linux-stable.git/drivers/media/platform, branch linux-4.8.y

s5p-mfc: fix failure path of s5p_mfc_alloc_memdev()

2017-01-09T07:21:59+00:00

commit 3467c9a7e7f9209a9ecd8f9db65b04a323a13932 upstream.

s5p_mfc_alloc_memdev() function lacks proper releasing
of allocated device in case of reserved memory initialization
failure. This results in NULL pointer dereference:

[    2.828457] Unable to handle kernel NULL pointer dereference at virtual address 00000001
[    2.835089] pgd = c0004000
[    2.837752] [00000001] *pgd=00000000
[    2.844696] Internal error: Oops: 5 [#1] PREEMPT SMP ARM
[    2.848680] Modules linked in:
[    2.851722] CPU: 1 PID: 1 Comm: swapper/0 Not tainted 4.8.0-rc6-00002-gafa1b97 #878
[    2.859357] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
[    2.865433] task: ef080000 task.stack: ef06c000
[    2.869952] PC is at strcmp+0x0/0x30
[    2.873508] LR is at platform_match+0x84/0xac
[    2.877847] pc : []    lr : []    psr: 20000013
[    2.877847] sp : ef06dea0  ip : 00000000  fp : 00000000
[    2.889303] r10: 00000000  r9 : c0b34848  r8 : c0b1e968
[    2.894511] r7 : 00000000  r6 : 00000001  r5 : c086e7fc  r4 : eeb8e010
[    2.901021] r3 : 0000006d  r2 : 00000000  r1 : c086e7fc  r0 : 00000001
[    2.907533] Flags: nzCv  IRQs on  FIQs on  Mode SVC_32  ISA ARM  Segment none
[    2.914649] Control: 10c5387d  Table: 4000404a  DAC: 00000051
[    2.920378] Process swapper/0 (pid: 1, stack limit = 0xef06c210)
[    2.926367] Stack: (0xef06dea0 to 0xef06e000)
[    2.930711] dea0: eeb8e010 c0c2d91c c03f4a6c c03f4a8c 00000000 c0c2d91c c03f4a6c c03f2fc8
[    2.938870] dec0: ef003274 ef10c4c0 c0c2d91c ef10cc80 c0c21270 c03f3fa4 c09c1be8 c0c2d91c
[    2.947028] dee0: 00000006 c0c2d91c 00000006 c0b3483c c0c47000 c03f5314 c0c2d908 c0b5fed8
[    2.955188] df00: 00000006 c010178c 60000013 c0a4ef14 00000000 c06feaa0 ef080000 60000013
[    2.963347] df20: 00000000 c0c095c8 efffca76 c0816b8c 000000d5 c0134098 c0b34848 c09d6cdc
[    2.971506] df40: c0a4de70 00000000 00000006 00000006 c0c09568 efffca40 c0b5fed8 00000006
[    2.979665] df60: c0b3483c c0c47000 000000d5 c0b34848 c0b005a4 c0b00d84 00000006 00000006
[    2.987824] df80: 00000000 c0b005a4 00000000 c06fb4d8 00000000 00000000 00000000 00000000
[    2.995983] dfa0: 00000000 c06fb4e0 00000000 c01079b8 00000000 00000000 00000000 00000000
[    3.004142] dfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[    3.012302] dfe0: 00000000 00000000 00000000 00000000 00000013 00000000 ffffffff ffffffff
[    3.020469] [] (strcmp) from [] (platform_match+0x84/0xac)
[    3.027672] [] (platform_match) from [] (__driver_attach+0x20/0xb0)
[    3.035654] [] (__driver_attach) from [] (bus_for_each_dev+0x54/0x88)
[    3.043812] [] (bus_for_each_dev) from [] (bus_add_driver+0xe8/0x1f4)
[    3.051971] [] (bus_add_driver) from [] (driver_register+0x78/0xf4)
[    3.059958] [] (driver_register) from [] (do_one_initcall+0x3c/0x16c)
[    3.068123] [] (do_one_initcall) from [] (kernel_init_freeable+0x120/0x1ec)
[    3.076802] [] (kernel_init_freeable) from [] (kernel_init+0x8/0x118)
[    3.084958] [] (kernel_init) from [] (ret_from_fork+0x14/0x3c)
[    3.092506] Code: 1afffffb e12fff1e e1a03000 eafffff7 (e4d03001)
[    3.098618] ---[ end trace 511bf9d750810709 ]---
[    3.103207] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b

This patch fixes this issue.

Fixes: c79667dd93b084fe412bcfe7fbf0ba43f7dec520 ("media: s5p-mfc: replace custom
	reserved memory handling code with generic one")

Signed-off-by: Marek Szyprowski 
Signed-off-by: Sylwester Nawrocki 
Signed-off-by: Mauro Carvalho Chehab 
Signed-off-by: Greg Kroah-Hartman

v4l: vsp1: Prevent pipelines from running when not streaming

2016-11-10T15:38:57+00:00

commit e4e70a147a48618a36ae1b81c641516cb9d45993 upstream.

Pipelines can only be run if all their video nodes are streaming. Commit
b4dfb9b35a19 ("[media] v4l: vsp1: Stop the pipeline upon the first
STREAMOFF") fixed the pipeline stop sequence, but introduced a race
condition that makes it possible to run a pipeline after stopping the
stream on a video node by queuing a buffer on the other side of the
pipeline.

Fix this by clearing the buffers ready flag when stopping the stream,
which will prevent the QBUF handler from running the pipeline.

Fixes: b4dfb9b35a19 ("[media] v4l: vsp1: Stop the pipeline upon the first STREAMOFF")

Reported-by: Kieran Bingham 
Tested-by: Kieran Bingham 
Signed-off-by: Laurent Pinchart 
Signed-off-by: Mauro Carvalho Chehab 
Signed-off-by: Greg Kroah-Hartman

[media] rcar-fcp: Make sure rcar_fcp_enable() returns 0 on success

2016-09-15T12:02:16+00:00

When resuming from suspend-to-RAM on r8a7795/salvator-x:

    dpm_run_callback(): pm_genpd_resume_noirq+0x0/0x90 returns 1
    PM: Device fe940000.fdp1 failed to resume noirq: error 1
    dpm_run_callback(): pm_genpd_resume_noirq+0x0/0x90 returns 1
    PM: Device fe944000.fdp1 failed to resume noirq: error 1
    dpm_run_callback(): pm_genpd_resume_noirq+0x0/0x90 returns 1
    PM: Device fe948000.fdp1 failed to resume noirq: error 1

According to its documentation, rcar_fcp_enable() returns 0 on success
or a negative error code if an error occurs.  Hence
fdp1_pm_runtime_resume() and vsp1_pm_runtime_resume() forward its return
value to their callers.

However, rcar_fcp_enable() forwards the return value of
pm_runtime_get_sync(), which can actually be 1 on success, leading to
the resume failure above.

To fix this, consider only negative values returned by
pm_runtime_get_sync() to be failures.

Fixes: 7b49235e83b2347c ("[media] v4l: Add Renesas R-Car FCP driver")

Signed-off-by: Geert Uytterhoeven 
Reviewed-by: Laurent Pinchart 
Signed-off-by: Mauro Carvalho Chehab

[media] vcodec:mediatek: Refine VP8 encoder driver

2016-08-24T14:28:36+00:00

This patch remove field and function that unused anymore

Signed-off-by: Tiffany Lin 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab

[media] vcodec:mediatek: Refine H264 encoder driver

2016-08-24T14:28:11+00:00

This patch :
1. remove field and function that unused anymore
2. add support V4L2_MPEG_VIDEO_H264_LEVEL_4_2

Signed-off-by: Tiffany Lin 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab

[media] vcodec:mediatek: change H264 profile default to profile high

2016-08-24T14:27:38+00:00

This patch change default H264 profile from V4L2_MPEG_VIDEO_H264_PROFILE_MAIN
to V4L2_MPEG_VIDEO_H264_PROFILE_HIGH

Signed-off-by: Tiffany Lin 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab

[media] vcodec:mediatek: Add timestamp and timecode copy for V4L2 Encoder

2016-08-24T14:27:10+00:00

This patch add copying timestamp and timecode from src buffer to dst buffer

Signed-off-by: Tiffany Lin 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab

[media] vcodec:mediatek: Fix visible_height larger than coded_height issue in s_fmt_out

2016-08-24T14:26:41+00:00

The original code add extra 32 line to visible_height.
It is incorrect, 32 line should be add to coded_height.
The purpose is that user space could calcuate real buffer size needed by using
coded_width * coded_height.
But this method will make v4l2-compliance test fail, since g_fmt != s_fmt(g_fmt)
So remove extend visible_height or coded_height, user space should just
use sizeimage to get real buffer size needed

Signed-off-by: Tiffany Lin 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab

[media] vcodec:mediatek: Fix fops_vcodec_release flow for V4L2 Encoder

2016-08-24T14:21:43+00:00

This patch fix that mtk_vcodec_venc_release should be called after v4l2_m2m_ctx_release

Signed-off-by: Tiffany Lin 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab

[media] vcodec:mediatek:code refine for v4l2 Encoder driver

2016-08-24T14:20:50+00:00

This patch remove unused header and define from haeder files

Signed-off-by: Tiffany Lin 
Signed-off-by: Hans Verkuil 
Signed-off-by: Mauro Carvalho Chehab