Hi, > From: Jaeryul Oh [mailto:jaeryul.oh@xxxxxxxxxxx] > > Hi, Kamil > I have a comment about s5p_mfc_stop_streaming()function. > > > -----Original Message----- > > From: linux-media-owner@xxxxxxxxxxxxxxx [mailto:linux-media- > > owner@xxxxxxxxxxxxxxx] On Behalf Of Kamil Debski > > Sent: Saturday, January 08, 2011 1:26 AM > > To: linux-media@xxxxxxxxxxxxxxx; linux-samsung-soc@xxxxxxxxxxxxxxx > > Cc: m.szyprowski@xxxxxxxxxxx; pawel@xxxxxxxxxx; > kyungmin.park@xxxxxxxxxxx; > > k.debski@xxxxxxxxxxx; jaeryul.oh@xxxxxxxxxxx; kgene.kim@xxxxxxxxxxx > > Subject: [RFC/PATCH v6 3/4] MFC: Add MFC 5.1 V4L2 driver > > > > Multi Format Codec 5.1 is capable of handling a range of video codecs > > and this driver provides V4L2 interface for video decoding. > > > > Signed-off-by: Kamil Debski <k.debski@xxxxxxxxxxx> > > Signed-off-by: Kyungmin Park <kyungmin.park@xxxxxxxxxxx> <snip> > > + > > +/* Thou shalt stream no more. */ > > +static int s5p_mfc_stop_streaming(struct vb2_queue *q) > > +{ > > + unsigned long flags; > > + struct s5p_mfc_ctx *ctx = q->drv_priv; > > + struct s5p_mfc_dev *dev = ctx->dev; > > + > > + if ((ctx->state == MFCINST_DEC_FINISHING || > > + ctx->state == MFCINST_DEC_RUNNING) && > > + dev->curr_ctx == ctx->num && dev->hw_lock) { > > + ctx->state = MFCINST_DEC_ABORT; > > + s5p_mfc_wait_for_done_ctx(ctx, > > S5P_FIMV_R2H_CMD_FRAME_DONE_RET, > > + 0); > > + } > > + ctx->state = MFCINST_DEC_FINISHED; > > + spin_lock_irqsave(&dev->irqlock, flags); > > + s5p_mfc_error_cleanup_queue(&ctx->dst_queue, > > + &ctx->vq_dst); > > + s5p_mfc_error_cleanup_queue(&ctx->src_queue, > > + &ctx->vq_src); > > + if (q->type == V4L2_BUF_TYPE_VIDEO_CAPTURE_MPLANE) > > + INIT_LIST_HEAD(&ctx->dst_queue); > > + if (q->type == V4L2_BUF_TYPE_VIDEO_OUTPUT_MPLANE) > > + INIT_LIST_HEAD(&ctx->src_queue); > > + spin_unlock_irqrestore(&dev->irqlock, flags); > > + return 0; > > +} > This function is called by __vb2_queue_cancel().and > __vb2_queue_cancel() > can be > called by vb2_queue_release() or vb2_streamoff(). > But, in this s5p_mfc_stop_streaming(),s5p_mfc_error_cleanup_queue() for > src/dst > is runned regardless of q->type. Is that right ? > and in case of streamoff, queued bufs should be removed, then qbuf is > needed > before streamon again, so ctx->dst_queue_cnt = 0; ctx->src_queue_cnt = > 0; > is required > what do you think about this ? > It has to be changed to support pause and dynamic resolution change. Best wishes, -- Kamil Debski Linux Platform Group Samsung Poland R&D Center -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
