Commit Graph

6 Commits

Author SHA1 Message Date
Ma Haijun 29f1cdb0d0 [media] videobuf-dma-contig: fix incorrect argument to vm_iomap_memory() call
The second argument should be physical address rather than virtual address.

Signed-off-by: Ma Haijun <mahaijuns@gmail.com>
Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
2014-05-23 12:54:09 -03:00
Hans Verkuil cca36e2eec [media] Revert "[media] videobuf_vm_{open,close} race fixes"
This reverts commit a242f42610.

That commit actually caused deadlocks, rather then fixing them.

If ext_lock is set to NULL (otherwise videobuf_queue_lock doesn't do
anything), then you get this deadlock:

The driver's mmap function calls videobuf_mmap_mapper which calls
videobuf_queue_lock on q. videobuf_mmap_mapper calls  __videobuf_mmap_mapper,
__videobuf_mmap_mapper calls videobuf_vm_open and videobuf_vm_open
calls videobuf_queue_lock on q (introduced by above patch): deadlocked.

This affects drivers using dma-contig and dma-vmalloc. Only dma-sg is
not affected since it doesn't call videobuf_vm_open from __videobuf_mmap_mapper.

Most drivers these days have a non-NULL ext_lock. Those that still use
NULL there are all fairly obscure drivers, which is why this hasn't been
seen earlier.

Since everything worked perfectly fine for many years I prefer to just
revert this patch rather than trying to fix it. videobuf is quite fragile
and I rather not touch it too much. Work is (slowly) progressing to move
everything over to vb2 or at the very least use non-NULL ext_lock in
videobuf.

Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com>
Cc: <stable@vger.kernel.org>      # for v3.11 and up
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Reported-by: Pete Eberlein <pete@sensoray.com>
Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
2014-02-04 06:29:46 -02:00
Mauro Carvalho Chehab 2a848b2c7f [media] videobuf-dma-contig: use vm_iomap_memory()
vm_iomap_memory() provides a better end user interface than
remap_pfn_range(), as it does the needed tests before doing
mmap.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2013-05-21 09:22:12 -03:00
Al Viro a242f42610 [media] videobuf_vm_{open,close} race fixes
just use videobuf_queue_lock(map->q) to protect map->count; vm_area_operations
->open() and ->close() are called just under vma->vm_mm->mmap_sem, which
doesn't help the drivers at all, since clonal VMAs are normally in different
address spaces...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2013-05-21 08:56:56 -03:00
Mauro Carvalho Chehab cb132cd5d7 [media] videobuf-dma-contig: remove support for cached mem
videobuf_queue_dma_contig_init_cached() is not used anywhere.
Drop support for it, cleaning up the code a little bit.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Acked-by: Hans Verkuil <hans.verkuil@cisco.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2013-04-25 09:50:19 -03:00
Mauro Carvalho Chehab 5bc3cb743b [media] v4l: move v4l2 core into a separate directory
Currently, the v4l2 core is mixed together with other non-core drivers.
Move them into a separate directory.

Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
2012-08-13 23:02:38 -03:00