Re: [PATCH 10/28] venus: vdec: call session_continue in insufficient event

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Vikash,

Thanks for the comments!

On  2.05.2018 09:26, Vikash Garodia wrote:
Hello Stanimir,

On 2018-04-24 18:14, Stanimir Varbanov wrote:
Call session_continue for Venus 4xx version even when the event
says that the buffer resources are not sufficient. Leaving a
comment with more information about the workaround.

Signed-off-by: Stanimir Varbanov <stanimir.varbanov@xxxxxxxxxx>
---
 drivers/media/platform/qcom/venus/vdec.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/media/platform/qcom/venus/vdec.c
b/drivers/media/platform/qcom/venus/vdec.c
index c45452634e7e..91c7384ff9c8 100644
--- a/drivers/media/platform/qcom/venus/vdec.c
+++ b/drivers/media/platform/qcom/venus/vdec.c
@@ -873,6 +873,14 @@ static void vdec_event_notify(struct venus_inst
*inst, u32 event,

             dev_dbg(dev, "event not sufficient resources (%ux%u)\n",
                 data->width, data->height);
+            /*
+             * Workaround: Even that the firmware send and event for
+             * insufficient buffer resources it is safe to call
+             * session_continue because actually the event says that
+             * the number of capture buffers is lower.
+             */
+            if (IS_V4(core))
+                hfi_session_continue(inst);
             break;
         case HFI_EVENT_RELEASE_BUFFER_REFERENCE:
             venus_helper_release_buf_ref(inst, data->tag);

Insufficient event from video firmware could be sent either,
1. due to insufficient buffer resources
2. due to lower capture buffers

It cannot be assumed that the event received by the host is due to lower capture buffers. Incase the buffer resource is insufficient, let say there is a bitstream resolution switch from 720p to 1080p, capture buffers needs to be reallocated.

I agree with you. I will rework this part and call session_continue only for case #2.


The driver should be sending the V4L2_EVENT_SOURCE_CHANGE to client instead of ignoring
the event from firmware.

The v4l2 event is sent always to v4l clients.

regards,
Stan



[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]

  Powered by Linux