summaryrefslogtreecommitdiff
path: root/drivers/media/platform
AgeCommit message (Collapse)AuthorFilesLines
2018-08-03media: vsp1: Clean up DLM objects on errorKieran Bingham1-1/+3
If there is an error allocating a display list within a DLM object the existing display lists are not free'd, and neither is the DL body pool. Use the existing vsp1_dlm_destroy() function to clean up on error. Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-03media: vsp1: Remove unused display list structure fieldKieran Bingham1-2/+0
The vsp1 reference in the vsp1_dl_body structure is not used. Remove it. Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-03media: vsp1: Rename dl_child to dl_nextKieran Bingham1-7/+7
Both vsp1_dl_list_commit() and __vsp1_dl_list_put() walk the display list chain referencing the nodes as children, when in reality they are siblings. Update the terminology to 'dl_next' to be consistent with the vsp1_video_pipeline_run() usage. Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-03media: vsp1: use kernel __packed for structuresKieran Bingham1-3/+3
The kernel provides a __packed definition to abstract away from the compiler specific attributes tag. Convert all packed structures in VSP1 to use it. The GCC documentation [0] describes this attribute as "the structure or union is placed to minimize the memory required". The Keil compiler documentation at [1] warns that the use of this attribute can cause a performance penalty in the event that the compiler can not deduce the allignment of each field. Careful examination of the object code generated both with and without this attribute shows that these structures are accessed identically and are not affected by any performance penalty. The structures are correctly aligned and padded to match the needs of the hardware already. This patch does not serve to make a decision as to the use of the attribute, but purely to clean up the code to use the kernel defined abstraction as per [2]. [0] https://gcc.gnu.org/onlinedocs/gcc/Common-Type-Attributes.html#index-packed-type-attribute [1] http://www.keil.com/support/man/docs/armcc/armcc_chr1359124230195.htm [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/include/linux/compiler-gcc.h?h=v4.16-rc5#n92 Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-03media: vsp1: drm: Fix minor grammar errorKieran Bingham1-1/+1
The pixel format is 'unsupported'. Fix the small debug message which incorrectly declares this. Signed-off-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-03media: davinci: vpif_display: Mix memory leak on probe error pathAnton Vasilyev1-8/+16
If vpif_probe() fails on v4l2_device_register() then memory allocated at initialize_vpif() for global vpif_obj.dev[i] become unreleased. The patch adds deallocation of vpif_obj.dev[i] on the error path and removes duplicated check on platform_data presence. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by: Anton Vasilyev <vasilyev@ispras.ru> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-03media: vimc: Remove redundant freeAnton Vasilyev1-1/+0
Commit 4a29b7090749 ("[media] vimc: Subdevices as modules") removes vimc allocation from vimc_probe(), so corresponding deallocation on the error path tries to free static memory. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by: Anton Vasilyev <vasilyev@ispras.ru> Acked-by: Helen Koike <helen.koike@collabora.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-03media: headers: fix linux/mod_devicetable.h inclusionsArnd Bergmann5-0/+5
A couple of drivers produced build errors after the mod_devicetable.h header was split out from the platform_device one, e.g. drivers/media/platform/davinci/vpbe_osd.c:42:40: error: array type has incomplete element type 'struct platform_device_id' drivers/media/platform/davinci/vpbe_venc.c:42:40: error: array type has incomplete element type 'struct platform_device_id' This adds the inclusion where needed. Fixes: ac3167257b9f ("headers: separate linux/mod_devicetable.h from linux/platform_device.h") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Randy Dunlap <rdunlap@infradead.org> Acked-by: Lad, Prabhakar <prabhakar.csengg@gmail.com> Acked-by: Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csid: Add support for events triggered by user controlsTodor Tomov1-1/+5
Changing a user control value can trigger an event to other users. Add support for that. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Acked-by: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Add support for 10-bit grayscale formatsTodor Tomov5-11/+56
Add support for 10-bit packed V4L2_PIX_FMT_Y10P (on 8x16 and 8x96) and unpacked V4L2_PIX_FMT_Y10 (on 8x96 only) pixel formats. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Add support for RAW MIPI14 on 8x96Todor Tomov5-0/+50
Add support for RAW MIPI14 format for RDI mode on 8x96. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csid: MIPI10 to Plain16 format conversionTodor Tomov4-1/+99
Use the PRDI mode on 8x96 to allow to configure RAW MIPI10 to Plain16 format conversion. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csid: Different format support on source padTodor Tomov1-13/+56
Usually the format on the source pad is the same as on the sink pad. However the CSID is able to do some format conversions. To support this make the format on the source pad selectable amongst a list of formats. This list can be different for each sink pad format. This is still not used but will be when the format conversions are implemented. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: vfe: Add support for UYVY output from VFE on 8x96Todor Tomov5-24/+152
Add support to output UYVY formats from the VFE (via the PIX interface). A configuration for the realign module in the VFE is added. As the realign module is present on 8x96 but not on 8x16, this is supported on 8x96 only. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: vfe: Different format support on source padTodor Tomov1-37/+135
Rework the format selection on the source pad. Make the format on the source pad selectable amongst a list of formats. This list can be different for each sink pad format. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Format configuration per hardware versionTodor Tomov9-211/+467
As the 8x16 and 8x96 support different formats, separate the arrays which contain the supported formats. For the VFE also add separate arrays for RDI and PIX subdevices. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: vfe: Add support for 8x96Todor Tomov5-0/+1064
Add VFE hardware dependent part for 8x96. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: vfe: Split to hardware dependent and independent partsTodor Tomov4-985/+1169
This will allow to add support for different hardware. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: ispif: Add support for 8x96Todor Tomov1-3/+73
ISPIF hardware modules on 8x16 and 8x96 are similar. However on 8x96 the ISPIF routes data to two VFE hardware modules. Add separate interrupt handler for 8x96 to handle the additional interrupts. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csid: Add support for 8x96Todor Tomov1-23/+37
CSID hardware modules on 8x16 and 8x96 are similar. There is no need to duplicate the code by adding separate versions. Just update the register macros to return the correct register addresses. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csiphy: Add support for 8x96Todor Tomov4-0/+260
Add CSIPHY hardware dependent part for 8x96. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csiphy: Unify lane handlingTodor Tomov3-26/+29
Restructure lane configuration so it is simpler and will allow similar (although not the same) handling for different hardware versions. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csiphy: Split to hardware dependent and independent partsTodor Tomov4-148/+213
This will allow to add support for different hardware. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> [hans.verkuil@cisco.com: remove trailing empty line] Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Add basic runtime PM supportTodor Tomov6-6/+139
There is a PM domain for each of the VFE hardware modules. Add support for basic runtime PM support to be able to control the PM domains. When a PM domain needs to be powered on - a device link is created. When a PM domain needs to be powered off - its device link is removed. This allows separate and independent control of the PM domains. Suspend/Resume is still not supported. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Add 8x96 resourcesTodor Tomov10-139/+390
Add structs with 8x96 resources. As the number of CSIPHY, CSID and VFE hardware modules is different on 8x16 and 8x96 select the number at runtime and allocate needed structures dynamically. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: vfe: Do not disable CAMIF when clearing its statusTodor Tomov1-1/+4
Use "no change" value when clearing CAMIF status and make sure this is done before configuring the new command. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: vfe: Get line pointer as container of video_outTodor Tomov1-34/+4
Simplify getting of the line pointer by using the container_of macro instead of traversing media controller links. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: vfe: Fix to_vfe() macro member nameTodor Tomov1-1/+1
Use the member name which is "line" instead of the pointer argument. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csid: Configure data type and decode format properlyTodor Tomov1-5/+11
The CSID decodes the input data stream. When the input comes from the Test Generator the format of the stream is set on the source media pad. When the input comes from the CSIPHY the format is the one on the sink media pad. Use the proper format for each case. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csiphy: Update settle count calculationTodor Tomov1-1/+1
Update settle count calculation as per specification. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Unify the clock namesTodor Tomov1-12/+8
Use more logical clock names - similar to the names in documentation. This will allow better handling of the clocks in the driver when support for more hardware versions is added - equivalent clocks on different hardware versions will have the same name. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: csiphy: Ensure clock mux config is done before the restTodor Tomov1-0/+1
Add a write memory barier after clock mux config and before the rest of the csiphy config. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Fix OF node usageTodor Tomov1-2/+3
of_graph_get_next_endpoint increases the refcount of the returned node and decreases the refcount of the passed node. Take this into account and use of_node_put properly. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: camss: Use SPDX license headersTodor Tomov12-108/+12
Use SPDX license headers for all files of the Qualcomm CAMSS driver. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-08-02media: Rename CAMSS driver pathTodor Tomov15-14/+14
Support for camera subsystem on QComm MSM8996/APQ8096 is to be added so remove hardware version from CAMSS driver's path. Signed-off-by: Todor Tomov <todor.tomov@linaro.org> Signed-off-by: Hans Verkuil <hansverk@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-30media: v4l: vsp1: Fix deadlock in VSPDL DRM pipelinesLaurent Pinchart1-3/+1
The VSP uses a lock to protect the BRU and BRS assignment when configuring pipelines. The lock is taken in vsp1_du_atomic_begin() and released in vsp1_du_atomic_flush(), as well as taken and released in vsp1_du_setup_lif(). This guards against multiple pipelines trying to assign the same BRU and BRS at the same time. The DRM framework calls the .atomic_begin() operations in a loop over all CRTCs included in an atomic commit. On a VSPDL (the only VSP type where this matters), a single VSP instance handles two CRTCs, with a single lock. This results in a deadlock when the .atomic_begin() operation is called on the second CRTC. The DRM framework serializes atomic commits that affect the same CRTCs, but doesn't know about two CRTCs sharing the same VSPDL. Two commits affecting the VSPDL LIF0 and LIF1 respectively can thus race each other, hence the need for a lock. This could be fixed on the DRM side by forcing serialization of commits affecting CRTCs backed by the same VSPDL, but that would negatively affect performances, as the locking is only needed when the BRU and BRS need to be reassigned, which is an uncommon case. The lock protects the whole .atomic_begin() to .atomic_flush() sequence. The only operation that can occur in-between is vsp1_du_atomic_update(), which doesn't touch the BRU and BRS, and thus doesn't need to be protected by the lock. We can thus only take the lock around the pipeline setup calls in vsp1_du_atomic_flush(), which fixes the deadlock. Fixes: f81f9adc4ee1 ("media: v4l: vsp1: Assign BRU and BRS to pipelines dynamically") Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: s5p-mfc: Fix buffer look up in s5p_mfc_handle_frame_{new, copy_time} ↵Sylwester Nawrocki1-11/+12
functions Look up of buffers in s5p_mfc_handle_frame_new, s5p_mfc_handle_frame_copy_time functions is not working properly for DMA addresses above 2 GiB. As a result flags and timestamp of returned buffers are not set correctly and it breaks operation of GStreamer/OMX plugins which rely on the CAPTURE buffer queue flags. Due to improper return type of the get_dec_y_adr, get_dspl_y_adr callbacks and sign bit extension these callbacks return incorrect address values, e.g. 0xfffffffffefc0000 instead of 0x00000000fefc0000. Then the statement: "if (vb2_dma_contig_plane_dma_addr(&dst_buf->b->vb2_buf, 0) == dec_y_addr)" is always false, which breaks looking up capture queue buffers. To ensure proper matching by address u32 type is used for the DMA addresses. This should work on all related SoCs, since the MFC DMA address width is not larger than 32-bit. Changes done in this patch are minimal as there is a larger patch series pending refactoring the whole driver. Signed-off-by: Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: rcar-csi2: update stream start for V3MNiklas Söderlund1-6/+14
Latest errata document updates the start procedure for V3M. This change in addition to adhering to the datasheet update fixes capture on early revisions of V3M. Signed-off-by: Niklas Söderlund <niklas.soderlund+renesas@ragnatech.se> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: rockchip/rga: Fix bad dma_free_attrs() parameterEzequiel Garcia1-1/+1
In rga_remove(), dma_free_attrs is being passed the wrong cpu address, which triggers an exception if the driver is removed. Fix it. Tested on a RK3399 platform, with a bind/unbind cycle. Signed-off-by: Ezequiel Garcia <ezequiel@collabora.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: vivid: Fix V4L2_FIELD_ALTERNATE new frame checkNicolas Dufresne1-1/+1
The vivid driver will overlay stream time on generated frames. Though, in interlacing mode V4L2_FIELD_ALTERNATE, each field is separate and must have the same time to ensure proper render. Though, this time was only updated every 2 frames as the code was checking against the wrong counter (frame counter rather then field counter). Signed-off-by: Nicolas Dufresne <nicolas.dufresne@collabora.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: coda: add SPS fixup code for frame sizes that are not multiples of 16Philipp Zabel3-0/+339
The CODA7541 firmware does not set the SPS frame cropping fields to properly describe coded h.264 streams with frame sizes that are not a multiple of the macroblock size. This adds RBSP parsing code and a SPS fixup routine to manually replace the cropping information in the headers produced by the firmware with the correct values. Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de> [hans.verkuil@cisco.com: added explanation of SPS RBSP to comment] Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: coda: let CODA960 firmware set frame cropping in SPS headerPhilipp Zabel1-0/+19
When encoding h.264, if visible resolution is not aligned to macroblock size, frame cropping has to be set in the SPS header to produce correct streams. The CODA960 firmware can do this on its own if asked to. Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: vsp1: Document vsp1_dl_body refcntKieran Bingham1-0/+1
In commit 2d9445db0ee9 ("media: vsp1: Use reference counting for bodies"), a new field was introduced to the vsp1_dl_body structure to account for usage tracking of the body. Document the newly added field in the kerneldoc. Signed-off-by: Kieran Bingham <kieran.bingham@ideasonboard.com> Reviewed-by: Simon Horman <horms+renesas@verge.net.au> Reviewed-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-27media: vicodec: current -> curHans Verkuil1-2/+2
'current' is also defined in asm-generic/current.h. When compiling this driver for older kernels with the media_build system, this header is included via compat.h and it no longer compiles. Rename current to cur. Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-25media: venus: add HEVC codec supportStanimir Varbanov5-0/+60
This add HEVC codec support for venus versions 3xx and 4xx. Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org> Reviewed-by: Tomasz Figa <tfiga@chromium.org> Reviewed-by: Alexandre Courbot <acourbot@chromium.org> Tested-by: Alexandre Courbot <acourbot@chromium.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-25media: venus: core: add sdm845 DT compatible and resource dataStanimir Varbanov1-0/+22
This adds sdm845 DT compatible string with it's resource data table. Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org> Reviewed-by: Rob Herring <robh@kernel.org> Reviewed-by: Tomasz Figa <tfiga@chromium.org> Reviewed-by: Alexandre Courbot <acourbot@chromium.org> Tested-by: Alexandre Courbot <acourbot@chromium.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-25media: venus: implementing multi-stream supportStanimir Varbanov5-4/+300
This is implementing multi-stream decoder support. The multi-stream will be used to enable/disable the primary/secondary decoder outputs. Depending on formats on both decoder outputs we could implement downscale, dithering and supporting UBWC (universal bandwidth compression) formats. The UBWC compressed raw format is used to optimize interconnect bandwidth for bigger resolutions like 4K and hence we will get some power-saving benefits as well. Both decoder outputs are distinguished by buffer_type field in the HFI packets. For example HFI_BUFFER_OUTPUT is the buffer type for primary decoder output and HFI_BUFFER_OUTPUT2 is for secondary decoder output. Starting from Venus 4xx the DPB buffers format must be UBWC, so the multi-stream becomes mandatory for this Venus version. That means that we need to allocate internally in the driver a set of DPB buffers (with UBWC NV12 format) and give them to the firmware. The other decoder output (we called it OPB) format will be NV12 linear format and with the same resolution (or smaller in case the user wants to downscale). The DPB buffers are used for decoder reference frames and those have to be in a specific format (UBWC). So one decoder output is used to fill those reference buffers while the other output is used to fill the userspace buffers with the user requested format. Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org> Reviewed-by: Tomasz Figa <tfiga@chromium.org> Reviewed-by: Alexandre Courbot <acourbot@chromium.org> Tested-by: Alexandre Courbot <acourbot@chromium.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-25media: venus: helpers: move frame size calculations on common placeStanimir Varbanov4-84/+126
This move the calculations of raw and compressed buffer sizes on common helper and make it identical for encoder and decoder. Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org> Reviewed-by: Tomasz Figa <tfiga@chromium.org> Reviewed-by: Alexandre Courbot <acourbot@chromium.org> Tested-by: Alexandre Courbot <acourbot@chromium.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-25media: venus: vdec: a new function for output configurationStanimir Varbanov1-15/+19
Make a new function vdec_output_conf() for decoder output configuration. vdec_output_conf() will set properties via HFI interface related to the output configuration, and keep vdec_set_properties() which will set properties related to decoding parameters. Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org> Reviewed-by: Tomasz Figa <tfiga@chromium.org> Reviewed-by: Alexandre Courbot <acourbot@chromium.org> Tested-by: Alexandre Courbot <acourbot@chromium.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
2018-07-25media: venus: vdec: get required input buffers as wellStanimir Varbanov1-17/+24
Rework and rename vdec_cap_num_buffers() to get the number of input buffers too. Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org> Reviewed-by: Tomasz Figa <tfiga@chromium.org> Reviewed-by: Alexandre Courbot <acourbot@chromium.org> Tested-by: Alexandre Courbot <acourbot@chromium.org> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab+samsung@kernel.org>
generated by cgit v1.2.3 (git 2.39.0) at 2025-02-14 23:59:58 +0300