linux.git/drivers/media/platform/mediatek, branch for-next

media: mediatek: vcodec: mark vdec_vp9_slice_map_counts_eob_coef noinline

2024-12-13T16:51:35+00:00

With KASAN enabled, clang fails to optimize the inline version of
vdec_vp9_slice_map_counts_eob_coef() properly, leading to kilobytes
of temporary values spilled to the stack:

drivers/media/platform/mediatek/vcodec/decoder/vdec/vdec_vp9_req_lat_if.c:1526:12: error: stack frame size (2160) exceeds limit (2048) in 'vdec_vp9_slice_update_prob' [-Werror,-Wframe-larger-than]

This seems to affect all versions of clang including the latest (clang-20),
but the degree of stack overhead is different per release.

Marking the function as noinline_for_stack is harmless here and avoids
the problem completely.

Signed-off-by: Arnd Bergmann 
Reviewed-by: Nathan Chancellor 
Signed-off-by: Sebastian Fricke 
Signed-off-by: Mauro Carvalho Chehab

media: platform: drop vb2_ops_wait_prepare/finish

2024-10-28T08:20:07+00:00

Since commit 88785982a19d ("media: vb2: use lock if wait_prepare/finish
are NULL") it is no longer needed to set the wait_prepare/finish
vb2_ops callbacks as long as the lock field in vb2_queue is set.

Since the vb2_ops_wait_prepare/finish callbacks already rely on that field,
we can safely drop these callbacks.

This simplifies the code and this is a step towards the goal of deleting
these callbacks.

Signed-off-by: Hans Verkuil 
Reviewed-by: Neil Armstrong  # for meson-ge2d
Acked-by: Andrzej Pietrasiewicz

media: Switch back to struct platform_driver::remove()

2024-10-12T14:28:25+00:00

After commit 0edb555a65d1 ("platform: Make platform_driver::remove()
return void") .remove() is (again) the right callback to implement for
platform drivers.

Convert all platform drivers below drivers/media to use .remove(), with
the eventual goal to drop struct platform_driver::remove_new(). As
.remove() and .remove_new() have the same prototypes, conversion is done
by just changing the structure member name in the driver initializer.

Signed-off-by: Uwe Kleine-König 
Signed-off-by: Hans Verkuil

media: mtk-jpeg: Fix null-ptr-deref during unload module

2024-10-12T14:28:25+00:00

The workqueue should be destroyed in mtk_jpeg_core.c since commit
09aea13ecf6f ("media: mtk-jpeg: refactor some variables"), otherwise
the below calltrace can be easily triggered.

[  677.862514] Unable to handle kernel paging request at virtual address dfff800000000023
[  677.863633] KASAN: null-ptr-deref in range [0x0000000000000118-0x000000000000011f]
...
[  677.879654] CPU: 6 PID: 1071 Comm: modprobe Tainted: G           O       6.8.12-mtk+gfa1a78e5d24b+ #17
...
[  677.882838] pc : destroy_workqueue+0x3c/0x770
[  677.883413] lr : mtk_jpegdec_destroy_workqueue+0x70/0x88 [mtk_jpeg_dec_hw]
[  677.884314] sp : ffff80008ad974f0
[  677.884744] x29: ffff80008ad974f0 x28: ffff0000d7115580 x27: ffff0000dd691070
[  677.885669] x26: ffff0000dd691408 x25: ffff8000844af3e0 x24: ffff80008ad97690
[  677.886592] x23: ffff0000e051d400 x22: ffff0000dd691010 x21: dfff800000000000
[  677.887515] x20: 0000000000000000 x19: 0000000000000000 x18: ffff800085397ac0
[  677.888438] x17: 0000000000000000 x16: ffff8000801b87c8 x15: 1ffff000115b2e10
[  677.889361] x14: 00000000f1f1f1f1 x13: 0000000000000000 x12: ffff7000115b2e4d
[  677.890285] x11: 1ffff000115b2e4c x10: ffff7000115b2e4c x9 : ffff80000aa43e90
[  677.891208] x8 : 00008fffeea4d1b4 x7 : ffff80008ad97267 x6 : 0000000000000001
[  677.892131] x5 : ffff80008ad97260 x4 : ffff7000115b2e4d x3 : 0000000000000000
[  677.893054] x2 : 0000000000000023 x1 : dfff800000000000 x0 : 0000000000000118
[  677.893977] Call trace:
[  677.894297]  destroy_workqueue+0x3c/0x770
[  677.894826]  mtk_jpegdec_destroy_workqueue+0x70/0x88 [mtk_jpeg_dec_hw]
[  677.895677]  devm_action_release+0x50/0x90
[  677.896211]  release_nodes+0xe8/0x170
[  677.896688]  devres_release_all+0xf8/0x178
[  677.897219]  device_unbind_cleanup+0x24/0x170
[  677.897785]  device_release_driver_internal+0x35c/0x480
[  677.898461]  device_release_driver+0x20/0x38
...
[  677.912665] ---[ end trace 0000000000000000 ]---

Fixes: 09aea13ecf6f ("media: mtk-jpeg: refactor some variables")
Cc: 
Signed-off-by: Guoqing Jiang 
Signed-off-by: Hans Verkuil

media: mediatek: vcodec: Constify struct vb2_ops

2024-08-28T08:01:10+00:00

"struct vb2_ops" are not modified in this driver.

Constifying this structure moves some data to a read-only section, so
increase overall security.

On a x86_64, with allmodconfig:
Before:
======
   text	   data	    bss	    dec	    hex	filename
  18059	   3096	     16	  21171	   52b3	drivers/media/platform/mediatek/vcodec/decoder/mtk_vcodec_dec_stateful.o

After:
=====
   text	   data	    bss	    dec	    hex	filename
  18171	   2968	     16	  21155	   52a3	drivers/media/platform/mediatek/vcodec/decoder/mtk_vcodec_dec_stateful.o

Signed-off-by: Christophe JAILLET 
Signed-off-by: Sebastian Fricke 
Signed-off-by: Hans Verkuil

media: mediatek: vcodec: Fix H264 stateless decoder smatch warning

2024-08-28T08:01:09+00:00

Fix a smatch static checker warning on vdec_h264_req_if.c.
Which leads to a kernel crash when fb is NULL.

Fixes: 06fa5f757dc5 ("media: mtk-vcodec: vdec: support stateless H.264 decoding")
Signed-off-by: Yunfei Dong 
Reviewed-by: AngeloGioacchino Del Regno 
Signed-off-by: Sebastian Fricke 
Signed-off-by: Hans Verkuil

media: mediatek: vcodec: Fix VP8 stateless decoder smatch warning

2024-08-28T08:01:09+00:00

Fix a smatch static checker warning on vdec_vp8_req_if.c.
Which leads to a kernel crash when fb is NULL.

Fixes: 7a7ae26fd458 ("media: mediatek: vcodec: support stateless VP8 decoding")
Signed-off-by: Yunfei Dong 
Reviewed-by: AngeloGioacchino Del Regno 
Signed-off-by: Sebastian Fricke 
Signed-off-by: Hans Verkuil

media: mediatek: vcodec: Fix H264 multi stateless decoder smatch warning

2024-08-28T08:01:09+00:00

Fix a smatch static checker warning on vdec_h264_req_multi_if.c.
Which leads to a kernel crash when fb is NULL.

Fixes: 397edc703a10 ("media: mediatek: vcodec: add h264 decoder driver for mt8186")
Signed-off-by: Yunfei Dong 
Reviewed-by: AngeloGioacchino Del Regno 
Signed-off-by: Sebastian Fricke 
Signed-off-by: Hans Verkuil

media: platform: mtk-mdp3: Add support for MT8188 MDP3 components

2024-06-20T09:47:48+00:00

MT8195 and MT8188 share a similar MDP3 macro-block, with minor
differences - as in, the latter supports a subset of the number
of components supported by the former, but are otherwise handled
in the same way.

Add driver data for MT8188, reusing the already present MT8195
data where possible.

Signed-off-by: AngeloGioacchino Del Regno 
Reviewed-by: Fei Shao 
Signed-off-by: Sebastian Fricke 
Signed-off-by: Hans Verkuil

media: mediatek: vcodec: Alloc DMA memory with DMA_ATTR_ALLOC_SINGLE_PAGES

2024-06-10T08:24:11+00:00

As talked about in commit 14d3ae2efeed ("ARM: 8507/1: dma-mapping: Use
DMA_ATTR_ALLOC_SINGLE_PAGES hint to optimize alloc"), it doesn't
really make sense to try to allocate contiguous chunks of memory for
video encoding/decoding. Let's switch the Mediatek vcodec driver to
pass DMA_ATTR_ALLOC_SINGLE_PAGES and take some of the stress off the
memory subsystem.

Signed-off-by: Douglas Anderson 
Tested-by: Fei Shao 
Reviewed-by: Fei Shao 
Reviewed-by: Nicolas Dufresne 
Signed-off-by: Sebastian Fricke 
Signed-off-by: Hans Verkuil