the bdi->wb_list is protected by - * RCU on the reader side - */ nr_wb = nr_dirty = nr_io = nr_more_io = 0; spin_lock(&inode_lock); - list_for_each_entry(wb, &bdi->wb_list, list) { - nr_wb++; - list_for_each_entry(inode, &wb->b_dirty, i_list) - nr_dirty++; - list_for_each_entry(inode, &wb->b_io, i_list) - nr_io++; - list_for_each_entry(inode, &wb->b_more_io, i_list) - nr_more_io++; - } + list_for_each_entry(inode, &wb->b_dirty, i_list) + nr_dirty++; + list_for_each_entry(inode, &wb->b_io, i_list) + nr_io++; + list_for_each_entry(inode, &wb->b_more_io, i_list) + nr_more_io++; spin_unlock(&inode_lock); get_dirty_limits(&background_thresh, &dirty_thresh, &bdi_thresh, bdi); @@ -98,19 +90,16 @@ static int bdi_debug_stats_show(struct seq_file *m, void *v) "BdiDirtyThresh: %8lu kB\n" "DirtyThresh: %8lu kB\n" "BackgroundThresh: %8lu kB\n" - "WritebackThreads: %8lu\n" "b_dirty: %8lu\n" "b_io: %8lu\n" "b_more_io: %8lu\n" "bdi_list: %8u\n" - "state: %8lx\n" - "wb_list: %8u\n", + "state: %8lx\n", (unsigned long) K(bdi_stat(bdi, BDI_WRITEBACK)), (unsigned long) K(bdi_stat(bdi, BDI_RECLAIMABLE)), K(bdi_thresh), K(dirty_thresh), - K(background_thresh), nr_wb, nr_dirty, nr_io, nr_more_io, - !list_empty(&bdi->bdi_list), bdi->state, - !list_empty(&bdi->wb_list)); + K(background_thresh), nr_dirty, nr_io, nr_more_io, + !list_empty(&bdi->bdi_list), bdi->state); #undef K return 0; @@ -247,7 +236,6 @@ static int __init default_bdi_init(void) sync_supers_tsk = kthread_run(bdi_sync_supers, NULL, "sync_supers"); BUG_ON(IS_ERR(sync_supers_tsk)); - init_timer(&sync_supers_timer); setup_timer(&sync_supers_timer, sync_supers_timer_fn, 0); bdi_arm_supers_timer(); @@ -259,77 +247,6 @@ static int __init default_bdi_init(void) } subsys_initcall(default_bdi_init); -static void bdi_wb_init(struct bdi_writeback *wb, struct backing_dev_info *bdi) -{ - memset(wb, 0, sizeof(*wb)); - - wb->bdi = bdi; - wb->last_old_flush = jiffies; - INIT_LIST_HEAD(&wb->b_dirty); - INIT_LIST_HEAD(&wb->b_io); - INIT_LIST_HEAD(&wb->b_more_io); -} - -static void bdi_task_init(struct backing_dev_info *bdi, - struct bdi_writeback *wb) -{ - struct task_struct *tsk = current; - - spin_lock(&bdi->wb_lock); - list_add_tail_rcu(&wb->list, &bdi->wb_list); - spin_unlock(&bdi->wb_lock); - - tsk->flags |= PF_FLUSHER | PF_SWAPWRITE; - set_freezable(); - - /* - * Our parent may run at a different priority, just set us to normal - */ - set_user_nice(tsk, 0); -} - -static int bdi_start_fn(void *ptr) -{ - struct bdi_writeback *wb = ptr; - struct backing_dev_info *bdi = wb->bdi; - int ret; - - /* - * Add us to the active bdi_list - */ - spin_lock_bh(&bdi_lock); - list_add_rcu(&bdi->bdi_list, &bdi_list); - spin_unlock_bh(&bdi_lock); - - bdi_task_init(bdi, wb); - - /* - * Clear pending bit and wakeup anybody waiting to tear us down - */ - clear_bit(BDI_pending, &bdi->state); - smp_mb__after_clear_bit(); - wake_up_bit(&bdi->state, BDI_pending); - - ret = bdi_writeback_task(wb); - - /* - * Remove us from the list - */ - spin_lock(&bdi->wb_lock); - list_del_rcu(&wb->list); - spin_unlock(&bdi->wb_lock); - - /* - * Flush any work that raced with us exiting. No new work - * will be added, since this bdi isn't discoverable anymore. - */ - if (!list_empty(&bdi->work_list)) - wb_do_writeback(wb, 1); - - wb->task = NULL; - return ret; -} - int bdi_has_dirty_io(struct backing_dev_info *bdi) { return wb_has_dirty_io(&bdi->wb); @@ -348,10 +265,10 @@ static void bdi_flush_io(struct backing_dev_info *bdi) } /* - * kupdated() used to do this. We cannot do it from the bdi_forker_task() + * kupdated() used to do this. We cannot do it from the bdi_forker_thread() * or we risk deadlocking on ->s_umount. The longer term solution would be * to implement sync_supers_bdi() or similar and simply do it from the - * bdi writeback tasks individually. + * bdi writeback thread individually. */ static int bdi_sync_supers(void *unused) { @@ -387,144 +304,198 @@ static void sync_supers_timer_fn(unsigned long unused) bdi_arm_supers_timer(); } -static int bdi_forker_task(void *ptr) +static void wakeup_timer_fn(unsigned long data) +{ + struct backing_dev_info *bdi = (struct backing_dev_info *)data; + + spin_lock_bh(&bdi->wb_lock); + if (bdi->wb.task) { + trace_writeback_wake_thread(bdi); + wake_up_process(bdi->wb.task); + } else { + /* + * When bdi tasks are inactive for long time, they are killed. + * In this case we have to wake-up the forker thread which + * should create and run the bdi thread. + */ + trace_writeback_wake_forker_thread(bdi); + wake_up_process(default_backing_dev_info.wb.task); + } + spin_unlock_bh(&bdi->wb_lock); +} + +/* + * This function is used when the first inode for this bdi is marked dirty. It + * wakes-up the corresponding bdi thread which should then take care of the + * periodic background write-out of dirty inodes. Since the write-out would + * starts only 'dirty_writeback_interval' centisecs from now anyway, we just + * set up a timer which wakes the bdi thread up later. + * + * Note, we wouldn't bother setting up the timer, but this function is on the + * fast-path (used by '__mark_inode_dirty()'), so we save few context switches + * by delaying the wake-up. + */ +void bdi_wakeup_thread_delayed(struct backing_dev_info *bdi) +{ + unsigned long timeout; + + timeout = msecs_to_jiffies(dirty_writeback_interval * 10); + mod_timer(&bdi->wb.wakeup_timer, jiffies + timeout); +} + +/* + * Calculate the longest interval (jiffies) bdi threads are allowed to be + * inactive. + */ +static unsigned long bdi_longest_inactive(void) +{ + unsigned long interval; + + interval = msecs_to_jiffies(dirty_writeback_interval * 10); + return max(5UL * 60 * HZ, interval); +} + +static int bdi_forker_thread(void *ptr) { struct bdi_writeback *me = ptr; - bdi_task_init(me->bdi, me); + current->flags |= PF_FLUSHER | PF_SWAPWRITE; + set_freezable(); + + /* + * Our parent may run at a different priority, just set us to normal + */ + set_user_nice(current, 0); for (;;) { - struct backing_dev_info *bdi, *tmp; - struct bdi_writeback *wb; + struct task_struct *task = NULL; + struct backing_dev_info *bdi; + enum { + NO_ACTION, /* Nothing to do */ + FORK_THREAD, /* Fork bdi thread */ + KILL_THREAD, /* Kill inactive bdi thread */ + } action = NO_ACTION; /* * Temporary measure, we want to make sure we don't see * dirty data on the default backing_dev_info */ - if (wb_has_dirty_io(me) || !list_empty(&me->bdi->work_list)) + if (wb_has_dirty_io(me) || !list_empty(&me->bdi->work_list)) { + del_timer(&me->wakeup_timer); wb_do_writeback(me, 0); + } spin_lock_bh(&bdi_lock); + set_current_state(TASK_INTERRUPTIBLE); - /* - * Check if any existing bdi's have dirty data without - * a thread registered. If so, set that up. - */ - list_for_each_entry_safe(bdi, tmp, &bdi_list, bdi_list) { - if (bdi->wb.task) - continue; - if (list_empty(&bdi->work_list) && - !bdi_has_dirty_io(bdi)) + list_for_each_entry(bdi, &bdi_list, bdi_list) { + bool have_dirty_io; + + if (!bdi_cap_writeback_dirty(bdi) || + bdi_cap_flush_forker(bdi)) continue; - bdi_add_default_flusher_task(bdi); - } + WARN(!test_bit(BDI_registered, &bdi->state), + "bdi %p/%s is not registered!\n", bdi, bdi->name); - set_current_state(TASK_INTERRUPTIBLE); + have_dirty_io = !list_empty(&bdi->work_list) || + wb_has_dirty_io(&bdi->wb); - if (list_empty(&bdi_pending_list)) { - unsigned long wait; + /* + * If the bdi has work to do, but the thread does not + * exist - create it. + */ + if (!bdi->wb.task && have_dirty_io) { + /* + * Set the pending bit - if someone will try to + * unregister this bdi - it'll wait on this bit. + */ + set_bit(BDI_pending, &bdi->state); + action = FORK_THREAD; + break; + } + + spin_lock(&bdi->wb_lock); + + /* + * If there is no work to do and the bdi thread was + * inactive long enough - kill it. The wb_lock is taken + * to make sure no-one adds more work to this bdi and + * wakes the bdi thread up. + */ + if (bdi->wb.task && !have_dirty_io && + time_after(jiffies, bdi->wb.last_active + + bdi_longest_inactive())) { + task = bdi->wb.task; + bdi->wb.task = NULL; + spin_unlock(&bdi->wb_lock); + set_bit(BDI_pending, &bdi->state); + action = KILL_THREAD; + break; + } + spin_unlock(&bdi->wb_lock); + } + spin_unlock_bh(&bdi_lock); - spin_unlock_bh(&bdi_lock); - wait = msecs_to_jiffies(dirty_writeback_interval * 10); - if (wait) - schedule_timeout(wait); + /* Keep working if default bdi still has things to do */ + if (!list_empty(&me->bdi->work_list)) + __set_current_state(TASK_RUNNING); + + switch (action) { + case FORK_THREAD: + __set_current_state(TASK_RUNNING); + task = kthread_run(bdi_writeback_thread, &bdi->wb, "flush-%s", + dev_name(bdi->dev)); + if (IS_ERR(task)) { + /* + * If thread creation fails, force writeout of + * the bdi from the thread. + */ + bdi_flush_io(bdi); + } else { + /* + * The spinlock makes sure we do not lose + * wake-ups when racing with 'bdi_queue_work()'. + */ + spin_lock_bh(&bdi->wb_lock); + bdi->wb.task = task; + spin_unlock_bh(&bdi->wb_lock); + } + break; + + case KILL_THREAD: + __set_current_state(TASK_RUNNING); + kthread_stop(task); + break; + + case NO_ACTION: + if (!wb_has_dirty_io(me) || !dirty_writeback_interval) + /* + * There are no dirty data. The only thing we + * should now care about is checking for + * inactive bdi threads and killing them. Thus, + * let's sleep for longer time, save energy and + * be friendly for battery-driven devices. + */ + schedule_timeout(bdi_longest_inactive()); else - schedule(); + schedule_timeout(msecs_to_jiffies(dirty_writeback_interval * 10)); try_to_freeze(); + /* Back to the main loop */ continue; } - __set_current_state(TASK_RUNNING); - - /* - * This is our real job - check for pending entries in - * bdi_pending_list, and create the tasks that got added - */ - bdi = list_entry(bdi_pending_list.next, struct backing_dev_info, - bdi_list); - list_del_init(&bdi->bdi_list); - spin_unlock_bh(&bdi_lock); - - wb = &bdi->wb; - wb->task = kthread_run(bdi_start_fn, wb, "flush-%s", - dev_name(bdi->dev)); /* - * If task creation fails, then readd the bdi to - * the pending list and force writeout of the bdi - * from this forker thread. That will free some memory - * and we can try again. + * Clear pending bit and wakeup anybody waiting to tear us down. */ - if (IS_ERR(wb->task)) { - wb->task = NULL; - - /* - * Add this 'bdi' to the back, so we get - * a chance to flush other bdi's to free - * memory. - */ - spin_lock_bh(&bdi_lock); - list_add_tail(&bdi->bdi_list, &bdi_pending_list); - spin_unlock_bh(&bdi_lock); - - bdi_flush_io(bdi); - } + clear_bit(BDI_pending, &bdi->state); + smp_mb__after_clear_bit(); + wake_up_bit(&bdi->state, BDI_pending); } return 0; } -static void bdi_add_to_pending(struct rcu_head *head) -{ - struct backing_dev_info *bdi; - - bdi = container_of(head, struct backing_dev_info, rcu_head); - INIT_LIST_HEAD(&bdi->bdi_list); - - spin_lock(&bdi_lock); - list_add_tail(&bdi->bdi_list, &bdi_pending_list); - spin_unlock(&bdi_lock); - - /* - * We are now on the pending list, wake up bdi_forker_task() - * to finish the job and add us back to the active bdi_list - */ - wake_up_process(default_backing_dev_info.wb.task); -} - -/* - * Add the default flusher task that gets created for any bdi - * that has dirty data pending writeout - */ -void static bdi_add_default_flusher_task(struct backing_dev_info *bdi) -{ - if (!bdi_cap_writeback_dirty(bdi)) - return; - - if (WARN_ON(!test_bit(BDI_registered, &bdi->state))) { - printk(KERN_ERR "bdi %p/%s is not registered!\n", - bdi, bdi->name); - return; - } - - /* - * Check with the helper whether to proceed adding a task. Will only - * abort if we two or more simultanous calls to - * bdi_add_default_flusher_task() occured, further additions will block - * waiting for previous additions to finish. - */ - if (!test_and_set_bit(BDI_pending, &bdi->state)) { - list_del_rcu(&bdi->bdi_list); - - /* - * We must wait for the current RCU period to end before - * moving to the pending list. So schedule that operation - * from an RCU callback. - */ - call_rcu(&bdi->rcu_head, bdi_add_to_pending); - } -} - /* * Remove bdi from bdi_list, and ensure that it is no longer visible */ @@ -541,23 +512,16 @@ int bdi_register(struct backing_dev_info *bdi, struct device *parent, const char *fmt, ...) { va_list args; - int ret = 0; struct device *dev; if (bdi->dev) /* The driver needs to use separate queues per device */ - goto exit; + return 0; va_start(args, fmt); dev = device_create_vargs(bdi_class, parent, MKDEV(0, 0), bdi, fmt, args); va_end(args); - if (IS_ERR(dev)) { - ret = PTR_ERR(dev); - goto exit; - } - - spin_lock_bh(&bdi_lock); - list_add_tail_rcu(&bdi->bdi_list, &bdi_list); - spin_unlock_bh(&bdi_lock); + if (IS_ERR(dev)) + return PTR_ERR(dev); bdi->dev = dev; @@ -569,21 +533,21 @@ int bdi_register(struct backing_dev_info *bdi, struct device *parent, if (bdi_cap_flush_forker(bdi)) { struct bdi_writeback *wb = &bdi->wb; - wb->task = kthread_run(bdi_forker_task, wb, "bdi-%s", + wb->task = kthread_run(bdi_forker_thread, wb, "bdi-%s", dev_name(dev)); - if (IS_ERR(wb->task)) { - wb->task = NULL; - ret = -ENOMEM; - - bdi_remove_from_list(bdi); - goto exit; - } + if (IS_ERR(wb->task)) + return PTR_ERR(wb->task); } bdi_debug_register(bdi, dev_name(dev)); set_bit(BDI_registered, &bdi->state); -exit: - return ret; + + spin_lock_bh(&bdi_lock); + list_add_tail_rcu(&bdi->bdi_list, &bdi_list); + spin_unlock_bh(&bdi_lock); + + trace_writeback_bdi_register(bdi); + return 0; } EXPORT_SYMBOL(bdi_register); @@ -598,31 +562,29 @@ EXPORT_SYMBOL(bdi_register_dev); */ static void bdi_wb_shutdown(struct backing_dev_info *bdi) { - struct bdi_writeback *wb; - if (!bdi_cap_writeback_dirty(bdi)) return; /* - * If setup is pending, wait for that to complete first + * Make sure nobody finds us on the bdi_list anymore */ - wait_on_bit(&bdi->state, BDI_pending, bdi_sched_wait, - TASK_UNINTERRUPTIBLE); + bdi_remove_from_list(bdi); /* - * Make sure nobody finds us on the bdi_list anymore + * If setup is pending, wait for that to complete first */ - bdi_remove_from_list(bdi); + wait_on_bit(&bdi->state, BDI_pending, bdi_sched_wait, + TASK_UNINTERRUPTIBLE); /* - * Finally, kill the kernel threads. We don't need to be RCU + * Finally, kill the kernel thread. We don't need to be RCU * safe anymore, since the bdi is gone from visibility. Force * unfreeze of the thread before calling kthread_stop(), otherwise * it would never exet if it is currently stuck in the refrigerator. */ - list_for_each_entry(wb, &bdi->wb_list, list) { - thaw_process(wb->task); - kthread_stop(wb->task); + if (bdi->wb.task) { + thaw_process(bdi->wb.task); + kthread_stop(bdi->wb.task); } } @@ -644,7 +606,9 @@ static void bdi_prune_sb(struct backing_dev_info *bdi) void bdi_unregister(struct backing_dev_info *bdi) { if (bdi->dev) { + trace_writeback_bdi_unregister(bdi); bdi_prune_sb(bdi); + del_timer_sync(&bdi->wb.wakeup_timer); if (!bdi_cap_flush_forker(bdi)) bdi_wb_shutdown(bdi); @@ -655,6 +619,18 @@ void bdi_unregister(struct backing_dev_info *bdi) } EXPORT_SYMBOL(bdi_unregister); +static void bdi_wb_init(struct bdi_writeback *wb, struct backing_dev_info *bdi) +{ + memset(wb, 0, sizeof(*wb)); + + wb->bdi = bdi; + wb->last_old_flush = jiffies; + INIT_LIST_HEAD(&wb->b_dirty); + INIT_LIST_HEAD(&wb->b_io); + INIT_LIST_HEAD(&wb->b_more_io); + setup_timer(&wb->wakeup_timer, wakeup_timer_fn, (unsigned long)bdi); +} + int bdi_init(struct backing_dev_info *bdi) { int i, err; @@ -666,7 +642,6 @@ int bdi_init(struct backing_dev_info *bdi) bdi->max_prop_frac = PROP_FRAC_BASE; spin_lock_init(&bdi->wb_lock); INIT_LIST_HEAD(&bdi->bdi_list); - INIT_LIST_HEAD(&bdi->wb_list); INIT_LIST_HEAD(&bdi->work_list); bdi_wb_init(&bdi->wb, bdi); diff --git a/mm/page-writeback.c b/mm/page-writeback.c index df8202e..0c6258b 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -34,6 +34,7 @@ #include <linux/syscalls.h> #include <linux/buffer_head.h> #include <linux/pagevec.h> +#include <trace/events/writeback.h> /* * After a CPU has dirtied this many pages, balance_dirty_pages_ratelimited @@ -535,11 +536,13 @@ static void balance_dirty_pages(struct address_space *mapping, * threshold otherwise wait until the disk writes catch * up. */ + trace_wbc_balance_dirty_start(&wbc, bdi); if (bdi_nr_reclaimable > bdi_thresh) { writeback_inodes_wb(&bdi->wb, &wbc); pages_written += write_chunk - wbc.nr_to_write; get_dirty_limits(&background_thresh, &dirty_thresh, &bdi_thresh, bdi); + trace_wbc_balance_dirty_written(&wbc, bdi); } /* @@ -565,6 +568,7 @@ static void balance_dirty_pages(struct address_space *mapping, if (pages_written >= write_chunk) break; /* We've done our duty */ + trace_wbc_balance_dirty_wait(&wbc, bdi); __set_current_state(TASK_INTERRUPTIBLE); io_schedule_timeout(pause); @@ -962,6 +966,7 @@ continue_unlock: if (!clear_page_dirty_for_io(page)) goto continue_unlock; + trace_wbc_writepage(wbc, mapping->backing_dev_info); ret = (*writepage)(page, wbc, data); if (unlikely(ret)) { if (ret == AOP_WRITEPAGE_ACTIVATE) { diff --git a/mm/page_io.c b/mm/page_io.c index 31a3b96..2dee975 100644 --- a/mm/page_io.c +++ b/mm/page_io.c @@ -106,7 +106,7 @@ int swap_writepage(struct page *page, struct writeback_control *wbc) goto out; } if (wbc->sync_mode == WB_SYNC_ALL) - rw |= (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_UNPLUG); + rw |= REQ_SYNC | REQ_UNPLUG; count_vm_event(PSWPOUT); set_page_writeback(page); unlock_page(page); |