linux-dev/drivers/media/common/saa7146, branch master

media: saa7146: deprecate hexium_gemini/orion, mxb and ttpci

2022-08-29T14:46:38Z

Deprecate the hexium_gemini, hexium_orion, mxb and ttpci saa7146-based drivers: these drivers do not use the vb2 framework for video streaming, instead it uses the old videobuf framework. We want to get rid of these old drivers, so deprecated these for future removal. [hverkuil: update MAINTAINERS file] Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: v4l: ioctl: Set bus_info in v4l_querycap()

2022-04-24T07:07:08Z

The bus_info field is set by most drivers based on the type of the device bus as well as the name of the device. Do this in v4l_querycap() so drivers don't need to. This keeps compatibility with non-default and silly bus_info. Signed-off-by: Sakari Ailus Reviewed-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab

media: Kconfig: cleanup VIDEO_DEV dependencies

2022-03-18T04:58:35Z

media Kconfig has two entries associated to V4L API: VIDEO_DEV and VIDEO_V4L2. On Kernel 2.6.x, there were two V4L APIs, each one with its own flag. VIDEO_DEV were meant to: 1) enable Video4Linux and make its Kconfig options to appear; 2) it makes the Kernel build the V4L core. while VIDEO_V4L2 where used to distinguish between drivers that implement the newer API and drivers that implemented the former one. With time, such meaning changed, specially after the removal of all V4L version 1 drivers. At the current implementation, VIDEO_DEV only does (1): it enables the media options related to V4L, that now has: menu "Video4Linux options" visible if VIDEO_DEV source "drivers/media/v4l2-core/Kconfig" endmenu but it doesn't affect anymore the V4L core drivers. The rationale is that the V4L2 core has a "soft" dependency at the I2C bus, and now requires to select a number of other Kconfig options: config VIDEO_V4L2 tristate depends on (I2C || I2C=n) && VIDEO_DEV select RATIONAL select VIDEOBUF2_V4L2 if VIDEOBUF2_CORE default (I2C || I2C=n) && VIDEO_DEV In the past, merging them would be tricky, but it seems that it is now possible to merge those symbols, in order to simplify V4L dependencies. Let's keep VIDEO_DEV, as this one is used on some make *defconfig configurations. Suggested-by: Laurent Pinchart Reviewed-by: Laurent Pinchart Reviewed-by: Jacopo Mondi Reviewed-by: Neil Armstrong # for meson-vdec & meson-ge2d Acked-by: Andrzej Pietrasiewicz Acked-by: Łukasz Stelmach Signed-off-by: Mauro Carvalho Chehab

media: saa7146: fix error logic at saa7146_vv_init()

2021-12-14T14:09:14Z

As the first thing this function does is to call v4l2_device_register(), it should call v4l2_device_unregister() if an error occurs, the same way as done at saa7146_vv_release(). Signed-off-by: Mauro Carvalho Chehab

media: saa7146: hexium_gemini: Fix a NULL pointer dereference in hexium_attach()

2021-12-14T14:09:14Z

In hexium_attach(dev, info), saa7146_vv_init() is called to allocate a new memory for dev->vv_data. saa7146_vv_release() will be called on failure of saa7146_register_device(). There is a dereference of dev->vv_data in saa7146_vv_release(), which could lead to a NULL pointer dereference on failure of saa7146_vv_init(). Fix this bug by adding a check of saa7146_vv_init(). This bug was found by a static analyzer. The analysis employs differential checking to identify inconsistent security operations (e.g., checks or kfrees) between two code paths and confirms that the inconsistent operations are not recovered in the current function or the callers, so they constitute bugs. Note that, as a bug found by static analysis, it can be a false positive or hard to trigger. Multiple researchers have cross-reviewed the bug. Builds with CONFIG_VIDEO_HEXIUM_GEMINI=m show no new warnings, and our static analyzer no longer warns about this code. Link: https://lore.kernel.org/linux-media/20211203154030.111210-1-zhou1615@umn.edu Signed-off-by: Zhou Qingyang Signed-off-by: Mauro Carvalho Chehab

media: saa7146: use sg_dma_len when building pgtable

2021-03-11T10:59:44Z

The new AMD IOMMU DMA implementation concatenates sglist entries under certain conditions, and because saa7146 accessed the length member directly, it did not support this scenario. This fixes IO_PAGE_FAULTs by using the sg_dma_len macro. Fixes: be62dbf554c5 ("iommu/amd: Convert AMD iommu driver to the dma-iommu api") Signed-off-by: Tasos Sahanidis Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: saa7146: switch from 'pci_' to 'dma_' API

2020-12-03T07:57:07Z

The wrappers in include/linux/pci-dma-compat.h should go away. The patch has been generated with the coccinelle script below and has been hand modified to replace GFP_ with a correct flag. It has been compile tested. When memory is allocated in 'saa7146_pgtable_alloc()' GFP_KERNEL can be used because the callers are either .buf_prepare functions or function that already use GFP_KERNEL (hidden in a 'vmalloc_32()' call). When memory is allocated in 'saa7146_init_one()' GFP_KERNEL can be used because it is probe function and no lock is taken in the between. When memory is allocated in 'saa7146_vv_init()' GFP_KERNEL can be used because this function already uses GFP_KERNEL and no lock is taken in the between. When memory is allocated in 'vbi_workaround()' GFP_KERNEL can be used because it is only called from a .open function. @@ @@ - PCI_DMA_BIDIRECTIONAL + DMA_BIDIRECTIONAL @@ @@ - PCI_DMA_TODEVICE + DMA_TO_DEVICE @@ @@ - PCI_DMA_FROMDEVICE + DMA_FROM_DEVICE @@ @@ - PCI_DMA_NONE + DMA_NONE @@ expression e1, e2, e3; @@ - pci_alloc_consistent(e1, e2, e3) + dma_alloc_coherent(&e1->dev, e2, e3, GFP_) @@ expression e1, e2, e3; @@ - pci_zalloc_consistent(e1, e2, e3) + dma_alloc_coherent(&e1->dev, e2, e3, GFP_) @@ expression e1, e2, e3, e4; @@ - pci_free_consistent(e1, e2, e3, e4) + dma_free_coherent(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_map_single(e1, e2, e3, e4) + dma_map_single(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_unmap_single(e1, e2, e3, e4) + dma_unmap_single(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4, e5; @@ - pci_map_page(e1, e2, e3, e4, e5) + dma_map_page(&e1->dev, e2, e3, e4, e5) @@ expression e1, e2, e3, e4; @@ - pci_unmap_page(e1, e2, e3, e4) + dma_unmap_page(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_map_sg(e1, e2, e3, e4) + dma_map_sg(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_unmap_sg(e1, e2, e3, e4) + dma_unmap_sg(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_dma_sync_single_for_cpu(e1, e2, e3, e4) + dma_sync_single_for_cpu(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_dma_sync_single_for_device(e1, e2, e3, e4) + dma_sync_single_for_device(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_dma_sync_sg_for_cpu(e1, e2, e3, e4) + dma_sync_sg_for_cpu(&e1->dev, e2, e3, e4) @@ expression e1, e2, e3, e4; @@ - pci_dma_sync_sg_for_device(e1, e2, e3, e4) + dma_sync_sg_for_device(&e1->dev, e2, e3, e4) @@ expression e1, e2; @@ - pci_dma_mapping_error(e1, e2) + dma_mapping_error(&e1->dev, e2) @@ expression e1, e2; @@ - pci_set_dma_mask(e1, e2) + dma_set_mask(&e1->dev, e2) @@ expression e1, e2; @@ - pci_set_consistent_dma_mask(e1, e2) + dma_set_coherent_mask(&e1->dev, e2) Signed-off-by: Christophe JAILLET Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: Bulk remove BUG_ON(in_interrupt())

2020-11-16T09:31:10Z

None of these BUG_ON()'s is justified. BUG_ON() should only be used when there is really no way to survive. If at all these could be replaced by lockdep_assert_preemption_enabled() to cover all invalid caller context and not just those covered by in_interrupt(). But all functions which are invoked from those places contain already debug mechanisms to catch wrong context, so having these extra checks is not having any advantage. Signed-off-by: Thomas Gleixner Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: v4l2: allocate v4l2_clip objects early

2020-11-16T09:31:06Z

The v4l2_format based ioctls can have an indirect pointer to an array of v4l2_clip structures for overlay mode, depending on the 'type' member. There are only five drivers that use the overlay mode and copy the data through the __user pointer. Change the five drivers to use memcpy() instead, and copy the data in common code using the check_array_args() helpers. This allows for a subsequent patch that use the same mechanism for compat ioctl handlers. Note that there is another pointer for a 'bitmap' that is only used in the 'vivid' driver and nowhere else. There is no easy way to use the same trick without adding complexity to the common code, so this remains a __user pointer. [hverkuil: fix: CHECK: spaces preferred around that '*' (ctx:VxV)] [hverkuil: fix: CHECK: Alignment should match open parenthesis] Signed-off-by: Arnd Bergmann Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: saa7146: drop double zeroing

2020-09-27T09:30:13Z

sg_init_table zeroes its first argument, so the allocation of that argument doesn't have to. the semantic patch that makes this change is as follows: (http://coccinelle.lip6.fr/) // @@ expression x,n,flags; @@ x = - kcalloc + kmalloc_array (n,sizeof(struct scatterlist),flags) ... sg_init_table(x,n) // Signed-off-by: Julia Lawall Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab