| Message ID | 20200213213007.17023-1-jkardatzke@google.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | media: venus: support frame rate control | expand |
Hi Jeff, Thanks for the patch! On 2/13/20 11:30 PM, Jeffrey Kardatzke wrote: > Frame rate control is always enabled in this driver, so make it silently > support the V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE. > > Signed-off-by: Jeffrey Kardatzke <jkardatzke@google.com> > --- > drivers/media/platform/qcom/venus/venc_ctrls.c | 6 ++++++ > 1 file changed, 6 insertions(+) > > diff --git a/drivers/media/platform/qcom/venus/venc_ctrls.c b/drivers/media/platform/qcom/venus/venc_ctrls.c > index 877c0b3299e9..9ede692f77c5 100644 > --- a/drivers/media/platform/qcom/venus/venc_ctrls.c > +++ b/drivers/media/platform/qcom/venus/venc_ctrls.c > @@ -199,6 +199,9 @@ static int venc_op_s_ctrl(struct v4l2_ctrl *ctrl) > } > mutex_unlock(&inst->lock); > break; > + case V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE: > + // Silently ignore, it's always enabled. Please, use C comments and follow the kernel coding style. I wonder shouldn't it be better to add rc_enable field in struct venc_controls and give the user choice to disable the rate control? We can keep the default to be "enabled". > + break; > default: > return -EINVAL; > } > @@ -351,6 +354,9 @@ int venc_ctrl_init(struct venus_inst *inst) > v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops, > V4L2_CID_MPEG_VIDEO_FORCE_KEY_FRAME, 0, 0, 0, 0); > > + v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops, > + V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE, 0, 1, 1, 1); > + you forgot to increment the number of controls in the call to v4l2_ctrl_handler_init. > ret = inst->ctrl_handler.error; > if (ret) > goto err; >
Sorry for the duplicate, accidentally used HTML format and it got bounced from the mailing lists so resending. On Mon, Feb 17, 2020 at 2:15 AM Stanimir Varbanov <stanimir.varbanov@linaro.org> wrote: > > Hi Jeff, > > Thanks for the patch! > > On 2/13/20 11:30 PM, Jeffrey Kardatzke wrote: > > Frame rate control is always enabled in this driver, so make it silently > > support the V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE. > > > > Signed-off-by: Jeffrey Kardatzke <jkardatzke@google.com> > > --- > > drivers/media/platform/qcom/venus/venc_ctrls.c | 6 ++++++ > > 1 file changed, 6 insertions(+) > > > > diff --git a/drivers/media/platform/qcom/venus/venc_ctrls.c b/drivers/media/platform/qcom/venus/venc_ctrls.c > > index 877c0b3299e9..9ede692f77c5 100644 > > --- a/drivers/media/platform/qcom/venus/venc_ctrls.c > > +++ b/drivers/media/platform/qcom/venus/venc_ctrls.c > > @@ -199,6 +199,9 @@ static int venc_op_s_ctrl(struct v4l2_ctrl *ctrl) > > } > > mutex_unlock(&inst->lock); > > break; > > + case V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE: > > + // Silently ignore, it's always enabled. > > Please, use C comments and follow the kernel coding style. OK, hopefully I've got that now. I didn't see any issues aside from the comment style though. I'll upload a new patch shortly. > > > I wonder shouldn't it be better to add rc_enable field in struct > venc_controls and give the user choice to disable the rate control? We > can keep the default to be "enabled". > That'd be fine. Is there a way to actually disable the rate control though? > > > + break; > > default: > > return -EINVAL; > > } > > @@ -351,6 +354,9 @@ int venc_ctrl_init(struct venus_inst *inst) > > v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops, > > V4L2_CID_MPEG_VIDEO_FORCE_KEY_FRAME, 0, 0, 0, 0); > > > > + v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops, > > + V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE, 0, 1, 1, 1); > > + > > you forgot to increment the number of controls in the call to > v4l2_ctrl_handler_init. > Done, thanks. > > > ret = inst->ctrl_handler.error; > > if (ret) > > goto err; > > > > -- > regards, > Stan
Hi Jeff, On 2/18/20 9:09 PM, Jeffrey Kardatzke wrote: > Sorry for the duplicate, accidentally used HTML format and it got > bounced from the mailing lists so resending. > > On Mon, Feb 17, 2020 at 2:15 AM Stanimir Varbanov > <stanimir.varbanov@linaro.org> wrote: >> >> Hi Jeff, >> >> Thanks for the patch! >> >> On 2/13/20 11:30 PM, Jeffrey Kardatzke wrote: >>> Frame rate control is always enabled in this driver, so make it silently >>> support the V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE. >>> >>> Signed-off-by: Jeffrey Kardatzke <jkardatzke@google.com> >>> --- >>> drivers/media/platform/qcom/venus/venc_ctrls.c | 6 ++++++ >>> 1 file changed, 6 insertions(+) >>> >>> diff --git a/drivers/media/platform/qcom/venus/venc_ctrls.c b/drivers/media/platform/qcom/venus/venc_ctrls.c >>> index 877c0b3299e9..9ede692f77c5 100644 >>> --- a/drivers/media/platform/qcom/venus/venc_ctrls.c >>> +++ b/drivers/media/platform/qcom/venus/venc_ctrls.c >>> @@ -199,6 +199,9 @@ static int venc_op_s_ctrl(struct v4l2_ctrl *ctrl) >>> } >>> mutex_unlock(&inst->lock); >>> break; >>> + case V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE: >>> + // Silently ignore, it's always enabled. >> >> Please, use C comments and follow the kernel coding style. > > OK, hopefully I've got that now. I didn't see any issues aside from > the comment style though. > I'll upload a new patch shortly. >> >> >> I wonder shouldn't it be better to add rc_enable field in struct >> venc_controls and give the user choice to disable the rate control? We >> can keep the default to be "enabled". >> > That'd be fine. Is there a way to actually disable the rate control though? The rate control property values are here [1], and [2] is where we set the control.
Got it, thanks. I'll submit a new patch with that. On Tue, Feb 18, 2020 at 11:56 PM Stanimir Varbanov <stanimir.varbanov@linaro.org> wrote: > > Hi Jeff, > > On 2/18/20 9:09 PM, Jeffrey Kardatzke wrote: > > Sorry for the duplicate, accidentally used HTML format and it got > > bounced from the mailing lists so resending. > > > > On Mon, Feb 17, 2020 at 2:15 AM Stanimir Varbanov > > <stanimir.varbanov@linaro.org> wrote: > >> > >> Hi Jeff, > >> > >> Thanks for the patch! > >> > >> On 2/13/20 11:30 PM, Jeffrey Kardatzke wrote: > >>> Frame rate control is always enabled in this driver, so make it silently > >>> support the V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE. > >>> > >>> Signed-off-by: Jeffrey Kardatzke <jkardatzke@google.com> > >>> --- > >>> drivers/media/platform/qcom/venus/venc_ctrls.c | 6 ++++++ > >>> 1 file changed, 6 insertions(+) > >>> > >>> diff --git a/drivers/media/platform/qcom/venus/venc_ctrls.c b/drivers/media/platform/qcom/venus/venc_ctrls.c > >>> index 877c0b3299e9..9ede692f77c5 100644 > >>> --- a/drivers/media/platform/qcom/venus/venc_ctrls.c > >>> +++ b/drivers/media/platform/qcom/venus/venc_ctrls.c > >>> @@ -199,6 +199,9 @@ static int venc_op_s_ctrl(struct v4l2_ctrl *ctrl) > >>> } > >>> mutex_unlock(&inst->lock); > >>> break; > >>> + case V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE: > >>> + // Silently ignore, it's always enabled. > >> > >> Please, use C comments and follow the kernel coding style. > > > > OK, hopefully I've got that now. I didn't see any issues aside from > > the comment style though. > > I'll upload a new patch shortly. > >> > >> > >> I wonder shouldn't it be better to add rc_enable field in struct > >> venc_controls and give the user choice to disable the rate control? We > >> can keep the default to be "enabled". > >> > > That'd be fine. Is there a way to actually disable the rate control though? > > The rate control property values are here [1], and [2] is where we set > the control. > > -- > regards, > Stan > > [1] > https://elixir.bootlin.com/linux/v5.6-rc2/source/drivers/media/platform/qcom/venus/hfi_helper.h#L229 > [2] > https://elixir.bootlin.com/linux/v5.6-rc2/source/drivers/media/platform/qcom/venus/venc.c#L734
diff --git a/drivers/media/platform/qcom/venus/venc_ctrls.c b/drivers/media/platform/qcom/venus/venc_ctrls.c index 877c0b3299e9..9ede692f77c5 100644 --- a/drivers/media/platform/qcom/venus/venc_ctrls.c +++ b/drivers/media/platform/qcom/venus/venc_ctrls.c @@ -199,6 +199,9 @@ static int venc_op_s_ctrl(struct v4l2_ctrl *ctrl) } mutex_unlock(&inst->lock); break; + case V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE: + // Silently ignore, it's always enabled. + break; default: return -EINVAL; } @@ -351,6 +354,9 @@ int venc_ctrl_init(struct venus_inst *inst) v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops, V4L2_CID_MPEG_VIDEO_FORCE_KEY_FRAME, 0, 0, 0, 0); + v4l2_ctrl_new_std(&inst->ctrl_handler, &venc_ctrl_ops, + V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE, 0, 1, 1, 1); + ret = inst->ctrl_handler.error; if (ret) goto err;
Frame rate control is always enabled in this driver, so make it silently support the V4L2_CID_MPEG_VIDEO_FRAME_RC_ENABLE. Signed-off-by: Jeffrey Kardatzke <jkardatzke@google.com> --- drivers/media/platform/qcom/venus/venc_ctrls.c | 6 ++++++ 1 file changed, 6 insertions(+)