diff options
| author | Alan Cox <alan@linux.intel.com> | 2009-06-05 11:56:18 +0100 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2009-06-05 08:05:23 -0700 |
| commit | e9e10124e269a39de089c5e0d9936fae2ff889b2 (patch) | |
| tree | ca2327be762018a8fe0a5c1a45a37478a000d931 /drivers | |
| parent | edaba2c5334492f82d39ec35637c6dea5176a977 (diff) | |
| download | kernel_samsung_espresso10-e9e10124e269a39de089c5e0d9936fae2ff889b2.zip kernel_samsung_espresso10-e9e10124e269a39de089c5e0d9936fae2ff889b2.tar.gz kernel_samsung_espresso10-e9e10124e269a39de089c5e0d9936fae2ff889b2.tar.bz2 | |
ivtv: Fix PCI DMA direction
The ivtv stream buffers may be for receive or for send but the attached
sg handle is always destined cpu->device. We flush it correctly but the
allocation is wrongly done with the same type as the buffers.
See bug: http://bugzilla.kernel.org/show_bug.cgi?id=13385
(Note this doesn't close the bug - it fixes the ivtv part and in turn
the logging next shows up some rather alarming DMA sg list warnings in
libata)
Signed-off-by: Alan Cox <alan@linux.intel.com>
Acked-by: Hans Verkuil <hverkuil@xs4all.nl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'drivers')
| -rw-r--r-- | drivers/media/video/ivtv/ivtv-queue.c | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/drivers/media/video/ivtv/ivtv-queue.c b/drivers/media/video/ivtv/ivtv-queue.c index ff7b7de..7fde36e 100644 --- a/drivers/media/video/ivtv/ivtv-queue.c +++ b/drivers/media/video/ivtv/ivtv-queue.c @@ -230,7 +230,8 @@ int ivtv_stream_alloc(struct ivtv_stream *s) return -ENOMEM; } if (ivtv_might_use_dma(s)) { - s->sg_handle = pci_map_single(itv->pdev, s->sg_dma, sizeof(struct ivtv_sg_element), s->dma); + s->sg_handle = pci_map_single(itv->pdev, s->sg_dma, + sizeof(struct ivtv_sg_element), PCI_DMA_TODEVICE); ivtv_stream_sync_for_cpu(s); } |