From 32a90b6e65792260d6212ac52e8f5be140b6f5be Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Fri, 8 Apr 2011 12:01:51 +0200 Subject: ARM: imx: fix usb related build failure for mach-vpr200 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This was broken by 4bd597b (ARM i.MX ehci: do ehci init in board specific functions) and fixes: CC arch/arm/mach-mx3/mach-vpr200.o arch/arm/mach-mx3/mach-vpr200.c:263: error: unknown field 'flags' specified in initializer arch/arm/mach-mx3/mach-vpr200.c:264: warning: initialization makes pointer from integer without a cast by just applying the change to mach-vpr200.c that the other machine files got by 4bd597b. LAKML-Reference: 1302257029-17397-1-git-send-email-u.kleine-koenig@pengutronix.de Acked-by: Marc Reilly Signed-off-by: Uwe Kleine-König Signed-off-by: Sascha Hauer --- arch/arm/mach-mx3/mach-vpr200.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-mx3/mach-vpr200.c b/arch/arm/mach-mx3/mach-vpr200.c index 2cf390f..47a69cb 100644 --- a/arch/arm/mach-mx3/mach-vpr200.c +++ b/arch/arm/mach-mx3/mach-vpr200.c @@ -257,11 +257,16 @@ static const struct fsl_usb2_platform_data otg_device_pdata __initconst = { .workaround = FLS_USB2_WORKAROUND_ENGCM09152, }; +static int vpr200_usbh_init(struct platform_device *pdev) +{ + return mx35_initialize_usb_hw(pdev->id, + MXC_EHCI_INTERFACE_SINGLE_UNI | MXC_EHCI_INTERNAL_PHY); +} + /* USB HOST config */ static const struct mxc_usbh_platform_data usb_host_pdata __initconst = { - .portsc = MXC_EHCI_MODE_SERIAL, - .flags = MXC_EHCI_INTERFACE_SINGLE_UNI | - MXC_EHCI_INTERNAL_PHY, + .init = vpr200_usbh_init, + .portsc = MXC_EHCI_MODE_SERIAL, }; static struct platform_device *devices[] __initdata = { -- cgit v1.1 From f61b9fc27e5b61dbc330696f040cc66ba1dbcbaa Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Mon, 4 Apr 2011 16:06:12 +0200 Subject: ARM: mxs/clock-mx28: fix up name##_set_rate MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For the lcdif clock get_rate looks as follows: read div from HW_CLKCTRL_DIS_LCDIF.DIV return clk_get_rate(clk->parent) / div with clk->parent being ref_pix_clk on my system. ref_pix_clk's rate depends on HW_CLKCTRL_FRAC1.PIXFRAC. The set_rate function for lcdif does: parent_rate = clk_get_rate(clk->parent); based on that calculate frac and div such that parent_rate * 18 / frac / div is near the requested rate. HW_CLKCTRL_FRAC1.PIXFRAC is updated with frac HW_CLKCTRL_DIS_LCDIF.DIV is updated with div For this calculation to be correct parent_rate needs to be initialized not with the clock rate of lcdif's parent (i.e. ref_pix) but that of its grandparent (i.e. ref_pix' parent == pll0_clk). The obvious downside of this patch is that now set_rate(lcdif) changes its parent's rate, too. Still this is better than a wrong rate. Acked-by: Shawn Guo LAKML-Reference: 20110225084950.GA13684@S2101-09.ap.freescale.net Signed-off-by: Uwe Kleine-König Signed-off-by: Sascha Hauer --- arch/arm/mach-mxs/clock-mx28.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-mxs/clock-mx28.c b/arch/arm/mach-mxs/clock-mx28.c index 1ad97fe..5dcc59d 100644 --- a/arch/arm/mach-mxs/clock-mx28.c +++ b/arch/arm/mach-mxs/clock-mx28.c @@ -295,11 +295,11 @@ static int name##_set_rate(struct clk *clk, unsigned long rate) \ unsigned long diff, parent_rate, calc_rate; \ int i; \ \ - parent_rate = clk_get_rate(clk->parent); \ div_max = BM_CLKCTRL_##dr##_DIV >> BP_CLKCTRL_##dr##_DIV; \ bm_busy = BM_CLKCTRL_##dr##_BUSY; \ \ if (clk->parent == &ref_xtal_clk) { \ + parent_rate = clk_get_rate(clk->parent); \ div = DIV_ROUND_UP(parent_rate, rate); \ if (clk == &cpu_clk) { \ div_max = BM_CLKCTRL_CPU_DIV_XTAL >> \ @@ -309,6 +309,11 @@ static int name##_set_rate(struct clk *clk, unsigned long rate) \ if (div == 0 || div > div_max) \ return -EINVAL; \ } else { \ + /* \ + * hack alert: this block modifies clk->parent, too, \ + * so the base to use it the grand parent. \ + */ \ + parent_rate = clk_get_rate(clk->parent->parent); \ rate >>= PARENT_RATE_SHIFT; \ parent_rate >>= PARENT_RATE_SHIFT; \ diff = parent_rate; \ -- cgit v1.1 From 0575b4b83edb0a766a9c1518a5da57780f386340 Mon Sep 17 00:00:00 2001 From: Dave Martin Date: Tue, 16 Nov 2010 13:13:37 +0000 Subject: ARM: mxc: Correct data alignment in headsmp.S for CONFIG_THUMB2_KERNEL MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Directives such as .long and .word do not magically cause the assembler location counter to become aligned in gas. As a result, using these directives in code sections can result in misaligned data words when building a Thumb-2 kernel (CONFIG_THUMB2_KERNEL). This is a Bad Thing, since the ABI permits the compiler to assume that fundamental types of word size or above are word- aligned when accessing them from C. If the data is not really word-aligned, this can cause impaired performance and stray alignment faults in some circumstances. In general, the following rules should be applied when using data word declaration directives inside code sections: * .quad and .double: .align 3 * .long, .word, .single, .float: .align (or .align 2) * .short: No explicit alignment required, since Thumb-2 instructions are always 2 or 4 bytes in size. immediately after an instruction. Signed-off-by: Dave Martin Signed-off-by: Sascha Hauer LAKML-Reference: 1289913217-8672-1-git-send-email-dave.martin@linaro.org Signed-off-by: Uwe Kleine-König Signed-off-by: Sascha Hauer --- arch/arm/plat-mxc/ssi-fiq.S | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch') diff --git a/arch/arm/plat-mxc/ssi-fiq.S b/arch/arm/plat-mxc/ssi-fiq.S index 4ddce56..8397a2d 100644 --- a/arch/arm/plat-mxc/ssi-fiq.S +++ b/arch/arm/plat-mxc/ssi-fiq.S @@ -124,6 +124,8 @@ imx_ssi_fiq_start: 1: @ return from FIQ subs pc, lr, #4 + + .align imx_ssi_fiq_base: .word 0x0 imx_ssi_fiq_rx_buffer: -- cgit v1.1 From b5eee2fdefc89df312034e5e0e6f5dd55c4e9bae Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Mon, 4 Apr 2011 14:29:58 +0200 Subject: ARM: mxc: Add missing lockdep annotation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The irq_set_wake() function of the gpio irq_chip calls enable/disable_irq_wake() on the demultiplex interrupt. That leads to a lockdep warning "INFO: possible recursive locking detected" because irq_set_type() is called under irq_desc->lock and the *_irq_wake() calls take irq_desc->lock of the demux interrupt. Tell lockdep that the gpio irqs are in a different lock class. Documentation/SubmitChecklist: 15: All codepaths have been exercised with all lockdep features enabled. That's a non-optional requirement, AFAICT. Reported-and-tested-by: Arnaud Patard Signed-off-by: Thomas Gleixner LAKML-Reference: alpine.LFD.2.00.1104041416290.19945@localhost6.localdomain6 Signed-off-by: Uwe Kleine-König Signed-off-by: Sascha Hauer --- arch/arm/plat-mxc/gpio.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'arch') diff --git a/arch/arm/plat-mxc/gpio.c b/arch/arm/plat-mxc/gpio.c index 7a10724..6cd6d7f 100644 --- a/arch/arm/plat-mxc/gpio.c +++ b/arch/arm/plat-mxc/gpio.c @@ -295,6 +295,12 @@ static int mxc_gpio_direction_output(struct gpio_chip *chip, return 0; } +/* + * This lock class tells lockdep that GPIO irqs are in a different + * category than their parents, so it won't report false recursion. + */ +static struct lock_class_key gpio_lock_class; + int __init mxc_gpio_init(struct mxc_gpio_port *port, int cnt) { int i, j; @@ -311,6 +317,7 @@ int __init mxc_gpio_init(struct mxc_gpio_port *port, int cnt) __raw_writel(~0, port[i].base + GPIO_ISR); for (j = port[i].virtual_irq_start; j < port[i].virtual_irq_start + 32; j++) { + irq_set_lockdep_class(j, &gpio_lock_class); irq_set_chip_and_handler(j, &gpio_irq_chip, handle_level_irq); set_irq_flags(j, IRQF_VALID); -- cgit v1.1 From 6caa15d0b84d2ea688fd31f4f172c8353463e109 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Tue, 12 Apr 2011 22:51:04 +0200 Subject: ARM: s3c2440: gta02; Register dfbmcs320 device for BT audio interface Register the dfbmcs320 device which provides the PCM DAI for the bluetooth module. Signed-off-by: Lars-Peter Clausen Acked-by: Liam Girdwood Signed-off-by: Mark Brown --- arch/arm/mach-s3c2440/mach-gta02.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'arch') diff --git a/arch/arm/mach-s3c2440/mach-gta02.c b/arch/arm/mach-s3c2440/mach-gta02.c index 0db2411..7166620 100644 --- a/arch/arm/mach-s3c2440/mach-gta02.c +++ b/arch/arm/mach-s3c2440/mach-gta02.c @@ -409,6 +409,10 @@ struct platform_device s3c24xx_pwm_device = { .num_resources = 0, }; +static struct platform_device gta02_dfbmcs320_device = { + .name = "dfbmcs320", +}; + static struct i2c_board_info gta02_i2c_devs[] __initdata = { { I2C_BOARD_INFO("pcf50633", 0x73), @@ -523,6 +527,7 @@ static struct platform_device *gta02_devices[] __initdata = { &s3c_device_iis, &samsung_asoc_dma, &s3c_device_i2c0, + >a02_dfbmcs320_device, >a02_buttons_device, &s3c_device_adc, &s3c_device_ts, -- cgit v1.1 From 0d58a2824d777923b2438107053c6e073c9c5ec1 Mon Sep 17 00:00:00 2001 From: Russell King Date: Fri, 15 Apr 2011 13:26:40 +0100 Subject: ARM: Add new syscalls Add syscalls for name_to_handle_at, open_by_handle_at, clock_adjtime and syncfs. Signed-off-by: Russell King --- arch/arm/include/asm/unistd.h | 4 ++++ arch/arm/kernel/calls.S | 4 ++++ 2 files changed, 8 insertions(+) (limited to 'arch') diff --git a/arch/arm/include/asm/unistd.h b/arch/arm/include/asm/unistd.h index c891eb7..87dbe3e 100644 --- a/arch/arm/include/asm/unistd.h +++ b/arch/arm/include/asm/unistd.h @@ -396,6 +396,10 @@ #define __NR_fanotify_init (__NR_SYSCALL_BASE+367) #define __NR_fanotify_mark (__NR_SYSCALL_BASE+368) #define __NR_prlimit64 (__NR_SYSCALL_BASE+369) +#define __NR_name_to_handle_at (__NR_SYSCALL_BASE+370) +#define __NR_open_by_handle_at (__NR_SYSCALL_BASE+371) +#define __NR_clock_adjtime (__NR_SYSCALL_BASE+372) +#define __NR_syncfs (__NR_SYSCALL_BASE+373) /* * The following SWIs are ARM private. diff --git a/arch/arm/kernel/calls.S b/arch/arm/kernel/calls.S index 5c26ecc..7fbf28c 100644 --- a/arch/arm/kernel/calls.S +++ b/arch/arm/kernel/calls.S @@ -379,6 +379,10 @@ CALL(sys_fanotify_init) CALL(sys_fanotify_mark) CALL(sys_prlimit64) +/* 370 */ CALL(sys_name_to_handle_at) + CALL(sys_open_by_handle_at) + CALL(sys_clock_adjtime) + CALL(sys_syncfs) #ifndef syscalls_counted .equ syscalls_padding, ((NR_syscalls + 3) & ~3) - NR_syscalls #define syscalls_counted -- cgit v1.1 From e9569c1511d2590a27b46b94bafb7acece034e5c Mon Sep 17 00:00:00 2001 From: Jonathan Cameron Date: Thu, 14 Apr 2011 12:11:42 +0100 Subject: ARM: 6881/1: cputype.h uses __attribute_const__ which requires including kernel.h Issue manifests as: In file included from arch/arm/mach-pxa/include/mach/hardware.h:62, from arch/arm/mach-pxa/include/mach/gpio.h:28, from /home/jic23/src/kernel/temp-remove/arch/arm/include/asm/gpio.h:5, from include/linux/gpio.h:7, from drivers/staging/iio/gyro/adis16080_core.c:8: /home/jic23/src/kernel/temp-remove/arch/arm/include/asm/cputype.h:57: error: expected '=', ',', ';', 'asm' or '__attribute__' before 'read_cpuid_id' ... Signed-off-by: Jonathan Cameron Signed-off-by: Russell King --- arch/arm/include/asm/cputype.h | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/arm/include/asm/cputype.h b/arch/arm/include/asm/cputype.h index ed5bc9e..cd4458f 100644 --- a/arch/arm/include/asm/cputype.h +++ b/arch/arm/include/asm/cputype.h @@ -2,6 +2,7 @@ #define __ASM_ARM_CPUTYPE_H #include +#include #define CPUID_ID 0 #define CPUID_CACHETYPE 1 -- cgit v1.1 From af289bfe15fc92ecfbf6d8312713815b33e452c0 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 18 Apr 2011 15:45:44 +0200 Subject: x86, gart: Convert spaces to tabs in enable_gart_translation Probably by copy&paste this function was indented by spaces. Convert this to tabs. Signed-off-by: Joerg Roedel Link: http://lkml.kernel.org/r/1303134346-5805-3-git-send-email-joerg.roedel@amd.com Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/gart.h | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) (limited to 'arch') diff --git a/arch/x86/include/asm/gart.h b/arch/x86/include/asm/gart.h index 43085bf..88c1ebe 100644 --- a/arch/x86/include/asm/gart.h +++ b/arch/x86/include/asm/gart.h @@ -75,17 +75,17 @@ static inline void enable_gart_translation(struct pci_dev *dev, u64 addr) { u32 tmp, ctl; - /* address of the mappings table */ - addr >>= 12; - tmp = (u32) addr<<4; - tmp &= ~0xf; - pci_write_config_dword(dev, AMD64_GARTTABLEBASE, tmp); - - /* Enable GART translation for this hammer. */ - pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &ctl); - ctl |= GARTEN; - ctl &= ~(DISGARTCPU | DISGARTIO); - pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); + /* address of the mappings table */ + addr >>= 12; + tmp = (u32) addr<<4; + tmp &= ~0xf; + pci_write_config_dword(dev, AMD64_GARTTABLEBASE, tmp); + + /* Enable GART translation for this hammer. */ + pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &ctl); + ctl |= GARTEN; + ctl &= ~(DISGARTCPU | DISGARTIO); + pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); } static inline int aperture_valid(u64 aper_base, u32 aper_size, u32 min_size) -- cgit v1.1 From c34151a742d84ae65db2088ea30495063f697fbe Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 18 Apr 2011 15:45:45 +0200 Subject: x86, gart: Set DISTLBWALKPRB bit always The DISTLBWALKPRB bit must be set for the GART because the gatt table is mapped UC. But the current code does not set the bit at boot when the BIOS setup the aperture correctly. Fix that by setting this bit when enabling the GART instead of the other places. Cc: Cc: Borislav Petkov Signed-off-by: Joerg Roedel Link: http://lkml.kernel.org/r/1303134346-5805-4-git-send-email-joerg.roedel@amd.com Signed-off-by: H. Peter Anvin --- arch/x86/include/asm/gart.h | 4 ++-- arch/x86/kernel/aperture_64.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/x86/include/asm/gart.h b/arch/x86/include/asm/gart.h index 88c1ebe..156cd5d 100644 --- a/arch/x86/include/asm/gart.h +++ b/arch/x86/include/asm/gart.h @@ -66,7 +66,7 @@ static inline void gart_set_size_and_enable(struct pci_dev *dev, u32 order) * Don't enable translation but enable GART IO and CPU accesses. * Also, set DISTLBWALKPRB since GART tables memory is UC. */ - ctl = DISTLBWALKPRB | order << 1; + ctl = order << 1; pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); } @@ -83,7 +83,7 @@ static inline void enable_gart_translation(struct pci_dev *dev, u64 addr) /* Enable GART translation for this hammer. */ pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &ctl); - ctl |= GARTEN; + ctl |= GARTEN | DISTLBWALKPRB; ctl &= ~(DISGARTCPU | DISGARTIO); pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); } diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c index 86d1ad4..73fb469 100644 --- a/arch/x86/kernel/aperture_64.c +++ b/arch/x86/kernel/aperture_64.c @@ -499,7 +499,7 @@ out: * Don't enable translation yet but enable GART IO and CPU * accesses and set DISTLBWALKPRB since GART table memory is UC. */ - u32 ctl = DISTLBWALKPRB | aper_order << 1; + u32 ctl = aper_order << 1; bus = amd_nb_bus_dev_ranges[i].bus; dev_base = amd_nb_bus_dev_ranges[i].dev_base; -- cgit v1.1 From 665d3e2af83c8fbd149534db8f57d82fa6fa6753 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 18 Apr 2011 15:45:46 +0200 Subject: x86, gart: Make sure GART does not map physmem above 1TB The GART can only map physical memory below 1TB. Make sure the gart driver in the kernel does not try to map memory above 1TB. Cc: Signed-off-by: Joerg Roedel Link: http://lkml.kernel.org/r/1303134346-5805-5-git-send-email-joerg.roedel@amd.com Signed-off-by: H. Peter Anvin --- arch/x86/kernel/pci-gart_64.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index 82ada01..b117efd 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -81,6 +81,9 @@ static u32 gart_unmapped_entry; #define AGPEXTERN #endif +/* GART can only remap to physical addresses < 1TB */ +#define GART_MAX_PHYS_ADDR (1ULL << 40) + /* backdoor interface to AGP driver */ AGPEXTERN int agp_memory_reserved; AGPEXTERN __u32 *agp_gatt_table; @@ -212,9 +215,13 @@ static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem, size_t size, int dir, unsigned long align_mask) { unsigned long npages = iommu_num_pages(phys_mem, size, PAGE_SIZE); - unsigned long iommu_page = alloc_iommu(dev, npages, align_mask); + unsigned long iommu_page; int i; + if (unlikely(phys_mem + size > GART_MAX_PHYS_ADDR)) + return bad_dma_addr; + + iommu_page = alloc_iommu(dev, npages, align_mask); if (iommu_page == -1) { if (!nonforced_iommu(dev, phys_mem, size)) return phys_mem; -- cgit v1.1 From bcb22a94f68cde70e820dc7280d1c731e8e695f7 Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Sun, 3 Apr 2011 22:47:23 -0300 Subject: ARM: mx5/mx53_loco: Fix build warning related to gpio_keys_button structure Fix the following warning: CC arch/arm/mach-mx5/board-mx53_loco.o arch/arm/mach-mx5/board-mx53_loco.c:203: warning: initialization discards qualifiers from pointer target type Signed-off-by: Fabio Estevam Signed-off-by: Sascha Hauer --- arch/arm/mach-mx5/board-mx53_loco.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-mx5/board-mx53_loco.c b/arch/arm/mach-mx5/board-mx53_loco.c index 10a1bea..6206b11 100644 --- a/arch/arm/mach-mx5/board-mx53_loco.c +++ b/arch/arm/mach-mx5/board-mx53_loco.c @@ -193,7 +193,7 @@ static iomux_v3_cfg_t mx53_loco_pads[] = { .wakeup = wake, \ } -static const struct gpio_keys_button loco_buttons[] __initconst = { +static struct gpio_keys_button loco_buttons[] = { GPIO_BUTTON(MX53_LOCO_POWER, KEY_POWER, 1, "power", 0), GPIO_BUTTON(MX53_LOCO_UI1, KEY_VOLUMEUP, 1, "volume-up", 0), GPIO_BUTTON(MX53_LOCO_UI2, KEY_VOLUMEDOWN, 1, "volume-down", 0), -- cgit v1.1 From 83112e688f5f05dea1e63787db9a6c16b2887a1d Mon Sep 17 00:00:00 2001 From: Andre Przywara Date: Sat, 16 Apr 2011 02:27:53 +0200 Subject: perf, x86: Fix pre-defined cache-misses event for AMD family 15h cpus With AMD cpu family 15h a unit mask was introduced for the Data Cache Miss event (0x041/L1-dcache-load-misses). We need to enable bit 0 (first data cache miss or streaming store to a 64 B cache line) of this mask to proper count data cache misses. Now we set this bit for all families and models. In case a PMU does not implement a unit mask for event 0x041 the bit is ignored. Signed-off-by: Andre Przywara Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/1302913676-14352-2-git-send-email-robert.richter@amd.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_amd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 461f62b..4e16138 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -8,7 +8,7 @@ static __initconst const u64 amd_hw_cache_event_ids [ C(L1D) ] = { [ C(OP_READ) ] = { [ C(RESULT_ACCESS) ] = 0x0040, /* Data Cache Accesses */ - [ C(RESULT_MISS) ] = 0x0041, /* Data Cache Misses */ + [ C(RESULT_MISS) ] = 0x0141, /* Data Cache Misses */ }, [ C(OP_WRITE) ] = { [ C(RESULT_ACCESS) ] = 0x0142, /* Data Cache Refills :system */ -- cgit v1.1 From 855357a21744e488cbee23a47d2b124035160a87 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Sat, 16 Apr 2011 02:27:54 +0200 Subject: perf, x86: Fix AMD family 15h FPU event constraints Depending on the unit mask settings some FPU events may be scheduled only on cpu counter #3. This patch fixes this. Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra Cc: Stephane Eranian Link: http://lkml.kernel.org/r/1302913676-14352-3-git-send-email-robert.richter@amd.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_amd.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 4e16138..cf4e369 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -427,7 +427,9 @@ static __initconst const struct x86_pmu amd_pmu = { * * Exceptions: * + * 0x000 FP PERF_CTL[3], PERF_CTL[5:3] (*) * 0x003 FP PERF_CTL[3] + * 0x004 FP PERF_CTL[3], PERF_CTL[5:3] (*) * 0x00B FP PERF_CTL[3] * 0x00D FP PERF_CTL[3] * 0x023 DE PERF_CTL[2:0] @@ -448,6 +450,8 @@ static __initconst const struct x86_pmu amd_pmu = { * 0x0DF LS PERF_CTL[5:0] * 0x1D6 EX PERF_CTL[5:0] * 0x1D8 EX PERF_CTL[5:0] + * + * (*) depending on the umask all FPU counters may be used */ static struct event_constraint amd_f15_PMC0 = EVENT_CONSTRAINT(0, 0x01, 0); @@ -460,18 +464,28 @@ static struct event_constraint amd_f15_PMC53 = EVENT_CONSTRAINT(0, 0x38, 0); static struct event_constraint * amd_get_event_constraints_f15h(struct cpu_hw_events *cpuc, struct perf_event *event) { - unsigned int event_code = amd_get_event_code(&event->hw); + struct hw_perf_event *hwc = &event->hw; + unsigned int event_code = amd_get_event_code(hwc); switch (event_code & AMD_EVENT_TYPE_MASK) { case AMD_EVENT_FP: switch (event_code) { + case 0x000: + if (!(hwc->config & 0x0000F000ULL)) + break; + if (!(hwc->config & 0x00000F00ULL)) + break; + return &amd_f15_PMC3; + case 0x004: + if (hweight_long(hwc->config & ARCH_PERFMON_EVENTSEL_UMASK) <= 1) + break; + return &amd_f15_PMC3; case 0x003: case 0x00B: case 0x00D: return &amd_f15_PMC3; - default: - return &amd_f15_PMC53; } + return &amd_f15_PMC53; case AMD_EVENT_LS: case AMD_EVENT_DC: case AMD_EVENT_EX_LS: -- cgit v1.1 From 1286eeb2fd22ddb4c56390f957e854ec06bab9fd Mon Sep 17 00:00:00 2001 From: Benoit Cousson Date: Tue, 19 Apr 2011 10:15:36 -0600 Subject: OMAP2+: hwmod data: Fix wrong dma_system end address OMAP2420, 2430 and 3xxx were using the OMAP4 end address that unfortunately is not located at the same base address. Moreover the OMAP4 size was set to 256 instead of 4096. Change all .pa_end to set them to .pa_start + 0xfff Cc: "G, Manjunath Kondaiah" Cc: Benoit Cousson Reported-by: Michael Fillinger Signed-off-by: Benoit Cousson Signed-off-by: Paul Walmsley --- arch/arm/mach-omap2/omap_hwmod_2420_data.c | 2 +- arch/arm/mach-omap2/omap_hwmod_2430_data.c | 2 +- arch/arm/mach-omap2/omap_hwmod_3xxx_data.c | 2 +- arch/arm/mach-omap2/omap_hwmod_44xx_data.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/omap_hwmod_2420_data.c b/arch/arm/mach-omap2/omap_hwmod_2420_data.c index 8eb3ce1..f50e56a 100644 --- a/arch/arm/mach-omap2/omap_hwmod_2420_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_2420_data.c @@ -1782,7 +1782,7 @@ static struct omap_hwmod_irq_info omap2420_dma_system_irqs[] = { static struct omap_hwmod_addr_space omap2420_dma_system_addrs[] = { { .pa_start = 0x48056000, - .pa_end = 0x4a0560ff, + .pa_end = 0x48056fff, .flags = ADDR_TYPE_RT }, }; diff --git a/arch/arm/mach-omap2/omap_hwmod_2430_data.c b/arch/arm/mach-omap2/omap_hwmod_2430_data.c index e6e3810..96753f7 100644 --- a/arch/arm/mach-omap2/omap_hwmod_2430_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_2430_data.c @@ -1915,7 +1915,7 @@ static struct omap_hwmod_irq_info omap2430_dma_system_irqs[] = { static struct omap_hwmod_addr_space omap2430_dma_system_addrs[] = { { .pa_start = 0x48056000, - .pa_end = 0x4a0560ff, + .pa_end = 0x48056fff, .flags = ADDR_TYPE_RT }, }; diff --git a/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c b/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c index b98e2df..9bee155 100644 --- a/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c @@ -2386,7 +2386,7 @@ static struct omap_hwmod_irq_info omap3xxx_dma_system_irqs[] = { static struct omap_hwmod_addr_space omap3xxx_dma_system_addrs[] = { { .pa_start = 0x48056000, - .pa_end = 0x4a0560ff, + .pa_end = 0x48056fff, .flags = ADDR_TYPE_RT }, }; diff --git a/arch/arm/mach-omap2/omap_hwmod_44xx_data.c b/arch/arm/mach-omap2/omap_hwmod_44xx_data.c index 3e88dd3..abc548a 100644 --- a/arch/arm/mach-omap2/omap_hwmod_44xx_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_44xx_data.c @@ -885,7 +885,7 @@ static struct omap_hwmod_ocp_if *omap44xx_dma_system_masters[] = { static struct omap_hwmod_addr_space omap44xx_dma_system_addrs[] = { { .pa_start = 0x4a056000, - .pa_end = 0x4a0560ff, + .pa_end = 0x4a056fff, .flags = ADDR_TYPE_RT }, }; -- cgit v1.1 From 2ea4db65be3c4027ed39da73e1b6a59c8aa6c7c9 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 19 Apr 2011 22:52:58 +0200 Subject: xtensa: Fixup irq conversion fallout and nmi_count Some unnamed moron fatfingered the arguments of the irq chip callbacks to irq_chip instead of irq_data. While at it remove the nmi_count() print in arch_show_interrupts() which has been broken before the irq conversion already. Signed-off-by: Thomas Gleixner --- arch/xtensa/kernel/irq.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) (limited to 'arch') diff --git a/arch/xtensa/kernel/irq.c b/arch/xtensa/kernel/irq.c index d77089d..4340ee0 100644 --- a/arch/xtensa/kernel/irq.c +++ b/arch/xtensa/kernel/irq.c @@ -64,47 +64,41 @@ asmlinkage void do_IRQ(int irq, struct pt_regs *regs) int arch_show_interrupts(struct seq_file *p, int prec) { - int j; - - seq_printf(p, "%*s: ", prec, "NMI"); - for_each_online_cpu(j) - seq_printf(p, "%10u ", nmi_count(j)); - seq_putc(p, '\n'); seq_printf(p, "%*s: ", prec, "ERR"); seq_printf(p, "%10u\n", atomic_read(&irq_err_count)); return 0; } -static void xtensa_irq_mask(struct irq_chip *d) +static void xtensa_irq_mask(struct irq_data *d) { cached_irq_mask &= ~(1 << d->irq); set_sr (cached_irq_mask, INTENABLE); } -static void xtensa_irq_unmask(struct irq_chip *d) +static void xtensa_irq_unmask(struct irq_data *d) { cached_irq_mask |= 1 << d->irq; set_sr (cached_irq_mask, INTENABLE); } -static void xtensa_irq_enable(struct irq_chip *d) +static void xtensa_irq_enable(struct irq_data *d) { variant_irq_enable(d->irq); xtensa_irq_unmask(d->irq); } -static void xtensa_irq_disable(struct irq_chip *d) +static void xtensa_irq_disable(struct irq_data *d) { xtensa_irq_mask(d->irq); variant_irq_disable(d->irq); } -static void xtensa_irq_ack(struct irq_chip *d) +static void xtensa_irq_ack(struct irq_data *d) { set_sr(1 << d->irq, INTCLEAR); } -static int xtensa_irq_retrigger(struct irq_chip *d) +static int xtensa_irq_retrigger(struct irq_data *d) { set_sr (1 << d->irq, INTSET); return 1; -- cgit v1.1 From 19234c0819da0e043a02710488dfd9b242b42eba Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Wed, 20 Apr 2011 00:36:11 +0200 Subject: PM: Add missing syscore_suspend() and syscore_resume() calls Device suspend/resume infrastructure is used not only by the suspend and hibernate code in kernel/power, but also by APM, Xen and the kexec jump feature. However, commit 40dc166cb5dddbd36aa4ad11c03915ea (PM / Core: Introduce struct syscore_ops for core subsystems PM) failed to add syscore_suspend() and syscore_resume() calls to that code, which generally leads to breakage when the features in question are used. To fix this problem, add the missing syscore_suspend() and syscore_resume() calls to arch/x86/kernel/apm_32.c, kernel/kexec.c and drivers/xen/manage.c. Signed-off-by: Rafael J. Wysocki Acked-by: Greg Kroah-Hartman Acked-by: Ian Campbell --- arch/x86/kernel/apm_32.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'arch') diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index 0b4be43..adee12e 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -228,6 +228,7 @@ #include #include #include +#include #include #include @@ -1238,6 +1239,7 @@ static int suspend(int vetoable) local_irq_disable(); sysdev_suspend(PMSG_SUSPEND); + syscore_suspend(); local_irq_enable(); @@ -1255,6 +1257,7 @@ static int suspend(int vetoable) apm_error("suspend", err); err = (err == APM_SUCCESS) ? 0 : -EIO; + syscore_resume(); sysdev_resume(); local_irq_enable(); @@ -1280,6 +1283,7 @@ static void standby(void) local_irq_disable(); sysdev_suspend(PMSG_SUSPEND); + syscore_suspend(); local_irq_enable(); err = set_system_power_state(APM_STATE_STANDBY); @@ -1287,6 +1291,7 @@ static void standby(void) apm_error("standby", err); local_irq_disable(); + syscore_resume(); sysdev_resume(); local_irq_enable(); -- cgit v1.1 From 5cf4c80a145d79db928ff55226e731de55c87644 Mon Sep 17 00:00:00 2001 From: Kevin Hilman Date: Tue, 19 Apr 2011 16:52:56 -0700 Subject: davinci: fix DEBUG_LL code for p2v changes Fixup davinci UART low-level debug code for new ARM generic p2v changes. Based on OMAP changes by Tony Lindgren Cc: Tony Lindgren Signed-off-by: Kevin Hilman --- arch/arm/mach-davinci/include/mach/debug-macro.S | 13 ++++++++----- arch/arm/mach-davinci/include/mach/serial.h | 2 +- 2 files changed, 9 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-davinci/include/mach/debug-macro.S b/arch/arm/mach-davinci/include/mach/debug-macro.S index 9f1befc..f8b7ea4 100644 --- a/arch/arm/mach-davinci/include/mach/debug-macro.S +++ b/arch/arm/mach-davinci/include/mach/debug-macro.S @@ -24,6 +24,9 @@ #define UART_SHIFT 2 +#define davinci_uart_v2p(x) ((x) - PAGE_OFFSET + PLAT_PHYS_OFFSET) +#define davinci_uart_p2v(x) ((x) - PLAT_PHYS_OFFSET + PAGE_OFFSET) + .pushsection .data davinci_uart_phys: .word 0 davinci_uart_virt: .word 0 @@ -34,7 +37,7 @@ davinci_uart_virt: .word 0 /* Use davinci_uart_phys/virt if already configured */ 10: mrc p15, 0, \rp, c1, c0 tst \rp, #1 @ MMU enabled? - ldreq \rp, =__virt_to_phys(davinci_uart_phys) + ldreq \rp, =davinci_uart_v2p(davinci_uart_phys) ldrne \rp, =davinci_uart_phys add \rv, \rp, #4 @ davinci_uart_virt ldr \rp, [\rp, #0] @@ -48,18 +51,18 @@ davinci_uart_virt: .word 0 tst \rp, #1 @ MMU enabled? /* Copy uart phys address from decompressor uart info */ - ldreq \rv, =__virt_to_phys(davinci_uart_phys) + ldreq \rv, =davinci_uart_v2p(davinci_uart_phys) ldrne \rv, =davinci_uart_phys ldreq \rp, =DAVINCI_UART_INFO - ldrne \rp, =__phys_to_virt(DAVINCI_UART_INFO) + ldrne \rp, =davinci_uart_p2v(DAVINCI_UART_INFO) ldr \rp, [\rp, #0] str \rp, [\rv] /* Copy uart virt address from decompressor uart info */ - ldreq \rv, =__virt_to_phys(davinci_uart_virt) + ldreq \rv, =davinci_uart_v2p(davinci_uart_virt) ldrne \rv, =davinci_uart_virt ldreq \rp, =DAVINCI_UART_INFO - ldrne \rp, =__phys_to_virt(DAVINCI_UART_INFO) + ldrne \rp, =davinci_uart_p2v(DAVINCI_UART_INFO) ldr \rp, [\rp, #4] str \rp, [\rv] diff --git a/arch/arm/mach-davinci/include/mach/serial.h b/arch/arm/mach-davinci/include/mach/serial.h index 8051110..c9e6ce1 100644 --- a/arch/arm/mach-davinci/include/mach/serial.h +++ b/arch/arm/mach-davinci/include/mach/serial.h @@ -22,7 +22,7 @@ * * This area sits just below the page tables (see arch/arm/kernel/head.S). */ -#define DAVINCI_UART_INFO (PHYS_OFFSET + 0x3ff8) +#define DAVINCI_UART_INFO (PLAT_PHYS_OFFSET + 0x3ff8) #define DAVINCI_UART0_BASE (IO_PHYS + 0x20000) #define DAVINCI_UART1_BASE (IO_PHYS + 0x20400) -- cgit v1.1 From c708c57e247775928b9a6bce7b4d8d14883bf39b Mon Sep 17 00:00:00 2001 From: Jan Glauber Date: Wed, 20 Apr 2011 10:15:31 +0200 Subject: [S390] prng: prevent access beyond end of stack While initializing the state of the prng only the first 8 bytes of random data where used, the second 8 bytes were read from the memory after the stack. If only 64 bytes of the kernel stack are used and CONFIG_DEBUG_PAGEALLOC is enabled a kernel panic may occur because of the invalid page access. Use the correct multiplicator to stay within the random data buffer. Signed-off-by: Jan Glauber Signed-off-by: Martin Schwidefsky --- arch/s390/crypto/prng.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/s390/crypto/prng.c b/arch/s390/crypto/prng.c index 975e3ab..44bca3f 100644 --- a/arch/s390/crypto/prng.c +++ b/arch/s390/crypto/prng.c @@ -76,7 +76,7 @@ static void prng_seed(int nbytes) /* Add the entropy */ while (nbytes >= 8) { - *((__u64 *)parm_block) ^= *((__u64 *)buf+i*8); + *((__u64 *)parm_block) ^= *((__u64 *)buf+i); prng_add_entropy(); i += 8; nbytes -= 8; -- cgit v1.1 From e4c031b4f2515e9531d71c8aa779799231dbcd0c Mon Sep 17 00:00:00 2001 From: Jan Glauber Date: Wed, 20 Apr 2011 10:15:32 +0200 Subject: [S390] fix page table walk for changing page attributes The page table walk for changing page attributes used the wrong address for pgd/pud/pmd lookups if the range was bigger than a pmd entry. Fix the lookup by using the correct address. Signed-off-by: Jan Glauber Signed-off-by: Martin Schwidefsky --- arch/s390/mm/pageattr.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/s390/mm/pageattr.c b/arch/s390/mm/pageattr.c index 122ffbd..0607e4b 100644 --- a/arch/s390/mm/pageattr.c +++ b/arch/s390/mm/pageattr.c @@ -24,12 +24,13 @@ static void change_page_attr(unsigned long addr, int numpages, WARN_ON_ONCE(1); continue; } - ptep = pte_offset_kernel(pmdp, addr + i * PAGE_SIZE); + ptep = pte_offset_kernel(pmdp, addr); pte = *ptep; pte = set(pte); - ptep_invalidate(&init_mm, addr + i * PAGE_SIZE, ptep); + ptep_invalidate(&init_mm, addr, ptep); *ptep = pte; + addr += PAGE_SIZE; } } -- cgit v1.1 From e35c76cd47c244eaa7a74adaabde4d0a1cadb907 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 20 Apr 2011 10:15:34 +0200 Subject: [S390] pfault: fix token handling f6649a7e "[S390] cleanup lowcore access from external interrupts" changed handling of external interrupts. Instead of letting the external interrupt handlers accessing the per cpu lowcore the entry code of the kernel reads already all fields that are necessary and passes them to the handlers. The pfault interrupt handler was incorrectly converted. It tries to dereference a value which used to be a pointer to a lowcore field. After the conversion however it is not anymore the pointer to the field but its content. So instead of a dereference only a cast is needed to get the task pointer that caused the pfault. Fixes a NULL pointer dereference and a subsequent kernel crash: Unable to handle kernel pointer dereference at virtual kernel address (null) Oops: 0004 [#1] SMP Modules linked in: nfsd exportfs nfs lockd fscache nfs_acl auth_rpcgss sunrpc loop qeth_l3 qeth vmur ccwgroup ext3 jbd mbcache dm_mod dasd_eckd_mod dasd_diag_mod dasd_mod CPU: 0 Not tainted 2.6.38-2-s390x #1 Process cron (pid: 1106, task: 000000001f962f78, ksp: 000000001fa0f9d0) Krnl PSW : 0404200180000000 000000000002c03e (pfault_interrupt+0xa2/0x138) R:0 T:1 IO:0 EX:0 Key:0 M:1 W:0 P:0 AS:0 CC:2 PM:0 EA:3 Krnl GPRS: 0000000000000000 0000000000000001 0000000000000000 0000000000000001 000000001f962f78 0000000000518968 0000000090000002 000000001ff03280 0000000000000000 000000000064f000 000000001f962f78 0000000000002603 0000000006002603 0000000000000000 000000001ff7fe68 000000001ff7fe48 Krnl Code: 000000000002c036: 5820d010 l %r2,16(%r13) 000000000002c03a: 1832 lr %r3,%r2 000000000002c03c: 1a31 ar %r3,%r1 >000000000002c03e: ba23d010 cs %r2,%r3,16(%r13) 000000000002c042: a744fffc brc 4,2c03a 000000000002c046: a7290002 lghi %r2,2 000000000002c04a: e320d0000024 stg %r2,0(%r13) 000000000002c050: 07f0 bcr 15,%r0 Call Trace: ([<000000001f962f78>] 0x1f962f78) [<000000000001acda>] do_extint+0xf6/0x138 [<000000000039b6ca>] ext_no_vtime+0x30/0x34 [<000000007d706e04>] 0x7d706e04 Last Breaking-Event-Address: [<0000000000000000>] 0x0 For stable maintainers: the first kernel which contains this bug is 2.6.37. Reported-by: Stephen Powell Cc: Jonathan Nieder Cc: stable@kernel.org Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/mm/fault.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 9217e33..4cf85fe 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -558,9 +558,9 @@ static void pfault_interrupt(unsigned int ext_int_code, * Get the token (= address of the task structure of the affected task). */ #ifdef CONFIG_64BIT - tsk = *(struct task_struct **) param64; + tsk = (struct task_struct *) param64; #else - tsk = *(struct task_struct **) param32; + tsk = (struct task_struct *) param32; #endif if (subcode & 0x0080) { -- cgit v1.1 From 9ff4cfb3fcfd48b49fdd9be7381b3be340853aa4 Mon Sep 17 00:00:00 2001 From: Carsten Otte Date: Wed, 20 Apr 2011 10:15:36 +0200 Subject: [S390] kvm-390: Let kernel exit SIE instruction on work From: Christian Borntraeger This patch fixes the sie exit on interrupts. The low level interrupt handler returns to the PSW address in pt_regs and not to the PSW address in the lowcore. Without this fix a cpu bound guest might never leave guest state since the host interrupt handler would blindly return to the SIE instruction, even on need_resched and friends. Cc: stable@kernel.org Signed-off-by: Carsten Otte Signed-off-by: Christian Borntraeger Signed-off-by: Martin Schwidefsky --- arch/s390/kvm/sie64a.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/s390/kvm/sie64a.S b/arch/s390/kvm/sie64a.S index 7e9d30d..ab0e041 100644 --- a/arch/s390/kvm/sie64a.S +++ b/arch/s390/kvm/sie64a.S @@ -48,10 +48,10 @@ sie_irq_handler: tm __TI_flags+7(%r2),_TIF_EXIT_SIE jz 0f larl %r2,sie_exit # work pending, leave sie - stg %r2,__LC_RETURN_PSW+8 + stg %r2,SPI_PSW+8(0,%r15) br %r14 0: larl %r2,sie_reenter # re-enter with guest id - stg %r2,__LC_RETURN_PSW+8 + stg %r2,SPI_PSW+8(0,%r15) 1: br %r14 /* -- cgit v1.1 From 24bdb0b62cc82120924762ae6bc85afc8c3f2b26 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 12 Apr 2011 12:19:52 +0100 Subject: xen: do not create the extra e820 region at an addr lower than 4G Do not add the extra e820 region at a physical address lower than 4G because it breaks e820_end_of_low_ram_pfn(). It is OK for us to move the xen_extra_mem_start up and down because this is the index of the memory that can be ballooned in/out - it is memory not available to the kernel during bootup. Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/setup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c index fa0269a..90bac0a 100644 --- a/arch/x86/xen/setup.c +++ b/arch/x86/xen/setup.c @@ -227,7 +227,7 @@ char * __init xen_memory_setup(void) memcpy(map_raw, map, sizeof(map)); e820.nr_map = 0; - xen_extra_mem_start = mem_end; + xen_extra_mem_start = max((1ULL << 32), mem_end); for (i = 0; i < memmap.nr_entries; i++) { unsigned long long end; -- cgit v1.1 From ee176455e28469e2420032aab3db11ac2ae3eaa8 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 19 Apr 2011 14:47:31 +0100 Subject: xen: mask_rw_pte: do not apply the early_ioremap checks on x86_32 The two "is_early_ioremap_ptep" checks in mask_rw_pte are only used on x86_64, in fact early_ioremap is not used at all to setup the initial pagetable on x86_32. Moreover on x86_32 the two checks are wrong because the range pgt_buf_start..pgt_buf_end initially should be mapped RW because the pages in the range are not pagetable pages yet and haven't been cleared yet. Afterwards considering the pgt_buf_start..pgt_buf_end is part of the initial mapping, xen_alloc_pte is capable of turning the ptes RO when they become pagetable pages. Fix the issue and improve the readability of the code providing two different implementation of mask_rw_pte for x86_32 and x86_64. Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/mmu.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index a991b57..aef7af9 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1473,16 +1473,20 @@ static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd) #endif } +#ifdef CONFIG_X86_32 static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) { - unsigned long pfn = pte_pfn(pte); - -#ifdef CONFIG_X86_32 /* If there's an existing pte, then don't allow _PAGE_RW to be set */ if (pte_val_ma(*ptep) & _PAGE_PRESENT) pte = __pte_ma(((pte_val_ma(*ptep) & _PAGE_RW) | ~_PAGE_RW) & pte_val_ma(pte)); -#endif + + return pte; +} +#else /* CONFIG_X86_64 */ +static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) +{ + unsigned long pfn = pte_pfn(pte); /* * If the new pfn is within the range of the newly allocated @@ -1497,6 +1501,7 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) return pte; } +#endif /* CONFIG_X86_64 */ /* Init-time set_pte while constructing initial pagetables, which doesn't allow RO pagetable pages to be remapped RW */ -- cgit v1.1 From cf568c58eb192368f5e796df935704535b54f451 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Wed, 30 Mar 2011 14:31:42 +0200 Subject: mach-ux500: fix i2c0 device setup regression Adding two sets of I2C devices to the same bus doesn't quite work, atleast not anymore. Stash one array and determine how much of it shall be added instead. Signed-off-by: Linus Walleij --- arch/arm/mach-ux500/board-mop500.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-ux500/board-mop500.c b/arch/arm/mach-ux500/board-mop500.c index af91374..6e1907fa 100644 --- a/arch/arm/mach-ux500/board-mop500.c +++ b/arch/arm/mach-ux500/board-mop500.c @@ -178,16 +178,15 @@ static struct i2c_board_info __initdata mop500_i2c0_devices[] = { .irq = NOMADIK_GPIO_TO_IRQ(217), .platform_data = &mop500_tc35892_data, }, -}; - -/* I2C0 devices only available prior to HREFv60 */ -static struct i2c_board_info __initdata mop500_i2c0_old_devices[] = { + /* I2C0 devices only available prior to HREFv60 */ { I2C_BOARD_INFO("tps61052", 0x33), .platform_data = &mop500_tps61052_data, }, }; +#define NUM_PRE_V60_I2C0_DEVICES 1 + static struct i2c_board_info __initdata mop500_i2c2_devices[] = { { /* lp5521 LED driver, 1st device */ @@ -425,6 +424,8 @@ static void __init mop500_uart_init(void) static void __init mop500_init_machine(void) { + int i2c0_devs; + /* * The HREFv60 board removed a GPIO expander and routed * all these GPIO pins to the internal GPIO controller @@ -448,11 +449,11 @@ static void __init mop500_init_machine(void) platform_device_register(&ab8500_device); - i2c_register_board_info(0, mop500_i2c0_devices, - ARRAY_SIZE(mop500_i2c0_devices)); - if (!machine_is_hrefv60()) - i2c_register_board_info(0, mop500_i2c0_old_devices, - ARRAY_SIZE(mop500_i2c0_old_devices)); + i2c0_devs = ARRAY_SIZE(mop500_i2c0_devices); + if (machine_is_hrefv60()) + i2c0_devs -= NUM_PRE_V60_I2C0_DEVICES; + + i2c_register_board_info(0, mop500_i2c0_devices, i2c0_devs); i2c_register_board_info(2, mop500_i2c2_devices, ARRAY_SIZE(mop500_i2c2_devices)); } -- cgit v1.1 From 2df122f52fd31c327e0aa05caf6017ecd7867d5f Mon Sep 17 00:00:00 2001 From: Tomi Valkeinen Date: Mon, 4 Apr 2011 09:26:19 +0300 Subject: OMAP4: clock data: Change DSS clock aliases MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit DSS driver has used fck and ick clocks on OMAP2/3 to get DSS HW up and running, and also to get the pixel clock's source clock rate from the fck. On OMAP4 the clock data is set up in a different way, as there's no ick, dss_fck points to a fake clock which just affects DSS's MODULEMODE, and dss_dss_clk if the DSS_FCK. >From DSS driver's point of view the dss_fck sounds like an ick, and dss_dss_clk is the fck. While this is not entirely correct from HW point of view, especially for the ick, configuring the clock aliases that way makes DSS "just work" with OMAP4's clock setup. In the (hopefully near) future DSS driver will be reworked to use pm_runtime support which should clean up the clock code. Signed-off-by: Tomi Valkeinen Cc: Benoît Cousson Signed-off-by: Paul Walmsley --- arch/arm/mach-omap2/clock44xx_data.c | 9 ++------- 1 file changed, 2 insertions(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/clock44xx_data.c b/arch/arm/mach-omap2/clock44xx_data.c index 276992d..8c96567 100644 --- a/arch/arm/mach-omap2/clock44xx_data.c +++ b/arch/arm/mach-omap2/clock44xx_data.c @@ -3116,14 +3116,9 @@ static struct omap_clk omap44xx_clks[] = { CLK(NULL, "dsp_fck", &dsp_fck, CK_443X), CLK("omapdss_dss", "sys_clk", &dss_sys_clk, CK_443X), CLK("omapdss_dss", "tv_clk", &dss_tv_clk, CK_443X), - CLK("omapdss_dss", "dss_clk", &dss_dss_clk, CK_443X), CLK("omapdss_dss", "video_clk", &dss_48mhz_clk, CK_443X), - CLK("omapdss_dss", "fck", &dss_fck, CK_443X), - /* - * On OMAP4, DSS ick is a dummy clock; this is needed for compatibility - * with OMAP2/3. - */ - CLK("omapdss_dss", "ick", &dummy_ck, CK_443X), + CLK("omapdss_dss", "fck", &dss_dss_clk, CK_443X), + CLK("omapdss_dss", "ick", &dss_fck, CK_443X), CLK(NULL, "efuse_ctrl_cust_fck", &efuse_ctrl_cust_fck, CK_443X), CLK(NULL, "emif1_fck", &emif1_fck, CK_443X), CLK(NULL, "emif2_fck", &emif2_fck, CK_443X), -- cgit v1.1 From 8bc2e98bcb280009cb0f85ce64e5f79b1669f9ff Mon Sep 17 00:00:00 2001 From: Eduardo Valentin Date: Wed, 13 Apr 2011 18:21:06 +0300 Subject: OMAP2+: PM: Fix the saving of CM_AUTOIDLE_PLL register on scratchpad area The saving of CCR.CM_AUTOIDLE_PLL is done in scratchpad area. However, in current code, the saving is done for CM_AUTOIDLE2_PLL (offset 0x34) instead of CM_AUTOIDLE_PLL (offset 0x30). This patch changes the code to save the correct register. Signed-off-by: Eduardo Valentin Signed-off-by: Paul Walmsley --- arch/arm/mach-omap2/control.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/control.c b/arch/arm/mach-omap2/control.c index 6952794..df0c75c 100644 --- a/arch/arm/mach-omap2/control.c +++ b/arch/arm/mach-omap2/control.c @@ -317,7 +317,7 @@ void omap3_save_scratchpad_contents(void) prcm_block_contents.cm_clken_pll = omap2_cm_read_mod_reg(PLL_MOD, CM_CLKEN); prcm_block_contents.cm_autoidle_pll = - omap2_cm_read_mod_reg(PLL_MOD, OMAP3430_CM_AUTOIDLE_PLL); + omap2_cm_read_mod_reg(PLL_MOD, CM_AUTOIDLE); prcm_block_contents.cm_clksel1_pll = omap2_cm_read_mod_reg(PLL_MOD, OMAP3430_CM_CLKSEL1_PLL); prcm_block_contents.cm_clksel2_pll = -- cgit v1.1 From a8ae645c014bc01090367de84f7601ad11628971 Mon Sep 17 00:00:00 2001 From: Eduardo Valentin Date: Wed, 13 Apr 2011 18:21:07 +0300 Subject: OMAP3: PM: Do not rely on ROM code to restore CM_AUTOIDLE_PLL.AUTO_PERIPH_DPLL As per OMAP3 erratum (i671), ROM code adds extra latencies while restoring CM_AUTOIDLE_PLL register, if AUTO_PERIPH_DPLL is equal to 1. This patch stores 0's in scratchpad content area corresponding to AUTO_PERIPH_DPLL, to prevent ROM code to try to lock per DPLL, since it won't respect proper programing scheme. This register is then stored in prcm context. The saving and restore is now done by kernel side. Here follow the erratum description DESCRIPTION After OFF mode transition, among many restorations, the ROM Code restores the CM_AUTOIDLE_PLL register, and after that, it tries to relock the PER DPLL. In case the restoration data stored in scratchpad memory contains a field CM_AUTOIDLE_PLL.AUTO_PERIPH_DPLL = 1, then the way the ROM Code restores and locks the PER DPLL does not respect the PER DPLL programming scheme. In that case, the DPLL might not lock. Meanwhile, when trying to lock the PER DPLL, the ROM Code does not hang. Only extra latencies are introduced at wake-up. WORKAROUND When saving the context-restore structure in scratchpad memory, in order to respect the PER DPLL programming scheme, it is advised to store 0 in the CM_AUTOIDLE_PLL.AUTO_PERIPH_DPLL field of the saved structure. After wake-up, the application should store in CM_AUTOIDLE_PLL register the right desired value. Signed-off-by: Eduardo Valentin Signed-off-by: Paul Walmsley --- arch/arm/mach-omap2/cm2xxx_3xxx.c | 17 +++++++++++++++++ arch/arm/mach-omap2/control.c | 8 +++++++- 2 files changed, 24 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/cm2xxx_3xxx.c b/arch/arm/mach-omap2/cm2xxx_3xxx.c index 9d0dec8..38830d8 100644 --- a/arch/arm/mach-omap2/cm2xxx_3xxx.c +++ b/arch/arm/mach-omap2/cm2xxx_3xxx.c @@ -247,6 +247,7 @@ struct omap3_cm_regs { u32 per_cm_clksel; u32 emu_cm_clksel; u32 emu_cm_clkstctrl; + u32 pll_cm_autoidle; u32 pll_cm_autoidle2; u32 pll_cm_clksel4; u32 pll_cm_clksel5; @@ -319,6 +320,15 @@ void omap3_cm_save_context(void) omap2_cm_read_mod_reg(OMAP3430_EMU_MOD, CM_CLKSEL1); cm_context.emu_cm_clkstctrl = omap2_cm_read_mod_reg(OMAP3430_EMU_MOD, OMAP2_CM_CLKSTCTRL); + /* + * As per erratum i671, ROM code does not respect the PER DPLL + * programming scheme if CM_AUTOIDLE_PLL.AUTO_PERIPH_DPLL == 1. + * In this case, even though this register has been saved in + * scratchpad contents, we need to restore AUTO_PERIPH_DPLL + * by ourselves. So, we need to save it anyway. + */ + cm_context.pll_cm_autoidle = + omap2_cm_read_mod_reg(PLL_MOD, CM_AUTOIDLE); cm_context.pll_cm_autoidle2 = omap2_cm_read_mod_reg(PLL_MOD, CM_AUTOIDLE2); cm_context.pll_cm_clksel4 = @@ -441,6 +451,13 @@ void omap3_cm_restore_context(void) CM_CLKSEL1); omap2_cm_write_mod_reg(cm_context.emu_cm_clkstctrl, OMAP3430_EMU_MOD, OMAP2_CM_CLKSTCTRL); + /* + * As per erratum i671, ROM code does not respect the PER DPLL + * programming scheme if CM_AUTOIDLE_PLL.AUTO_PERIPH_DPLL == 1. + * In this case, we need to restore AUTO_PERIPH_DPLL by ourselves. + */ + omap2_cm_write_mod_reg(cm_context.pll_cm_autoidle, PLL_MOD, + CM_AUTOIDLE); omap2_cm_write_mod_reg(cm_context.pll_cm_autoidle2, PLL_MOD, CM_AUTOIDLE2); omap2_cm_write_mod_reg(cm_context.pll_cm_clksel4, PLL_MOD, diff --git a/arch/arm/mach-omap2/control.c b/arch/arm/mach-omap2/control.c index df0c75c..da53ba3 100644 --- a/arch/arm/mach-omap2/control.c +++ b/arch/arm/mach-omap2/control.c @@ -316,8 +316,14 @@ void omap3_save_scratchpad_contents(void) omap2_cm_read_mod_reg(WKUP_MOD, CM_CLKSEL); prcm_block_contents.cm_clken_pll = omap2_cm_read_mod_reg(PLL_MOD, CM_CLKEN); + /* + * As per erratum i671, ROM code does not respect the PER DPLL + * programming scheme if CM_AUTOIDLE_PLL..AUTO_PERIPH_DPLL == 1. + * Then, in anycase, clear these bits to avoid extra latencies. + */ prcm_block_contents.cm_autoidle_pll = - omap2_cm_read_mod_reg(PLL_MOD, CM_AUTOIDLE); + omap2_cm_read_mod_reg(PLL_MOD, CM_AUTOIDLE) & + ~OMAP3430_AUTO_PERIPH_DPLL_MASK; prcm_block_contents.cm_clksel1_pll = omap2_cm_read_mod_reg(PLL_MOD, OMAP3430_CM_CLKSEL1_PLL); prcm_block_contents.cm_clksel2_pll = -- cgit v1.1 From f95440ca5bdd3ed3e31c2fbad07b9056b31ad18c Mon Sep 17 00:00:00 2001 From: "Avinash.H.M" Date: Tue, 5 Apr 2011 21:10:15 +0530 Subject: OMAP2/3: hwmod: fix gpio-reset timeouts seen during bootup. GPIO module expects the debounce clocks to be enabled during reset. It doesn't reset properly and timeouts are seen, if this clock isn't enabled during reset. Add the HWMOD_CONTROL_OPT_CLKS_IN_RESET flags to the GPIO HWMODs, with which the debounce clocks are enabled during reset. Cc: Rajendra Nayak Cc: Paul Walmsley Cc: Benoit Cousson Cc: Kevin Hilman Signed-off-by: Avinash.H.M Signed-off-by: Paul Walmsley --- arch/arm/mach-omap2/omap_hwmod_2420_data.c | 4 ++++ arch/arm/mach-omap2/omap_hwmod_2430_data.c | 5 +++++ arch/arm/mach-omap2/omap_hwmod_3xxx_data.c | 6 ++++++ 3 files changed, 15 insertions(+) (limited to 'arch') diff --git a/arch/arm/mach-omap2/omap_hwmod_2420_data.c b/arch/arm/mach-omap2/omap_hwmod_2420_data.c index f50e56a..c4d0ae8 100644 --- a/arch/arm/mach-omap2/omap_hwmod_2420_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_2420_data.c @@ -1639,6 +1639,7 @@ static struct omap_hwmod_ocp_if *omap2420_gpio1_slaves[] = { static struct omap_hwmod omap2420_gpio1_hwmod = { .name = "gpio1", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap242x_gpio1_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap242x_gpio1_irqs), .main_clk = "gpios_fck", @@ -1669,6 +1670,7 @@ static struct omap_hwmod_ocp_if *omap2420_gpio2_slaves[] = { static struct omap_hwmod omap2420_gpio2_hwmod = { .name = "gpio2", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap242x_gpio2_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap242x_gpio2_irqs), .main_clk = "gpios_fck", @@ -1699,6 +1701,7 @@ static struct omap_hwmod_ocp_if *omap2420_gpio3_slaves[] = { static struct omap_hwmod omap2420_gpio3_hwmod = { .name = "gpio3", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap242x_gpio3_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap242x_gpio3_irqs), .main_clk = "gpios_fck", @@ -1729,6 +1732,7 @@ static struct omap_hwmod_ocp_if *omap2420_gpio4_slaves[] = { static struct omap_hwmod omap2420_gpio4_hwmod = { .name = "gpio4", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap242x_gpio4_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap242x_gpio4_irqs), .main_clk = "gpios_fck", diff --git a/arch/arm/mach-omap2/omap_hwmod_2430_data.c b/arch/arm/mach-omap2/omap_hwmod_2430_data.c index 96753f7..9682dd5 100644 --- a/arch/arm/mach-omap2/omap_hwmod_2430_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_2430_data.c @@ -1742,6 +1742,7 @@ static struct omap_hwmod_ocp_if *omap2430_gpio1_slaves[] = { static struct omap_hwmod omap2430_gpio1_hwmod = { .name = "gpio1", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap243x_gpio1_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap243x_gpio1_irqs), .main_clk = "gpios_fck", @@ -1772,6 +1773,7 @@ static struct omap_hwmod_ocp_if *omap2430_gpio2_slaves[] = { static struct omap_hwmod omap2430_gpio2_hwmod = { .name = "gpio2", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap243x_gpio2_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap243x_gpio2_irqs), .main_clk = "gpios_fck", @@ -1802,6 +1804,7 @@ static struct omap_hwmod_ocp_if *omap2430_gpio3_slaves[] = { static struct omap_hwmod omap2430_gpio3_hwmod = { .name = "gpio3", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap243x_gpio3_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap243x_gpio3_irqs), .main_clk = "gpios_fck", @@ -1832,6 +1835,7 @@ static struct omap_hwmod_ocp_if *omap2430_gpio4_slaves[] = { static struct omap_hwmod omap2430_gpio4_hwmod = { .name = "gpio4", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap243x_gpio4_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap243x_gpio4_irqs), .main_clk = "gpios_fck", @@ -1862,6 +1866,7 @@ static struct omap_hwmod_ocp_if *omap2430_gpio5_slaves[] = { static struct omap_hwmod omap2430_gpio5_hwmod = { .name = "gpio5", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap243x_gpio5_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap243x_gpio5_irqs), .main_clk = "gpio5_fck", diff --git a/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c b/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c index 9bee155..909a84d 100644 --- a/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c +++ b/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c @@ -2141,6 +2141,7 @@ static struct omap_hwmod_ocp_if *omap3xxx_gpio1_slaves[] = { static struct omap_hwmod omap3xxx_gpio1_hwmod = { .name = "gpio1", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap3xxx_gpio1_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap3xxx_gpio1_irqs), .main_clk = "gpio1_ick", @@ -2177,6 +2178,7 @@ static struct omap_hwmod_ocp_if *omap3xxx_gpio2_slaves[] = { static struct omap_hwmod omap3xxx_gpio2_hwmod = { .name = "gpio2", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap3xxx_gpio2_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap3xxx_gpio2_irqs), .main_clk = "gpio2_ick", @@ -2213,6 +2215,7 @@ static struct omap_hwmod_ocp_if *omap3xxx_gpio3_slaves[] = { static struct omap_hwmod omap3xxx_gpio3_hwmod = { .name = "gpio3", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap3xxx_gpio3_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap3xxx_gpio3_irqs), .main_clk = "gpio3_ick", @@ -2249,6 +2252,7 @@ static struct omap_hwmod_ocp_if *omap3xxx_gpio4_slaves[] = { static struct omap_hwmod omap3xxx_gpio4_hwmod = { .name = "gpio4", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap3xxx_gpio4_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap3xxx_gpio4_irqs), .main_clk = "gpio4_ick", @@ -2285,6 +2289,7 @@ static struct omap_hwmod_ocp_if *omap3xxx_gpio5_slaves[] = { static struct omap_hwmod omap3xxx_gpio5_hwmod = { .name = "gpio5", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap3xxx_gpio5_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap3xxx_gpio5_irqs), .main_clk = "gpio5_ick", @@ -2321,6 +2326,7 @@ static struct omap_hwmod_ocp_if *omap3xxx_gpio6_slaves[] = { static struct omap_hwmod omap3xxx_gpio6_hwmod = { .name = "gpio6", + .flags = HWMOD_CONTROL_OPT_CLKS_IN_RESET, .mpu_irqs = omap3xxx_gpio6_irqs, .mpu_irqs_cnt = ARRAY_SIZE(omap3xxx_gpio6_irqs), .main_clk = "gpio6_ick", -- cgit v1.1 From 37f8527dbfd05af0f670aa02370d0c4cca7fbda6 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 20 Apr 2011 19:19:10 -0700 Subject: Revert "x86, NUMA: Fix fakenuma boot failure" Andreas Herrmann reported that 7d6b46707f24 ("x86, NUMA: Fix fakenuma boot failure") causes certain physical NUMA topologies (for example AMD Magny-Cours) to move sibling cpus to a single node when in reality they are in separate domains. This may result in some nodes being completely void of cpus, which doesn't accurately represent the correct topology. The system will boot, but will have suboptimal NUMA performance. This commit was intended as a fix for NUMA emulation, but should not cause a regression for real NUMA machines as a side effect. ( There will be a separate fix for the numa-debug code, which will not affect physical topologies. ) Reported-by: Andreas Herrmann Signed-off-by: David Rientjes Acked-by: KOSAKI Motohiro Cc: Tejun Heo Cc: Linus Torvalds Link: http://lkml.kernel.org/r/alpine.DEB.2.00.1104201918110.12634@chino.kir.corp.google.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/smpboot.c | 23 ----------------------- 1 file changed, 23 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 8ed8908..c2871d3 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -312,26 +312,6 @@ void __cpuinit smp_store_cpu_info(int id) identify_secondary_cpu(c); } -static void __cpuinit check_cpu_siblings_on_same_node(int cpu1, int cpu2) -{ - int node1 = early_cpu_to_node(cpu1); - int node2 = early_cpu_to_node(cpu2); - - /* - * Our CPU scheduler assumes all logical cpus in the same physical cpu - * share the same node. But, buggy ACPI or NUMA emulation might assign - * them to different node. Fix it. - */ - if (node1 != node2) { - pr_warning("CPU %d in node %d and CPU %d in node %d are in the same physical CPU. forcing same node %d\n", - cpu1, node1, cpu2, node2, node2); - - numa_remove_cpu(cpu1); - numa_set_node(cpu1, node2); - numa_add_cpu(cpu1); - } -} - static void __cpuinit link_thread_siblings(int cpu1, int cpu2) { cpumask_set_cpu(cpu1, cpu_sibling_mask(cpu2)); @@ -340,7 +320,6 @@ static void __cpuinit link_thread_siblings(int cpu1, int cpu2) cpumask_set_cpu(cpu2, cpu_core_mask(cpu1)); cpumask_set_cpu(cpu1, cpu_llc_shared_mask(cpu2)); cpumask_set_cpu(cpu2, cpu_llc_shared_mask(cpu1)); - check_cpu_siblings_on_same_node(cpu1, cpu2); } @@ -382,12 +361,10 @@ void __cpuinit set_cpu_sibling_map(int cpu) per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i)) { cpumask_set_cpu(i, cpu_llc_shared_mask(cpu)); cpumask_set_cpu(cpu, cpu_llc_shared_mask(i)); - check_cpu_siblings_on_same_node(cpu, i); } if (c->phys_proc_id == cpu_data(i).phys_proc_id) { cpumask_set_cpu(i, cpu_core_mask(cpu)); cpumask_set_cpu(cpu, cpu_core_mask(i)); - check_cpu_siblings_on_same_node(cpu, i); /* * Does this new cpu bringup a new core? */ -- cgit v1.1 From 7a6c6547825a2324faa76cff856db11d78de075e Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 20 Apr 2011 19:19:13 -0700 Subject: x86, numa: Fix cpu nodemasks for NUMA emulation and CONFIG_DEBUG_PER_CPU_MAPS The cpu<->node mappings under CONFIG_DEBUG_PER_CPU_MAPS=y when NUMA emulation is enabled is currently broken because it does not iterate through every emulated node and bind cpus that have affinity to it. NUMA emulation should bind each cpu to every local node to accurately represent the true NUMA topology of the underlying machine. debug_cpumask_set_cpu() needs to be fixed at the same time so that the debugging information that it emits shows the new cpumask of the node being assigned when the cpu is being added or removed. It can now take responsibility of setting or clearing the cpu itself to remove the need for duplicate code. Also change its last parameter, "enable", to have the correct bool type since it can only be true or false. -v2: Fix the return statements, by Kosaki Motohiro Acked-and-Tested-by: KOSAKI Motohiro Signed-off-by: David Rientjes Cc: Andreas Herrmann Cc: Tejun Heo Cc: Linus Torvalds Link: http://lkml.kernel.org/r/alpine.DEB.2.00.1104201918470.12634@chino.kir.corp.google.com Signed-off-by: Ingo Molnar --- arch/x86/include/asm/numa.h | 2 +- arch/x86/mm/numa.c | 31 +++++++++++++------------------ arch/x86/mm/numa_emulation.c | 20 ++++++-------------- 3 files changed, 20 insertions(+), 33 deletions(-) (limited to 'arch') diff --git a/arch/x86/include/asm/numa.h b/arch/x86/include/asm/numa.h index 3d4dab4..a50fc9f 100644 --- a/arch/x86/include/asm/numa.h +++ b/arch/x86/include/asm/numa.h @@ -51,7 +51,7 @@ static inline void numa_remove_cpu(int cpu) { } #endif /* CONFIG_NUMA */ #ifdef CONFIG_DEBUG_PER_CPU_MAPS -struct cpumask __cpuinit *debug_cpumask_set_cpu(int cpu, int enable); +void debug_cpumask_set_cpu(int cpu, int node, bool enable); #endif #endif /* _ASM_X86_NUMA_H */ diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index 9559d36..745258d 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -213,53 +213,48 @@ int early_cpu_to_node(int cpu) return per_cpu(x86_cpu_to_node_map, cpu); } -struct cpumask __cpuinit *debug_cpumask_set_cpu(int cpu, int enable) +void debug_cpumask_set_cpu(int cpu, int node, bool enable) { - int node = early_cpu_to_node(cpu); struct cpumask *mask; char buf[64]; if (node == NUMA_NO_NODE) { /* early_cpu_to_node() already emits a warning and trace */ - return NULL; + return; } mask = node_to_cpumask_map[node]; if (!mask) { pr_err("node_to_cpumask_map[%i] NULL\n", node); dump_stack(); - return NULL; + return; } + if (enable) + cpumask_set_cpu(cpu, mask); + else + cpumask_clear_cpu(cpu, mask); + cpulist_scnprintf(buf, sizeof(buf), mask); printk(KERN_DEBUG "%s cpu %d node %d: mask now %s\n", enable ? "numa_add_cpu" : "numa_remove_cpu", cpu, node, buf); - return mask; + return; } # ifndef CONFIG_NUMA_EMU -static void __cpuinit numa_set_cpumask(int cpu, int enable) +static void __cpuinit numa_set_cpumask(int cpu, bool enable) { - struct cpumask *mask; - - mask = debug_cpumask_set_cpu(cpu, enable); - if (!mask) - return; - - if (enable) - cpumask_set_cpu(cpu, mask); - else - cpumask_clear_cpu(cpu, mask); + debug_cpumask_set_cpu(cpu, early_cpu_to_node(cpu), enable); } void __cpuinit numa_add_cpu(int cpu) { - numa_set_cpumask(cpu, 1); + numa_set_cpumask(cpu, true); } void __cpuinit numa_remove_cpu(int cpu) { - numa_set_cpumask(cpu, 0); + numa_set_cpumask(cpu, false); } # endif /* !CONFIG_NUMA_EMU */ diff --git a/arch/x86/mm/numa_emulation.c b/arch/x86/mm/numa_emulation.c index ad091e4..de84cc1 100644 --- a/arch/x86/mm/numa_emulation.c +++ b/arch/x86/mm/numa_emulation.c @@ -454,10 +454,9 @@ void __cpuinit numa_remove_cpu(int cpu) cpumask_clear_cpu(cpu, node_to_cpumask_map[i]); } #else /* !CONFIG_DEBUG_PER_CPU_MAPS */ -static void __cpuinit numa_set_cpumask(int cpu, int enable) +static void __cpuinit numa_set_cpumask(int cpu, bool enable) { - struct cpumask *mask; - int nid, physnid, i; + int nid, physnid; nid = early_cpu_to_node(cpu); if (nid == NUMA_NO_NODE) { @@ -467,28 +466,21 @@ static void __cpuinit numa_set_cpumask(int cpu, int enable) physnid = emu_nid_to_phys[nid]; - for_each_online_node(i) { + for_each_online_node(nid) { if (emu_nid_to_phys[nid] != physnid) continue; - mask = debug_cpumask_set_cpu(cpu, enable); - if (!mask) - return; - - if (enable) - cpumask_set_cpu(cpu, mask); - else - cpumask_clear_cpu(cpu, mask); + debug_cpumask_set_cpu(cpu, nid, enable); } } void __cpuinit numa_add_cpu(int cpu) { - numa_set_cpumask(cpu, 1); + numa_set_cpumask(cpu, true); } void __cpuinit numa_remove_cpu(int cpu) { - numa_set_cpumask(cpu, 0); + numa_set_cpumask(cpu, false); } #endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ -- cgit v1.1 From d9b41e0b54fd7e164daf1e9c539c1070398aa02e Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 20 Apr 2011 19:27:13 -0700 Subject: [PARISC] set memory ranges in N_NORMAL_MEMORY when onlined When a DISCONTIGMEM memory range is brought online as a NUMA node, it also needs to have its bet set in N_NORMAL_MEMORY. This is necessary for generic kernel code that utilizes N_NORMAL_MEMORY as a subset of N_ONLINE for memory savings. These types of hacks can hopefully be removed once DISCONTIGMEM is either removed or abstracted away from CONFIG_NUMA. Fixes a panic in the slub code which only initializes structures for N_NORMAL_MEMORY to save memory: Backtrace: [<000000004021c938>] add_partial+0x28/0x98 [<000000004021faa0>] __slab_free+0x1d0/0x1d8 [<000000004021fd04>] kmem_cache_free+0xc4/0x128 [<000000004033bf9c>] ida_get_new_above+0x21c/0x2c0 [<00000000402a8980>] sysfs_new_dirent+0xd0/0x238 [<00000000402a974c>] create_dir+0x5c/0x168 [<00000000402a9ab0>] sysfs_create_dir+0x98/0x128 [<000000004033d6c4>] kobject_add_internal+0x114/0x258 [<000000004033d9ac>] kobject_add_varg+0x7c/0xa0 [<000000004033df20>] kobject_add+0x50/0x90 [<000000004033dfb4>] kobject_create_and_add+0x54/0xc8 [<00000000407862a0>] cgroup_init+0x138/0x1f0 [<000000004077ce50>] start_kernel+0x5a0/0x840 [<000000004011fa3c>] start_parisc+0xa4/0xb8 [<00000000404bb034>] packet_ioctl+0x16c/0x208 [<000000004049ac30>] ip_mroute_setsockopt+0x260/0xf20 Signed-off-by: David Rientjes Cc: stable@kernel.org Signed-off-by: James Bottomley --- arch/parisc/mm/init.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c index b7ed8d7..b1d1262 100644 --- a/arch/parisc/mm/init.c +++ b/arch/parisc/mm/init.c @@ -266,8 +266,10 @@ static void __init setup_bootmem(void) } memset(pfnnid_map, 0xff, sizeof(pfnnid_map)); - for (i = 0; i < npmem_ranges; i++) + for (i = 0; i < npmem_ranges; i++) { + node_set_state(i, N_NORMAL_MEMORY); node_set_online(i); + } #endif /* -- cgit v1.1 From e2a85aecebc03d165bc2dcd233deadd5dd97ea9f Mon Sep 17 00:00:00 2001 From: Andrea Galbusera Date: Thu, 21 Apr 2011 02:21:21 +0000 Subject: powerpc: Fix multicast problem in fs_enet driver mac-fec.c was setting individual UDP address registers instead of multicast group address registers when joining a multicast group. This prevented from correctly receiving UDP multicast packets. According to datasheet, replaced hash_table_high and hash_table_low with grp_hash_table_high and grp_hash_table_low respectively. Also renamed hash_table_* with grp_hash_table_* in struct fec declaration for 8xx: these registers are used only for multicast there. Tested on a MPC5121 based board. Build tested also against mpc866_ads_defconfig. Signed-off-by: Andrea Galbusera Signed-off-by: David S. Miller --- arch/powerpc/include/asm/8xx_immap.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/powerpc/include/asm/8xx_immap.h b/arch/powerpc/include/asm/8xx_immap.h index 6b6dc20..bdf0563 100644 --- a/arch/powerpc/include/asm/8xx_immap.h +++ b/arch/powerpc/include/asm/8xx_immap.h @@ -393,8 +393,8 @@ typedef struct fec { uint fec_addr_low; /* lower 32 bits of station address */ ushort fec_addr_high; /* upper 16 bits of station address */ ushort res1; /* reserved */ - uint fec_hash_table_high; /* upper 32-bits of hash table */ - uint fec_hash_table_low; /* lower 32-bits of hash table */ + uint fec_grp_hash_table_high; /* upper 32-bits of hash table */ + uint fec_grp_hash_table_low; /* lower 32-bits of hash table */ uint fec_r_des_start; /* beginning of Rx descriptor ring */ uint fec_x_des_start; /* beginning of Tx descriptor ring */ uint fec_r_buff_size; /* Rx buffer size */ -- cgit v1.1 From b2508e828d71baacd9a743dd48dcbf85d96affdd Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 21 Apr 2011 16:48:35 -0700 Subject: perf: Support Xeon E7's via the Westmere PMU driver There's a new model number public, 47, for Xeon E7 (aka Westmere EX). Signed-off-by: Andi Kleen Cc: a.p.zijlstra@chello.nl Link: http://lkml.kernel.org/r/1303429715-10202-1-git-send-email-andi@firstfloor.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_intel.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 8fc2b2c..586cced 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -1425,6 +1425,7 @@ static __init int intel_pmu_init(void) case 37: /* 32 nm nehalem, "Clarkdale" */ case 44: /* 32 nm nehalem, "Gulftown" */ + case 47: /* 32 nm Xeon E7 */ memcpy(hw_cache_event_ids, westmere_hw_cache_event_ids, sizeof(hw_cache_event_ids)); memcpy(hw_cache_extra_regs, nehalem_hw_cache_extra_regs, -- cgit v1.1 From b52c55c6a25e4515b5e075a989ff346fc251ed09 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Fri, 22 Apr 2011 08:44:38 +0200 Subject: x86, perf event: Turn off unstructured raw event access to offcore registers Andi Kleen pointed out that the Intel offcore support patches were merged without user-space tool support to the functionality: | | The offcore_msr perf kernel code was merged into 2.6.39-rc*, but the | user space bits were not. This made it impossible to set the extra mask | and actually do the OFFCORE profiling | Andi submitted a preliminary patch for user-space support, as an extension to perf's raw event syntax: | | Some raw events -- like the Intel OFFCORE events -- support additional | parameters. These can be appended after a ':'. | | For example on a multi socket Intel Nehalem: | | perf stat -e r1b7:20ff -a sleep 1 | | Profile the OFFCORE_RESPONSE.ANY_REQUEST with event mask REMOTE_DRAM_0 | that measures any access to DRAM on another socket. | But this kind of usability is absolutely unacceptable - users should not be expected to type in magic, CPU and model specific incantations to get access to useful hardware functionality. The proper solution is to expose useful offcore functionality via generalized events - that way users do not have to care which specific CPU model they are using, they can use the conceptual event and not some model specific quirky hexa number. We already have such generalization in place for CPU cache events, and it's all very extensible. "Offcore" events measure general DRAM access patters along various parameters. They are particularly useful in NUMA systems. We want to support them via generalized DRAM events: either as the fourth level of cache (after the last-level cache), or as a separate generalization category. That way user-space support would be very obvious, memory access profiling could be done via self-explanatory commands like: perf record -e dram ./myapp perf record -e dram-remote ./myapp ... to measure DRAM accesses or more expensive cross-node NUMA DRAM accesses. These generalized events would work on all CPUs and architectures that have comparable PMU features. ( Note, these are just examples: actual implementation could have more sophistication and more parameter - as long as they center around similarly simple usecases. ) Now we do not want to revert *all* of the current offcore bits, as they are still somewhat useful for generic last-level-cache events, implemented in this commit: e994d7d23a0b: perf: Fix LLC-* events on Intel Nehalem/Westmere But we definitely do not yet want to expose the unstructured raw events to user-space, until better generalization and usability is implemented for these hardware event features. ( Note: after generalization has been implemented raw offcore events can be supported as well: there can always be an odd event that is marginally useful but not useful enough to generalize. DRAM profiling is definitely *not* such a category so generalization must be done first. ) Furthermore, PERF_TYPE_RAW access to these registers was not intended to go upstream without proper support - it was a side-effect of the above e994d7d23a0b commit, not mentioned in the changelog. As v2.6.39 is nearing release we go for the simplest approach: disable the PERF_TYPE_RAW offcore hack for now, before it escapes into a released kernel and becomes an ABI. Once proper structure is implemented for these hardware events and users are offered usable solutions we can revisit this issue. Reported-by: Andi Kleen Acked-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Thomas Gleixner Cc: Linus Torvalds Link: http://lkml.kernel.org/r/1302658203-4239-1-git-send-email-andi@firstfloor.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index eed3673a..632e5dc 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -586,8 +586,12 @@ static int x86_setup_perfctr(struct perf_event *event) return -EOPNOTSUPP; } + /* + * Do not allow config1 (extended registers) to propagate, + * there's no sane user-space generalization yet: + */ if (attr->type == PERF_TYPE_RAW) - return x86_pmu_extra_regs(event->attr.config, event); + return 0; if (attr->type == PERF_TYPE_HW_CACHE) return set_ext_hw_attr(hwc, event); -- cgit v1.1 From 1ea5a6afd95a4803900c97ed63a47a883ebe7b3e Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Thu, 21 Apr 2011 11:03:21 -0400 Subject: perf, x86: P4 PMU - Don't forget to clear cpuc->active_mask on overflow It's not enough to simply disable event on overflow the cpuc->active_mask should be cleared as well otherwise counter may stall in "active" even in real being already disabled (which potentially may lead to the situation that user may not use this counter further). Don pointed out that: " I also noticed this patch fixed some unknown NMIs on a P4 when I stressed the box". Tested-by: Lin Ming Signed-off-by: Cyrill Gorcunov Acked-by: Don Zickus Signed-off-by: Don Zickus Cc: Cyrill Gorcunov Link: http://lkml.kernel.org/r/1303398203-2918-3-git-send-email-dzickus@redhat.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_p4.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index c2520e1..d1f77e2 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -947,7 +947,7 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) if (!x86_perf_event_set_period(event)) continue; if (perf_event_overflow(event, 1, &data, regs)) - p4_pmu_disable_event(event); + x86_pmu_stop(event, 0); } if (handled) { -- cgit v1.1 From f4929bd37208540c2c6f416e9035ff1938f2dbc6 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Apr 2011 13:39:56 +0200 Subject: perf, x86: Update/fix Intel Nehalem cache events Change the Nehalem cache events to use retired memory instruction counters (similar to Westmere), this greatly improves the provided stats. Using: main () { int i; for (i = 0; i < 1000000000; i++) { asm("mov (%%rsp), %%rbx;" "mov %%rbx, (%%rsp);" : : : "rbx"); } } We find: $ perf stat --repeat 10 -e instructions:u -e l1-dcache-loads:u -e l1-dcache-stores:u ./loop_1b_loads+stores Performance counter stats for './loop_1b_loads+stores' (10 runs): 4,000,081,056 instructions:u # 0.000 IPC ( +- 0.000% ) 4,999,502,846 l1-dcache-loads:u ( +- 0.008% ) 1,000,034,832 l1-dcache-stores:u ( +- 0.000% ) 1.565184942 seconds time elapsed ( +- 0.005% ) The 5b is surprising - we'd expect 1b: $ perf stat --repeat 10 -e instructions:u -e r10b:u -e l1-dcache-stores:u ./loop_1b_loads+stores Performance counter stats for './loop_1b_loads+stores' (10 runs): 4,000,081,054 instructions:u # 0.000 IPC ( +- 0.000% ) 1,000,021,961 r10b:u ( +- 0.000% ) 1,000,030,951 l1-dcache-stores:u ( +- 0.000% ) 1.565055422 seconds time elapsed ( +- 0.003% ) Which this patch thus fixes. Signed-off-by: Peter Zijlstra Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Mike Galbraith Cc: Steven Rostedt Cc: Stephane Eranian Cc: Lin Ming Cc: Cyrill Gorcunov Link: http://lkml.kernel.org/n/tip-q9rtru7b7840tws75xzboapv@git.kernel.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_intel.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 586cced..43fa20b 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -391,12 +391,12 @@ static __initconst const u64 nehalem_hw_cache_event_ids { [ C(L1D) ] = { [ C(OP_READ) ] = { - [ C(RESULT_ACCESS) ] = 0x0f40, /* L1D_CACHE_LD.MESI */ - [ C(RESULT_MISS) ] = 0x0140, /* L1D_CACHE_LD.I_STATE */ + [ C(RESULT_ACCESS) ] = 0x010b, /* MEM_INST_RETIRED.LOADS */ + [ C(RESULT_MISS) ] = 0x0151, /* L1D.REPL */ }, [ C(OP_WRITE) ] = { - [ C(RESULT_ACCESS) ] = 0x0f41, /* L1D_CACHE_ST.MESI */ - [ C(RESULT_MISS) ] = 0x0141, /* L1D_CACHE_ST.I_STATE */ + [ C(RESULT_ACCESS) ] = 0x020b, /* MEM_INST_RETURED.STORES */ + [ C(RESULT_MISS) ] = 0x0251, /* L1D.M_REPL */ }, [ C(OP_PREFETCH) ] = { [ C(RESULT_ACCESS) ] = 0x014e, /* L1D_PREFETCH.REQUESTS */ -- cgit v1.1 From 568aa7508c99952ef18e4bfed87545e66f9e042f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Tue, 19 Apr 2011 10:21:20 +0200 Subject: ARM: at91: AT91CAP9 has a macb device MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit ee621dd (net: atmel_macb Kconfig: remove long dependency line) replaced a list of several explicit machines in the dependencies of MACB by a single symbol that is selected by the respective machines. ee621dd missed to let ARCH_AT91CAP9 select HAVE_NET_MACB though which is fixed here. Signed-off-by: Uwe Kleine-König Acked-by: Andrew Victor Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD --- arch/arm/mach-at91/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/arm/mach-at91/Kconfig b/arch/arm/mach-at91/Kconfig index 1939023..2d299bf 100644 --- a/arch/arm/mach-at91/Kconfig +++ b/arch/arm/mach-at91/Kconfig @@ -83,6 +83,7 @@ config ARCH_AT91CAP9 select CPU_ARM926T select GENERIC_CLOCKEVENTS select HAVE_FB_ATMEL + select HAVE_NET_MACB config ARCH_AT572D940HF bool "AT572D940HF" -- cgit v1.1 From 0312e826a48168761dbbe10a1e8cbc22ebc99767 Mon Sep 17 00:00:00 2001 From: Greg Ungerer Date: Wed, 30 Mar 2011 22:31:15 +1000 Subject: arm: at91: minimal defconfig for at91x40 SoC A minimal defconfig for build testing the AT91x40 SoC. Signed-off-by: Greg Ungerer Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD --- arch/arm/configs/at91x40_defconfig | 48 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 48 insertions(+) create mode 100644 arch/arm/configs/at91x40_defconfig (limited to 'arch') diff --git a/arch/arm/configs/at91x40_defconfig b/arch/arm/configs/at91x40_defconfig new file mode 100644 index 0000000..c55e921 --- /dev/null +++ b/arch/arm/configs/at91x40_defconfig @@ -0,0 +1,48 @@ +CONFIG_EXPERIMENTAL=y +CONFIG_LOG_BUF_SHIFT=14 +CONFIG_EMBEDDED=y +# CONFIG_HOTPLUG is not set +# CONFIG_ELF_CORE is not set +# CONFIG_FUTEX is not set +# CONFIG_TIMERFD is not set +# CONFIG_VM_EVENT_COUNTERS is not set +# CONFIG_COMPAT_BRK is not set +CONFIG_SLAB=y +# CONFIG_LBDAF is not set +# CONFIG_BLK_DEV_BSG is not set +# CONFIG_IOSCHED_DEADLINE is not set +# CONFIG_IOSCHED_CFQ is not set +# CONFIG_MMU is not set +CONFIG_ARCH_AT91=y +CONFIG_ARCH_AT91X40=y +CONFIG_MACH_AT91EB01=y +CONFIG_AT91_EARLY_USART0=y +CONFIG_CPU_ARM7TDMI=y +CONFIG_SET_MEM_PARAM=y +CONFIG_DRAM_BASE=0x01000000 +CONFIG_DRAM_SIZE=0x00400000 +CONFIG_FLASH_MEM_BASE=0x01400000 +CONFIG_PROCESSOR_ID=0x14000040 +CONFIG_ZBOOT_ROM_TEXT=0x0 +CONFIG_ZBOOT_ROM_BSS=0x0 +CONFIG_BINFMT_FLAT=y +# CONFIG_SUSPEND is not set +# CONFIG_FW_LOADER is not set +CONFIG_MTD=y +CONFIG_MTD_PARTITIONS=y +CONFIG_MTD_CHAR=y +CONFIG_MTD_BLOCK=y +CONFIG_MTD_RAM=y +CONFIG_MTD_ROM=y +CONFIG_BLK_DEV_RAM=y +# CONFIG_INPUT is not set +# CONFIG_SERIO is not set +# CONFIG_VT is not set +# CONFIG_DEVKMEM is not set +# CONFIG_HW_RANDOM is not set +# CONFIG_HWMON is not set +# CONFIG_USB_SUPPORT is not set +CONFIG_EXT2_FS=y +# CONFIG_DNOTIFY is not set +CONFIG_ROMFS_FS=y +# CONFIG_ENABLE_MUST_CHECK is not set -- cgit v1.1 From 91a2f4d3cd2a2f1a2c6830896bd2403ca0130137 Mon Sep 17 00:00:00 2001 From: Greg Ungerer Date: Wed, 30 Mar 2011 14:43:32 +1000 Subject: arm: at91: fix compiler warning for eb01 board build Fix compiler warning when building for AT91EB01 board: arch/arm/mach-at91/board-eb01.c:41: warning: initialisation from incompatible pointer type Signed-off-by: Greg Ungerer Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD --- arch/arm/mach-at91/board-eb01.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-at91/board-eb01.c b/arch/arm/mach-at91/board-eb01.c index 1f9d3cb..d8df59a 100644 --- a/arch/arm/mach-at91/board-eb01.c +++ b/arch/arm/mach-at91/board-eb01.c @@ -30,6 +30,11 @@ #include #include "generic.h" +static void __init at91eb01_init_irq(void) +{ + at91x40_init_interrupts(NULL); +} + static void __init at91eb01_map_io(void) { at91x40_initialize(40000000); @@ -38,7 +43,7 @@ static void __init at91eb01_map_io(void) MACHINE_START(AT91EB01, "Atmel AT91 EB01") /* Maintainer: Greg Ungerer */ .timer = &at91x40_timer, - .init_irq = at91x40_init_interrupts, + .init_irq = at91eb01_init_irq, .map_io = at91eb01_map_io, MACHINE_END -- cgit v1.1 From 9baeb7e47aed8e399d15d2ea8c032efe3680f20b Mon Sep 17 00:00:00 2001 From: Jean-Christophe PLAGNIOL-VILLARD Date: Sat, 23 Apr 2011 10:52:16 +0800 Subject: at91: Add ARCH_ID and basic cpu macros definition for 5series chips family. Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD Signed-off-by: Nicolas Ferre --- arch/arm/mach-at91/include/mach/cpu.h | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) (limited to 'arch') diff --git a/arch/arm/mach-at91/include/mach/cpu.h b/arch/arm/mach-at91/include/mach/cpu.h index 3bef931..0700f21 100644 --- a/arch/arm/mach-at91/include/mach/cpu.h +++ b/arch/arm/mach-at91/include/mach/cpu.h @@ -27,6 +27,7 @@ #define ARCH_ID_AT91SAM9G45 0x819b05a0 #define ARCH_ID_AT91SAM9G45MRL 0x819b05a2 /* aka 9G45-ES2 & non ES lots */ #define ARCH_ID_AT91SAM9G45ES 0x819b05a1 /* 9G45-ES (Engineering Sample) */ +#define ARCH_ID_AT91SAM9X5 0x819a05a0 #define ARCH_ID_AT91CAP9 0x039A03A0 #define ARCH_ID_AT91SAM9XE128 0x329973a0 @@ -55,6 +56,12 @@ static inline unsigned long at91_cpu_fully_identify(void) #define ARCH_EXID_AT91SAM9G46 0x00000003 #define ARCH_EXID_AT91SAM9G45 0x00000004 +#define ARCH_EXID_AT91SAM9G15 0x00000000 +#define ARCH_EXID_AT91SAM9G35 0x00000001 +#define ARCH_EXID_AT91SAM9X35 0x00000002 +#define ARCH_EXID_AT91SAM9G25 0x00000003 +#define ARCH_EXID_AT91SAM9X25 0x00000004 + static inline unsigned long at91_exid_identify(void) { return at91_sys_read(AT91_DBGU_EXID); @@ -143,6 +150,27 @@ static inline unsigned long at91cap9_rev_identify(void) #define cpu_is_at91sam9m11() (0) #endif +#ifdef CONFIG_ARCH_AT91SAM9X5 +#define cpu_is_at91sam9x5() (at91_cpu_identify() == ARCH_ID_AT91SAM9X5) +#define cpu_is_at91sam9g15() (cpu_is_at91sam9x5() && \ + (at91_exid_identify() == ARCH_EXID_AT91SAM9G15)) +#define cpu_is_at91sam9g35() (cpu_is_at91sam9x5() && \ + (at91_exid_identify() == ARCH_EXID_AT91SAM9G35)) +#define cpu_is_at91sam9x35() (cpu_is_at91sam9x5() && \ + (at91_exid_identify() == ARCH_EXID_AT91SAM9X35)) +#define cpu_is_at91sam9g25() (cpu_is_at91sam9x5() && \ + (at91_exid_identify() == ARCH_EXID_AT91SAM9G25)) +#define cpu_is_at91sam9x25() (cpu_is_at91sam9x5() && \ + (at91_exid_identify() == ARCH_EXID_AT91SAM9X25)) +#else +#define cpu_is_at91sam9x5() (0) +#define cpu_is_at91sam9g15() (0) +#define cpu_is_at91sam9g35() (0) +#define cpu_is_at91sam9x35() (0) +#define cpu_is_at91sam9g25() (0) +#define cpu_is_at91sam9x25() (0) +#endif + #ifdef CONFIG_ARCH_AT91CAP9 #define cpu_is_at91cap9() (at91_cpu_identify() == ARCH_ID_AT91CAP9) #define cpu_is_at91cap9_revB() (at91cap9_rev_identify() == ARCH_REVISION_CAP9_B) -- cgit v1.1 From 745b1f4f62852340e97a67d8c414d52fa1529185 Mon Sep 17 00:00:00 2001 From: Paul Parsons Date: Fri, 15 Apr 2011 14:39:27 +0000 Subject: ARM: pxa/hx4700: bq24022 regulator needs to be enabled Add REGULATOR_CHANGE_STATUS flag to hx4700 bq24022 regulator. Without this flag the bq24022 cannot be enabled and the battery will not charge. Signed-off-by: Paul Parsons Cc: Philipp Zabel Tested-by: Dmitry Artamonow Signed-off-by: Eric Miao --- arch/arm/mach-pxa/hx4700.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-pxa/hx4700.c b/arch/arm/mach-pxa/hx4700.c index 6de0ad0..9cdcca5 100644 --- a/arch/arm/mach-pxa/hx4700.c +++ b/arch/arm/mach-pxa/hx4700.c @@ -711,7 +711,7 @@ static struct regulator_consumer_supply bq24022_consumers[] = { static struct regulator_init_data bq24022_init_data = { .constraints = { .max_uA = 500000, - .valid_ops_mask = REGULATOR_CHANGE_CURRENT, + .valid_ops_mask = REGULATOR_CHANGE_CURRENT|REGULATOR_CHANGE_STATUS, }, .num_consumer_supplies = ARRAY_SIZE(bq24022_consumers), .consumer_supplies = bq24022_consumers, -- cgit v1.1 From e454d1632000b6e86003209811bda1186b34f8f3 Mon Sep 17 00:00:00 2001 From: Philipp Zabel Date: Wed, 20 Apr 2011 15:41:29 +0200 Subject: ARM: pxa/magician: bq24022 regulator needs to be enabled Add REGULATOR_CHANGE_STATUS flag to magician bq24022 regulator to enable charging. Signed-off-by: Philipp Zabel Signed-off-by: Eric Miao --- arch/arm/mach-pxa/magician.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-pxa/magician.c b/arch/arm/mach-pxa/magician.c index a72993d..9984ef7 100644 --- a/arch/arm/mach-pxa/magician.c +++ b/arch/arm/mach-pxa/magician.c @@ -599,7 +599,7 @@ static struct regulator_consumer_supply bq24022_consumers[] = { static struct regulator_init_data bq24022_init_data = { .constraints = { .max_uA = 500000, - .valid_ops_mask = REGULATOR_CHANGE_CURRENT, + .valid_ops_mask = REGULATOR_CHANGE_CURRENT | REGULATOR_CHANGE_STATUS, }, .num_consumer_supplies = ARRAY_SIZE(bq24022_consumers), .consumer_supplies = bq24022_consumers, -- cgit v1.1 From ad10e1051ec97a7e56e001eb1e9721b3d883e030 Mon Sep 17 00:00:00 2001 From: Michael Williamson Date: Wed, 23 Mar 2011 12:15:41 +0000 Subject: davinci: mityomapl138: Use correct id for NAND controller For the MityDSP-L138/MityARM-1808 SOMS, the NAND controller id (which needs to correspond to the chipselect, and is used for controlling the HW ECC computation) is not correct. Fix it. Signed-off-by: Michael Williamson Signed-off-by: Sekhar Nori --- arch/arm/mach-davinci/board-mityomapl138.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-davinci/board-mityomapl138.c b/arch/arm/mach-davinci/board-mityomapl138.c index 2aa79c5..e5d554c 100644 --- a/arch/arm/mach-davinci/board-mityomapl138.c +++ b/arch/arm/mach-davinci/board-mityomapl138.c @@ -414,7 +414,7 @@ static struct resource mityomapl138_nandflash_resource[] = { static struct platform_device mityomapl138_nandflash_device = { .name = "davinci_nand", - .id = 0, + .id = 1, .dev = { .platform_data = &mityomapl138_nandflash_data, }, -- cgit v1.1 From 336f402790defe4c4d33838b41bcd239537a5431 Mon Sep 17 00:00:00 2001 From: Michael Williamson Date: Wed, 23 Mar 2011 12:15:42 +0000 Subject: davinci: mityomapl138: Use auto-probe to determine attached PHY ID Current board configurations involving the MityDSP-L138 and MityARM-1808 only have one attached PHY, but it's address may not be the same. Default the behavior to auto-probe for the PHY and use the first one found. Signed-off-by: Michael Williamson Signed-off-by: Sekhar Nori --- arch/arm/mach-davinci/board-mityomapl138.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-davinci/board-mityomapl138.c b/arch/arm/mach-davinci/board-mityomapl138.c index e5d554c..606a6f2 100644 --- a/arch/arm/mach-davinci/board-mityomapl138.c +++ b/arch/arm/mach-davinci/board-mityomapl138.c @@ -29,7 +29,7 @@ #include #include -#define MITYOMAPL138_PHY_ID "0:03" +#define MITYOMAPL138_PHY_ID "" #define FACTORY_CONFIG_MAGIC 0x012C0138 #define FACTORY_CONFIG_VERSION 0x00010001 -- cgit v1.1 From 9e7d24f622fc57ff5d58af3d44ba503430148354 Mon Sep 17 00:00:00 2001 From: Sergei Shtylyov Date: Wed, 6 Apr 2011 17:17:21 +0000 Subject: DA830: fix SPI1 base address Commit 54ce6883d29630ff334bee4256a25e3f8719a181 (davinci: da8xx: add spi resources and registration routine) wrongly assumed that SPI1 is mapped at the same address on DA830/OMAP-L137 and DA850/OMAP-L138; actually, the base address was valid only for the latter SoC. Teach the code to pass the correct SPI1 memory resource for both SoCs... Signed-off-by: Sergei Shtylyov Signed-off-by: Sekhar Nori --- arch/arm/mach-davinci/devices-da8xx.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-davinci/devices-da8xx.c b/arch/arm/mach-davinci/devices-da8xx.c index 625d4b6..58a02dc 100644 --- a/arch/arm/mach-davinci/devices-da8xx.c +++ b/arch/arm/mach-davinci/devices-da8xx.c @@ -39,7 +39,8 @@ #define DA8XX_GPIO_BASE 0x01e26000 #define DA8XX_I2C1_BASE 0x01e28000 #define DA8XX_SPI0_BASE 0x01c41000 -#define DA8XX_SPI1_BASE 0x01f0e000 +#define DA830_SPI1_BASE 0x01e12000 +#define DA850_SPI1_BASE 0x01f0e000 #define DA8XX_EMAC_CTRL_REG_OFFSET 0x3000 #define DA8XX_EMAC_MOD_REG_OFFSET 0x2000 @@ -762,8 +763,8 @@ static struct resource da8xx_spi0_resources[] = { static struct resource da8xx_spi1_resources[] = { [0] = { - .start = DA8XX_SPI1_BASE, - .end = DA8XX_SPI1_BASE + SZ_4K - 1, + .start = DA830_SPI1_BASE, + .end = DA830_SPI1_BASE + SZ_4K - 1, .flags = IORESOURCE_MEM, }, [1] = { @@ -832,5 +833,10 @@ int __init da8xx_register_spi(int instance, struct spi_board_info *info, da8xx_spi_pdata[instance].num_chipselect = len; + if (instance == 1 && cpu_is_davinci_da850()) { + da8xx_spi1_resources[0].start = DA850_SPI1_BASE; + da8xx_spi1_resources[0].end = DA850_SPI1_BASE + SZ_4K - 1; + } + return platform_device_register(&da8xx_spi_device[instance]); } -- cgit v1.1 From 45b146d746ea1b7f87b023a79d5186d0e87793eb Mon Sep 17 00:00:00 2001 From: Russell King - ARM Linux Date: Fri, 15 Apr 2011 12:32:40 +0000 Subject: ARM: Davinci: Fix I2C build errors Several Davinci platforms select the I2C EEPROM support, but don't select I2C support. This causes I2C EEPROM support to be built into the kernel, but I2C support may not be configured to be built in. This leads to linker errors due to missing I2C symbols. Arrange for I2C to be selected whenever EEPROM_AT24 is selected. Signed-off-by: Russell King Signed-off-by: Sekhar Nori --- arch/arm/mach-davinci/Kconfig | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'arch') diff --git a/arch/arm/mach-davinci/Kconfig b/arch/arm/mach-davinci/Kconfig index 32f1479..c0deaca 100644 --- a/arch/arm/mach-davinci/Kconfig +++ b/arch/arm/mach-davinci/Kconfig @@ -63,6 +63,7 @@ config MACH_DAVINCI_EVM depends on ARCH_DAVINCI_DM644x select MISC_DEVICES select EEPROM_AT24 + select I2C help Configure this option to specify the whether the board used for development is a DM644x EVM @@ -72,6 +73,7 @@ config MACH_SFFSDR depends on ARCH_DAVINCI_DM644x select MISC_DEVICES select EEPROM_AT24 + select I2C help Say Y here to select the Lyrtech Small Form Factor Software Defined Radio (SFFSDR) board. @@ -105,6 +107,7 @@ config MACH_DAVINCI_DM6467_EVM select MACH_DAVINCI_DM6467TEVM select MISC_DEVICES select EEPROM_AT24 + select I2C help Configure this option to specify the whether the board used for development is a DM6467 EVM @@ -118,6 +121,7 @@ config MACH_DAVINCI_DM365_EVM depends on ARCH_DAVINCI_DM365 select MISC_DEVICES select EEPROM_AT24 + select I2C help Configure this option to specify whether the board used for development is a DM365 EVM @@ -129,6 +133,7 @@ config MACH_DAVINCI_DA830_EVM select GPIO_PCF857X select MISC_DEVICES select EEPROM_AT24 + select I2C help Say Y here to select the TI DA830/OMAP-L137/AM17x Evaluation Module. @@ -205,6 +210,7 @@ config MACH_MITYOMAPL138 depends on ARCH_DAVINCI_DA850 select MISC_DEVICES select EEPROM_AT24 + select I2C help Say Y here to select the Critical Link MityDSP-L138/MityARM-1808 System on Module. Information on this SoM may be found at -- cgit v1.1 From 87dc669ba25777b67796d7262c569429e58b1ed4 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 8 Apr 2011 17:29:36 +0200 Subject: x86, hw_breakpoints: Fix racy access to ptrace breakpoints While the tracer accesses ptrace breakpoints, the child task may concurrently exit due to a SIGKILL and thus release its breakpoints at the same time. We can then dereference some freed pointers. To fix this, hold a reference on the child breakpoints before manipulating them. Reported-by: Oleg Nesterov Signed-off-by: Frederic Weisbecker Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Will Deacon Cc: Prasad Cc: Paul Mundt Cc: v2.6.33.. Link: http://lkml.kernel.org/r/1302284067-7860-3-git-send-email-fweisbec@gmail.com --- arch/x86/kernel/ptrace.c | 36 ++++++++++++++++++++++++++---------- 1 file changed, 26 insertions(+), 10 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 45892dc..f65e5b5 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -608,6 +608,9 @@ static int ptrace_write_dr7(struct task_struct *tsk, unsigned long data) unsigned len, type; struct perf_event *bp; + if (ptrace_get_breakpoints(tsk) < 0) + return -ESRCH; + data &= ~DR_CONTROL_RESERVED; old_dr7 = ptrace_get_dr7(thread->ptrace_bps); restore: @@ -655,6 +658,9 @@ restore: } goto restore; } + + ptrace_put_breakpoints(tsk); + return ((orig_ret < 0) ? orig_ret : rc); } @@ -668,10 +674,17 @@ static unsigned long ptrace_get_debugreg(struct task_struct *tsk, int n) if (n < HBP_NUM) { struct perf_event *bp; + + if (ptrace_get_breakpoints(tsk) < 0) + return -ESRCH; + bp = thread->ptrace_bps[n]; if (!bp) - return 0; - val = bp->hw.info.address; + val = 0; + else + val = bp->hw.info.address; + + ptrace_put_breakpoints(tsk); } else if (n == 6) { val = thread->debugreg6; } else if (n == 7) { @@ -686,6 +699,10 @@ static int ptrace_set_breakpoint_addr(struct task_struct *tsk, int nr, struct perf_event *bp; struct thread_struct *t = &tsk->thread; struct perf_event_attr attr; + int err = 0; + + if (ptrace_get_breakpoints(tsk) < 0) + return -ESRCH; if (!t->ptrace_bps[nr]) { ptrace_breakpoint_init(&attr); @@ -709,24 +726,23 @@ static int ptrace_set_breakpoint_addr(struct task_struct *tsk, int nr, * writing for the user. And anyway this is the previous * behaviour. */ - if (IS_ERR(bp)) - return PTR_ERR(bp); + if (IS_ERR(bp)) { + err = PTR_ERR(bp); + goto put; + } t->ptrace_bps[nr] = bp; } else { - int err; - bp = t->ptrace_bps[nr]; attr = bp->attr; attr.bp_addr = addr; err = modify_user_hw_breakpoint(bp, &attr); - if (err) - return err; } - - return 0; +put: + ptrace_put_breakpoints(tsk); + return err; } /* -- cgit v1.1 From 07fa7a0a8a586c01a8b416358c7012dcb9dc688d Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 8 Apr 2011 17:29:36 +0200 Subject: powerpc, hw_breakpoints: Fix racy access to ptrace breakpoints While the tracer accesses ptrace breakpoints, the child task may concurrently exit due to a SIGKILL and thus release its breakpoints at the same time. We can then dereference some freed pointers. To fix this, hold a reference on the child breakpoints before manipulating them. Reported-by: Oleg Nesterov Signed-off-by: Frederic Weisbecker Acked-by: Prasad Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Will Deacon Cc: Paul Mundt Cc: v2.6.33.. Link: http://lkml.kernel.org/r/1302284067-7860-4-git-send-email-fweisbec@gmail.com --- arch/powerpc/kernel/ptrace.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch') diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c index 55613e3..4edeeb3 100644 --- a/arch/powerpc/kernel/ptrace.c +++ b/arch/powerpc/kernel/ptrace.c @@ -1591,7 +1591,10 @@ long arch_ptrace(struct task_struct *child, long request, } case PTRACE_SET_DEBUGREG: + if (ptrace_get_breakpoints(child) < 0) + return -ESRCH; ret = ptrace_set_debugreg(child, addr, data); + ptrace_put_breakpoints(child); break; #ifdef CONFIG_PPC64 -- cgit v1.1 From bf0b8f4b55e591ba417c2dbaff42769e1fc773b0 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 8 Apr 2011 17:29:36 +0200 Subject: arm, hw_breakpoints: Fix racy access to ptrace breakpoints While the tracer accesses ptrace breakpoints, the child task may concurrently exit due to a SIGKILL and thus release its breakpoints at the same time. We can then dereference some freed pointers. To fix this, hold a reference on the child breakpoints before manipulating them. Reported-by: Oleg Nesterov Signed-off-by: Frederic Weisbecker Acked-by: Will Deacon Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Prasad Cc: Paul Mundt Link: http://lkml.kernel.org/r/1302284067-7860-5-git-send-email-fweisbec@gmail.com --- arch/arm/kernel/ptrace.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'arch') diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index 2bf27f3..8182f45 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -767,12 +767,20 @@ long arch_ptrace(struct task_struct *child, long request, #ifdef CONFIG_HAVE_HW_BREAKPOINT case PTRACE_GETHBPREGS: + if (ptrace_get_breakpoints(child) < 0) + return -ESRCH; + ret = ptrace_gethbpregs(child, addr, (unsigned long __user *)data); + ptrace_put_breakpoints(child); break; case PTRACE_SETHBPREGS: + if (ptrace_get_breakpoints(child) < 0) + return -ESRCH; + ret = ptrace_sethbpregs(child, addr, (unsigned long __user *)data); + ptrace_put_breakpoints(child); break; #endif -- cgit v1.1 From e0ac8457d020c0289ea566917267da9e5e6d9865 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 8 Apr 2011 17:29:36 +0200 Subject: sh, hw_breakpoints: Fix racy access to ptrace breakpoints While the tracer accesses ptrace breakpoints, the child task may concurrently exit due to a SIGKILL and thus release its breakpoints at the same time. We can then dereference some freed pointers. To fix this, hold a reference on the child breakpoints before manipulating them. Reported-by: Oleg Nesterov Signed-off-by: Frederic Weisbecker Acked-by: Paul Mundt Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Will Deacon Cc: Prasad Link: http://lkml.kernel.org/r/1302284067-7860-6-git-send-email-fweisbec@gmail.com --- arch/sh/kernel/ptrace_32.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch') diff --git a/arch/sh/kernel/ptrace_32.c b/arch/sh/kernel/ptrace_32.c index 2130ca6..3d7b209 100644 --- a/arch/sh/kernel/ptrace_32.c +++ b/arch/sh/kernel/ptrace_32.c @@ -117,7 +117,11 @@ void user_enable_single_step(struct task_struct *child) set_tsk_thread_flag(child, TIF_SINGLESTEP); + if (ptrace_get_breakpoints(child) < 0) + return; + set_single_step(child, pc); + ptrace_put_breakpoints(child); } void user_disable_single_step(struct task_struct *child) -- cgit v1.1 From 39b68976ac653cfdc7f872a293e8b7928de2dcc6 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Mon, 25 Apr 2011 14:52:37 -0700 Subject: x86, setup: When probing memory with e801, use ax/bx as a pair When we use BIOS function e801 to probe memory, we should use ax/bx (or cx/dx) as a pair, not mix and match. This was a typo during the translation from assembly code, and breaks at least one set of machines in the field (which return cx = dx = 0). Reported-and-tested-by: Chris Samuel Fix-proposed-by: Thomas Meyer Link: http://lkml.kernel.org/r/1303566747.12067.10.camel@localhost.localdomain --- arch/x86/boot/memory.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/boot/memory.c b/arch/x86/boot/memory.c index cae3feb..db75d07 100644 --- a/arch/x86/boot/memory.c +++ b/arch/x86/boot/memory.c @@ -91,7 +91,7 @@ static int detect_memory_e801(void) if (oreg.ax > 15*1024) { return -1; /* Bogus! */ } else if (oreg.ax == 15*1024) { - boot_params.alt_mem_k = (oreg.dx << 6) + oreg.ax; + boot_params.alt_mem_k = (oreg.bx << 6) + oreg.ax; } else { /* * This ignores memory above 16MB if we have a memory -- cgit v1.1 From 6ba5932ca4b610d036cb89d0ce2a465d06504c4d Mon Sep 17 00:00:00 2001 From: Oskar Andero Date: Tue, 26 Apr 2011 02:24:50 -0700 Subject: arm: omap2: enable smc instruction for sleep34xx This fixes broken build when using binutils 2.21. Signed-off-by: Oskar Andero Signed-off-by: Tony Lindgren --- arch/arm/mach-omap2/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/Makefile b/arch/arm/mach-omap2/Makefile index a45cd64..512b152 100644 --- a/arch/arm/mach-omap2/Makefile +++ b/arch/arm/mach-omap2/Makefile @@ -68,7 +68,7 @@ obj-$(CONFIG_OMAP_SMARTREFLEX) += sr_device.o smartreflex.o obj-$(CONFIG_OMAP_SMARTREFLEX_CLASS3) += smartreflex-class3.o AFLAGS_sleep24xx.o :=-Wa,-march=armv6 -AFLAGS_sleep34xx.o :=-Wa,-march=armv7-a +AFLAGS_sleep34xx.o :=-Wa,-march=armv7-a$(plus_sec) ifeq ($(CONFIG_PM_VERBOSE),y) CFLAGS_pm_bus.o += -DDEBUG -- cgit v1.1 From bc16b3777ec3749c086a17f81c99f8643f4a6576 Mon Sep 17 00:00:00 2001 From: omar ramirez Date: Tue, 26 Apr 2011 02:24:50 -0700 Subject: OMAP3: l3: fix for "irq 10: nobody cared" message If an error occurs in the L3 on any other initiator than MPU, the interrupt goes unhandled given that the 'base' register was calculated with the initialized err_source value (which coincidentally points to MPU) and not with the actual source of the error. Removed parenthesis that are not needed for the touched lines. Signed-off-by: Omar Ramirez Luna Acked-by: Santosh Shilimkar Signed-off-by: Tony Lindgren --- arch/arm/mach-omap2/omap_l3_smx.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/omap_l3_smx.c b/arch/arm/mach-omap2/omap_l3_smx.c index 5f2da756..4321e79 100644 --- a/arch/arm/mach-omap2/omap_l3_smx.c +++ b/arch/arm/mach-omap2/omap_l3_smx.c @@ -196,11 +196,11 @@ static irqreturn_t omap3_l3_app_irq(int irq, void *_l3) /* No timeout error for debug sources */ } - base = ((l3->rt) + (*(omap3_l3_bases[int_type] + err_source))); - /* identify the error source */ for (err_source = 0; !(status & (1 << err_source)); err_source++) ; + + base = l3->rt + *(omap3_l3_bases[int_type] + err_source); error = omap3_l3_readll(base, L3_ERROR_LOG); if (error) { -- cgit v1.1 From 26a064d5246e8ac51241d6ec2792aebf24f3b41a Mon Sep 17 00:00:00 2001 From: Felipe Contreras Date: Tue, 26 Apr 2011 02:45:28 -0700 Subject: omap: rx51: mark reserved memory earlier So that omap_vram_set_sdram_vram() is called before omap_vram_reserve_sdram_memblock(). Signed-off-by: Felipe Contreras Acked-by: Tomi Valkeinen Signed-off-by: Tony Lindgren --- arch/arm/mach-omap2/board-rx51.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/board-rx51.c b/arch/arm/mach-omap2/board-rx51.c index e964895..f8ba20a 100644 --- a/arch/arm/mach-omap2/board-rx51.c +++ b/arch/arm/mach-omap2/board-rx51.c @@ -141,14 +141,19 @@ static void __init rx51_init(void) static void __init rx51_map_io(void) { omap2_set_globals_3xxx(); - rx51_video_mem_init(); omap34xx_map_common_io(); } +static void __init rx51_reserve(void) +{ + rx51_video_mem_init(); + omap_reserve(); +} + MACHINE_START(NOKIA_RX51, "Nokia RX-51 board") /* Maintainer: Lauri Leukkunen */ .boot_params = 0x80000100, - .reserve = omap_reserve, + .reserve = rx51_reserve, .map_io = rx51_map_io, .init_early = rx51_init_early, .init_irq = omap_init_irq, -- cgit v1.1 From 919686458fabc67a13ffa412f9e5a8fed46d10b8 Mon Sep 17 00:00:00 2001 From: Shweta Gulati Date: Tue, 26 Apr 2011 02:32:26 -0700 Subject: OMAP4: Intialize IVA Device in addition to DSP device. OMAP4 has two different Devices IVA and DSP. DSP is bound with IVA for DVFS. The registration of IVA dev in API 'omap2_init_processor_devices' was missing. Init dev for 'iva_dev' is added. This also fixes the following error seen during boot as omap2_set_init_voltage can now find the iva device omap2_set_init_voltage: Invalid parameters! omap2_set_init_voltage: Unable to put vdd_iva to its init voltage Signed-off-by: Shweta Gulati Acked-by: Nishanth Menon Signed-off-by: Tony Lindgren --- arch/arm/mach-omap2/pm.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch') diff --git a/arch/arm/mach-omap2/pm.c b/arch/arm/mach-omap2/pm.c index 30af335..49486f5 100644 --- a/arch/arm/mach-omap2/pm.c +++ b/arch/arm/mach-omap2/pm.c @@ -89,6 +89,7 @@ static void omap2_init_processor_devices(void) if (cpu_is_omap44xx()) { _init_omap_device("l3_main_1", &l3_dev); _init_omap_device("dsp", &dsp_dev); + _init_omap_device("iva", &iva_dev); } else { _init_omap_device("l3_main", &l3_dev); } -- cgit v1.1 From 3f126087ee143775961947b39416aad03044c988 Mon Sep 17 00:00:00 2001 From: Nishanth Menon Date: Tue, 26 Apr 2011 02:33:10 -0700 Subject: OMAP3+: voltage: remove initial voltage Blindly setting 1.2V in the initial structure may not even match the default voltages stored in the voltage table which are supported for the domain. For example, OMAP3430 core domain does not use 1.2V and ends up generating a warning on the first transition. Further, since omap2_set_init_voltage is called as part of the pm framework's initialization sequence to configure the voltage required for the current OPP, the call does(and has to) setup the system voltage(curr_volt as a result) using the right mechanisms appropriate for the system at that point of time. This also overrides initialization we are currently doing in voltage.c making it redundant. So, remove the wrong and redundant initialization. Signed-off-by: Nishanth Menon Signed-off-by: Kevin Hilman Signed-off-by: Tony Lindgren --- arch/arm/mach-omap2/voltage.c | 1 - 1 file changed, 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mach-omap2/voltage.c b/arch/arm/mach-omap2/voltage.c index 6fb5209..0c1552d 100644 --- a/arch/arm/mach-omap2/voltage.c +++ b/arch/arm/mach-omap2/voltage.c @@ -114,7 +114,6 @@ static int __init _config_common_vdd_data(struct omap_vdd_info *vdd) sys_clk_speed /= 1000; /* Generic voltage parameters */ - vdd->curr_volt = 1200000; vdd->volt_scale = vp_forceupdate_scale_voltage; vdd->vp_enabled = false; -- cgit v1.1 From 18a073a3acd3a47fbb5e23333df7fad28d576345 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 26 Apr 2011 13:24:33 +0200 Subject: perf, x86: Fix BTS condition Currently the x86 backend incorrectly assumes that any BRANCH_INSN with sample_period==1 is a BTS request. This is not true when we do frequency driven profiling such as 'perf record -e branches'. Solves this error: $ perf record -e branches ./array Error: sys_perf_event_open() syscall returned with 95 (Operation not supported). Signed-off-by: Peter Zijlstra Reported-by: Ingo Molnar Cc: "Metzger, Markus T" Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Link: http://lkml.kernel.org/n/tip-rd2y4ct71hjawzz6fpvsy9hg@git.kernel.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event.c | 4 ++-- arch/x86/kernel/cpu/perf_event_intel.c | 3 +++ 2 files changed, 5 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 632e5dc..fac0654 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -613,8 +613,8 @@ static int x86_setup_perfctr(struct perf_event *event) /* * Branch tracing: */ - if ((attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS) && - (hwc->sample_period == 1)) { + if (attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS && + !attr->freq && hwc->sample_period == 1) { /* BTS is not supported by this architecture. */ if (!x86_pmu.bts_active) return -EOPNOTSUPP; diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 43fa20b..9194b06 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -998,6 +998,9 @@ intel_bts_constraints(struct perf_event *event) struct hw_perf_event *hwc = &event->hw; unsigned int hw_event, bts_event; + if (event->attr.freq) + return NULL; + hw_event = hwc->config & INTEL_ARCH_EVENT_MASK; bts_event = x86_pmu.event_map(PERF_COUNT_HW_BRANCH_INSTRUCTIONS); -- cgit v1.1 From ec75a71634dabe439db91c1ef51d5099f4493808 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Wed, 27 Apr 2011 11:51:41 +0200 Subject: perf events, x86: Work around the Nehalem AAJ80 erratum On Nehalem CPUs the retired branch-misses event can be completely bogus, when there are no branch-misses occuring. When there are a lot of branch misses then the count is pretty accurate. Still, this leaves us with an event that over-counts a lot. Detect this erratum and work it around by using BR_MISP_EXEC.ANY events. These will also count speculated branches but still it's a lot more precise in practice than the architectural event. Acked-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Link: http://lkml.kernel.org/n/tip-yyfg0bxo9jsqxd6a0ovfny27@git.kernel.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_intel.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 9194b06..9ae4a2a 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -25,7 +25,7 @@ struct intel_percore { /* * Intel PerfMon, used on Core and later. */ -static const u64 intel_perfmon_event_map[] = +static u64 intel_perfmon_event_map[PERF_COUNT_HW_MAX] __read_mostly = { [PERF_COUNT_HW_CPU_CYCLES] = 0x003c, [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, @@ -1308,7 +1308,7 @@ static void intel_clovertown_quirks(void) * AJ106 could possibly be worked around by not allowing LBR * usage from PEBS, including the fixup. * AJ68 could possibly be worked around by always programming - * a pebs_event_reset[0] value and coping with the lost events. + * a pebs_event_reset[0] value and coping with the lost events. * * But taken together it might just make sense to not enable PEBS on * these chips. @@ -1412,6 +1412,18 @@ static __init int intel_pmu_init(void) x86_pmu.percore_constraints = intel_nehalem_percore_constraints; x86_pmu.enable_all = intel_pmu_nhm_enable_all; x86_pmu.extra_regs = intel_nehalem_extra_regs; + + if (ebx & 0x40) { + /* + * Erratum AAJ80 detected, we work it around by using + * the BR_MISP_EXEC.ANY event. This will over-count + * branch-misses, but it's still much better than the + * architectural event which is often completely bogus: + */ + intel_perfmon_event_map[PERF_COUNT_HW_BRANCH_MISSES] = 0x7f89; + + pr_cont("erratum AAJ80 worked around, "); + } pr_cont("Nehalem events, "); break; -- cgit v1.1 From e9c549998dc24209847007e1f209f3b6c88d21ba Mon Sep 17 00:00:00 2001 From: Lucas De Marchi Date: Tue, 26 Apr 2011 23:28:26 -0700 Subject: Revert wrong fixes for common misspellings These changes were incorrectly fixed by codespell. They were now manually corrected. Signed-off-by: Lucas De Marchi --- arch/arm/mach-davinci/dm355.c | 2 +- arch/arm/mach-davinci/dm644x.c | 2 +- arch/powerpc/include/asm/uninorth.h | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/arm/mach-davinci/dm355.c b/arch/arm/mach-davinci/dm355.c index f680122..a3a94e9 100644 --- a/arch/arm/mach-davinci/dm355.c +++ b/arch/arm/mach-davinci/dm355.c @@ -314,7 +314,7 @@ static struct clk timer2_clk = { .name = "timer2", .parent = &pll1_aux_clk, .lpsc = DAVINCI_LPSC_TIMER2, - .usecount = 1, /* REVISIT: why can't' this be disabled? */ + .usecount = 1, /* REVISIT: why can't this be disabled? */ }; static struct clk timer3_clk = { diff --git a/arch/arm/mach-davinci/dm644x.c b/arch/arm/mach-davinci/dm644x.c index 5f8a654..4c82c27 100644 --- a/arch/arm/mach-davinci/dm644x.c +++ b/arch/arm/mach-davinci/dm644x.c @@ -274,7 +274,7 @@ static struct clk timer2_clk = { .name = "timer2", .parent = &pll1_aux_clk, .lpsc = DAVINCI_LPSC_TIMER2, - .usecount = 1, /* REVISIT: why can't' this be disabled? */ + .usecount = 1, /* REVISIT: why can't this be disabled? */ }; static struct clk_lookup dm644x_clks[] = { diff --git a/arch/powerpc/include/asm/uninorth.h b/arch/powerpc/include/asm/uninorth.h index ae9c899..d12b11d 100644 --- a/arch/powerpc/include/asm/uninorth.h +++ b/arch/powerpc/include/asm/uninorth.h @@ -60,7 +60,7 @@ * * Obviously, the GART is not cache coherent and so any change to it * must be flushed to memory (or maybe just make the GART space non - * cachable). AGP memory itself does't seem to be cache coherent neither. + * cachable). AGP memory itself doesn't seem to be cache coherent neither. * * In order to invalidate the GART (which is probably necessary to inval * the bridge internal TLBs), the following sequence has to be written, -- cgit v1.1 From ed961581a7ca91d6a4852af2e44333e983100505 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Wed, 27 Apr 2011 09:34:33 +0200 Subject: [S390] prng: fix pointer arithmetic The git commit c708c57e247775928b9a6bce7b4d8d14883bf39b fixed the access beyond the end of the stack in prng_seed but the pointer arithmetic is still incorrect. The calculation has been off by a factor of 64, now it is only off by a factor of 8. prng_seed is called with a maximum of 16 for nbytes, small enough that the incorrect calculation stays insides the limits of the stack. Place parentheses for correct pointer arithmetic. Signed-off-by: Martin Schwidefsky --- arch/s390/crypto/prng.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/s390/crypto/prng.c b/arch/s390/crypto/prng.c index 44bca3f..8b16c47 100644 --- a/arch/s390/crypto/prng.c +++ b/arch/s390/crypto/prng.c @@ -76,7 +76,7 @@ static void prng_seed(int nbytes) /* Add the entropy */ while (nbytes >= 8) { - *((__u64 *)parm_block) ^= *((__u64 *)buf+i); + *((__u64 *)parm_block) ^= *((__u64 *)(buf+i)); prng_add_entropy(); i += 8; nbytes -= 8; -- cgit v1.1 From 4aac0b4815ba592052758f4b468f253d383dc9d6 Mon Sep 17 00:00:00 2001 From: Michael Schmitz Date: Tue, 26 Apr 2011 14:51:53 +1200 Subject: m68k/mm: Set all online nodes in N_NORMAL_MEMORY For m68k, N_NORMAL_MEMORY represents all nodes that have present memory since it does not support HIGHMEM. This patch sets the bit at the time node_present_pages has been set by free_area_init_node. At the time the node is brought online, the node state would have to be done unconditionally since information about present memory has not yet been recorded. If N_NORMAL_MEMORY is not accurate, slub may encounter errors since it uses this nodemask to setup per-cache kmem_cache_node data structures. This pach is an alternative to the one proposed by David Rientjes attempting to set node state immediately when bringing the node online. Signed-off-by: Michael Schmitz Tested-by: Thorsten Glaser Signed-off-by: Geert Uytterhoeven CC: stable@kernel.org --- arch/m68k/mm/motorola.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch') diff --git a/arch/m68k/mm/motorola.c b/arch/m68k/mm/motorola.c index 02b7a03..8b3db1c 100644 --- a/arch/m68k/mm/motorola.c +++ b/arch/m68k/mm/motorola.c @@ -300,6 +300,8 @@ void __init paging_init(void) zones_size[ZONE_DMA] = m68k_memory[i].size >> PAGE_SHIFT; free_area_init_node(i, zones_size, m68k_memory[i].addr >> PAGE_SHIFT, NULL); + if (node_present_pages(i)) + node_set_state(i, N_NORMAL_MEMORY); } } -- cgit v1.1 From 2bce5daca28346f19c190dbdb5542c9fe3e8c6e6 Mon Sep 17 00:00:00 2001 From: Don Zickus Date: Wed, 27 Apr 2011 06:32:33 -0400 Subject: perf, x86, nmi: Move LVT un-masking into irq handlers It was noticed that P4 machines were generating double NMIs for each perf event. These extra NMIs lead to 'Dazed and confused' messages on the screen. I tracked this down to a P4 quirk that said the overflow bit had to be cleared before re-enabling the apic LVT mask. My first attempt was to move the un-masking inside the perf nmi handler from before the chipset NMI handler to after. This broke Nehalem boxes that seem to like the unmasking before the counters themselves are re-enabled. In order to keep this change simple for 2.6.39, I decided to just simply move the apic LVT un-masking to the beginning of all the chipset NMI handlers, with the exception of Pentium4's to fix the double NMI issue. Later on we can move the un-masking to later in the handlers to save a number of 'extra' NMIs on those particular chipsets. I tested this change on a P4 machine, an AMD machine, a Nehalem box, and a core2quad box. 'perf top' worked correctly along with various other small 'perf record' runs. Anything high stress breaks all the machines but that is a different problem. Thanks to various people for testing different versions of this patch. Reported-and-tested-by: Shaun Ruffell Signed-off-by: Don Zickus Cc: Cyrill Gorcunov Link: http://lkml.kernel.org/r/1303900353-10242-1-git-send-email-dzickus@redhat.com Signed-off-by: Ingo Molnar CC: Cyrill Gorcunov --- arch/x86/kernel/cpu/perf_event.c | 12 ++++++++++-- arch/x86/kernel/cpu/perf_event_intel.c | 10 ++++++++++ arch/x86/kernel/cpu/perf_event_p4.c | 17 +++++++++++++---- 3 files changed, 33 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index fac0654..e638689 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1288,6 +1288,16 @@ static int x86_pmu_handle_irq(struct pt_regs *regs) cpuc = &__get_cpu_var(cpu_hw_events); + /* + * Some chipsets need to unmask the LVTPC in a particular spot + * inside the nmi handler. As a result, the unmasking was pushed + * into all the nmi handlers. + * + * This generic handler doesn't seem to have any issues where the + * unmasking occurs so it was left at the top. + */ + apic_write(APIC_LVTPC, APIC_DM_NMI); + for (idx = 0; idx < x86_pmu.num_counters; idx++) { if (!test_bit(idx, cpuc->active_mask)) { /* @@ -1374,8 +1384,6 @@ perf_event_nmi_handler(struct notifier_block *self, return NOTIFY_DONE; } - apic_write(APIC_LVTPC, APIC_DM_NMI); - handled = x86_pmu.handle_irq(args->regs); if (!handled) return NOTIFY_DONE; diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 9ae4a2a..e61539b 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -933,6 +933,16 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) cpuc = &__get_cpu_var(cpu_hw_events); + /* + * Some chipsets need to unmask the LVTPC in a particular spot + * inside the nmi handler. As a result, the unmasking was pushed + * into all the nmi handlers. + * + * This handler doesn't seem to have any issues with the unmasking + * so it was left at the top. + */ + apic_write(APIC_LVTPC, APIC_DM_NMI); + intel_pmu_disable_all(); handled = intel_pmu_drain_bts_buffer(); status = intel_pmu_get_status(); diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index d1f77e2..e93fcd5 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -950,11 +950,20 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) x86_pmu_stop(event, 0); } - if (handled) { - /* p4 quirk: unmask it again */ - apic_write(APIC_LVTPC, apic_read(APIC_LVTPC) & ~APIC_LVT_MASKED); + if (handled) inc_irq_stat(apic_perf_irqs); - } + + /* + * When dealing with the unmasking of the LVTPC on P4 perf hw, it has + * been observed that the OVF bit flag has to be cleared first _before_ + * the LVTPC can be unmasked. + * + * The reason is the NMI line will continue to be asserted while the OVF + * bit is set. This causes a second NMI to generate if the LVTPC is + * unmasked before the OVF bit is cleared, leading to unknown NMI + * messages. + */ + apic_write(APIC_LVTPC, APIC_DM_NMI); return handled; } -- cgit v1.1 From 860ad7823fdc00cd61dc70e7f35e07fb327cc9a4 Mon Sep 17 00:00:00 2001 From: Sonny Rao Date: Mon, 18 Apr 2011 22:12:59 +0100 Subject: ARM: 6884/1: Fix infinite loop in ARM user perf_event backtrace code The ARM user backtrace code can get into an infinite loop if it runs into an invalid stack frame which points back to itself. This situation has been observed in practice. Fix it by capping the number of entries in the backtrace. This is also what other architectures do in their backtrace code. Signed-off-by: Sonny Rao Acked-by: Jamie Iles Acked-by: Olof Johansson Signed-off-by: Russell King --- arch/arm/kernel/perf_event.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index 979da39..139e3c8 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -746,7 +746,8 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) tail = (struct frame_tail __user *)regs->ARM_fp - 1; - while (tail && !((unsigned long)tail & 0x3)) + while ((entry->nr < PERF_MAX_STACK_DEPTH) && + tail && !((unsigned long)tail & 0x3)) tail = user_backtrace(tail, entry); } -- cgit v1.1 From a8d2518c2a7235f0b772c1f1bd30218130e631a7 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Wed, 27 Apr 2011 01:34:27 +0100 Subject: ARM: 6887/1: Mark broadcast_timer_setup() __cpuinit This function is only called by percpu_timer_setup() which is also __cpuinit marked. Thus it's safe to mark this function as __cpuinit as well. Signed-off-by: Stephen Boyd Signed-off-by: Russell King --- arch/arm/kernel/smp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c index 8fe05ad..f29b8a2 100644 --- a/arch/arm/kernel/smp.c +++ b/arch/arm/kernel/smp.c @@ -479,7 +479,7 @@ static void broadcast_timer_set_mode(enum clock_event_mode mode, { } -static void broadcast_timer_setup(struct clock_event_device *evt) +static void __cpuinit broadcast_timer_setup(struct clock_event_device *evt) { evt->name = "dummy_timer"; evt->features = CLOCK_EVT_FEAT_ONESHOT | -- cgit v1.1 From 20443598d9bdfe3563f901e27fd482a3f5d3d231 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Wed, 27 Apr 2011 16:30:52 +0200 Subject: x86: devicetree: Configure IOAPIC pin only once We use io_apic_setup_irq_pin() in order to configure pin's interrupt number polarity and type. This is done on every irq_create_of_mapping() which happens for instance during pci enable calls. Level typed interrupts are masked by default, edge are unmasked. On the first ->xlate() call the level interrupt is configured and masked. The driver calls request_irq() and the line is unmasked. Lets assume the interrupt line is shared with another device and we call pci_enable_device() for this device. The ->xlate() configures the pin again and it is masked. request_irq() does not unmask the line because it _is_ already unmasked according to its internal state. So the interrupt will never be unmasked again. This patch is based on an earlier work by Torben Hohn and solves the problem by configuring the pin only once. Since all devices must agree on the same type and polarity there is no point in configuring the pin more than once. [ tglx: Split out the ce4100 part into a separate patch ] Cc: Torben Hohn Signed-off-by: Sebastian Andrzej Siewior Link: http://lkml.kernel.org/r/%3C20110427143052.GA15211%40linutronix.de%3E Signed-off-by: Thomas Gleixner --- arch/x86/include/asm/io_apic.h | 2 +- arch/x86/kernel/apic/io_apic.c | 10 +++++----- arch/x86/kernel/devicetree.c | 2 +- 3 files changed, 7 insertions(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/x86/include/asm/io_apic.h b/arch/x86/include/asm/io_apic.h index c4bd267..a97a240 100644 --- a/arch/x86/include/asm/io_apic.h +++ b/arch/x86/include/asm/io_apic.h @@ -150,7 +150,7 @@ void setup_IO_APIC_irq_extra(u32 gsi); extern void ioapic_and_gsi_init(void); extern void ioapic_insert_resources(void); -int io_apic_setup_irq_pin(unsigned int irq, int node, struct io_apic_irq_attr *attr); +int io_apic_setup_irq_pin_once(unsigned int irq, int node, struct io_apic_irq_attr *attr); extern struct IO_APIC_route_entry **alloc_ioapic_entries(void); extern void free_ioapic_entries(struct IO_APIC_route_entry **ioapic_entries); diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 68df09b..45fd33d 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -128,8 +128,8 @@ static int __init parse_noapic(char *str) } early_param("noapic", parse_noapic); -static int io_apic_setup_irq_pin_once(unsigned int irq, int node, - struct io_apic_irq_attr *attr); +static int io_apic_setup_irq_pin(unsigned int irq, int node, + struct io_apic_irq_attr *attr); /* Will be called in mpparse/acpi/sfi codes for saving IRQ info */ void mp_save_irq(struct mpc_intsrc *m) @@ -3570,7 +3570,7 @@ int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev) } #endif /* CONFIG_HT_IRQ */ -int +static int io_apic_setup_irq_pin(unsigned int irq, int node, struct io_apic_irq_attr *attr) { struct irq_cfg *cfg = alloc_irq_and_cfg_at(irq, node); @@ -3585,8 +3585,8 @@ io_apic_setup_irq_pin(unsigned int irq, int node, struct io_apic_irq_attr *attr) return ret; } -static int io_apic_setup_irq_pin_once(unsigned int irq, int node, - struct io_apic_irq_attr *attr) +int io_apic_setup_irq_pin_once(unsigned int irq, int node, + struct io_apic_irq_attr *attr) { unsigned int id = attr->ioapic, pin = attr->ioapic_pin; int ret; diff --git a/arch/x86/kernel/devicetree.c b/arch/x86/kernel/devicetree.c index 706a9fb..e90f084 100644 --- a/arch/x86/kernel/devicetree.c +++ b/arch/x86/kernel/devicetree.c @@ -391,7 +391,7 @@ static int ioapic_xlate(struct irq_domain *id, const u32 *intspec, u32 intsize, set_io_apic_irq_attr(&attr, idx, line, it->trigger, it->polarity); - return io_apic_setup_irq_pin(*out_hwirq, cpu_to_node(0), &attr); + return io_apic_setup_irq_pin_once(*out_hwirq, cpu_to_node(0), &attr); } static void __init ioapic_add_ofnode(struct device_node *np) -- cgit v1.1 From 1ff42c32c7614c2e810ed388fd1ba04a5626b74c Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Wed, 27 Apr 2011 16:30:52 +0200 Subject: x86: ce4100: Configure IOAPIC pins for USB and SATA to level type The USB and SATA ioapic interrrupt pins are configured as edge type, but need to be level type interrupts to work correctly. [ tglx: Split out from the combo patch ] Cc: Torben Hohn Signed-off-by: Sebastian Andrzej Siewior Link: http://lkml.kernel.org/r/%3C20110427143052.GA15211%40linutronix.de%3E Signed-off-by: Thomas Gleixner --- arch/x86/platform/ce4100/falconfalls.dts | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/x86/platform/ce4100/falconfalls.dts b/arch/x86/platform/ce4100/falconfalls.dts index 2d6d226..e70be38 100644 --- a/arch/x86/platform/ce4100/falconfalls.dts +++ b/arch/x86/platform/ce4100/falconfalls.dts @@ -347,7 +347,7 @@ "pciclass0c03"; reg = <0x16800 0x0 0x0 0x0 0x0>; - interrupts = <22 3>; + interrupts = <22 1>; }; usb@d,1 { @@ -357,7 +357,7 @@ "pciclass0c03"; reg = <0x16900 0x0 0x0 0x0 0x0>; - interrupts = <22 3>; + interrupts = <22 1>; }; sata@e,0 { @@ -367,7 +367,7 @@ "pciclass0106"; reg = <0x17000 0x0 0x0 0x0 0x0>; - interrupts = <23 3>; + interrupts = <23 1>; }; flash@f,0 { -- cgit v1.1 From 83a5d2d1b4cc06ce17b44873e004fc0b55552183 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Sat, 12 Mar 2011 03:27:22 +0000 Subject: ARM: Fix .size directive for xscale_dma_a0_map_area gas used to accept (and ignore?) .size directives which referred to undefined symbols, as this does. In binutils 2.21 these are treated as fatal errors. Signed-off-by: Ben Hutchings Signed-off-by: Eric Miao --- arch/arm/mm/proc-xscale.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/mm/proc-xscale.S b/arch/arm/mm/proc-xscale.S index ce233bc..42af976 100644 --- a/arch/arm/mm/proc-xscale.S +++ b/arch/arm/mm/proc-xscale.S @@ -395,7 +395,7 @@ ENTRY(xscale_dma_a0_map_area) teq r2, #DMA_TO_DEVICE beq xscale_dma_clean_range b xscale_dma_flush_range -ENDPROC(xscsale_dma_a0_map_area) +ENDPROC(xscale_dma_a0_map_area) /* * dma_unmap_area(start, size, dir) -- cgit v1.1 From 57d8e02e3cd21bccf2b84b26b42feb79e1f0f83e Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Wed, 27 Apr 2011 15:26:51 -0700 Subject: um: mdd support for 64 bit atomic operations This adds support for 64 bit atomic operations on 32 bit UML systems. XFS needs them since 2.6.38. $ make ARCH=um SUBARCH=i386 ... LD .tmp_vmlinux1 fs/built-in.o: In function `xlog_regrant_reserve_log_space': xfs_log.c:(.text+0xd8584): undefined reference to `atomic64_read_386' xfs_log.c:(.text+0xd85ac): undefined reference to `cmpxchg8b_emu' ... Addresses https://bugzilla.kernel.org/show_bug.cgi?id=32812 Reported-by: Martin Walch Tested-by: Martin Walch Cc: Martin Walch Cc: [2.6.38.x 084189a: um: disable CONFIG_CMPXCHG_LOCAL] Signed-off-by: Richard Weinberger Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/sys-i386/Makefile | 2 +- arch/um/sys-i386/atomic64_cx8_32.S | 225 +++++++++++++++++++++++++++++++++++++ 2 files changed, 226 insertions(+), 1 deletion(-) create mode 100644 arch/um/sys-i386/atomic64_cx8_32.S (limited to 'arch') diff --git a/arch/um/sys-i386/Makefile b/arch/um/sys-i386/Makefile index 804b28d..b1da91c 100644 --- a/arch/um/sys-i386/Makefile +++ b/arch/um/sys-i386/Makefile @@ -4,7 +4,7 @@ obj-y = bug.o bugs.o checksum.o delay.o fault.o ksyms.o ldt.o ptrace.o \ ptrace_user.o setjmp.o signal.o stub.o stub_segv.o syscalls.o sysrq.o \ - sys_call_table.o tls.o + sys_call_table.o tls.o atomic64_cx8_32.o obj-$(CONFIG_BINFMT_ELF) += elfcore.o diff --git a/arch/um/sys-i386/atomic64_cx8_32.S b/arch/um/sys-i386/atomic64_cx8_32.S new file mode 100644 index 0000000..1e901d3 --- /dev/null +++ b/arch/um/sys-i386/atomic64_cx8_32.S @@ -0,0 +1,225 @@ +/* + * atomic64_t for 586+ + * + * Copied from arch/x86/lib/atomic64_cx8_32.S + * + * Copyright © 2010 Luca Barbieri + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + */ + +#include +#include +#include + +.macro SAVE reg + pushl_cfi %\reg + CFI_REL_OFFSET \reg, 0 +.endm + +.macro RESTORE reg + popl_cfi %\reg + CFI_RESTORE \reg +.endm + +.macro read64 reg + movl %ebx, %eax + movl %ecx, %edx +/* we need LOCK_PREFIX since otherwise cmpxchg8b always does the write */ + LOCK_PREFIX + cmpxchg8b (\reg) +.endm + +ENTRY(atomic64_read_cx8) + CFI_STARTPROC + + read64 %ecx + ret + CFI_ENDPROC +ENDPROC(atomic64_read_cx8) + +ENTRY(atomic64_set_cx8) + CFI_STARTPROC + +1: +/* we don't need LOCK_PREFIX since aligned 64-bit writes + * are atomic on 586 and newer */ + cmpxchg8b (%esi) + jne 1b + + ret + CFI_ENDPROC +ENDPROC(atomic64_set_cx8) + +ENTRY(atomic64_xchg_cx8) + CFI_STARTPROC + + movl %ebx, %eax + movl %ecx, %edx +1: + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + + ret + CFI_ENDPROC +ENDPROC(atomic64_xchg_cx8) + +.macro addsub_return func ins insc +ENTRY(atomic64_\func\()_return_cx8) + CFI_STARTPROC + SAVE ebp + SAVE ebx + SAVE esi + SAVE edi + + movl %eax, %esi + movl %edx, %edi + movl %ecx, %ebp + + read64 %ebp +1: + movl %eax, %ebx + movl %edx, %ecx + \ins\()l %esi, %ebx + \insc\()l %edi, %ecx + LOCK_PREFIX + cmpxchg8b (%ebp) + jne 1b + +10: + movl %ebx, %eax + movl %ecx, %edx + RESTORE edi + RESTORE esi + RESTORE ebx + RESTORE ebp + ret + CFI_ENDPROC +ENDPROC(atomic64_\func\()_return_cx8) +.endm + +addsub_return add add adc +addsub_return sub sub sbb + +.macro incdec_return func ins insc +ENTRY(atomic64_\func\()_return_cx8) + CFI_STARTPROC + SAVE ebx + + read64 %esi +1: + movl %eax, %ebx + movl %edx, %ecx + \ins\()l $1, %ebx + \insc\()l $0, %ecx + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + +10: + movl %ebx, %eax + movl %ecx, %edx + RESTORE ebx + ret + CFI_ENDPROC +ENDPROC(atomic64_\func\()_return_cx8) +.endm + +incdec_return inc add adc +incdec_return dec sub sbb + +ENTRY(atomic64_dec_if_positive_cx8) + CFI_STARTPROC + SAVE ebx + + read64 %esi +1: + movl %eax, %ebx + movl %edx, %ecx + subl $1, %ebx + sbb $0, %ecx + js 2f + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + +2: + movl %ebx, %eax + movl %ecx, %edx + RESTORE ebx + ret + CFI_ENDPROC +ENDPROC(atomic64_dec_if_positive_cx8) + +ENTRY(atomic64_add_unless_cx8) + CFI_STARTPROC + SAVE ebp + SAVE ebx +/* these just push these two parameters on the stack */ + SAVE edi + SAVE esi + + movl %ecx, %ebp + movl %eax, %esi + movl %edx, %edi + + read64 %ebp +1: + cmpl %eax, 0(%esp) + je 4f +2: + movl %eax, %ebx + movl %edx, %ecx + addl %esi, %ebx + adcl %edi, %ecx + LOCK_PREFIX + cmpxchg8b (%ebp) + jne 1b + + movl $1, %eax +3: + addl $8, %esp + CFI_ADJUST_CFA_OFFSET -8 + RESTORE ebx + RESTORE ebp + ret +4: + cmpl %edx, 4(%esp) + jne 2b + xorl %eax, %eax + jmp 3b + CFI_ENDPROC +ENDPROC(atomic64_add_unless_cx8) + +ENTRY(atomic64_inc_not_zero_cx8) + CFI_STARTPROC + SAVE ebx + + read64 %esi +1: + testl %eax, %eax + je 4f +2: + movl %eax, %ebx + movl %edx, %ecx + addl $1, %ebx + adcl $0, %ecx + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + + movl $1, %eax +3: + RESTORE ebx + ret +4: + testl %edx, %edx + jne 2b + jmp 3b + CFI_ENDPROC +ENDPROC(atomic64_inc_not_zero_cx8) -- cgit v1.1 From 365a0deae2b6743b3263a71871bbd8c9f66ac34c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 27 Apr 2011 15:26:53 -0700 Subject: uml: fix hppfs build Make HoneyPot ProcFS depend on CONFIG_PROC_FS so that it will build. Recommended by Christoph Hellwig. Addresses https://bugzilla.kernel.org/show_bug.cgi?id=33692 Reported-by: Simon Danner Signed-off-by: Randy Dunlap Cc: Jeff Dike Cc: Christoph Hellwig Signed-off-by: Richard Weinberger Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/Kconfig.um | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/um/Kconfig.um b/arch/um/Kconfig.um index 90a438a..b5e675e 100644 --- a/arch/um/Kconfig.um +++ b/arch/um/Kconfig.um @@ -47,7 +47,7 @@ config HOSTFS config HPPFS tristate "HoneyPot ProcFS (EXPERIMENTAL)" - depends on EXPERIMENTAL + depends on EXPERIMENTAL && PROC_FS help hppfs (HoneyPot ProcFS) is a filesystem which allows UML /proc entries to be overridden, removed, or fabricated from the host. -- cgit v1.1 From 534e3adbd22efa327e6ff27cf2d8ebaad8382ecd Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Wed, 27 Apr 2011 15:26:54 -0700 Subject: um: adjust current_thread_info() for newer gcc versions In some cases gcc >= 4.5.2 will optimize away current_thread_info(). To prevent gcc from doing so the stack address has to be obtained via inline asm. Signed-off-by: Richard Weinberger Acked-by: Kirill A. Shutemov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/include/asm/thread_info.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/um/include/asm/thread_info.h b/arch/um/include/asm/thread_info.h index e2cf786..5bd1bad 100644 --- a/arch/um/include/asm/thread_info.h +++ b/arch/um/include/asm/thread_info.h @@ -49,7 +49,10 @@ static inline struct thread_info *current_thread_info(void) { struct thread_info *ti; unsigned long mask = THREAD_SIZE - 1; - ti = (struct thread_info *) (((unsigned long) &ti) & ~mask); + void *p; + + asm volatile ("" : "=r" (p) : "0" (&ti)); + ti = (struct thread_info *) (((unsigned long)p) & ~mask); return ti; } -- cgit v1.1 From cf3cc1aa9b6d0bf1750143af65829f4368d77492 Mon Sep 17 00:00:00 2001 From: Viktor Rosendahl Date: Mon, 28 Mar 2011 18:56:05 +0300 Subject: kprobes/arm: Fix ldrd/strd emulation Currently emulate_ldrd and emulate_strd don't even have the adjustment of the PC value, so in case of Rn == PC, it will not update the PC incorrectly but instead load/store from the wrong address. Let's add both the adjustment of the PC value and the check for PC == PC. Signed-off-by: Viktor Rosendahl Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 2389131..3b0cf90 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -540,9 +540,12 @@ static void __kprobes emulate_ldrd(struct kprobe *p, struct pt_regs *regs) { insn_2arg_fn_t *i_fn = (insn_2arg_fn_t *)&p->ainsn.insn[0]; kprobe_opcode_t insn = p->opcode; + long ppc = (long)p->addr + 8; int rd = (insn >> 12) & 0xf; int rn = (insn >> 16) & 0xf; int rm = insn & 0xf; /* rm may be invalid, don't care. */ + long rmv = (rm == 15) ? ppc : regs->uregs[rm]; + long rnv = (rn == 15) ? ppc : regs->uregs[rn]; /* Not following the C calling convention here, so need asm(). */ __asm__ __volatile__ ( @@ -554,29 +557,36 @@ static void __kprobes emulate_ldrd(struct kprobe *p, struct pt_regs *regs) "str r0, %[rn] \n\t" /* in case of writeback */ "str r2, %[rd0] \n\t" "str r3, %[rd1] \n\t" - : [rn] "+m" (regs->uregs[rn]), + : [rn] "+m" (rnv), [rd0] "=m" (regs->uregs[rd]), [rd1] "=m" (regs->uregs[rd+1]) - : [rm] "m" (regs->uregs[rm]), + : [rm] "m" (rmv), [cpsr] "r" (regs->ARM_cpsr), [i_fn] "r" (i_fn) : "r0", "r1", "r2", "r3", "lr", "cc" ); + if (rn != 15) + regs->uregs[rn] = rnv; /* Save Rn in case of writeback. */ } static void __kprobes emulate_strd(struct kprobe *p, struct pt_regs *regs) { insn_4arg_fn_t *i_fn = (insn_4arg_fn_t *)&p->ainsn.insn[0]; kprobe_opcode_t insn = p->opcode; + long ppc = (long)p->addr + 8; int rd = (insn >> 12) & 0xf; int rn = (insn >> 16) & 0xf; int rm = insn & 0xf; - long rnv = regs->uregs[rn]; - long rmv = regs->uregs[rm]; /* rm/rmv may be invalid, don't care. */ + long rnv = (rn == 15) ? ppc : regs->uregs[rn]; + /* rm/rmv may be invalid, don't care. */ + long rmv = (rm == 15) ? ppc : regs->uregs[rm]; + long rnv_wb; - regs->uregs[rn] = insnslot_4arg_rflags(rnv, rmv, regs->uregs[rd], + rnv_wb = insnslot_4arg_rflags(rnv, rmv, regs->uregs[rd], regs->uregs[rd+1], regs->ARM_cpsr, i_fn); + if (rn != 15) + regs->uregs[rn] = rnv_wb; /* Save Rn in case of writeback. */ } static void __kprobes emulate_ldr(struct kprobe *p, struct pt_regs *regs) -- cgit v1.1 From 073090cb701148396b1130be81f8ac84a41f196d Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Wed, 6 Apr 2011 11:17:09 +0100 Subject: ARM: kprobes: Fix probing of conditionally executed instructions When a kprobe is placed onto conditionally executed ARM instructions, many of the emulation routines used to single step them produce corrupt register results. Rather than fix all of these cases we modify the framework which calls them to test the relevant condition flags and, if the test fails, skip calling the emulation code. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/include/asm/kprobes.h | 3 ++ arch/arm/kernel/kprobes-decode.c | 91 ++++++++++++++++++++++++++++++++++++++++ arch/arm/kernel/kprobes.c | 3 +- 3 files changed, 96 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/include/asm/kprobes.h b/arch/arm/include/asm/kprobes.h index bb8a19b..e46bdd0 100644 --- a/arch/arm/include/asm/kprobes.h +++ b/arch/arm/include/asm/kprobes.h @@ -39,10 +39,13 @@ typedef u32 kprobe_opcode_t; struct kprobe; typedef void (kprobe_insn_handler_t)(struct kprobe *, struct pt_regs *); +typedef unsigned long (kprobe_check_cc)(unsigned long); + /* Architecture specific copy of original instruction. */ struct arch_specific_insn { kprobe_opcode_t *insn; kprobe_insn_handler_t *insn_handler; + kprobe_check_cc *insn_check_cc; }; struct prev_kprobe { diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 3b0cf90..2e84169 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1396,6 +1396,96 @@ space_cccc_111x(kprobe_opcode_t insn, struct arch_specific_insn *asi) return INSN_GOOD; } +static unsigned long __kprobes __check_eq(unsigned long cpsr) +{ + return cpsr & PSR_Z_BIT; +} + +static unsigned long __kprobes __check_ne(unsigned long cpsr) +{ + return (~cpsr) & PSR_Z_BIT; +} + +static unsigned long __kprobes __check_cs(unsigned long cpsr) +{ + return cpsr & PSR_C_BIT; +} + +static unsigned long __kprobes __check_cc(unsigned long cpsr) +{ + return (~cpsr) & PSR_C_BIT; +} + +static unsigned long __kprobes __check_mi(unsigned long cpsr) +{ + return cpsr & PSR_N_BIT; +} + +static unsigned long __kprobes __check_pl(unsigned long cpsr) +{ + return (~cpsr) & PSR_N_BIT; +} + +static unsigned long __kprobes __check_vs(unsigned long cpsr) +{ + return cpsr & PSR_V_BIT; +} + +static unsigned long __kprobes __check_vc(unsigned long cpsr) +{ + return (~cpsr) & PSR_V_BIT; +} + +static unsigned long __kprobes __check_hi(unsigned long cpsr) +{ + cpsr &= ~(cpsr >> 1); /* PSR_C_BIT &= ~PSR_Z_BIT */ + return cpsr & PSR_C_BIT; +} + +static unsigned long __kprobes __check_ls(unsigned long cpsr) +{ + cpsr &= ~(cpsr >> 1); /* PSR_C_BIT &= ~PSR_Z_BIT */ + return (~cpsr) & PSR_C_BIT; +} + +static unsigned long __kprobes __check_ge(unsigned long cpsr) +{ + cpsr ^= (cpsr << 3); /* PSR_N_BIT ^= PSR_V_BIT */ + return (~cpsr) & PSR_N_BIT; +} + +static unsigned long __kprobes __check_lt(unsigned long cpsr) +{ + cpsr ^= (cpsr << 3); /* PSR_N_BIT ^= PSR_V_BIT */ + return cpsr & PSR_N_BIT; +} + +static unsigned long __kprobes __check_gt(unsigned long cpsr) +{ + unsigned long temp = cpsr ^ (cpsr << 3); /* PSR_N_BIT ^= PSR_V_BIT */ + temp |= (cpsr << 1); /* PSR_N_BIT |= PSR_Z_BIT */ + return (~temp) & PSR_N_BIT; +} + +static unsigned long __kprobes __check_le(unsigned long cpsr) +{ + unsigned long temp = cpsr ^ (cpsr << 3); /* PSR_N_BIT ^= PSR_V_BIT */ + temp |= (cpsr << 1); /* PSR_N_BIT |= PSR_Z_BIT */ + return temp & PSR_N_BIT; +} + +static unsigned long __kprobes __check_al(unsigned long cpsr) +{ + return true; +} + +static kprobe_check_cc * const condition_checks[16] = { + &__check_eq, &__check_ne, &__check_cs, &__check_cc, + &__check_mi, &__check_pl, &__check_vs, &__check_vc, + &__check_hi, &__check_ls, &__check_ge, &__check_lt, + &__check_gt, &__check_le, &__check_al, &__check_al +}; + /* Return: * INSN_REJECTED If instruction is one not allowed to kprobe, * INSN_GOOD If instruction is supported and uses instruction slot, @@ -1411,6 +1501,7 @@ space_cccc_111x(kprobe_opcode_t insn, struct arch_specific_insn *asi) enum kprobe_insn __kprobes arm_kprobe_decode_insn(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + asi->insn_check_cc = condition_checks[insn>>28]; asi->insn[1] = KPROBE_RETURN_INSTRUCTION; if ((insn & 0xf0000000) == 0xf0000000) { diff --git a/arch/arm/kernel/kprobes.c b/arch/arm/kernel/kprobes.c index 2ba7deb..1656c87 100644 --- a/arch/arm/kernel/kprobes.c +++ b/arch/arm/kernel/kprobes.c @@ -134,7 +134,8 @@ static void __kprobes singlestep(struct kprobe *p, struct pt_regs *regs, struct kprobe_ctlblk *kcb) { regs->ARM_pc += 4; - p->ainsn.insn_handler(p, regs); + if (p->ainsn.insn_check_cc(regs->ARM_cpsr)) + p->ainsn.insn_handler(p, regs); } /* -- cgit v1.1 From a539f5d46c868cb1f37b92e62e040b400571aed4 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Wed, 6 Apr 2011 11:17:10 +0100 Subject: ARM: kprobes: Remove redundant condition checks from simulation routines Now we have the framework code handling conditionally executed instructions we can remove redundant checks in individual simulation routines. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 30 +++--------------------------- 1 file changed, 3 insertions(+), 27 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 2e84169..f52fac0 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -71,10 +71,6 @@ #define PSR_fs (PSR_f|PSR_s) #define KPROBE_RETURN_INSTRUCTION 0xe1a0f00e /* mov pc, lr */ -#define SET_R0_TRUE_INSTRUCTION 0xe3a00001 /* mov r0, #1 */ - -#define truecc_insn(insn) (((insn) & 0xf0000000) | \ - (SET_R0_TRUE_INSTRUCTION & 0x0fffffff)) typedef long (insn_0arg_fn_t)(void); typedef long (insn_1arg_fn_t)(long); @@ -419,14 +415,10 @@ insnslot_llret_4arg_rwflags(long r0, long r1, long r2, long r3, long *cpsr, static void __kprobes simulate_bbl(struct kprobe *p, struct pt_regs *regs) { - insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; kprobe_opcode_t insn = p->opcode; long iaddr = (long)p->addr; int disp = branch_displacement(insn); - if (!insnslot_1arg_rflags(0, regs->ARM_cpsr, i_fn)) - return; - if (insn & (1 << 24)) regs->ARM_lr = iaddr + 4; @@ -446,14 +438,10 @@ static void __kprobes simulate_blx1(struct kprobe *p, struct pt_regs *regs) static void __kprobes simulate_blx2bx(struct kprobe *p, struct pt_regs *regs) { - insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; kprobe_opcode_t insn = p->opcode; int rm = insn & 0xf; long rmv = regs->uregs[rm]; - if (!insnslot_1arg_rflags(0, regs->ARM_cpsr, i_fn)) - return; - if (insn & (1 << 5)) regs->ARM_lr = (long)p->addr + 4; @@ -465,7 +453,6 @@ static void __kprobes simulate_blx2bx(struct kprobe *p, struct pt_regs *regs) static void __kprobes simulate_ldm1stm1(struct kprobe *p, struct pt_regs *regs) { - insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; kprobe_opcode_t insn = p->opcode; int rn = (insn >> 16) & 0xf; int lbit = insn & (1 << 20); @@ -476,9 +463,6 @@ static void __kprobes simulate_ldm1stm1(struct kprobe *p, struct pt_regs *regs) int reg_bit_vector; int reg_count; - if (!insnslot_1arg_rflags(0, regs->ARM_cpsr, i_fn)) - return; - reg_count = 0; reg_bit_vector = insn & 0xffff; while (reg_bit_vector) { @@ -510,11 +494,6 @@ static void __kprobes simulate_ldm1stm1(struct kprobe *p, struct pt_regs *regs) static void __kprobes simulate_stm1_pc(struct kprobe *p, struct pt_regs *regs) { - insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; - - if (!insnslot_1arg_rflags(0, regs->ARM_cpsr, i_fn)) - return; - regs->ARM_pc = (long)p->addr + str_pc_offset; simulate_ldm1stm1(p, regs); regs->ARM_pc = (long)p->addr + 4; @@ -1056,9 +1035,8 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* BLX(2) : cccc 0001 0010 xxxx xxxx xxxx 0011 xxxx */ /* BX : cccc 0001 0010 xxxx xxxx xxxx 0001 xxxx */ if ((insn & 0x0ff000d0) == 0x01200010) { - asi->insn[0] = truecc_insn(insn); asi->insn_handler = simulate_blx2bx; - return INSN_GOOD; + return INSN_GOOD_NO_SLOT; } /* CLZ : cccc 0001 0110 xxxx xxxx xxxx 0001 xxxx */ @@ -1333,10 +1311,9 @@ space_cccc_100x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* LDM(1) : cccc 100x x0x1 xxxx xxxx xxxx xxxx xxxx */ /* STM(1) : cccc 100x x0x0 xxxx xxxx xxxx xxxx xxxx */ - asi->insn[0] = truecc_insn(insn); asi->insn_handler = ((insn & 0x108000) == 0x008000) ? /* STM & R15 */ simulate_stm1_pc : simulate_ldm1stm1; - return INSN_GOOD; + return INSN_GOOD_NO_SLOT; } static enum kprobe_insn __kprobes @@ -1344,9 +1321,8 @@ space_cccc_101x(kprobe_opcode_t insn, struct arch_specific_insn *asi) { /* B : cccc 1010 xxxx xxxx xxxx xxxx xxxx xxxx */ /* BL : cccc 1011 xxxx xxxx xxxx xxxx xxxx xxxx */ - asi->insn[0] = truecc_insn(insn); asi->insn_handler = simulate_bbl; - return INSN_GOOD; + return INSN_GOOD_NO_SLOT; } static enum kprobe_insn __kprobes -- cgit v1.1 From ad111ce46674a473370d302325db8f418ac4fe92 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Wed, 6 Apr 2011 11:17:11 +0100 Subject: ARM: kprobes: Fix emulation of CMP, CMN, TST and TEQ instructions. Probing these instructions was corrupting R0 because the emulation code didn't account for the fact that they don't write a result to a register. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 55 ++++++++++++++++++++++++++++++++++++++-- 1 file changed, 53 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index f52fac0..ee29d33 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -808,6 +808,17 @@ emulate_alu_imm_rwflags(struct kprobe *p, struct pt_regs *regs) } static void __kprobes +emulate_alu_tests_imm(struct kprobe *p, struct pt_regs *regs) +{ + insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; + kprobe_opcode_t insn = p->opcode; + int rn = (insn >> 16) & 0xf; + long rnv = (rn == 15) ? (long)p->addr + 8 : regs->uregs[rn]; + + insnslot_1arg_rwflags(rnv, ®s->ARM_cpsr, i_fn); +} + +static void __kprobes emulate_alu_rflags(struct kprobe *p, struct pt_regs *regs) { insn_3arg_fn_t *i_fn = (insn_3arg_fn_t *)&p->ainsn.insn[0]; @@ -843,6 +854,22 @@ emulate_alu_rwflags(struct kprobe *p, struct pt_regs *regs) insnslot_3arg_rwflags(rnv, rmv, rsv, ®s->ARM_cpsr, i_fn); } +static void __kprobes +emulate_alu_tests(struct kprobe *p, struct pt_regs *regs) +{ + insn_3arg_fn_t *i_fn = (insn_3arg_fn_t *)&p->ainsn.insn[0]; + kprobe_opcode_t insn = p->opcode; + long ppc = (long)p->addr + 8; + int rn = (insn >> 16) & 0xf; + int rs = (insn >> 8) & 0xf; /* rs/rsv may be invalid, don't care. */ + int rm = insn & 0xf; + long rnv = (rn == 15) ? ppc : regs->uregs[rn]; + long rmv = (rm == 15) ? ppc : regs->uregs[rm]; + long rsv = regs->uregs[rs]; + + insnslot_3arg_rwflags(rnv, rmv, rsv, ®s->ARM_cpsr, i_fn); +} + static enum kprobe_insn __kprobes prep_emulate_ldr_str(kprobe_opcode_t insn, struct arch_specific_insn *asi) { @@ -1142,8 +1169,20 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) insn |= 0x00000200; /* Rs = r2 */ } asi->insn[0] = insn; - asi->insn_handler = (insn & (1 << 20)) ? /* S-bit */ + + if ((insn & 0x0f900000) == 0x01100000) { + /* + * TST : cccc 0001 0001 xxxx xxxx xxxx xxxx xxxx + * TEQ : cccc 0001 0011 xxxx xxxx xxxx xxxx xxxx + * CMP : cccc 0001 0101 xxxx xxxx xxxx xxxx xxxx + * CMN : cccc 0001 0111 xxxx xxxx xxxx xxxx xxxx + */ + asi->insn_handler = emulate_alu_tests; + } else { + /* ALU ops which write to Rd */ + asi->insn_handler = (insn & (1 << 20)) ? /* S-bit */ emulate_alu_rwflags : emulate_alu_rflags; + } return INSN_GOOD; } @@ -1170,8 +1209,20 @@ space_cccc_001x(kprobe_opcode_t insn, struct arch_specific_insn *asi) */ insn &= 0xffff0fff; /* Rd = r0 */ asi->insn[0] = insn; - asi->insn_handler = (insn & (1 << 20)) ? /* S-bit */ + + if ((insn & 0x0f900000) == 0x03100000) { + /* + * TST : cccc 0011 0001 xxxx xxxx xxxx xxxx xxxx + * TEQ : cccc 0011 0011 xxxx xxxx xxxx xxxx xxxx + * CMP : cccc 0011 0101 xxxx xxxx xxxx xxxx xxxx + * CMN : cccc 0011 0111 xxxx xxxx xxxx xxxx xxxx + */ + asi->insn_handler = emulate_alu_tests_imm; + } else { + /* ALU ops which write to Rd */ + asi->insn_handler = (insn & (1 << 20)) ? /* S-bit */ emulate_alu_imm_rwflags : emulate_alu_imm_rflags; + } return INSN_GOOD; } -- cgit v1.1 From 896a74e19d0131413a96502429994bc8e6bbbe5a Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Wed, 6 Apr 2011 11:17:12 +0100 Subject: ARM: kprobes: Fix emulation of Data-processing (immediate) instructions Emulation of instructions like "ADD rd, rn, #" would result in a corrupted value for rd. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index ee29d33..baf053e 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1207,7 +1207,7 @@ space_cccc_001x(kprobe_opcode_t insn, struct arch_specific_insn *asi) * *S (bit 20) updates condition codes * ADC/SBC/RSC reads the C flag */ - insn &= 0xffff0fff; /* Rd = r0 */ + insn &= 0xfff00fff; /* Rn = r0 and Rd = r0 */ asi->insn[0] = insn; if ((insn & 0x0f900000) == 0x03100000) { -- cgit v1.1 From 51468ea91efad9c7e6dbae43cd8bdc423ec61709 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Thu, 7 Apr 2011 13:25:15 +0100 Subject: ARM: kprobes: Reject probing MRS instructions which read SPSR We need to reject probing of instructions which read SPSR because we can't handle this as the value in SPSR is lost when the exception handler for the probe breakpoint first runs. This patch also fixes the bitmask for MRS instructions decoding to include checking bits 5-7. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index baf053e..e5bc576 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1026,14 +1026,16 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* cccc 0001 0xx0 xxxx xxxx xxxx xxxx xxx0 xxxx */ if ((insn & 0x0f900010) == 0x01000000) { - /* BXJ : cccc 0001 0010 xxxx xxxx xxxx 0010 xxxx */ - /* MSR : cccc 0001 0x10 xxxx xxxx xxxx 0000 xxxx */ + /* BXJ : cccc 0001 0010 xxxx xxxx xxxx 0010 xxxx */ + /* MSR : cccc 0001 0x10 xxxx xxxx xxxx 0000 xxxx */ + /* MRS spsr : cccc 0001 0100 xxxx xxxx xxxx 0000 xxxx */ if ((insn & 0x0ff000f0) == 0x01200020 || - (insn & 0x0fb000f0) == 0x01200000) + (insn & 0x0fb000f0) == 0x01200000 || + (insn & 0x0ff000f0) == 0x01400000) return INSN_REJECTED; - /* MRS : cccc 0001 0x00 xxxx xxxx xxxx 0000 xxxx */ - if ((insn & 0x0fb00010) == 0x01000000) + /* MRS cpsr : cccc 0001 0000 xxxx xxxx xxxx 0000 xxxx */ + if ((insn & 0x0ff000f0) == 0x01000000) return prep_emulate_rd12(insn, asi); /* SMLALxy : cccc 0001 0100 xxxx xxxx xxxx 1xx0 xxxx */ -- cgit v1.1 From c412aba2a1243192a4d53736805a96bdda915608 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Thu, 7 Apr 2011 13:25:16 +0100 Subject: ARM: kprobes: Fix emulation of MRS instruction The MRS instruction should set mode and interrupt bits in the read value so it is simpler to use a new simulation routine (simulate_mrs) rather than some modified emulation. prep_emulate_rd12 is now unused and removed. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index e5bc576..4ab83f4 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -451,6 +451,14 @@ static void __kprobes simulate_blx2bx(struct kprobe *p, struct pt_regs *regs) regs->ARM_cpsr |= PSR_T_BIT; } +static void __kprobes simulate_mrs(struct kprobe *p, struct pt_regs *regs) +{ + kprobe_opcode_t insn = p->opcode; + int rd = (insn >> 12) & 0xf; + unsigned long mask = 0xf8ff03df; /* Mask out execution state */ + regs->uregs[rd] = regs->ARM_cpsr & mask; +} + static void __kprobes simulate_ldm1stm1(struct kprobe *p, struct pt_regs *regs) { kprobe_opcode_t insn = p->opcode; @@ -896,15 +904,6 @@ prep_emulate_rd12rm0(kprobe_opcode_t insn, struct arch_specific_insn *asi) } static enum kprobe_insn __kprobes -prep_emulate_rd12(kprobe_opcode_t insn, struct arch_specific_insn *asi) -{ - insn &= 0xffff0fff; /* Rd = r0 */ - asi->insn[0] = insn; - asi->insn_handler = emulate_rd12; - return INSN_GOOD; -} - -static enum kprobe_insn __kprobes prep_emulate_rd12rn16rm0_wflags(kprobe_opcode_t insn, struct arch_specific_insn *asi) { @@ -1035,8 +1034,10 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) return INSN_REJECTED; /* MRS cpsr : cccc 0001 0000 xxxx xxxx xxxx 0000 xxxx */ - if ((insn & 0x0ff000f0) == 0x01000000) - return prep_emulate_rd12(insn, asi); + if ((insn & 0x0ff000f0) == 0x01000000) { + asi->insn_handler = simulate_mrs; + return INSN_GOOD_NO_SLOT; + } /* SMLALxy : cccc 0001 0100 xxxx xxxx xxxx 1xx0 xxxx */ if ((insn & 0x0ff00090) == 0x01400080) -- cgit v1.1 From 983ebd9365096c3386f6ed0978333e15f66024f5 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Thu, 7 Apr 2011 13:25:17 +0100 Subject: ARM: kprobes: Reject probing of instructions which write to PC unpredictably. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 4ab83f4..6d09db9 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -68,6 +68,8 @@ #define branch_displacement(insn) sign_extend(((insn) & 0xffffff) << 2, 25) +#define is_r15(insn, bitpos) (((insn) & (0xf << bitpos)) == (0xf << bitpos)) + #define PSR_fs (PSR_f|PSR_s) #define KPROBE_RETURN_INSTRUCTION 0xe1a0f00e /* mov pc, lr */ @@ -897,6 +899,9 @@ prep_emulate_ldr_str(kprobe_opcode_t insn, struct arch_specific_insn *asi) static enum kprobe_insn __kprobes prep_emulate_rd12rm0(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ + insn &= 0xffff0ff0; /* Rd = r0, Rm = r0 */ asi->insn[0] = insn; asi->insn_handler = emulate_rd12rm0; @@ -907,6 +912,9 @@ static enum kprobe_insn __kprobes prep_emulate_rd12rn16rm0_wflags(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ + insn &= 0xfff00ff0; /* Rd = r0, Rn = r0 */ insn |= 0x00000001; /* Rm = r1 */ asi->insn[0] = insn; @@ -918,6 +926,9 @@ static enum kprobe_insn __kprobes prep_emulate_rd16rs8rm0_wflags(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + if (is_r15(insn, 16)) + return INSN_REJECTED; /* Rd is PC */ + insn &= 0xfff0f0f0; /* Rd = r0, Rs = r0 */ insn |= 0x00000001; /* Rm = r1 */ asi->insn[0] = insn; @@ -929,6 +940,9 @@ static enum kprobe_insn __kprobes prep_emulate_rd16rn12rs8rm0_wflags(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + if (is_r15(insn, 16)) + return INSN_REJECTED; /* Rd is PC */ + insn &= 0xfff000f0; /* Rd = r0, Rn = r0 */ insn |= 0x00000102; /* Rs = r1, Rm = r2 */ asi->insn[0] = insn; @@ -940,6 +954,9 @@ static enum kprobe_insn __kprobes prep_emulate_rdhi16rdlo12rs8rm0_wflags(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + if (is_r15(insn, 16) || is_r15(insn, 12)) + return INSN_REJECTED; /* RdHi or RdLo is PC */ + insn &= 0xfff000f0; /* RdHi = r0, RdLo = r1 */ insn |= 0x00001203; /* Rs = r2, Rm = r3 */ asi->insn[0] = insn; @@ -1035,6 +1052,8 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* MRS cpsr : cccc 0001 0000 xxxx xxxx xxxx 0000 xxxx */ if ((insn & 0x0ff000f0) == 0x01000000) { + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ asi->insn_handler = simulate_mrs; return INSN_GOOD_NO_SLOT; } @@ -1065,6 +1084,8 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* BLX(2) : cccc 0001 0010 xxxx xxxx xxxx 0011 xxxx */ /* BX : cccc 0001 0010 xxxx xxxx xxxx 0001 xxxx */ if ((insn & 0x0ff000d0) == 0x01200010) { + if ((insn & 0x0ff000ff) == 0x0120003f) + return INSN_REJECTED; /* BLX pc */ asi->insn_handler = simulate_blx2bx; return INSN_GOOD_NO_SLOT; } @@ -1234,6 +1255,8 @@ space_cccc_0110__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) { /* SEL : cccc 0110 1000 xxxx xxxx xxxx 1011 xxxx GE: !!! */ if ((insn & 0x0ff000f0) == 0x068000b0) { + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ insn &= 0xfff00ff0; /* Rd = r0, Rn = r0 */ insn |= 0x00000001; /* Rm = r1 */ asi->insn[0] = insn; @@ -1247,6 +1270,8 @@ space_cccc_0110__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* USAT16 : cccc 0110 1110 xxxx xxxx xxxx 0011 xxxx :Q */ if ((insn & 0x0fa00030) == 0x06a00010 || (insn & 0x0fb000f0) == 0x06a00030) { + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ insn &= 0xffff0ff0; /* Rd = r0, Rm = r0 */ asi->insn[0] = insn; asi->insn_handler = emulate_sat; @@ -1384,6 +1409,9 @@ space_cccc_1100_010x(kprobe_opcode_t insn, struct arch_specific_insn *asi) { /* MCRR : cccc 1100 0100 xxxx xxxx xxxx xxxx xxxx : (Rd!=Rn) */ /* MRRC : cccc 1100 0101 xxxx xxxx xxxx xxxx xxxx : (Rd!=Rn) */ + if (is_r15(insn, 16) || is_r15(insn, 12)) + return INSN_REJECTED; /* Rn or Rd is PC */ + insn &= 0xfff00fff; insn |= 0x00001000; /* Rn = r0, Rd = r1 */ asi->insn[0] = insn; -- cgit v1.1 From 75539aea4cb96823171cabac7391f1a26fbf9c1b Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Thu, 7 Apr 2011 13:25:18 +0100 Subject: ARM: kprobes: Fix error in comment Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 6d09db9..54c1759 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1069,7 +1069,7 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) return prep_emulate_rd16rs8rm0_wflags(insn, asi); /* SMLAxy : cccc 0001 0000 xxxx xxxx xxxx 1xx0 xxxx : Q */ - /* SMLAWy : cccc 0001 0010 xxxx xxxx xxxx 0x00 xxxx : Q */ + /* SMLAWy : cccc 0001 0010 xxxx xxxx xxxx 1x00 xxxx : Q */ return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); } -- cgit v1.1 From ba48d40713ae5afe758dd5525a0f8740cce71d22 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Thu, 7 Apr 2011 13:25:19 +0100 Subject: ARM: kprobes: Reject probing of undefined multiply instructions The instructions space for 'Multiply and multiply-accumulate' instructions contains some undefined patterns. We need to reject probing of these because they may in future become defined and the kprobes code may then emulate them faultily. This has already happened with the new MLS instruction which this patch also adds correct decoding for as well as tightening up other decoding tests. (Before this patch the wrong emulation routine was being called for MLS though it still produced correct results.) Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 54c1759..8f3f03e 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1102,13 +1102,16 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) } /* cccc 0000 xxxx xxxx xxxx xxxx xxxx 1001 xxxx */ - else if ((insn & 0x0f000090) == 0x00000090) { + else if ((insn & 0x0f0000f0) == 0x00000090) { /* MUL : cccc 0000 0000 xxxx xxxx xxxx 1001 xxxx : */ /* MULS : cccc 0000 0001 xxxx xxxx xxxx 1001 xxxx :cc */ /* MLA : cccc 0000 0010 xxxx xxxx xxxx 1001 xxxx : */ /* MLAS : cccc 0000 0011 xxxx xxxx xxxx 1001 xxxx :cc */ /* UMAAL : cccc 0000 0100 xxxx xxxx xxxx 1001 xxxx : */ + /* undef : cccc 0000 0101 xxxx xxxx xxxx 1001 xxxx : */ + /* MLS : cccc 0000 0110 xxxx xxxx xxxx 1001 xxxx : */ + /* undef : cccc 0000 0111 xxxx xxxx xxxx 1001 xxxx : */ /* UMULL : cccc 0000 1000 xxxx xxxx xxxx 1001 xxxx : */ /* UMULLS : cccc 0000 1001 xxxx xxxx xxxx 1001 xxxx :cc */ /* UMLAL : cccc 0000 1010 xxxx xxxx xxxx 1001 xxxx : */ @@ -1117,9 +1120,11 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* SMULLS : cccc 0000 1101 xxxx xxxx xxxx 1001 xxxx :cc */ /* SMLAL : cccc 0000 1110 xxxx xxxx xxxx 1001 xxxx : */ /* SMLALS : cccc 0000 1111 xxxx xxxx xxxx 1001 xxxx :cc */ - if ((insn & 0x0fe000f0) == 0x00000090) { + if ((insn & 0x00d00000) == 0x00500000) { + return INSN_REJECTED; + } else if ((insn & 0x00e00000) == 0x00000000) { return prep_emulate_rd16rs8rm0_wflags(insn, asi); - } else if ((insn & 0x0fe000f0) == 0x00200090) { + } else if ((insn & 0x00a00000) == 0x00200000) { return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); } else { return prep_emulate_rdhi16rdlo12rs8rm0_wflags(insn, asi); -- cgit v1.1 From ec58d7f2373cec47f30b773d40a89f18bf6fc489 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Fri, 8 Apr 2011 15:32:53 +0100 Subject: ARM: kprobes: Reject probing of STREX and LDREX instructions The emulation code for STREX and LDREX instructions is faulty, however, rather than attempting to fix this we reject probes of these instructions. We do this because they can never succeed in gaining exclusive access as the exception framework clears the exclusivity monitor when a probes breakpoint is hit. (This is a general problem when probing all instructions executing between a LDREX and its corresponding STREX and can lead to infinite retry loops.) Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 27 ++++++++++++++++++++++----- 1 file changed, 22 insertions(+), 5 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 8f3f03e..c24e21e 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1136,17 +1136,34 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* SWP : cccc 0001 0000 xxxx xxxx xxxx 1001 xxxx */ /* SWPB : cccc 0001 0100 xxxx xxxx xxxx 1001 xxxx */ - /* LDRD : cccc 000x xxx0 xxxx xxxx xxxx 1101 xxxx */ - /* STRD : cccc 000x xxx0 xxxx xxxx xxxx 1111 xxxx */ + /* ??? : cccc 0001 0x01 xxxx xxxx xxxx 1001 xxxx */ + /* ??? : cccc 0001 0x10 xxxx xxxx xxxx 1001 xxxx */ + /* ??? : cccc 0001 0x11 xxxx xxxx xxxx 1001 xxxx */ /* STREX : cccc 0001 1000 xxxx xxxx xxxx 1001 xxxx */ /* LDREX : cccc 0001 1001 xxxx xxxx xxxx 1001 xxxx */ + /* STREXD: cccc 0001 1010 xxxx xxxx xxxx 1001 xxxx */ + /* LDREXD: cccc 0001 1011 xxxx xxxx xxxx 1001 xxxx */ + /* STREXB: cccc 0001 1100 xxxx xxxx xxxx 1001 xxxx */ + /* LDREXB: cccc 0001 1101 xxxx xxxx xxxx 1001 xxxx */ + /* STREXH: cccc 0001 1110 xxxx xxxx xxxx 1001 xxxx */ + /* LDREXH: cccc 0001 1111 xxxx xxxx xxxx 1001 xxxx */ + + /* LDRD : cccc 000x xxx0 xxxx xxxx xxxx 1101 xxxx */ + /* STRD : cccc 000x xxx0 xxxx xxxx xxxx 1111 xxxx */ /* LDRH : cccc 000x xxx1 xxxx xxxx xxxx 1011 xxxx */ /* STRH : cccc 000x xxx0 xxxx xxxx xxxx 1011 xxxx */ /* LDRSB : cccc 000x xxx1 xxxx xxxx xxxx 1101 xxxx */ /* LDRSH : cccc 000x xxx1 xxxx xxxx xxxx 1111 xxxx */ - if ((insn & 0x0fb000f0) == 0x01000090) { - /* SWP/SWPB */ - return prep_emulate_rd12rn16rm0_wflags(insn, asi); + if ((insn & 0x0f0000f0) == 0x01000090) { + if ((insn & 0x0fb000f0) == 0x01000090) { + /* SWP/SWPB */ + return prep_emulate_rd12rn16rm0_wflags(insn, + asi); + } else { + /* STREX/LDREX variants and unallocaed space */ + return INSN_REJECTED; + } + } else if ((insn & 0x0e1000d0) == 0x00000d0) { /* STRD/LDRD */ insn &= 0xfff00fff; -- cgit v1.1 From 6823fc85fcfba11675f2027aadf2d5291c6f351b Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Fri, 8 Apr 2011 15:32:54 +0100 Subject: ARM: kprobes: Fix emulation of LDRH, STRH, LDRSB and LDRSH instructions The decoding of these instructions got the register indexed and immediate indexed forms the wrong way around, causing incorrect emulation. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index c24e21e..348ff47 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -883,11 +883,12 @@ emulate_alu_tests(struct kprobe *p, struct pt_regs *regs) static enum kprobe_insn __kprobes prep_emulate_ldr_str(kprobe_opcode_t insn, struct arch_specific_insn *asi) { - int ibit = (insn & (1 << 26)) ? 25 : 22; + int not_imm = (insn & (1 << 26)) ? (insn & (1 << 25)) + : (~insn & (1 << 22)); insn &= 0xfff00fff; insn |= 0x00001000; /* Rn = r0, Rd = r1 */ - if (insn & (1 << ibit)) { + if (not_imm) { insn &= ~0xf; insn |= 2; /* Rm = r2 */ } -- cgit v1.1 From 54823accfcfc715e9e757a621afb40dabc01d033 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Fri, 8 Apr 2011 15:32:55 +0100 Subject: ARM: kprobes: Reject probing of LDR/STR instructions which update PC unpredictably Using PC as an base register with writeback is UNPREDICTABLE, as is non word-sized loads or stores of PC. (We only really care about preventing loads to PC but it keeps the code simpler if we also exclude stores.) Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 348ff47..a4dba1f 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -70,6 +70,12 @@ #define is_r15(insn, bitpos) (((insn) & (0xf << bitpos)) == (0xf << bitpos)) +/* + * Test if load/store instructions writeback the address register. + * if P (bit 24) == 0 or W (bit 21) == 1 + */ +#define is_writeback(insn) ((insn ^ 0x01000000) & 0x01200000) + #define PSR_fs (PSR_f|PSR_s) #define KPROBE_RETURN_INSTRUCTION 0xe1a0f00e /* mov pc, lr */ @@ -886,6 +892,9 @@ prep_emulate_ldr_str(kprobe_opcode_t insn, struct arch_specific_insn *asi) int not_imm = (insn & (1 << 26)) ? (insn & (1 << 25)) : (~insn & (1 << 22)); + if (is_writeback(insn) && is_r15(insn, 16)) + return INSN_REJECTED; /* Writeback to PC */ + insn &= 0xfff00fff; insn |= 0x00001000; /* Rn = r0, Rd = r1 */ if (not_imm) { @@ -1167,6 +1176,11 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) } else if ((insn & 0x0e1000d0) == 0x00000d0) { /* STRD/LDRD */ + if ((insn & 0x0000e000) == 0x0000e000) + return INSN_REJECTED; /* Rd is LR or PC */ + if (is_writeback(insn) && is_r15(insn, 16)) + return INSN_REJECTED; /* Writeback to PC */ + insn &= 0xfff00fff; insn |= 0x00002000; /* Rn = r0, Rd = r2 */ if (insn & (1 << 22)) { @@ -1180,6 +1194,9 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) return INSN_GOOD; } + /* LDRH/STRH/LDRSB/LDRSH */ + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ return prep_emulate_ldr_str(insn, asi); } -- cgit v1.1 From 5c6b76fc7d8220e8f00e7a49fb56ca852d7fb661 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Fri, 8 Apr 2011 15:32:56 +0100 Subject: ARM: kprobes: Fix emulation of LDRD and STRD instructions The decoding of these instructions got the register indexed and immediate indexed forms the wrong way around, causing incorrect emulation. Instructions like "LDRD Rx, [Rx]" were corrupting Rx because the base register writeback was being performed unconditionally, overwriting the value just loaded from memory. The fix is to only writeback the base register when that form of the instruction is used. Note, now that we reject probing writeback with PC the emulation code doesn't need the check rn!=15. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index a4dba1f..826abc1 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -560,8 +560,8 @@ static void __kprobes emulate_ldrd(struct kprobe *p, struct pt_regs *regs) [i_fn] "r" (i_fn) : "r0", "r1", "r2", "r3", "lr", "cc" ); - if (rn != 15) - regs->uregs[rn] = rnv; /* Save Rn in case of writeback. */ + if (is_writeback(insn)) + regs->uregs[rn] = rnv; } static void __kprobes emulate_strd(struct kprobe *p, struct pt_regs *regs) @@ -580,8 +580,8 @@ static void __kprobes emulate_strd(struct kprobe *p, struct pt_regs *regs) rnv_wb = insnslot_4arg_rflags(rnv, rmv, regs->uregs[rd], regs->uregs[rd+1], regs->ARM_cpsr, i_fn); - if (rn != 15) - regs->uregs[rn] = rnv_wb; /* Save Rn in case of writeback. */ + if (is_writeback(insn)) + regs->uregs[rn] = rnv_wb; } static void __kprobes emulate_ldr(struct kprobe *p, struct pt_regs *regs) @@ -1183,8 +1183,8 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) insn &= 0xfff00fff; insn |= 0x00002000; /* Rn = r0, Rd = r2 */ - if (insn & (1 << 22)) { - /* I bit */ + if (!(insn & (1 << 22))) { + /* Register index */ insn &= ~0xf; insn |= 1; /* Rm = r1 */ } -- cgit v1.1 From 81ff5720b9561b48e3dc640ca15799ba3919f5a6 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 12 Apr 2011 07:45:21 +0100 Subject: ARM: kprobes: Reject probing of LDRB instructions which load PC These instructions are specified as UNPREDICTABLE. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 826abc1..a37745f 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1416,6 +1416,10 @@ space_cccc_01xx(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* STRB : cccc 01xx x1x0 xxxx xxxx xxxx xxxx xxxx */ /* STRBT : cccc 01x0 x110 xxxx xxxx xxxx xxxx xxxx */ /* STRT : cccc 01x0 x010 xxxx xxxx xxxx xxxx xxxx */ + + if ((insn & 0x00500000) == 0x00500000 && is_r15(insn, 12)) + return INSN_REJECTED; /* LDRB into PC */ + return prep_emulate_ldr_str(insn, asi); } -- cgit v1.1 From 0e384ed164bdc2ad832270e81dbd14a17c143e78 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 12 Apr 2011 07:45:22 +0100 Subject: ARM: kprobes: Add emulation of RBIT instruction The v6T2 RBIT instruction was accidentally being emulated correctly, this patch adds correct decoding for the instruction. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index a37745f..b6bbc0b 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1320,9 +1320,10 @@ space_cccc_0110__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* REV : cccc 0110 1011 xxxx xxxx xxxx 0011 xxxx */ /* REV16 : cccc 0110 1011 xxxx xxxx xxxx 1011 xxxx */ + /* RBIT : cccc 0110 1111 xxxx xxxx xxxx 0011 xxxx */ /* REVSH : cccc 0110 1111 xxxx xxxx xxxx 1011 xxxx */ if ((insn & 0x0ff00070) == 0x06b00030 || - (insn & 0x0ff000f0) == 0x06f000b0) + (insn & 0x0ff00070) == 0x06f00030) return prep_emulate_rd12rm0(insn, asi); /* SADD16 : cccc 0110 0001 xxxx xxxx xxxx 0001 xxxx :GE */ -- cgit v1.1 From 780b5c1162286cc75ef2dcc0f14215d9f9f06230 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 12 Apr 2011 07:45:23 +0100 Subject: ARM: kprobes: Reject probing of undefined media instructions The instructions space for media instructions contains some undefined patterns. We need to reject probing of these because they may in future become defined and the kprobes code may then emulate them faultily. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 36 +++++++++++++++++++++++++++++++++++- 1 file changed, 35 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index b6bbc0b..f0ca7f4 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1326,52 +1326,86 @@ space_cccc_0110__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) (insn & 0x0ff00070) == 0x06f00030) return prep_emulate_rd12rm0(insn, asi); + /* ??? : cccc 0110 0000 xxxx xxxx xxxx xxx1 xxxx : */ /* SADD16 : cccc 0110 0001 xxxx xxxx xxxx 0001 xxxx :GE */ /* SADDSUBX : cccc 0110 0001 xxxx xxxx xxxx 0011 xxxx :GE */ /* SSUBADDX : cccc 0110 0001 xxxx xxxx xxxx 0101 xxxx :GE */ /* SSUB16 : cccc 0110 0001 xxxx xxxx xxxx 0111 xxxx :GE */ /* SADD8 : cccc 0110 0001 xxxx xxxx xxxx 1001 xxxx :GE */ + /* ??? : cccc 0110 0001 xxxx xxxx xxxx 1011 xxxx : */ + /* ??? : cccc 0110 0001 xxxx xxxx xxxx 1101 xxxx : */ /* SSUB8 : cccc 0110 0001 xxxx xxxx xxxx 1111 xxxx :GE */ /* QADD16 : cccc 0110 0010 xxxx xxxx xxxx 0001 xxxx : */ /* QADDSUBX : cccc 0110 0010 xxxx xxxx xxxx 0011 xxxx : */ /* QSUBADDX : cccc 0110 0010 xxxx xxxx xxxx 0101 xxxx : */ /* QSUB16 : cccc 0110 0010 xxxx xxxx xxxx 0111 xxxx : */ /* QADD8 : cccc 0110 0010 xxxx xxxx xxxx 1001 xxxx : */ + /* ??? : cccc 0110 0010 xxxx xxxx xxxx 1011 xxxx : */ + /* ??? : cccc 0110 0010 xxxx xxxx xxxx 1101 xxxx : */ /* QSUB8 : cccc 0110 0010 xxxx xxxx xxxx 1111 xxxx : */ /* SHADD16 : cccc 0110 0011 xxxx xxxx xxxx 0001 xxxx : */ /* SHADDSUBX : cccc 0110 0011 xxxx xxxx xxxx 0011 xxxx : */ /* SHSUBADDX : cccc 0110 0011 xxxx xxxx xxxx 0101 xxxx : */ /* SHSUB16 : cccc 0110 0011 xxxx xxxx xxxx 0111 xxxx : */ /* SHADD8 : cccc 0110 0011 xxxx xxxx xxxx 1001 xxxx : */ + /* ??? : cccc 0110 0011 xxxx xxxx xxxx 1011 xxxx : */ + /* ??? : cccc 0110 0011 xxxx xxxx xxxx 1101 xxxx : */ /* SHSUB8 : cccc 0110 0011 xxxx xxxx xxxx 1111 xxxx : */ + /* ??? : cccc 0110 0100 xxxx xxxx xxxx xxx1 xxxx : */ /* UADD16 : cccc 0110 0101 xxxx xxxx xxxx 0001 xxxx :GE */ /* UADDSUBX : cccc 0110 0101 xxxx xxxx xxxx 0011 xxxx :GE */ /* USUBADDX : cccc 0110 0101 xxxx xxxx xxxx 0101 xxxx :GE */ /* USUB16 : cccc 0110 0101 xxxx xxxx xxxx 0111 xxxx :GE */ /* UADD8 : cccc 0110 0101 xxxx xxxx xxxx 1001 xxxx :GE */ + /* ??? : cccc 0110 0101 xxxx xxxx xxxx 1011 xxxx : */ + /* ??? : cccc 0110 0101 xxxx xxxx xxxx 1101 xxxx : */ /* USUB8 : cccc 0110 0101 xxxx xxxx xxxx 1111 xxxx :GE */ /* UQADD16 : cccc 0110 0110 xxxx xxxx xxxx 0001 xxxx : */ /* UQADDSUBX : cccc 0110 0110 xxxx xxxx xxxx 0011 xxxx : */ /* UQSUBADDX : cccc 0110 0110 xxxx xxxx xxxx 0101 xxxx : */ /* UQSUB16 : cccc 0110 0110 xxxx xxxx xxxx 0111 xxxx : */ /* UQADD8 : cccc 0110 0110 xxxx xxxx xxxx 1001 xxxx : */ + /* ??? : cccc 0110 0110 xxxx xxxx xxxx 1011 xxxx : */ + /* ??? : cccc 0110 0110 xxxx xxxx xxxx 1101 xxxx : */ /* UQSUB8 : cccc 0110 0110 xxxx xxxx xxxx 1111 xxxx : */ /* UHADD16 : cccc 0110 0111 xxxx xxxx xxxx 0001 xxxx : */ /* UHADDSUBX : cccc 0110 0111 xxxx xxxx xxxx 0011 xxxx : */ /* UHSUBADDX : cccc 0110 0111 xxxx xxxx xxxx 0101 xxxx : */ /* UHSUB16 : cccc 0110 0111 xxxx xxxx xxxx 0111 xxxx : */ /* UHADD8 : cccc 0110 0111 xxxx xxxx xxxx 1001 xxxx : */ + /* ??? : cccc 0110 0111 xxxx xxxx xxxx 1011 xxxx : */ + /* ??? : cccc 0110 0111 xxxx xxxx xxxx 1101 xxxx : */ /* UHSUB8 : cccc 0110 0111 xxxx xxxx xxxx 1111 xxxx : */ + if ((insn & 0x0f800010) == 0x06000010) { + if ((insn & 0x00300000) == 0x00000000 || + (insn & 0x000000e0) == 0x000000a0 || + (insn & 0x000000e0) == 0x000000c0) + return INSN_REJECTED; /* Unallocated space */ + return prep_emulate_rd12rn16rm0_wflags(insn, asi); + } + /* PKHBT : cccc 0110 1000 xxxx xxxx xxxx x001 xxxx : */ /* PKHTB : cccc 0110 1000 xxxx xxxx xxxx x101 xxxx : */ + if ((insn & 0x0ff00030) == 0x06800010) + return prep_emulate_rd12rn16rm0_wflags(insn, asi); + /* SXTAB16 : cccc 0110 1000 xxxx xxxx xxxx 0111 xxxx : */ /* SXTB : cccc 0110 1010 xxxx xxxx xxxx 0111 xxxx : */ + /* ??? : cccc 0110 1001 xxxx xxxx xxxx 0111 xxxx : */ /* SXTAB : cccc 0110 1010 xxxx xxxx xxxx 0111 xxxx : */ /* SXTAH : cccc 0110 1011 xxxx xxxx xxxx 0111 xxxx : */ /* UXTAB16 : cccc 0110 1100 xxxx xxxx xxxx 0111 xxxx : */ + /* ??? : cccc 0110 1101 xxxx xxxx xxxx 0111 xxxx : */ /* UXTAB : cccc 0110 1110 xxxx xxxx xxxx 0111 xxxx : */ /* UXTAH : cccc 0110 1111 xxxx xxxx xxxx 0111 xxxx : */ - return prep_emulate_rd12rn16rm0_wflags(insn, asi); + if ((insn & 0x0f8000f0) == 0x06800070) { + if ((insn & 0x00300000) == 0x00100000) + return INSN_REJECTED; /* Unallocated space */ + return prep_emulate_rd12rn16rm0_wflags(insn, asi); + } + + /* Other instruction encodings aren't yet defined */ + return INSN_REJECTED; } static enum kprobe_insn __kprobes -- cgit v1.1 From 8dd7cfbed83c74b1fb991fae264944e041e22e62 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 12 Apr 2011 07:45:24 +0100 Subject: ARM: kprobes: Fix emulation of SXTB16, SXTB, SXTH, UXTB16, UXTB and UXTH instructions These sign extension instructions are encoded as extend-and-add instructions where the register to add is specified as r15. The decoding routines weren't checking for this and were using the incorrect emulation code, giving incorrect results. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index f0ca7f4..1e413a9 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1390,18 +1390,28 @@ space_cccc_0110__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) return prep_emulate_rd12rn16rm0_wflags(insn, asi); /* SXTAB16 : cccc 0110 1000 xxxx xxxx xxxx 0111 xxxx : */ - /* SXTB : cccc 0110 1010 xxxx xxxx xxxx 0111 xxxx : */ + /* SXTB16 : cccc 0110 1000 1111 xxxx xxxx 0111 xxxx : */ /* ??? : cccc 0110 1001 xxxx xxxx xxxx 0111 xxxx : */ /* SXTAB : cccc 0110 1010 xxxx xxxx xxxx 0111 xxxx : */ + /* SXTB : cccc 0110 1010 1111 xxxx xxxx 0111 xxxx : */ /* SXTAH : cccc 0110 1011 xxxx xxxx xxxx 0111 xxxx : */ + /* SXTH : cccc 0110 1011 1111 xxxx xxxx 0111 xxxx : */ /* UXTAB16 : cccc 0110 1100 xxxx xxxx xxxx 0111 xxxx : */ + /* UXTB16 : cccc 0110 1100 1111 xxxx xxxx 0111 xxxx : */ /* ??? : cccc 0110 1101 xxxx xxxx xxxx 0111 xxxx : */ /* UXTAB : cccc 0110 1110 xxxx xxxx xxxx 0111 xxxx : */ + /* UXTB : cccc 0110 1110 1111 xxxx xxxx 0111 xxxx : */ /* UXTAH : cccc 0110 1111 xxxx xxxx xxxx 0111 xxxx : */ + /* UXTH : cccc 0110 1111 1111 xxxx xxxx 0111 xxxx : */ if ((insn & 0x0f8000f0) == 0x06800070) { if ((insn & 0x00300000) == 0x00100000) return INSN_REJECTED; /* Unallocated space */ - return prep_emulate_rd12rn16rm0_wflags(insn, asi); + + if ((insn & 0x000f0000) == 0x000f0000) { + return prep_emulate_rd12rm0(insn, asi); + } else { + return prep_emulate_rd12rn16rm0_wflags(insn, asi); + } } /* Other instruction encodings aren't yet defined */ -- cgit v1.1 From 038c3839c917e3eea1150a1dc55607b9bde2d5ac Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 12 Apr 2011 07:45:25 +0100 Subject: ARM: kprobes: Fix emulation of SMUAD, SMUSD and SMMUL instructions The signed multiply instructions were being decoded incorrectly. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 22 +++++++++++++++------- 1 file changed, 15 insertions(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 1e413a9..068e5c8 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1436,18 +1436,26 @@ space_cccc_0111__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) return prep_emulate_rdhi16rdlo12rs8rm0_wflags(insn, asi); /* SMLAD : cccc 0111 0000 xxxx xxxx xxxx 00x1 xxxx :Q */ + /* SMUAD : cccc 0111 0000 xxxx 1111 xxxx 00x1 xxxx :Q */ /* SMLSD : cccc 0111 0000 xxxx xxxx xxxx 01x1 xxxx :Q */ + /* SMUSD : cccc 0111 0000 xxxx 1111 xxxx 01x1 xxxx : */ /* SMMLA : cccc 0111 0101 xxxx xxxx xxxx 00x1 xxxx : */ - /* SMMLS : cccc 0111 0101 xxxx xxxx xxxx 11x1 xxxx : */ + /* SMMUL : cccc 0111 0101 xxxx 1111 xxxx 00x1 xxxx : */ if ((insn & 0x0ff00090) == 0x07000010 || - (insn & 0x0ff000d0) == 0x07500010 || - (insn & 0x0ff000d0) == 0x075000d0) + (insn & 0x0ff000d0) == 0x07500010) { + + if ((insn & 0x0000f000) == 0x0000f000) { + return prep_emulate_rd16rs8rm0_wflags(insn, asi); + } else { + return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); + } + } + + /* SMMLS : cccc 0111 0101 xxxx xxxx xxxx 11x1 xxxx : */ + if ((insn & 0x0ff000d0) == 0x075000d0) return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); - /* SMUSD : cccc 0111 0000 xxxx xxxx xxxx 01x1 xxxx : */ - /* SMUAD : cccc 0111 0000 xxxx 1111 xxxx 00x1 xxxx :Q */ - /* SMMUL : cccc 0111 0101 xxxx 1111 xxxx 00x1 xxxx : */ - return prep_emulate_rd16rs8rm0_wflags(insn, asi); + return INSN_REJECTED; } static enum kprobe_insn __kprobes -- cgit v1.1 From c6e4ae32911c2a0ad02d47ee59ccba352a74c38e Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 12 Apr 2011 07:45:26 +0100 Subject: ARM: kprobes: Fix emulation of USAD8 instructions The USAD8 instruction wasn't being explicitly decoded leading to the incorrect emulation routine being called. It can be correctly decoded in the same way as the signed multiply instructions so we move the decoding there. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 068e5c8..8391dac 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1425,11 +1425,6 @@ space_cccc_0111__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) if ((insn & 0x0ff000f0) == 0x03f000f0) return INSN_REJECTED; - /* USADA8 : cccc 0111 1000 xxxx xxxx xxxx 0001 xxxx */ - /* USAD8 : cccc 0111 1000 xxxx 1111 xxxx 0001 xxxx */ - if ((insn & 0x0ff000f0) == 0x07800010) - return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); - /* SMLALD : cccc 0111 0100 xxxx xxxx xxxx 00x1 xxxx */ /* SMLSLD : cccc 0111 0100 xxxx xxxx xxxx 01x1 xxxx */ if ((insn & 0x0ff00090) == 0x07400010) @@ -1441,8 +1436,11 @@ space_cccc_0111__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* SMUSD : cccc 0111 0000 xxxx 1111 xxxx 01x1 xxxx : */ /* SMMLA : cccc 0111 0101 xxxx xxxx xxxx 00x1 xxxx : */ /* SMMUL : cccc 0111 0101 xxxx 1111 xxxx 00x1 xxxx : */ + /* USADA8 : cccc 0111 1000 xxxx xxxx xxxx 0001 xxxx : */ + /* USAD8 : cccc 0111 1000 xxxx 1111 xxxx 0001 xxxx : */ if ((insn & 0x0ff00090) == 0x07000010 || - (insn & 0x0ff000d0) == 0x07500010) { + (insn & 0x0ff000d0) == 0x07500010 || + (insn & 0x0ff000f0) == 0x07800010) { if ((insn & 0x0000f000) == 0x0000f000) { return prep_emulate_rd16rs8rm0_wflags(insn, asi); -- cgit v1.1 From fa1a03b429b3fd5f28e7fdd20ce99ca572bfd236 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Mon, 18 Apr 2011 08:53:54 +0100 Subject: ARM: kprobes: Reject probing of all coprocessor instructions The kernel doesn't currently support VFP or Neon code, and probing of code with CP15 operations is fraught with bad consequences. Therefore we don't need the ability to probe coprocessor instructions and the code to support this can be removed. The removed code also had at least two bugs: - MRC into R15 should set CPSR not trash PC - LDC and STC which use PC as base register needed the address offset by 8 Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 113 ++++----------------------------------- 1 file changed, 11 insertions(+), 102 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 8391dac..ff39707 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -520,17 +520,6 @@ static void __kprobes simulate_mov_ipsp(struct kprobe *p, struct pt_regs *regs) regs->uregs[12] = regs->uregs[13]; } -static void __kprobes emulate_ldcstc(struct kprobe *p, struct pt_regs *regs) -{ - insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; - kprobe_opcode_t insn = p->opcode; - int rn = (insn >> 16) & 0xf; - long rnv = regs->uregs[rn]; - - /* Save Rn in case of writeback. */ - regs->uregs[rn] = insnslot_1arg_rflags(rnv, regs->ARM_cpsr, i_fn); -} - static void __kprobes emulate_ldrd(struct kprobe *p, struct pt_regs *regs) { insn_2arg_fn_t *i_fn = (insn_2arg_fn_t *)&p->ainsn.insn[0]; @@ -635,31 +624,6 @@ static void __kprobes emulate_str(struct kprobe *p, struct pt_regs *regs) regs->uregs[rn] = rnv_wb; /* Save Rn in case of writeback. */ } -static void __kprobes emulate_mrrc(struct kprobe *p, struct pt_regs *regs) -{ - insn_llret_0arg_fn_t *i_fn = (insn_llret_0arg_fn_t *)&p->ainsn.insn[0]; - kprobe_opcode_t insn = p->opcode; - union reg_pair fnr; - int rd = (insn >> 12) & 0xf; - int rn = (insn >> 16) & 0xf; - - fnr.dr = insnslot_llret_0arg_rflags(regs->ARM_cpsr, i_fn); - regs->uregs[rn] = fnr.r0; - regs->uregs[rd] = fnr.r1; -} - -static void __kprobes emulate_mcrr(struct kprobe *p, struct pt_regs *regs) -{ - insn_2arg_fn_t *i_fn = (insn_2arg_fn_t *)&p->ainsn.insn[0]; - kprobe_opcode_t insn = p->opcode; - int rd = (insn >> 12) & 0xf; - int rn = (insn >> 16) & 0xf; - long rnv = regs->uregs[rn]; - long rdv = regs->uregs[rd]; - - insnslot_2arg_rflags(rnv, rdv, regs->ARM_cpsr, i_fn); -} - static void __kprobes emulate_sat(struct kprobe *p, struct pt_regs *regs) { insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; @@ -693,24 +657,6 @@ static void __kprobes emulate_none(struct kprobe *p, struct pt_regs *regs) insnslot_0arg_rflags(regs->ARM_cpsr, i_fn); } -static void __kprobes emulate_rd12(struct kprobe *p, struct pt_regs *regs) -{ - insn_0arg_fn_t *i_fn = (insn_0arg_fn_t *)&p->ainsn.insn[0]; - kprobe_opcode_t insn = p->opcode; - int rd = (insn >> 12) & 0xf; - - regs->uregs[rd] = insnslot_0arg_rflags(regs->ARM_cpsr, i_fn); -} - -static void __kprobes emulate_ird12(struct kprobe *p, struct pt_regs *regs) -{ - insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; - kprobe_opcode_t insn = p->opcode; - int ird = (insn >> 12) & 0xf; - - insnslot_1arg_rflags(regs->uregs[ird], regs->ARM_cpsr, i_fn); -} - static void __kprobes emulate_rn16(struct kprobe *p, struct pt_regs *regs) { insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; @@ -1010,40 +956,22 @@ space_1111(kprobe_opcode_t insn, struct arch_specific_insn *asi) } /* SETEND : 1111 0001 0000 0001 xxxx xxxx 0000 xxxx */ - /* CDP2 : 1111 1110 xxxx xxxx xxxx xxxx xxx0 xxxx */ - if ((insn & 0xffff00f0) == 0xf1010000 || - (insn & 0xff000010) == 0xfe000000) { + if ((insn & 0xffff00f0) == 0xf1010000) { asi->insn[0] = insn; asi->insn_handler = emulate_none; return INSN_GOOD; } + /* Coprocessor instructions... */ /* MCRR2 : 1111 1100 0100 xxxx xxxx xxxx xxxx xxxx : (Rd != Rn) */ /* MRRC2 : 1111 1100 0101 xxxx xxxx xxxx xxxx xxxx : (Rd != Rn) */ - if ((insn & 0xffe00000) == 0xfc400000) { - insn &= 0xfff00fff; /* Rn = r0 */ - insn |= 0x00001000; /* Rd = r1 */ - asi->insn[0] = insn; - asi->insn_handler = - (insn & (1 << 20)) ? emulate_mrrc : emulate_mcrr; - return INSN_GOOD; - } + /* LDC2 : 1111 110x xxx1 xxxx xxxx xxxx xxxx xxxx */ + /* STC2 : 1111 110x xxx0 xxxx xxxx xxxx xxxx xxxx */ + /* CDP2 : 1111 1110 xxxx xxxx xxxx xxxx xxx0 xxxx */ + /* MCR2 : 1111 1110 xxx0 xxxx xxxx xxxx xxx1 xxxx */ + /* MRC2 : 1111 1110 xxx1 xxxx xxxx xxxx xxx1 xxxx */ - /* LDC2 : 1111 110x xxx1 xxxx xxxx xxxx xxxx xxxx */ - /* STC2 : 1111 110x xxx0 xxxx xxxx xxxx xxxx xxxx */ - if ((insn & 0xfe000000) == 0xfc000000) { - insn &= 0xfff0ffff; /* Rn = r0 */ - asi->insn[0] = insn; - asi->insn_handler = emulate_ldcstc; - return INSN_GOOD; - } - - /* MCR2 : 1111 1110 xxx0 xxxx xxxx xxxx xxx1 xxxx */ - /* MRC2 : 1111 1110 xxx1 xxxx xxxx xxxx xxx1 xxxx */ - insn &= 0xffff0fff; /* Rd = r0 */ - asi->insn[0] = insn; - asi->insn_handler = (insn & (1 << 20)) ? emulate_rd12 : emulate_ird12; - return INSN_GOOD; + return INSN_REJECTED; } static enum kprobe_insn __kprobes @@ -1504,14 +1432,7 @@ space_cccc_1100_010x(kprobe_opcode_t insn, struct arch_specific_insn *asi) { /* MCRR : cccc 1100 0100 xxxx xxxx xxxx xxxx xxxx : (Rd!=Rn) */ /* MRRC : cccc 1100 0101 xxxx xxxx xxxx xxxx xxxx : (Rd!=Rn) */ - if (is_r15(insn, 16) || is_r15(insn, 12)) - return INSN_REJECTED; /* Rn or Rd is PC */ - - insn &= 0xfff00fff; - insn |= 0x00001000; /* Rn = r0, Rd = r1 */ - asi->insn[0] = insn; - asi->insn_handler = (insn & (1 << 20)) ? emulate_mrrc : emulate_mcrr; - return INSN_GOOD; + return INSN_REJECTED; } static enum kprobe_insn __kprobes @@ -1519,10 +1440,7 @@ space_cccc_110x(kprobe_opcode_t insn, struct arch_specific_insn *asi) { /* LDC : cccc 110x xxx1 xxxx xxxx xxxx xxxx xxxx */ /* STC : cccc 110x xxx0 xxxx xxxx xxxx xxxx xxxx */ - insn &= 0xfff0ffff; /* Rn = r0 */ - asi->insn[0] = insn; - asi->insn_handler = emulate_ldcstc; - return INSN_GOOD; + return INSN_REJECTED; } static enum kprobe_insn __kprobes @@ -1535,18 +1453,9 @@ space_cccc_111x(kprobe_opcode_t insn, struct arch_specific_insn *asi) return INSN_REJECTED; /* CDP : cccc 1110 xxxx xxxx xxxx xxxx xxx0 xxxx */ - if ((insn & 0x0f000010) == 0x0e000000) { - asi->insn[0] = insn; - asi->insn_handler = emulate_none; - return INSN_GOOD; - } - /* MCR : cccc 1110 xxx0 xxxx xxxx xxxx xxx1 xxxx */ /* MRC : cccc 1110 xxx1 xxxx xxxx xxxx xxx1 xxxx */ - insn &= 0xffff0fff; /* Rd = r0 */ - asi->insn[0] = insn; - asi->insn_handler = (insn & (1 << 20)) ? emulate_rd12 : emulate_ird12; - return INSN_GOOD; + return INSN_REJECTED; } static unsigned long __kprobes __check_eq(unsigned long cpsr) -- cgit v1.1 From ac211c6994fb5f1f282745054c00d29e53639cb1 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Mon, 18 Apr 2011 08:53:55 +0100 Subject: ARM: kprobes: Consolidate stub decoding functions Following the change to remove support for coprocessor instructions we are left with three stub functions which can be consolidated. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 40 +++++++++------------------------------- 1 file changed, 9 insertions(+), 31 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index ff39707..fb81897 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1428,33 +1428,19 @@ space_cccc_101x(kprobe_opcode_t insn, struct arch_specific_insn *asi) } static enum kprobe_insn __kprobes -space_cccc_1100_010x(kprobe_opcode_t insn, struct arch_specific_insn *asi) +space_cccc_11xx(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + /* Coprocessor instructions... */ /* MCRR : cccc 1100 0100 xxxx xxxx xxxx xxxx xxxx : (Rd!=Rn) */ /* MRRC : cccc 1100 0101 xxxx xxxx xxxx xxxx xxxx : (Rd!=Rn) */ - return INSN_REJECTED; -} + /* LDC : cccc 110x xxx1 xxxx xxxx xxxx xxxx xxxx */ + /* STC : cccc 110x xxx0 xxxx xxxx xxxx xxxx xxxx */ + /* CDP : cccc 1110 xxxx xxxx xxxx xxxx xxx0 xxxx */ + /* MCR : cccc 1110 xxx0 xxxx xxxx xxxx xxx1 xxxx */ + /* MRC : cccc 1110 xxx1 xxxx xxxx xxxx xxx1 xxxx */ -static enum kprobe_insn __kprobes -space_cccc_110x(kprobe_opcode_t insn, struct arch_specific_insn *asi) -{ - /* LDC : cccc 110x xxx1 xxxx xxxx xxxx xxxx xxxx */ - /* STC : cccc 110x xxx0 xxxx xxxx xxxx xxxx xxxx */ - return INSN_REJECTED; -} + /* SVC : cccc 1111 xxxx xxxx xxxx xxxx xxxx xxxx */ -static enum kprobe_insn __kprobes -space_cccc_111x(kprobe_opcode_t insn, struct arch_specific_insn *asi) -{ - /* BKPT : 1110 0001 0010 xxxx xxxx xxxx 0111 xxxx */ - /* SWI : cccc 1111 xxxx xxxx xxxx xxxx xxxx xxxx */ - if ((insn & 0xfff000f0) == 0xe1200070 || - (insn & 0x0f000000) == 0x0f000000) - return INSN_REJECTED; - - /* CDP : cccc 1110 xxxx xxxx xxxx xxxx xxx0 xxxx */ - /* MCR : cccc 1110 xxx0 xxxx xxxx xxxx xxx1 xxxx */ - /* MRC : cccc 1110 xxx1 xxxx xxxx xxxx xxx1 xxxx */ return INSN_REJECTED; } @@ -1598,17 +1584,9 @@ arm_kprobe_decode_insn(kprobe_opcode_t insn, struct arch_specific_insn *asi) return space_cccc_101x(insn, asi); - } else if ((insn & 0x0fe00000) == 0x0c400000) { - - return space_cccc_1100_010x(insn, asi); - - } else if ((insn & 0x0e000000) == 0x0c000000) { - - return space_cccc_110x(insn, asi); - } - return space_cccc_111x(insn, asi); + return space_cccc_11xx(insn, asi); } void __init arm_kprobe_decode_init(void) -- cgit v1.1 From f0aeb8bff0fe9de50e1e4093ef86ff8f17a9b1b0 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Mon, 18 Apr 2011 08:53:56 +0100 Subject: ARM: kprobes: Reject probing of SETEND instructions The emulation of SETEND was broken as it changed the endianess for the running kprobes handling code. Rather than adding a new simulation routine to fix this we'll just reject probing of SETEND as these should be very rare in the kernel. Note, the function emulate_none is now unused but it is left in the source code as future patches will use it. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 5 ----- 1 file changed, 5 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index fb81897..b81fbfb 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -956,11 +956,6 @@ space_1111(kprobe_opcode_t insn, struct arch_specific_insn *asi) } /* SETEND : 1111 0001 0000 0001 xxxx xxxx 0000 xxxx */ - if ((insn & 0xffff00f0) == 0xf1010000) { - asi->insn[0] = insn; - asi->insn_handler = emulate_none; - return INSN_GOOD; - } /* Coprocessor instructions... */ /* MCRR2 : 1111 1100 0100 xxxx xxxx xxxx xxxx xxxx : (Rd != Rn) */ -- cgit v1.1 From 41713d1396312a027ec02abc4767041627c178ef Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Mon, 18 Apr 2011 08:53:57 +0100 Subject: ARM: kprobes: Fix emulation of PLD instructions The PLD instructions wasn't being decoded correctly and the emulation code wasn't adjusting PC correctly. As the PLD instruction is only a performance hint we emulate it as a simple nop, and we can broaden the instruction decoding to take into account newer PLI and PLDW instructions. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 21 ++++++++------------- 1 file changed, 8 insertions(+), 13 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index b81fbfb..bddf8d0 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -657,14 +657,8 @@ static void __kprobes emulate_none(struct kprobe *p, struct pt_regs *regs) insnslot_0arg_rflags(regs->ARM_cpsr, i_fn); } -static void __kprobes emulate_rn16(struct kprobe *p, struct pt_regs *regs) +static void __kprobes emulate_nop(struct kprobe *p, struct pt_regs *regs) { - insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; - kprobe_opcode_t insn = p->opcode; - int rn = (insn >> 16) & 0xf; - long rnv = regs->uregs[rn]; - - insnslot_1arg_rflags(rnv, regs->ARM_cpsr, i_fn); } static void __kprobes emulate_rd12rm0(struct kprobe *p, struct pt_regs *regs) @@ -941,12 +935,13 @@ space_1111(kprobe_opcode_t insn, struct arch_specific_insn *asi) (insn & 0xfe5f0f00) == 0xf84d0500) return INSN_REJECTED; - /* PLD : 1111 01x1 x101 xxxx xxxx xxxx xxxx xxxx : */ - if ((insn & 0xfd700000) == 0xf4500000) { - insn &= 0xfff0ffff; /* Rn = r0 */ - asi->insn[0] = insn; - asi->insn_handler = emulate_rn16; - return INSN_GOOD; + /* memory hint : 1111 0100 x001 xxxx xxxx xxxx xxxx xxxx : */ + /* PLDI : 1111 0100 x101 xxxx xxxx xxxx xxxx xxxx : */ + /* PLDW : 1111 0101 x001 xxxx xxxx xxxx xxxx xxxx : */ + /* PLD : 1111 0101 x101 xxxx xxxx xxxx xxxx xxxx : */ + if ((insn & 0xfe300000) == 0xf4100000) { + asi->insn_handler = emulate_nop; + return INSN_GOOD_NO_SLOT; } /* BLX(1) : 1111 101x xxxx xxxx xxxx xxxx xxxx xxxx : */ -- cgit v1.1 From 72c2bab2be734d63dee4342e67b1754c54fded70 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Mon, 18 Apr 2011 08:53:58 +0100 Subject: ARM: kprobes: Remove redundant code in space_1111 The tests to explicitly reject probing CPS, RFE and SRS instructions are redundant as the default case is now to reject undecoded patterns. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index bddf8d0..dc315c1 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -927,14 +927,6 @@ prep_emulate_rdhi16rdlo12rs8rm0_wflags(kprobe_opcode_t insn, static enum kprobe_insn __kprobes space_1111(kprobe_opcode_t insn, struct arch_specific_insn *asi) { - /* CPS mmod == 1 : 1111 0001 0000 xx10 xxxx xxxx xx0x xxxx */ - /* RFE : 1111 100x x0x1 xxxx xxxx 1010 xxxx xxxx */ - /* SRS : 1111 100x x1x0 1101 xxxx 0101 xxxx xxxx */ - if ((insn & 0xfff30020) == 0xf1020000 || - (insn & 0xfe500f00) == 0xf8100a00 || - (insn & 0xfe5f0f00) == 0xf84d0500) - return INSN_REJECTED; - /* memory hint : 1111 0100 x001 xxxx xxxx xxxx xxxx xxxx : */ /* PLDI : 1111 0100 x101 xxxx xxxx xxxx xxxx xxxx : */ /* PLDW : 1111 0101 x001 xxxx xxxx xxxx xxxx xxxx : */ @@ -950,7 +942,11 @@ space_1111(kprobe_opcode_t insn, struct arch_specific_insn *asi) return INSN_GOOD_NO_SLOT; } - /* SETEND : 1111 0001 0000 0001 xxxx xxxx 0000 xxxx */ + /* CPS : 1111 0001 0000 xxx0 xxxx xxxx xx0x xxxx */ + /* SETEND: 1111 0001 0000 0001 xxxx xxxx 0000 xxxx */ + + /* SRS : 1111 100x x1x0 xxxx xxxx xxxx xxxx xxxx */ + /* RFE : 1111 100x x0x1 xxxx xxxx xxxx xxxx xxxx */ /* Coprocessor instructions... */ /* MCRR2 : 1111 1100 0100 xxxx xxxx xxxx xxxx xxxx : (Rd != Rn) */ -- cgit v1.1 From f704a6e25bd07e1381af81f19fb1d123975807b1 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 19 Apr 2011 10:52:16 +0100 Subject: ARM: kprobes: Reject probing of undefined data processing instructions The instruction decoding in space_cccc_000x needs to reject probing of instructions with undefined patterns as they may in future become defined and then emulated faultily - as has already happened with the SMC instruction. This fix is achieved by testing for the instruction patterns we want to probe and making the the default fall-through paths reject probes. This also allows us to remove some explicit tests for instructions that we wish to reject, as that is now the default action. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 31 +++++++++++++++++-------------- 1 file changed, 17 insertions(+), 14 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index dc315c1..4715537 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -966,14 +966,6 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* cccc 0001 0xx0 xxxx xxxx xxxx xxxx xxx0 xxxx */ if ((insn & 0x0f900010) == 0x01000000) { - /* BXJ : cccc 0001 0010 xxxx xxxx xxxx 0010 xxxx */ - /* MSR : cccc 0001 0x10 xxxx xxxx xxxx 0000 xxxx */ - /* MRS spsr : cccc 0001 0100 xxxx xxxx xxxx 0000 xxxx */ - if ((insn & 0x0ff000f0) == 0x01200020 || - (insn & 0x0fb000f0) == 0x01200000 || - (insn & 0x0ff000f0) == 0x01400000) - return INSN_REJECTED; - /* MRS cpsr : cccc 0001 0000 xxxx xxxx xxxx 0000 xxxx */ if ((insn & 0x0ff000f0) == 0x01000000) { if (is_r15(insn, 12)) @@ -994,17 +986,21 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* SMLAxy : cccc 0001 0000 xxxx xxxx xxxx 1xx0 xxxx : Q */ /* SMLAWy : cccc 0001 0010 xxxx xxxx xxxx 1x00 xxxx : Q */ - return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); + if ((insn & 0x0ff00090) == 0x01000080 || + (insn & 0x0ff000b0) == 0x01200080) + return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); + + /* BXJ : cccc 0001 0010 xxxx xxxx xxxx 0010 xxxx */ + /* MSR : cccc 0001 0x10 xxxx xxxx xxxx 0000 xxxx */ + /* MRS spsr : cccc 0001 0100 xxxx xxxx xxxx 0000 xxxx */ + /* Other instruction encodings aren't yet defined */ + return INSN_REJECTED; } /* cccc 0001 0xx0 xxxx xxxx xxxx xxxx 0xx1 xxxx */ else if ((insn & 0x0f900090) == 0x01000010) { - /* BKPT : 1110 0001 0010 xxxx xxxx xxxx 0111 xxxx */ - if ((insn & 0xfff000f0) == 0xe1200070) - return INSN_REJECTED; - /* BLX(2) : cccc 0001 0010 xxxx xxxx xxxx 0011 xxxx */ /* BX : cccc 0001 0010 xxxx xxxx xxxx 0001 xxxx */ if ((insn & 0x0ff000d0) == 0x01200010) { @@ -1022,7 +1018,14 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* QSUB : cccc 0001 0010 xxxx xxxx xxxx 0101 xxxx :Q */ /* QDADD : cccc 0001 0100 xxxx xxxx xxxx 0101 xxxx :Q */ /* QDSUB : cccc 0001 0110 xxxx xxxx xxxx 0101 xxxx :Q */ - return prep_emulate_rd12rn16rm0_wflags(insn, asi); + if ((insn & 0x0f9000f0) == 0x01000050) + return prep_emulate_rd12rn16rm0_wflags(insn, asi); + + /* BKPT : 1110 0001 0010 xxxx xxxx xxxx 0111 xxxx */ + /* SMC : cccc 0001 0110 xxxx xxxx xxxx 0111 xxxx */ + + /* Other instruction encodings aren't yet defined */ + return INSN_REJECTED; } /* cccc 0000 xxxx xxxx xxxx xxxx xxxx 1001 xxxx */ -- cgit v1.1 From c9836777d5f22e64eefc759d682511560b6d6d78 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 19 Apr 2011 10:52:17 +0100 Subject: ARM: kprobes: Add emulation of MOVW and MOVT instructions The MOVW and MOVT instructions account for approximately 7% of all instructions in a ARMv7 kernel as GCC uses them instead of a literal pool. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 30 ++++++++++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 4715537..a063f15 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -661,6 +661,17 @@ static void __kprobes emulate_nop(struct kprobe *p, struct pt_regs *regs) { } +static void __kprobes +emulate_rd12_modify(struct kprobe *p, struct pt_regs *regs) +{ + insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; + kprobe_opcode_t insn = p->opcode; + int rd = (insn >> 12) & 0xf; + long rdv = regs->uregs[rd]; + + regs->uregs[rd] = insnslot_1arg_rflags(rdv, regs->ARM_cpsr, i_fn); +} + static void __kprobes emulate_rd12rm0(struct kprobe *p, struct pt_regs *regs) { insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; @@ -847,6 +858,18 @@ prep_emulate_ldr_str(kprobe_opcode_t insn, struct arch_specific_insn *asi) } static enum kprobe_insn __kprobes +prep_emulate_rd12_modify(kprobe_opcode_t insn, struct arch_specific_insn *asi) +{ + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ + + insn &= 0xffff0fff; /* Rd = r0 */ + asi->insn[0] = insn; + asi->insn_handler = emulate_rd12_modify; + return INSN_GOOD; +} + +static enum kprobe_insn __kprobes prep_emulate_rd12rm0(kprobe_opcode_t insn, struct arch_specific_insn *asi) { if (is_r15(insn, 12)) @@ -1170,14 +1193,17 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) static enum kprobe_insn __kprobes space_cccc_001x(kprobe_opcode_t insn, struct arch_specific_insn *asi) { + /* MOVW : cccc 0011 0000 xxxx xxxx xxxx xxxx xxxx */ + /* MOVT : cccc 0011 0100 xxxx xxxx xxxx xxxx xxxx */ + if ((insn & 0x0fb00000) == 0x03000000) + return prep_emulate_rd12_modify(insn, asi); + /* * MSR : cccc 0011 0x10 xxxx xxxx xxxx xxxx xxxx - * Undef : cccc 0011 0100 xxxx xxxx xxxx xxxx xxxx * ALU op with S bit and Rd == 15 : * cccc 001x xxx1 xxxx 1111 xxxx xxxx xxxx */ if ((insn & 0x0fb00000) == 0x03200000 || /* MSR */ - (insn & 0x0ff00000) == 0x03400000 || /* Undef */ (insn & 0x0e10f000) == 0x0210f000) /* ALU s-bit, R15 */ return INSN_REJECTED; -- cgit v1.1 From 20e8155e24ba5c04558780d0a8da13c39f98c002 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 19 Apr 2011 10:52:18 +0100 Subject: ARM: kprobes: Add emulation of SBFX, UBFX, BFI and BFC instructions These bit field manipulation instructions occur several thousand times in an ARMv7 kernel. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 42 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 42 insertions(+) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index a063f15..6bed8b0 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -672,6 +672,19 @@ emulate_rd12_modify(struct kprobe *p, struct pt_regs *regs) regs->uregs[rd] = insnslot_1arg_rflags(rdv, regs->ARM_cpsr, i_fn); } +static void __kprobes +emulate_rd12rn0_modify(struct kprobe *p, struct pt_regs *regs) +{ + insn_2arg_fn_t *i_fn = (insn_2arg_fn_t *)&p->ainsn.insn[0]; + kprobe_opcode_t insn = p->opcode; + int rd = (insn >> 12) & 0xf; + int rn = insn & 0xf; + long rdv = regs->uregs[rd]; + long rnv = regs->uregs[rn]; + + regs->uregs[rd] = insnslot_2arg_rflags(rdv, rnv, regs->ARM_cpsr, i_fn); +} + static void __kprobes emulate_rd12rm0(struct kprobe *p, struct pt_regs *regs) { insn_1arg_fn_t *i_fn = (insn_1arg_fn_t *)&p->ainsn.insn[0]; @@ -870,6 +883,20 @@ prep_emulate_rd12_modify(kprobe_opcode_t insn, struct arch_specific_insn *asi) } static enum kprobe_insn __kprobes +prep_emulate_rd12rn0_modify(kprobe_opcode_t insn, + struct arch_specific_insn *asi) +{ + if (is_r15(insn, 12)) + return INSN_REJECTED; /* Rd is PC */ + + insn &= 0xffff0ff0; /* Rd = r0 */ + insn |= 0x00000001; /* Rn = r1 */ + asi->insn[0] = insn; + asi->insn_handler = emulate_rd12rn0_modify; + return INSN_GOOD; +} + +static enum kprobe_insn __kprobes prep_emulate_rd12rm0(kprobe_opcode_t insn, struct arch_specific_insn *asi) { if (is_r15(insn, 12)) @@ -1396,6 +1423,21 @@ space_cccc_0111__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) if ((insn & 0x0ff000d0) == 0x075000d0) return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); + /* SBFX : cccc 0111 101x xxxx xxxx xxxx x101 xxxx : */ + /* UBFX : cccc 0111 111x xxxx xxxx xxxx x101 xxxx : */ + if ((insn & 0x0fa00070) == 0x07a00050) + return prep_emulate_rd12rm0(insn, asi); + + /* BFI : cccc 0111 110x xxxx xxxx xxxx x001 xxxx : */ + /* BFC : cccc 0111 110x xxxx xxxx xxxx x001 1111 : */ + if ((insn & 0x0fe00070) == 0x07c00010) { + + if ((insn & 0x0000000f) == 0x0000000f) + return prep_emulate_rd12_modify(insn, asi); + else + return prep_emulate_rd12rn0_modify(insn, asi); + } + return INSN_REJECTED; } -- cgit v1.1 From 94254930786216106100e9a28c9a244df58be61f Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 19 Apr 2011 10:52:19 +0100 Subject: ARM: kprobes: Add emulation of hint instructions like NOP and WFI Being able to probe NOP instructions is useful for hard-coding probeable locations and is used by the kprobes test code. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 6bed8b0..03b85ad 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -1225,6 +1225,30 @@ space_cccc_001x(kprobe_opcode_t insn, struct arch_specific_insn *asi) if ((insn & 0x0fb00000) == 0x03000000) return prep_emulate_rd12_modify(insn, asi); + /* hints : cccc 0011 0010 0000 xxxx xxxx xxxx xxxx */ + if ((insn & 0x0fff0000) == 0x03200000) { + unsigned op2 = insn & 0x000000ff; + if (op2 == 0x01 || op2 == 0x04) { + /* YIELD : cccc 0011 0010 0000 xxxx xxxx 0000 0001 */ + /* SEV : cccc 0011 0010 0000 xxxx xxxx 0000 0100 */ + asi->insn[0] = insn; + asi->insn_handler = emulate_none; + return INSN_GOOD; + } else if (op2 <= 0x03) { + /* NOP : cccc 0011 0010 0000 xxxx xxxx 0000 0000 */ + /* WFE : cccc 0011 0010 0000 xxxx xxxx 0000 0010 */ + /* WFI : cccc 0011 0010 0000 xxxx xxxx 0000 0011 */ + /* + * We make WFE and WFI true NOPs to avoid stalls due + * to missing events whilst processing the probe. + */ + asi->insn_handler = emulate_nop; + return INSN_GOOD_NO_SLOT; + } + /* For DBG and unallocated hints it's safest to reject them */ + return INSN_REJECTED; + } + /* * MSR : cccc 0011 0x10 xxxx xxxx xxxx xxxx xxxx * ALU op with S bit and Rd == 15 : -- cgit v1.1 From cdc253611582f08036ccb6e10efe95b8e760a7e2 Mon Sep 17 00:00:00 2001 From: Jon Medhurst Date: Tue, 19 Apr 2011 10:52:20 +0100 Subject: ARM: kprobes: Tidy-up kprobes-decode.c - Remove coding standard violations reported by checkpatch.pl - Delete comment about handling of conditional branches which is no longer true. - Delete comment at end of file which lists all ARM instructions. This duplicates data available in the ARM ARM and seems like an unnecessary maintenance burden to keep this up to date and accurate. Signed-off-by: Jon Medhurst Signed-off-by: Nicolas Pitre --- arch/arm/kernel/kprobes-decode.c | 131 +++++++-------------------------------- 1 file changed, 23 insertions(+), 108 deletions(-) (limited to 'arch') diff --git a/arch/arm/kernel/kprobes-decode.c b/arch/arm/kernel/kprobes-decode.c index 03b85ad..15eeff6 100644 --- a/arch/arm/kernel/kprobes-decode.c +++ b/arch/arm/kernel/kprobes-decode.c @@ -34,9 +34,6 @@ * * *) If the PC is written to by the instruction, the * instruction must be fully simulated in software. - * If it is a conditional instruction, the handler - * will use insn[0] to copy its condition code to - * set r0 to 1 and insn[1] to "mov pc, lr" to return. * * *) Otherwise, a modified form of the instruction is * directly executed. Its handler calls the @@ -1026,7 +1023,8 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* SMLALxy : cccc 0001 0100 xxxx xxxx xxxx 1xx0 xxxx */ if ((insn & 0x0ff00090) == 0x01400080) - return prep_emulate_rdhi16rdlo12rs8rm0_wflags(insn, asi); + return prep_emulate_rdhi16rdlo12rs8rm0_wflags(insn, + asi); /* SMULWy : cccc 0001 0010 xxxx xxxx xxxx 1x10 xxxx */ /* SMULxy : cccc 0001 0110 xxxx xxxx xxxx 1xx0 xxxx */ @@ -1097,15 +1095,15 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* SMULLS : cccc 0000 1101 xxxx xxxx xxxx 1001 xxxx :cc */ /* SMLAL : cccc 0000 1110 xxxx xxxx xxxx 1001 xxxx : */ /* SMLALS : cccc 0000 1111 xxxx xxxx xxxx 1001 xxxx :cc */ - if ((insn & 0x00d00000) == 0x00500000) { + if ((insn & 0x00d00000) == 0x00500000) return INSN_REJECTED; - } else if ((insn & 0x00e00000) == 0x00000000) { - return prep_emulate_rd16rs8rm0_wflags(insn, asi); - } else if ((insn & 0x00a00000) == 0x00200000) { - return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); - } else { - return prep_emulate_rdhi16rdlo12rs8rm0_wflags(insn, asi); - } + else if ((insn & 0x00e00000) == 0x00000000) + return prep_emulate_rd16rs8rm0_wflags(insn, asi); + else if ((insn & 0x00a00000) == 0x00200000) + return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); + else + return prep_emulate_rdhi16rdlo12rs8rm0_wflags(insn, + asi); } /* cccc 000x xxxx xxxx xxxx xxxx xxxx 1xx1 xxxx */ @@ -1171,7 +1169,7 @@ space_cccc_000x(kprobe_opcode_t insn, struct arch_specific_insn *asi) /* * ALU op with S bit and Rd == 15 : - * cccc 000x xxx1 xxxx 1111 xxxx xxxx xxxx + * cccc 000x xxx1 xxxx 1111 xxxx xxxx xxxx */ if ((insn & 0x0e10f000) == 0x0010f000) return INSN_REJECTED; @@ -1401,11 +1399,10 @@ space_cccc_0110__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) if ((insn & 0x00300000) == 0x00100000) return INSN_REJECTED; /* Unallocated space */ - if ((insn & 0x000f0000) == 0x000f0000) { + if ((insn & 0x000f0000) == 0x000f0000) return prep_emulate_rd12rm0(insn, asi); - } else { + else return prep_emulate_rd12rn16rm0_wflags(insn, asi); - } } /* Other instruction encodings aren't yet defined */ @@ -1436,11 +1433,10 @@ space_cccc_0111__1(kprobe_opcode_t insn, struct arch_specific_insn *asi) (insn & 0x0ff000d0) == 0x07500010 || (insn & 0x0ff000f0) == 0x07800010) { - if ((insn & 0x0000f000) == 0x0000f000) { + if ((insn & 0x0000f000) == 0x0000f000) return prep_emulate_rd16rs8rm0_wflags(insn, asi); - } else { + else return prep_emulate_rd16rn12rs8rm0_wflags(insn, asi); - } } /* SMMLS : cccc 0111 0101 xxxx xxxx xxxx 11x1 xxxx : */ @@ -1633,40 +1629,38 @@ arm_kprobe_decode_insn(kprobe_opcode_t insn, struct arch_specific_insn *asi) asi->insn_check_cc = condition_checks[insn>>28]; asi->insn[1] = KPROBE_RETURN_INSTRUCTION; - if ((insn & 0xf0000000) == 0xf0000000) { + if ((insn & 0xf0000000) == 0xf0000000) return space_1111(insn, asi); - } else if ((insn & 0x0e000000) == 0x00000000) { + else if ((insn & 0x0e000000) == 0x00000000) return space_cccc_000x(insn, asi); - } else if ((insn & 0x0e000000) == 0x02000000) { + else if ((insn & 0x0e000000) == 0x02000000) return space_cccc_001x(insn, asi); - } else if ((insn & 0x0f000010) == 0x06000010) { + else if ((insn & 0x0f000010) == 0x06000010) return space_cccc_0110__1(insn, asi); - } else if ((insn & 0x0f000010) == 0x07000010) { + else if ((insn & 0x0f000010) == 0x07000010) return space_cccc_0111__1(insn, asi); - } else if ((insn & 0x0c000000) == 0x04000000) { + else if ((insn & 0x0c000000) == 0x04000000) return space_cccc_01xx(insn, asi); - } else if ((insn & 0x0e000000) == 0x08000000) { + else if ((insn & 0x0e000000) == 0x08000000) return space_cccc_100x(insn, asi); - } else if ((insn & 0x0e000000) == 0x0a000000) { + else if ((insn & 0x0e000000) == 0x0a000000) return space_cccc_101x(insn, asi); - } - return space_cccc_11xx(insn, asi); } @@ -1674,82 +1668,3 @@ void __init arm_kprobe_decode_init(void) { find_str_pc_offset(); } - - -/* - * All ARM instructions listed below. - * - * Instructions and their general purpose registers are given. - * If a particular register may not use R15, it is prefixed with a "!". - * If marked with a "*" means the value returned by reading R15 - * is implementation defined. - * - * ADC/ADD/AND/BIC/CMN/CMP/EOR/MOV/MVN/ORR/RSB/RSC/SBC/SUB/TEQ - * TST: Rd, Rn, Rm, !Rs - * BX: Rm - * BLX(2): !Rm - * BX: Rm (R15 legal, but discouraged) - * BXJ: !Rm, - * CLZ: !Rd, !Rm - * CPY: Rd, Rm - * LDC/2,STC/2 immediate offset & unindex: Rn - * LDC/2,STC/2 immediate pre/post-indexed: !Rn - * LDM(1/3): !Rn, register_list - * LDM(2): !Rn, !register_list - * LDR,STR,PLD immediate offset: Rd, Rn - * LDR,STR,PLD register offset: Rd, Rn, !Rm - * LDR,STR,PLD scaled register offset: Rd, !Rn, !Rm - * LDR,STR immediate pre/post-indexed: Rd, !Rn - * LDR,STR register pre/post-indexed: Rd, !Rn, !Rm - * LDR,STR scaled register pre/post-indexed: Rd, !Rn, !Rm - * LDRB,STRB immediate offset: !Rd, Rn - * LDRB,STRB register offset: !Rd, Rn, !Rm - * LDRB,STRB scaled register offset: !Rd, !Rn, !Rm - * LDRB,STRB immediate pre/post-indexed: !Rd, !Rn - * LDRB,STRB register pre/post-indexed: !Rd, !Rn, !Rm - * LDRB,STRB scaled register pre/post-indexed: !Rd, !Rn, !Rm - * LDRT,LDRBT,STRBT immediate pre/post-indexed: !Rd, !Rn - * LDRT,LDRBT,STRBT register pre/post-indexed: !Rd, !Rn, !Rm - * LDRT,LDRBT,STRBT scaled register pre/post-indexed: !Rd, !Rn, !Rm - * LDRH/SH/SB/D,STRH/SH/SB/D immediate offset: !Rd, Rn - * LDRH/SH/SB/D,STRH/SH/SB/D register offset: !Rd, Rn, !Rm - * LDRH/SH/SB/D,STRH/SH/SB/D immediate pre/post-indexed: !Rd, !Rn - * LDRH/SH/SB/D,STRH/SH/SB/D register pre/post-indexed: !Rd, !Rn, !Rm - * LDREX: !Rd, !Rn - * MCR/2: !Rd - * MCRR/2,MRRC/2: !Rd, !Rn - * MLA: !Rd, !Rn, !Rm, !Rs - * MOV: Rd - * MRC/2: !Rd (if Rd==15, only changes cond codes, not the register) - * MRS,MSR: !Rd - * MUL: !Rd, !Rm, !Rs - * PKH{BT,TB}: !Rd, !Rn, !Rm - * QDADD,[U]QADD/16/8/SUBX: !Rd, !Rm, !Rn - * QDSUB,[U]QSUB/16/8/ADDX: !Rd, !Rm, !Rn - * REV/16/SH: !Rd, !Rm - * RFE: !Rn - * {S,U}[H]ADD{16,8,SUBX},{S,U}[H]SUB{16,8,ADDX}: !Rd, !Rn, !Rm - * SEL: !Rd, !Rn, !Rm - * SMLA,SMLA{D,W},SMLSD,SMML{A,S}: !Rd, !Rn, !Rm, !Rs - * SMLAL,SMLA{D,LD},SMLSLD,SMMULL,SMULW: !RdHi, !RdLo, !Rm, !Rs - * SMMUL,SMUAD,SMUL,SMUSD: !Rd, !Rm, !Rs - * SSAT/16: !Rd, !Rm - * STM(1/2): !Rn, register_list* (R15 in reg list not recommended) - * STRT immediate pre/post-indexed: Rd*, !Rn - * STRT register pre/post-indexed: Rd*, !Rn, !Rm - * STRT scaled register pre/post-indexed: Rd*, !Rn, !Rm - * STREX: !Rd, !Rn, !Rm - * SWP/B: !Rd, !Rn, !Rm - * {S,U}XTA{B,B16,H}: !Rd, !Rn, !Rm - * {S,U}XT{B,B16,H}: !Rd, !Rm - * UM{AA,LA,UL}L: !RdHi, !RdLo, !Rm, !Rs - * USA{D8,A8,T,T16}: !Rd, !Rm, !Rs - * - * May transfer control by writing R15 (possible mode changes or alternate - * mode accesses marked by "*"): - * ALU op (* with s-bit), B, BL, BKPT, BLX(1/2), BX, BXJ, CPS*, CPY, - * LDM(1), LDM(2/3)*, LDR, MOV, RFE*, SWI* - * - * Instructions that do not take general registers, nor transfer control: - * CDP/2, SETEND, SRS* - */ -- cgit v1.1 From a9851832857dc1e4efefca1713f5cff3e168a25c Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 29 Apr 2011 10:42:19 +0200 Subject: [S390] irqstats: fix counting of pfault, dasd diag and virtio irqs pfault, dasd diag and virtio all use the same external interrupt number. The respective interrupt handlers decide by the subcode if they are meant to handle the interrupt. Counting is currently done before looking at the subcode which means each handler counts an interrupt even if it is not handling it. Fix this by moving the kstat code after the code which looks at the subcode. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/mm/fault.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 4cf85fe..ab98813 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -543,7 +543,6 @@ static void pfault_interrupt(unsigned int ext_int_code, struct task_struct *tsk; __u16 subcode; - kstat_cpu(smp_processor_id()).irqs[EXTINT_PFL]++; /* * Get the external interruption subcode & pfault * initial/completion signal bit. VM stores this @@ -553,6 +552,7 @@ static void pfault_interrupt(unsigned int ext_int_code, subcode = ext_int_code >> 16; if ((subcode & 0xff00) != __SUBCODE_MASK) return; + kstat_cpu(smp_processor_id()).irqs[EXTINT_PFL]++; /* * Get the token (= address of the task structure of the affected task). -- cgit v1.1 From 0f22072ab50cac7983f9660d33974b45184da4f9 Mon Sep 17 00:00:00 2001 From: Dan Rosenberg Date: Fri, 29 Apr 2011 15:48:07 +0100 Subject: ARM: 6891/1: prevent heap corruption in OABI semtimedop When CONFIG_OABI_COMPAT is set, the wrapper for semtimedop does not bound the nsops argument. A sufficiently large value will cause an integer overflow in allocation size, followed by copying too much data into the allocated buffer. Fix this by restricting nsops to SEMOPM. Untested. Cc: stable@kernel.org Signed-off-by: Dan Rosenberg Signed-off-by: Russell King --- arch/arm/kernel/sys_oabi-compat.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/arm/kernel/sys_oabi-compat.c b/arch/arm/kernel/sys_oabi-compat.c index 4ad8da1..af0aaeb 100644 --- a/arch/arm/kernel/sys_oabi-compat.c +++ b/arch/arm/kernel/sys_oabi-compat.c @@ -311,7 +311,7 @@ asmlinkage long sys_oabi_semtimedop(int semid, long err; int i; - if (nsops < 1) + if (nsops < 1 || nsops > SEMOPM) return -EINVAL; sops = kmalloc(sizeof(*sops) * nsops, GFP_KERNEL); if (!sops) -- cgit v1.1 From e20a2d205c05cef6b5783df339a7d54adeb50962 Mon Sep 17 00:00:00 2001 From: Boris Ostrovsky Date: Fri, 29 Apr 2011 17:47:43 -0400 Subject: x86, AMD: Fix APIC timer erratum 400 affecting K8 Rev.A-E processors Older AMD K8 processors (Revisions A-E) are affected by erratum 400 (APIC timer interrupts don't occur in C states greater than C1). This, for example, means that X86_FEATURE_ARAT flag should not be set for these parts. This addresses regression introduced by commit b87cf80af3ba4b4c008b4face3c68d604e1715c6 ("x86, AMD: Set ARAT feature on AMD processors") where the system may become unresponsive until external interrupt (such as keyboard input) occurs. This results, for example, in time not being reported correctly, lack of progress on the system and other lockups. Reported-by: Joerg-Volker Peetz Tested-by: Joerg-Volker Peetz Acked-by: Borislav Petkov Signed-off-by: Boris Ostrovsky Cc: stable@kernel.org Link: http://lkml.kernel.org/r/1304113663-6586-1-git-send-email-ostr@amd64.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/amd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 3532d3b..bb9eb29 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -698,7 +698,7 @@ cpu_dev_register(amd_cpu_dev); */ const int amd_erratum_400[] = - AMD_OSVW_ERRATUM(1, AMD_MODEL_RANGE(0xf, 0x41, 0x2, 0xff, 0xf), + AMD_OSVW_ERRATUM(1, AMD_MODEL_RANGE(0x0f, 0x4, 0x2, 0xff, 0xf), AMD_MODEL_RANGE(0x10, 0x2, 0x1, 0xff, 0xf)); EXPORT_SYMBOL_GPL(amd_erratum_400); -- cgit v1.1 From 2be19102b71c1a45d37fec50303791daa1a06869 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Sun, 1 May 2011 19:12:04 +0200 Subject: x86, NUMA: Fix empty memblk detection in numa_cleanup_meminfo() numa_cleanup_meminfo() trims each memblk between low (0) and high (max_pfn) limits and discards empty ones. However, the emptiness detection incorrectly used equality test. If the start of a memblk is higher than max_pfn, it is empty but fails the equality test and doesn't get discarded. The condition triggers when max_pfn is lower than start of a NUMA node and results in memory misconfiguration - leading to WARN_ON()s and other funnies. The bug was discovered in devel branch where 32bit too uses this code path for NUMA init. If a node is above the addressing limit, max_pfn ends up lower than the node triggering this problem. The failure hasn't been observed on x86-64 but is still possible with broken hardware e820/NUMA info. As the fix is very low risk, it would be better to apply it even for 64bit. Fix it by using >= instead of ==. Signed-off-by: Yinghai Lu [ Extracted the actual fix from the original patch and rewrote patch description. ] Signed-off-by: Tejun Heo Link: http://lkml.kernel.org/r/20110501171204.GO29280@htj.dyndns.org Signed-off-by: Ingo Molnar --- arch/x86/mm/numa_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index e8c00cc..85b52fc 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -306,7 +306,7 @@ int __init numa_cleanup_meminfo(struct numa_meminfo *mi) bi->end = min(bi->end, high); /* and there's no empty block */ - if (bi->start == bi->end) { + if (bi->start >= bi->end) { numa_remove_memblk_from(i--, mi); continue; } -- cgit v1.1 From a38647837a411f7df79623128421eef2118b5884 Mon Sep 17 00:00:00 2001 From: Konrad Rzeszutek Wilk Date: Fri, 29 Apr 2011 11:34:00 -0400 Subject: xen/mmu: Add workaround "x86-64, mm: Put early page table high" As a consequence of the commit: commit 4b239f458c229de044d6905c2b0f9fe16ed9e01e Author: Yinghai Lu Date: Fri Dec 17 16:58:28 2010 -0800 x86-64, mm: Put early page table high it causes the Linux kernel to crash under Xen: mapping kernel into physical memory Xen: setup ISA identity maps about to get started... (XEN) mm.c:2466:d0 Bad type (saw 7400000000000001 != exp 1000000000000000) for mfn b1d89 (pfn bacf7) (XEN) mm.c:3027:d0 Error while pinning mfn b1d89 (XEN) traps.c:481:d0 Unhandled invalid opcode fault/trap [#6] on VCPU 0 [ec=0000] (XEN) domain_crash_sync called from entry.S (XEN) Domain 0 (vcpu#0) crashed on cpu#0: ... The reason is that at some point init_memory_mapping is going to reach the pagetable pages area and map those pages too (mapping them as normal memory that falls in the range of addresses passed to init_memory_mapping as argument). Some of those pages are already pagetable pages (they are in the range pgt_buf_start-pgt_buf_end) therefore they are going to be mapped RO and everything is fine. Some of these pages are not pagetable pages yet (they fall in the range pgt_buf_end-pgt_buf_top; for example the page at pgt_buf_end) so they are going to be mapped RW. When these pages become pagetable pages and are hooked into the pagetable, xen will find that the guest has already a RW mapping of them somewhere and fail the operation. The reason Xen requires pagetables to be RO is that the hypervisor needs to verify that the pagetables are valid before using them. The validation operations are called "pinning" (more details in arch/x86/xen/mmu.c). In order to fix the issue we mark all the pages in the entire range pgt_buf_start-pgt_buf_top as RO, however when the pagetable allocation is completed only the range pgt_buf_start-pgt_buf_end is reserved by init_memory_mapping. Hence the kernel is going to crash as soon as one of the pages in the range pgt_buf_end-pgt_buf_top is reused (b/c those ranges are RO). For this reason, this function is introduced which is called _after_ the init_memory_mapping has completed (in a perfect world we would call this function from init_memory_mapping, but lets ignore that). Because we are called _after_ init_memory_mapping the pgt_buf_[start, end,top] have all changed to new values (b/c another init_memory_mapping is called). Hence, the first time we enter this function, we save away the pgt_buf_start value and update the pgt_buf_[end,top]. When we detect that the "old" pgt_buf_start through pgt_buf_end PFNs have been reserved (so memblock_x86_reserve_range has been called), we immediately set out to RW the "old" pgt_buf_end through pgt_buf_top. And then we update those "old" pgt_buf_[end|top] with the new ones so that we can redo this on the next pagetable. Acked-by: "H. Peter Anvin" Reviewed-by: Jeremy Fitzhardinge [v1: Updated with Jeremy's comments] [v2: Added the crash output] Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/mmu.c | 123 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 123 insertions(+) (limited to 'arch') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index aef7af9..1bca25f 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1463,6 +1463,119 @@ static int xen_pgd_alloc(struct mm_struct *mm) return ret; } +#ifdef CONFIG_X86_64 +static __initdata u64 __last_pgt_set_rw = 0; +static __initdata u64 __pgt_buf_start = 0; +static __initdata u64 __pgt_buf_end = 0; +static __initdata u64 __pgt_buf_top = 0; +/* + * As a consequence of the commit: + * + * commit 4b239f458c229de044d6905c2b0f9fe16ed9e01e + * Author: Yinghai Lu + * Date: Fri Dec 17 16:58:28 2010 -0800 + * + * x86-64, mm: Put early page table high + * + * at some point init_memory_mapping is going to reach the pagetable pages + * area and map those pages too (mapping them as normal memory that falls + * in the range of addresses passed to init_memory_mapping as argument). + * Some of those pages are already pagetable pages (they are in the range + * pgt_buf_start-pgt_buf_end) therefore they are going to be mapped RO and + * everything is fine. + * Some of these pages are not pagetable pages yet (they fall in the range + * pgt_buf_end-pgt_buf_top; for example the page at pgt_buf_end) so they + * are going to be mapped RW. When these pages become pagetable pages and + * are hooked into the pagetable, xen will find that the guest has already + * a RW mapping of them somewhere and fail the operation. + * The reason Xen requires pagetables to be RO is that the hypervisor needs + * to verify that the pagetables are valid before using them. The validation + * operations are called "pinning". + * + * In order to fix the issue we mark all the pages in the entire range + * pgt_buf_start-pgt_buf_top as RO, however when the pagetable allocation + * is completed only the range pgt_buf_start-pgt_buf_end is reserved by + * init_memory_mapping. Hence the kernel is going to crash as soon as one + * of the pages in the range pgt_buf_end-pgt_buf_top is reused (b/c those + * ranges are RO). + * + * For this reason, 'mark_rw_past_pgt' is introduced which is called _after_ + * the init_memory_mapping has completed (in a perfect world we would + * call this function from init_memory_mapping, but lets ignore that). + * + * Because we are called _after_ init_memory_mapping the pgt_buf_[start, + * end,top] have all changed to new values (b/c init_memory_mapping + * is called and setting up another new page-table). Hence, the first time + * we enter this function, we save away the pgt_buf_start value and update + * the pgt_buf_[end,top]. + * + * When we detect that the "old" pgt_buf_start through pgt_buf_end + * PFNs have been reserved (so memblock_x86_reserve_range has been called), + * we immediately set out to RW the "old" pgt_buf_end through pgt_buf_top. + * + * And then we update those "old" pgt_buf_[end|top] with the new ones + * so that we can redo this on the next pagetable. + */ +static __init void mark_rw_past_pgt(void) { + + if (pgt_buf_end > pgt_buf_start) { + u64 addr, size; + + /* Save it away. */ + if (!__pgt_buf_start) { + __pgt_buf_start = pgt_buf_start; + __pgt_buf_end = pgt_buf_end; + __pgt_buf_top = pgt_buf_top; + return; + } + /* If we get the range that starts at __pgt_buf_end that means + * the range is reserved, and that in 'init_memory_mapping' + * the 'memblock_x86_reserve_range' has been called with the + * outdated __pgt_buf_start, __pgt_buf_end (the "new" + * pgt_buf_[start|end|top] refer now to a new pagetable. + * Note: we are called _after_ the pgt_buf_[..] have been + * updated.*/ + + addr = memblock_x86_find_in_range_size(PFN_PHYS(__pgt_buf_start), + &size, PAGE_SIZE); + + /* Still not reserved, meaning 'memblock_x86_reserve_range' + * hasn't been called yet. Update the _end and _top.*/ + if (addr == PFN_PHYS(__pgt_buf_start)) { + __pgt_buf_end = pgt_buf_end; + __pgt_buf_top = pgt_buf_top; + return; + } + + /* OK, the area is reserved, meaning it is time for us to + * set RW for the old end->top PFNs. */ + + /* ..unless we had already done this. */ + if (__pgt_buf_end == __last_pgt_set_rw) + return; + + addr = PFN_PHYS(__pgt_buf_end); + + /* set as RW the rest */ + printk(KERN_DEBUG "xen: setting RW the range %llx - %llx\n", + PFN_PHYS(__pgt_buf_end), PFN_PHYS(__pgt_buf_top)); + + while (addr < PFN_PHYS(__pgt_buf_top)) { + make_lowmem_page_readwrite(__va(addr)); + addr += PAGE_SIZE; + } + /* And update everything so that we are ready for the next + * pagetable (the one created for regions past 4GB) */ + __last_pgt_set_rw = __pgt_buf_end; + __pgt_buf_start = pgt_buf_start; + __pgt_buf_end = pgt_buf_end; + __pgt_buf_top = pgt_buf_top; + } + return; +} +#else +static __init void mark_rw_past_pgt(void) { } +#endif static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd) { #ifdef CONFIG_X86_64 @@ -1489,6 +1602,14 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) unsigned long pfn = pte_pfn(pte); /* + * A bit of optimization. We do not need to call the workaround + * when xen_set_pte_init is called with a PTE with 0 as PFN. + * That is b/c the pagetable at that point are just being populated + * with empty values and we can save some cycles by not calling + * the 'memblock' code.*/ + if (pfn) + mark_rw_past_pgt(); + /* * If the new pfn is within the range of the newly allocated * kernel pagetable, and it isn't being mapped into an * early_ioremap fixmap slot as a freshly allocated page, make sure @@ -1997,6 +2118,8 @@ __init void xen_ident_map_ISA(void) static __init void xen_post_allocator_init(void) { + mark_rw_past_pgt(); + #ifdef CONFIG_XEN_DEBUG pv_mmu_ops.make_pte = PV_CALLEE_SAVE(xen_make_pte_debug); #endif -- cgit v1.1 From b9269dc7bfdf8c985971c09f2dcb2aa04ad7986d Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 12 Apr 2011 12:19:49 +0100 Subject: xen: mask_rw_pte mark RO all pagetable pages up to pgt_buf_top mask_rw_pte is currently checking if a pfn is a pagetable page if it falls in the range pgt_buf_start - pgt_buf_end but that is incorrect because pgt_buf_end is a moving target: pgt_buf_top is the real boundary. Acked-by: "H. Peter Anvin" Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk --- arch/x86/xen/mmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index 1bca25f..55c965b 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1616,7 +1616,7 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) * it is RO. */ if (((!is_early_ioremap_ptep(ptep) && - pfn >= pgt_buf_start && pfn < pgt_buf_end)) || + pfn >= pgt_buf_start && pfn < pgt_buf_top)) || (is_early_ioremap_ptep(ptep) && pfn != (pgt_buf_end - 1))) pte = pte_wrprotect(pte); -- cgit v1.1 From 7806a49ab625ebeb1709e5e87299b64932b807a7 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Mon, 2 May 2011 14:33:24 -0700 Subject: x86, reboot: Fix relocations in reboot_32.S The use of base for %ebx in this file is arbitrary, *except* that we also use it to compute the real-mode segment. Therefore, make it so that r_base really is the true address to which %ebx points. This resolves kernel bugzilla 33302. Reported-and-tested-by: Alexey Zaytsev Signed-off-by: H. Peter Anvin Link: http://lkml.kernel.org/n/tip-08os5wi3yq1no0y4i5m4z7he@git.kernel.org --- arch/x86/kernel/reboot_32.S | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/reboot_32.S b/arch/x86/kernel/reboot_32.S index 29092b3..1d5c46d 100644 --- a/arch/x86/kernel/reboot_32.S +++ b/arch/x86/kernel/reboot_32.S @@ -21,26 +21,26 @@ r_base = . /* Get our own relocated address */ call 1f 1: popl %ebx - subl $1b, %ebx + subl $(1b - r_base), %ebx /* Compute the equivalent real-mode segment */ movl %ebx, %ecx shrl $4, %ecx /* Patch post-real-mode segment jump */ - movw dispatch_table(%ebx,%eax,2),%ax - movw %ax, 101f(%ebx) - movw %cx, 102f(%ebx) + movw (dispatch_table - r_base)(%ebx,%eax,2),%ax + movw %ax, (101f - r_base)(%ebx) + movw %cx, (102f - r_base)(%ebx) /* Set up the IDT for real mode. */ - lidtl machine_real_restart_idt(%ebx) + lidtl (machine_real_restart_idt - r_base)(%ebx) /* * Set up a GDT from which we can load segment descriptors for real * mode. The GDT is not used in real mode; it is just needed here to * prepare the descriptors. */ - lgdtl machine_real_restart_gdt(%ebx) + lgdtl (machine_real_restart_gdt - r_base)(%ebx) /* * Load the data segment registers with 16-bit compatible values -- cgit v1.1 From 925f83c085e1bb08435556c5b4844a60de002e31 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 6 May 2011 01:53:18 +0200 Subject: hw_breakpoints, powerpc: Fix CONFIG_HAVE_HW_BREAKPOINT off-case in ptrace_set_debugreg() We make use of ptrace_get_breakpoints() / ptrace_put_breakpoints() to protect ptrace_set_debugreg() even if CONFIG_HAVE_HW_BREAKPOINT if off. However in this case, these APIs are not implemented. To fix this, push the protection down inside the relevant ifdef. Best would be to export the code inside CONFIG_HAVE_HW_BREAKPOINT into a standalone function to cleanup the ifdefury there and call the breakpoint ref API inside. But as it is more invasive, this should be rather made in an -rc1. Fixes this build error: arch/powerpc/kernel/ptrace.c:1594: error: implicit declaration of function 'ptrace_get_breakpoints' make[2]: *** Reported-by: Ingo Molnar Signed-off-by: Frederic Weisbecker Cc: LPPC Cc: Prasad Cc: v2.6.33.. Link: http://lkml.kernel.org/r/1304639598-4707-1-git-send-email-fweisbec@gmail.com Signed-off-by: Ingo Molnar --- arch/powerpc/kernel/ptrace.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) (limited to 'arch') diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c index 4edeeb3..a6ae1cf 100644 --- a/arch/powerpc/kernel/ptrace.c +++ b/arch/powerpc/kernel/ptrace.c @@ -933,12 +933,16 @@ int ptrace_set_debugreg(struct task_struct *task, unsigned long addr, if (data && !(data & DABR_TRANSLATION)) return -EIO; #ifdef CONFIG_HAVE_HW_BREAKPOINT + if (ptrace_get_breakpoints(task) < 0) + return -ESRCH; + bp = thread->ptrace_bps[0]; if ((!data) || !(data & (DABR_DATA_WRITE | DABR_DATA_READ))) { if (bp) { unregister_hw_breakpoint(bp); thread->ptrace_bps[0] = NULL; } + ptrace_put_breakpoints(task); return 0; } if (bp) { @@ -948,9 +952,12 @@ int ptrace_set_debugreg(struct task_struct *task, unsigned long addr, (DABR_DATA_WRITE | DABR_DATA_READ), &attr.bp_type); ret = modify_user_hw_breakpoint(bp, &attr); - if (ret) + if (ret) { + ptrace_put_breakpoints(task); return ret; + } thread->ptrace_bps[0] = bp; + ptrace_put_breakpoints(task); thread->dabr = data; return 0; } @@ -965,9 +972,12 @@ int ptrace_set_debugreg(struct task_struct *task, unsigned long addr, ptrace_triggered, task); if (IS_ERR(bp)) { thread->ptrace_bps[0] = NULL; + ptrace_put_breakpoints(task); return PTR_ERR(bp); } + ptrace_put_breakpoints(task); + #endif /* CONFIG_HAVE_HW_BREAKPOINT */ /* Move contents to the DABR register */ @@ -1591,10 +1601,7 @@ long arch_ptrace(struct task_struct *child, long request, } case PTRACE_SET_DEBUGREG: - if (ptrace_get_breakpoints(child) < 0) - return -ESRCH; ret = ptrace_set_debugreg(child, addr, data); - ptrace_put_breakpoints(child); break; #ifdef CONFIG_PPC64 -- cgit v1.1 From 63b6a6758eede2f9283c3594265b6e32e75d7456 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sat, 23 Apr 2011 00:57:42 +0200 Subject: perf events, x86: Fix Intel Nehalem and Westmere last level cache event definitions The Intel Nehalem offcore bits implemented in: e994d7d23a0b: perf: Fix LLC-* events on Intel Nehalem/Westmere ... are wrong: they implemented _ACCESS as _HIT and counted OTHER_CORE_HIT* as MISS even though its clearly documented as an L3 hit ... Fix them and the Westmere definitions as well. Cc: Andi Kleen Cc: Lin Ming Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Mike Galbraith Cc: Steven Rostedt Link: http://lkml.kernel.org/r/1299119690-13991-3-git-send-email-ming.m.lin@intel.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/cpu/perf_event_intel.c | 87 ++++++++++++++++++++-------------- 1 file changed, 52 insertions(+), 35 deletions(-) (limited to 'arch') diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index e61539b..447a28d 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -184,26 +184,23 @@ static __initconst const u64 snb_hw_cache_event_ids }, }, [ C(LL ) ] = { - /* - * TBD: Need Off-core Response Performance Monitoring support - */ [ C(OP_READ) ] = { - /* OFFCORE_RESPONSE_0.ANY_DATA.LOCAL_CACHE */ + /* OFFCORE_RESPONSE.ANY_DATA.LOCAL_CACHE */ [ C(RESULT_ACCESS) ] = 0x01b7, - /* OFFCORE_RESPONSE_1.ANY_DATA.ANY_LLC_MISS */ - [ C(RESULT_MISS) ] = 0x01bb, + /* OFFCORE_RESPONSE.ANY_DATA.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, }, [ C(OP_WRITE) ] = { - /* OFFCORE_RESPONSE_0.ANY_RFO.LOCAL_CACHE */ + /* OFFCORE_RESPONSE.ANY_RFO.LOCAL_CACHE */ [ C(RESULT_ACCESS) ] = 0x01b7, - /* OFFCORE_RESPONSE_1.ANY_RFO.ANY_LLC_MISS */ - [ C(RESULT_MISS) ] = 0x01bb, + /* OFFCORE_RESPONSE.ANY_RFO.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, }, [ C(OP_PREFETCH) ] = { - /* OFFCORE_RESPONSE_0.PREFETCH.LOCAL_CACHE */ + /* OFFCORE_RESPONSE.PREFETCH.LOCAL_CACHE */ [ C(RESULT_ACCESS) ] = 0x01b7, - /* OFFCORE_RESPONSE_1.PREFETCH.ANY_LLC_MISS */ - [ C(RESULT_MISS) ] = 0x01bb, + /* OFFCORE_RESPONSE.PREFETCH.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, }, }, [ C(DTLB) ] = { @@ -285,26 +282,26 @@ static __initconst const u64 westmere_hw_cache_event_ids }, [ C(LL ) ] = { [ C(OP_READ) ] = { - /* OFFCORE_RESPONSE_0.ANY_DATA.LOCAL_CACHE */ + /* OFFCORE_RESPONSE.ANY_DATA.LOCAL_CACHE */ [ C(RESULT_ACCESS) ] = 0x01b7, - /* OFFCORE_RESPONSE_1.ANY_DATA.ANY_LLC_MISS */ - [ C(RESULT_MISS) ] = 0x01bb, + /* OFFCORE_RESPONSE.ANY_DATA.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, }, /* * Use RFO, not WRITEBACK, because a write miss would typically occur * on RFO. */ [ C(OP_WRITE) ] = { - /* OFFCORE_RESPONSE_1.ANY_RFO.LOCAL_CACHE */ - [ C(RESULT_ACCESS) ] = 0x01bb, - /* OFFCORE_RESPONSE_0.ANY_RFO.ANY_LLC_MISS */ + /* OFFCORE_RESPONSE.ANY_RFO.LOCAL_CACHE */ + [ C(RESULT_ACCESS) ] = 0x01b7, + /* OFFCORE_RESPONSE.ANY_RFO.ANY_LLC_MISS */ [ C(RESULT_MISS) ] = 0x01b7, }, [ C(OP_PREFETCH) ] = { - /* OFFCORE_RESPONSE_0.PREFETCH.LOCAL_CACHE */ + /* OFFCORE_RESPONSE.PREFETCH.LOCAL_CACHE */ [ C(RESULT_ACCESS) ] = 0x01b7, - /* OFFCORE_RESPONSE_1.PREFETCH.ANY_LLC_MISS */ - [ C(RESULT_MISS) ] = 0x01bb, + /* OFFCORE_RESPONSE.PREFETCH.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, }, }, [ C(DTLB) ] = { @@ -352,16 +349,36 @@ static __initconst const u64 westmere_hw_cache_event_ids }; /* - * OFFCORE_RESPONSE MSR bits (subset), See IA32 SDM Vol 3 30.6.1.3 + * Nehalem/Westmere MSR_OFFCORE_RESPONSE bits; + * See IA32 SDM Vol 3B 30.6.1.3 */ -#define DMND_DATA_RD (1 << 0) -#define DMND_RFO (1 << 1) -#define DMND_WB (1 << 3) -#define PF_DATA_RD (1 << 4) -#define PF_DATA_RFO (1 << 5) -#define RESP_UNCORE_HIT (1 << 8) -#define RESP_MISS (0xf600) /* non uncore hit */ +#define NHM_DMND_DATA_RD (1 << 0) +#define NHM_DMND_RFO (1 << 1) +#define NHM_DMND_IFETCH (1 << 2) +#define NHM_DMND_WB (1 << 3) +#define NHM_PF_DATA_RD (1 << 4) +#define NHM_PF_DATA_RFO (1 << 5) +#define NHM_PF_IFETCH (1 << 6) +#define NHM_OFFCORE_OTHER (1 << 7) +#define NHM_UNCORE_HIT (1 << 8) +#define NHM_OTHER_CORE_HIT_SNP (1 << 9) +#define NHM_OTHER_CORE_HITM (1 << 10) + /* reserved */ +#define NHM_REMOTE_CACHE_FWD (1 << 12) +#define NHM_REMOTE_DRAM (1 << 13) +#define NHM_LOCAL_DRAM (1 << 14) +#define NHM_NON_DRAM (1 << 15) + +#define NHM_ALL_DRAM (NHM_REMOTE_DRAM|NHM_LOCAL_DRAM) + +#define NHM_DMND_READ (NHM_DMND_DATA_RD) +#define NHM_DMND_WRITE (NHM_DMND_RFO|NHM_DMND_WB) +#define NHM_DMND_PREFETCH (NHM_PF_DATA_RD|NHM_PF_DATA_RFO) + +#define NHM_L3_HIT (NHM_UNCORE_HIT|NHM_OTHER_CORE_HIT_SNP|NHM_OTHER_CORE_HITM) +#define NHM_L3_MISS (NHM_NON_DRAM|NHM_ALL_DRAM|NHM_REMOTE_CACHE_FWD) +#define NHM_L3_ACCESS (NHM_L3_HIT|NHM_L3_MISS) static __initconst const u64 nehalem_hw_cache_extra_regs [PERF_COUNT_HW_CACHE_MAX] @@ -370,16 +387,16 @@ static __initconst const u64 nehalem_hw_cache_extra_regs { [ C(LL ) ] = { [ C(OP_READ) ] = { - [ C(RESULT_ACCESS) ] = DMND_DATA_RD|RESP_UNCORE_HIT, - [ C(RESULT_MISS) ] = DMND_DATA_RD|RESP_MISS, + [ C(RESULT_ACCESS) ] = NHM_DMND_READ|NHM_L3_ACCESS, + [ C(RESULT_MISS) ] = NHM_DMND_READ|NHM_L3_MISS, }, [ C(OP_WRITE) ] = { - [ C(RESULT_ACCESS) ] = DMND_RFO|DMND_WB|RESP_UNCORE_HIT, - [ C(RESULT_MISS) ] = DMND_RFO|DMND_WB|RESP_MISS, + [ C(RESULT_ACCESS) ] = NHM_DMND_WRITE|NHM_L3_ACCESS, + [ C(RESULT_MISS) ] = NHM_DMND_WRITE|NHM_L3_MISS, }, [ C(OP_PREFETCH) ] = { - [ C(RESULT_ACCESS) ] = PF_DATA_RD|PF_DATA_RFO|RESP_UNCORE_HIT, - [ C(RESULT_MISS) ] = PF_DATA_RD|PF_DATA_RFO|RESP_MISS, + [ C(RESULT_ACCESS) ] = NHM_DMND_PREFETCH|NHM_L3_ACCESS, + [ C(RESULT_MISS) ] = NHM_DMND_PREFETCH|NHM_L3_MISS, }, } }; -- cgit v1.1