Dear Nicolas Dufresne:
Thanks for your patches!
The Media CI robot had a great time going through them, and it could
not find any issue.
For more details, check the full report at:
https://linux-media.pages.freedesktop.org/-/users/patchwork/-/jobs/75887122/artifacts/report.htm
.
Best regards,
On Thu Apr 24 10:05:35 2025 +0300, Tomi Valkeinen wrote:
> On Gen4 we need to set VNMC's EXINF to a different value (1) than in
> Gen3 (0). Add a define for this, and set the bit for Gen4.
>
> Signed-off-by: Tomi Valkeinen
> Reviewed-by: Niklas Söderlund
> Tested-by: Niklas Söderlund
> Link:
On Tue Apr 1 16:22:02 2025 +0200, Jacopo Mondi wrote:
> The vsp1_dl library allows to program a display list and feed it
> continuously to the VSP2. As an alternative operation mode, the library
> allows to program the VSP2 in 'single shot' mode, where a display list
> is submitted to the VSP on re
On Wed Apr 23 18:31:13 2025 +0200, Niklas Söderlund wrote:
> Extend the device tree parsing to optionally parse the cs memory region
> by name. The change is backward compatible with the device tree model
> where a single unnamed region describes only the ISP channel select
> function.
>
> Signed
On Wed Apr 30 02:28:56 2025 +0300, Laurent Pinchart wrote:
> The VSP1 driver is missing the ability to enumerate pixel formats on its
> video nodes, which is supposed to be supported according to the V4L2
> API. Implement the enumeration to fix this issue.
>
> As the device is media controller-cen
On Tue Apr 1 16:22:01 2025 +0200, Jacopo Mondi wrote:
> The IIF (ISP InterFace) is a VSP2 function that transfers data
> to the ISP by reading from external memory through two RPF
> instances.
>
> Add support for it in the vsp1 driver by introducing a new entity
> type. The sole required operation
On Wed Apr 30 02:29:00 2025 +0300, Laurent Pinchart wrote:
> The vsp1 driver implements very partial colour space support: it
> hardcodes the colorspace field on all video devices and subdevices to
> V4L2_COLORSPACE_SRGB, regardless of the configured format. The
> xfer_func, ycbcr_enc and quantizat
On Tue Apr 1 16:22:03 2025 +0200, Jacopo Mondi wrote:
> vsp1_wpf.c calls vsp1_rwpf_init_ctrls() to initialize controls that
> are common between RPF and WPF.
>
> However, the vsp1_wpf.c implementation does not check for the function
> call return value. Fix this by propagating to the caller the re
On Thu Apr 24 14:20:45 2025 +0530, Vikash Garodia wrote:
> Document the IRIS video decoder/encoder accelerator found in the QCS8300
> platform. It belongs to same iris v3 family as that of SM8550 but is a
> downscaled version of SM8550. It has 2 frame processing hardware blocks
> while SM8550 has 4
On Wed Apr 23 18:31:07 2025 +0200, Niklas Söderlund wrote:
> Some R-Car ISP instances have in addition to the channel selector (CS)
> an ISP core (CORE) to perform operations on an image stream. The core
> function is mapped to a different memory region and has a separate
> interrupt than CS, exte
On Thu Apr 17 16:59:04 2025 +0200, Neil Armstrong wrote:
> The vpu33 HW is very close to vpu3, and shares most of the
> operations, so rename file to vpu3x since we'll handle all vpu3
> variants in it.
>
> Reviewed-by: Dikshita Agarwal
> Tested-by: Bryan O'Donoghue # x1e Dell
> Reviewed-by: Vika
On Thu Apr 24 10:05:33 2025 +0300, Tomi Valkeinen wrote:
> Currently Gen4 VINs are marked as RCAN_GEN3 models. Add a new enum
> value, RCAR_GEN4, and use it for Gen4 VINs. No functional changes in
> this patch.
>
> Signed-off-by: Tomi Valkeinen
> Reviewed-by: Niklas Söderlund
> Tested-by: Nikla
On Wed Apr 23 18:31:12 2025 +0200, Niklas Söderlund wrote:
> Prepare for extending the driver to in addition to supporting the
> channel selector (CS) also support the core ISP. The two different
> functions have different base addresses so the driver needs to
> distinguish between them.
>
> Prep
On Sun Mar 30 10:17:35 2025 +0100, Biju Das wrote:
> Replace '%08x->%p4cc' printk modifier to print FourCC codes for pixel
> formats.
>
> Signed-off-by: Biju Das
> Reviewed-by: Laurent Pinchart
> Reviewed-by: Kieran Bingham
> Link:
> https://lore.kernel.org/r/20250330091738.27052-1-biju.das...
On Thu Apr 24 14:20:47 2025 +0530, Vikash Garodia wrote:
> Add platform data for QCS8300, which has different capabilities compared
> to SM8550. Introduce a QCS8300 header that defines these capabilities.
>
> Reviewed-by: Bryan O'Donoghue
> Signed-off-by: Vikash Garodia
> Reviewed-by: Dikshita A
On Thu Apr 17 16:59:06 2025 +0200, Neil Armstrong wrote:
> In order to prepare for supporting the SM8650 SoC, move the
> iris_platform_sm8550.c file into iris_platform_gen2.c that will
> contain all the common HFI GEN2x structures.
>
> Reviewed-by: Bryan O'Donoghue
> Tested-by: Bryan O'Donoghue
On Thu Apr 24 14:20:46 2025 +0530, Vikash Garodia wrote:
> Fix the order of compatible strings to make it in alpha numeric order.
>
> Reviewed-by: Bryan O'Donoghue
> Signed-off-by: Vikash Garodia
> Reviewed-by: Dikshita Agarwal
> Signed-off-by: Bryan O'Donoghue
> Signed-off-by: Hans Verkuil
On Thu Apr 17 16:59:02 2025 +0200, Neil Armstrong wrote:
> In order to support the SM8650 iris33 hardware, we need to provide a
> specific constoller power off sequences via the vpu_ops callbacks.
>
> Add the callback, and use the current helper for currently supported
> platforms.
>
> Reviewed-b
On Tue Apr 1 16:22:04 2025 +0200, Jacopo Mondi wrote:
> When the RPF/WPF units are used for ISP interfacing through
> the IIF, the set of accessible registers is limited compared to
> the regular VSPD operations.
>
> Support ISP interfacing in the rpf and wpf entities by checking if
> the pipe fea
On Thu Apr 24 10:05:36 2025 +0300, Tomi Valkeinen wrote:
> Fix the following to get RAW10 formats working:
>
> In rvin_formats, the bpp is set to 4 for RAW10. As VIN unpacks RAW10 to
> 16-bit containers, the bpp should be 2.
>
> Don't set VNDMR_YC_THR to the VNDMR register. The YC_THR is "YC Data
On Thu Apr 24 10:05:34 2025 +0300, Tomi Valkeinen wrote:
> Remove unnecessary checks wrt. formats and interfaces in rvin_setup().
> The validity of the formats has already been checked earlier.
>
> Signed-off-by: Tomi Valkeinen
> Reviewed-by: Niklas Söderlund
> Tested-by: Niklas Söderlund
> L
On Wed Apr 30 02:28:57 2025 +0300, Laurent Pinchart wrote:
> Not all VSP instance incorporate the HSI and HST modules. Add a
> VSP1_HAS_HSIT feature flag, and create the modules only on VSP instances
> that implement them.
>
> Signed-off-by: Laurent Pinchart
> Reviewed-by: Tomi Valkeinen
> Link:
On Thu Apr 17 16:59:03 2025 +0200, Neil Armstrong wrote:
> Introduce an optional controller_rst_tbl use to store reset lines
> used to reset part of the controller.
>
> This is necessary for the vpu3 support, when the xo reset line
> must be asserted separately from the other reset line
> on power
On Thu Apr 17 16:59:05 2025 +0200, Neil Armstrong wrote:
> The IRIS acceleration found in the SM8650 platforms uses the vpu33
> hardware version, and requires a slighly different reset and power off
> sequences in order to properly get out of runtime suspend.
>
> Tested-by: Bryan O'Donoghue # x1e
On Wed Apr 23 18:31:11 2025 +0200, Niklas Söderlund wrote:
> Before extending the driver with functions from the R-Car ISP core that
> will span multiple files move the existing driver to a separate
> directory. While at it rename the single source file to allow future
> files to be grouped by fun
On Thu Apr 17 16:59:01 2025 +0200, Neil Armstrong wrote:
> Document the IRIS video decoder and encoder accelerator found in the
> SM8650 platform, it requires 2 more reset lines in addition to the
> properties required for the SM8550 platform.
>
> Reviewed-by: Rob Herring (Arm)
> Reviewed-by: Vik
On Thu Apr 17 16:59:07 2025 +0200, Neil Armstrong wrote:
> Add support for the SM8650 platform by re-using the SM8550
> definitions and using the vpu33 ops.
>
> Move the Sm8650 reset tables that differs in a per-SoC platform
> header, that will contain mode SoC specific data when
> more codecs are
On Wed Apr 30 02:28:58 2025 +0300, Laurent Pinchart wrote:
> The HSV formats are not restricted to Gen2 platforms, but to VSP
> instances that implement the HSI and HST modules. Make it conditional to
> the VSP1_HAS_HSIT feature flag.
>
> Signed-off-by: Laurent Pinchart
> Reviewed-by: Tomi Valkei
On Wed Apr 30 02:28:59 2025 +0300, Laurent Pinchart wrote:
> The RWPF source pad media bus code can only be different from the sink
> pad code when enabling color space conversion, which can only convert
> between RGB and YUV. If the sink pad code is HSV, no conversion is
> possible. Fix the pad se
On Wed Apr 30 02:29:01 2025 +0300, Laurent Pinchart wrote:
> The RPF and WPF support different encodings and quantizations when
> converting between RGB and YUV formats. Allow setting the corresponding
> format parameters from userspace, and configure the hardware
> accordingly.
>
> Signed-off-by:
On Wed Apr 30 02:29:02 2025 +0300, Laurent Pinchart wrote:
> The vsp1_drm structure defines an anonymous nested structure to store
> per-input data. In preparation for extending that structure, give it a
> name and is it through the driver. This improves code readability.
>
> Signed-off-by: Lauren
On Wed Apr 30 02:29:03 2025 +0300, Laurent Pinchart wrote:
> Now that the VSP1 driver supports color spaces, expose them through the
> API used by the DU driver. This allows configuring the YCbCr encoding
> and quantization used by each plane, ensuring correct color rendering.
>
> Signed-off-by: L
32 matches
Mail list logo
