[libav-devel] [PATCH v3] qsvvpp: Fix to perform full init only when needed

Maxym Dmytrychenko maxim.d33 at gmail.com
Tue Jul 17 00:13:57 CEST 2018


On Tue, Jul 17, 2018 at 12:00 AM Mark Thompson <sw at jkqxz.net> wrote:

> On 16/07/18 14:26, Maxym Dmytrychenko wrote:
> > Not used VPP sessions, like for hwupload/hwdownload handling,
> > can increase CPU utilization and this patch fixes it.
> > thank you,Joe, for the contribution.
> >
> > Signed-off-by: Maxym Dmytrychenko <maxim.d33 at gmail.com>
> > ---
> >  libavutil/hwcontext_qsv.c | 35 ++++++++++++++++++++++++++---------
> >  1 file changed, 26 insertions(+), 9 deletions(-)
>
> This makes sense, but it looks like it might need some sort of 'once'
> construction for thread-safety?
>
> I believe the current API intent is that performing simultaneous transfer
> operations on different frames in the same frames context should be safe.
>
> good point!
So far, I see pretty much always the same thread, so like single threaded
usage, unless I am missing something.
should we consider this implementation ok and to remember: if
multithreading support - to be adjusted?


>
> > diff --git a/libavutil/hwcontext_qsv.c b/libavutil/hwcontext_qsv.c
> > index b3eb4a3ea..3e6c38037 100644
> > --- a/libavutil/hwcontext_qsv.c
> > +++ b/libavutil/hwcontext_qsv.c
> > @@ -56,7 +56,9 @@ typedef struct QSVDeviceContext {
> >
> >  typedef struct QSVFramesContext {
> >      mfxSession session_download;
> > +    int session_download_init;
> >      mfxSession session_upload;
> > +    int session_upload_init;
> >
> >      AVBufferRef *child_frames_ref;
> >      mfxFrameSurface1 *surfaces_internal;
> > @@ -146,13 +148,15 @@ static void qsv_frames_uninit(AVHWFramesContext
> *ctx)
> >          MFXVideoVPP_Close(s->session_download);
> >          MFXClose(s->session_download);
> >      }
> > -    s->session_download = NULL;
> > +    s->session_download      = NULL;
> > +    s->session_download_init = 0;
> >
> >      if (s->session_upload) {
> >          MFXVideoVPP_Close(s->session_upload);
> >          MFXClose(s->session_upload);
> >      }
> > -    s->session_upload = NULL;
> > +    s->session_upload      = NULL;
> > +    s->session_upload_init = 0;
> >
> >      av_freep(&s->mem_ids);
> >      av_freep(&s->surface_ptrs);
> > @@ -535,13 +539,10 @@ static int qsv_frames_init(AVHWFramesContext *ctx)
> >              s->mem_ids[i] = frames_hwctx->surfaces[i].Data.MemId;
> >      }
> >
> > -    ret = qsv_init_internal_session(ctx, &s->session_download, 0);
> > -    if (ret < 0)
> > -        return ret;
> > -
> > -    ret = qsv_init_internal_session(ctx, &s->session_upload, 1);
> > -    if (ret < 0)
> > -        return ret;
> > +    s->session_download      = NULL;
> > +    s->session_upload        = NULL;
> > +    s->session_download_init = 0;
> > +    s->session_upload_init   = 0;
> >
> >      return 0;
> >  }
> > @@ -740,6 +741,14 @@ static int qsv_transfer_data_from(AVHWFramesContext
> *ctx, AVFrame *dst,
> >
> >      mfxSyncPoint sync = NULL;
> >      mfxStatus err;
> > +    int ret = -1;
>
> The initialisation is redundant?  The -1 confused me, but I don't think
> it's ever read anywhere.
>
> can change this, sure.


> > +
> > +    if (!s->session_download_init) {
> > +        s->session_download_init = 1;
> > +        ret = qsv_init_internal_session(ctx, &s->session_download, 0);
> > +        if (ret < 0)
> > +            return ret;
> > +    }
> >
> >      if (!s->session_download) {
> >          if (s->child_frames_ref)
> > @@ -787,6 +796,14 @@ static int qsv_transfer_data_to(AVHWFramesContext
> *ctx, AVFrame *dst,
> >
> >      mfxSyncPoint sync = NULL;
> >      mfxStatus err;
> > +    int ret = -1;
> Likewise this one.
>
> as above as well - can change.


> > +
> > +    if (!s->session_upload_init) {
> > +        s->session_upload_init = 1;
> > +        ret = qsv_init_internal_session(ctx, &s->session_upload, 1);
> > +        if (ret < 0)
> > +            return ret;
> > +    }
> >
> >      if (!s->session_upload) {
> >          if (s->child_frames_ref)
> >
>
> Thanks,
>
> thank you for the review


> - Mark
> _______________________________________________
> libav-devel mailing list
> libav-devel at libav.org
> https://lists.libav.org/mailman/listinfo/libav-devel


Max


More information about the libav-devel mailing list