| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Mon, 10 Oct 2022 14:14:05 +0800
From: "mingjia.zhang@...iatek.com" <mingjia.zhang@...iatek.com>
To: Hans Verkuil <hverkuil-cisco@...all.nl>,
Yunfei Dong <yunfei.dong@...iatek.com>,
Alexandre Courbot <acourbot@...omium.org>,
Nicolas Dufresne <nicolas@...fresne.ca>,
AngeloGioacchino Del Regno
<angelogioacchino.delregno@...labora.com>,
Benjamin Gaignard <benjamin.gaignard@...labora.com>,
Tiffany Lin <tiffany.lin@...iatek.com>,
Andrew-CT Chen <andrew-ct.chen@...iatek.com>,
Mauro Carvalho Chehab <mchehab@...nel.org>,
Rob Herring <robh+dt@...nel.org>,
Matthias Brugger <matthias.bgg@...il.com>,
Tomasz Figa <tfiga@...gle.com>
CC: Irui Wang <irui.wang@...iatek.com>,
George Sun <george.sun@...iatek.com>,
Steve Cho <stevecho@...omium.org>,
<devicetree@...r.kernel.org>,
<Project_Global_Chrome_Upstream_Group@...iatek.com>,
<linux-kernel@...r.kernel.org>,
dri-devel <dri-devel@...ts.freedesktop.org>,
Xiaoyong Lu <xiaoyong.lu@...iatek.com>,
<linux-mediatek@...ts.infradead.org>,
Hsin-Yi Wang <hsinyi@...omium.org>,
Fritz Koenig <frkoenig@...omium.org>,
<linux-arm-kernel@...ts.infradead.org>,
<linux-media@...r.kernel.org>
Subject: Re: [PATCH, v2] media: mediatek: vcodec: Add to support VP9 inner
racing mode
Hi Hans,
Thanks for you reply and useful comments.
After reviewing the error handling in vdec_vp9_slice_core_decode(), I
found that the problem you pointed out does exist in error handle part.
'instance' may be used uninitialized in this function.
I have fixed this build warning and reviewed other code, thanks.
Thanks,
mingjia
On Wed, 2022-08-24 at 15:02 +0200, Hans Verkuil wrote:
> Hi Mingjia,
>
> On 27/07/2022 08:13, Mingjia Zhang wrote:
> > In order to reduce decoder latency, enable VP9 inner racing mode.
> > Send lat trans buffer information to core when trigger lat to work,
> > need not to wait until lat decode done.
> >
> > Signed-off-by: mingjia zhang <mingjia.zhang@...iatek.com>
>
> I'm getting this compile warning:
>
> drivers/media/platform/mediatek/vcodec/vdec/vdec_vp9_req_lat_if.c: In
> function 'vdec_vp9_slice_core_decode':
> drivers/media/platform/mediatek/vcodec/vdec/vdec_vp9_req_lat_if.c:221
> 8:50: warning: 'instance' may be used uninitialized in this function
> [-Wmaybe-uninitialized]
> 2218 | if (IS_VDEC_INNER_RACING(instance->ctx->dev-
> >dec_capability))
> | ^~
>
> I think you need to take a close look at the error handling in
> vdec_vp9_slice_core_decode().
>
> After each error there is a 'goto err;' and that will run the new
> code, and that doesn't
> feel right.
>
> Regards,
>
> Hans
>
> > ---
> > 1. CTS/GTS test pass
> > 2. Fluster result: Ran 240/303 tests successfully
> > ---
> > .../vcodec/vdec/vdec_vp9_req_lat_if.c | 64 ++++++++++++---
> > ----
> > 1 file changed, 40 insertions(+), 24 deletions(-)
> >
> > diff --git
> > a/drivers/media/platform/mediatek/vcodec/vdec/vdec_vp9_req_lat_if.c
> > b/drivers/media/platform/mediatek/vcodec/vdec/vdec_vp9_req_lat_if.c
> > index fb1c36a3592d..92b47f0fdf40 100644
> > ---
> > a/drivers/media/platform/mediatek/vcodec/vdec/vdec_vp9_req_lat_if.c
> > +++
> > b/drivers/media/platform/mediatek/vcodec/vdec/vdec_vp9_req_lat_if.c
> > @@ -436,6 +436,7 @@ struct vdec_vp9_slice_ref {
> > * @frame_ctx: 4 frame context according to VP9 Spec
> > * @frame_ctx_helper: 4 frame context according to newest
> > kernel spec
> > * @dirty: state of each frame context
> > + * @local_vsi: local instance vsi information
> > * @init_vsi: vsi used for initialized VP9 instance
> > * @vsi: vsi used for decoding/flush ...
> > * @core_vsi: vsi used for Core stage
> > @@ -482,6 +483,8 @@ struct vdec_vp9_slice_instance {
> > struct v4l2_vp9_frame_context frame_ctx_helper;
> > unsigned char dirty[4];
> >
> > + struct vdec_vp9_slice_vsi local_vsi;
> > +
> > /* MicroP vsi */
> > union {
> > struct vdec_vp9_slice_init_vsi *init_vsi;
> > @@ -1616,16 +1619,10 @@ static int
> > vdec_vp9_slice_update_single(struct vdec_vp9_slice_instance
> > *instance
> > }
> >
> > static int vdec_vp9_slice_update_lat(struct
> > vdec_vp9_slice_instance *instance,
> > - struct vdec_lat_buf *lat_buf,
> > - struct vdec_vp9_slice_pfc *pfc)
> > + struct vdec_vp9_slice_vsi *vsi)
> > {
> > - struct vdec_vp9_slice_vsi *vsi;
> > -
> > - vsi = &pfc->vsi;
> > - memcpy(&pfc->state[0], &vsi->state, sizeof(vsi->state));
> > -
> > mtk_vcodec_debug(instance, "Frame %u LAT CRC 0x%08x %lx %lx\n",
> > - pfc->seq, vsi->state.crc[0],
> > + (instance->seq - 1), vsi->state.crc[0],
> > (unsigned long)vsi->trans.dma_addr,
> > (unsigned long)vsi->trans.dma_addr_end);
> >
> > @@ -2090,6 +2087,13 @@ static int vdec_vp9_slice_lat_decode(void
> > *h_vdec, struct mtk_vcodec_mem *bs,
> > return ret;
> > }
> >
> > + if (IS_VDEC_INNER_RACING(instance->ctx->dev->dec_capability)) {
> > + vdec_vp9_slice_vsi_from_remote(vsi, instance->vsi, 0);
> > + memcpy(&instance->local_vsi, vsi, sizeof(*vsi));
> > + vdec_msg_queue_qbuf(&ctx->dev->msg_queue_core_ctx,
> > lat_buf);
> > + vsi = &instance->local_vsi;
> > + }
> > +
> > if (instance->irq) {
> > ret = mtk_vcodec_wait_for_done_ctx(ctx, MTK_INST_IR
> > Q_RECEIVED,
> > WAIT_INTR_TIMEOUT_MS
> > , MTK_VDEC_LAT0);
> > @@ -2102,22 +2106,25 @@ static int vdec_vp9_slice_lat_decode(void
> > *h_vdec, struct mtk_vcodec_mem *bs,
> > }
> >
> > vdec_vp9_slice_vsi_from_remote(vsi, instance->vsi, 0);
> > - ret = vdec_vp9_slice_update_lat(instance, lat_buf, pfc);
> > + ret = vdec_vp9_slice_update_lat(instance, vsi);
> >
> > - /* LAT trans full, no more UBE or decode timeout */
> > - if (ret) {
> > - mtk_vcodec_err(instance, "VP9 decode error: %d\n",
> > ret);
> > - return ret;
> > - }
> > + if (!IS_VDEC_INNER_RACING(instance->ctx->dev->dec_capability))
> > + /* LAT trans full, no more UBE or decode timeout */
> > + if (ret) {
> > + mtk_vcodec_err(instance, "frame[%d] decode
> > error: %d\n",
> > + ret, (instance->seq - 1));
> > + return ret;
> > + }
> >
> > - mtk_vcodec_debug(instance, "lat dma addr: 0x%lx 0x%lx\n",
> > - (unsigned long)pfc->vsi.trans.dma_addr,
> > - (unsigned long)pfc->vsi.trans.dma_addr_end);
> >
> > - vdec_msg_queue_update_ube_wptr(&ctx->msg_queue,
> > - vsi->trans.dma_addr_end +
> > - ctx-
> > >msg_queue.wdma_addr.dma_addr);
> > - vdec_msg_queue_qbuf(&ctx->dev->msg_queue_core_ctx, lat_buf);
> > + vsi->trans.dma_addr_end += ctx->msg_queue.wdma_addr.dma_addr;
> > + vdec_msg_queue_update_ube_wptr(&ctx->msg_queue, vsi-
> > >trans.dma_addr_end);
> > + if (!IS_VDEC_INNER_RACING(instance->ctx->dev->dec_capability))
> > + vdec_msg_queue_qbuf(&ctx->dev->msg_queue_core_ctx,
> > lat_buf);
> > +
> > + mtk_vcodec_debug(instance, "lat trans end addr(0x%lx), ube
> > start addr(0x%lx)\n",
> > + (unsigned long)vsi->trans.dma_addr_end,
> > + (unsigned long)ctx-
> > >msg_queue.wdma_addr.dma_addr);
> >
> > return 0;
> > }
> > @@ -2193,10 +2200,14 @@ static int
> > vdec_vp9_slice_core_decode(struct vdec_lat_buf *lat_buf)
> > goto err;
> > }
> >
> > - pfc->vsi.trans.dma_addr_end += ctx-
> > >msg_queue.wdma_addr.dma_addr;
> > mtk_vcodec_debug(instance, "core dma_addr_end 0x%lx\n",
> > (unsigned long)pfc->vsi.trans.dma_addr_end);
> > - vdec_msg_queue_update_ube_rptr(&ctx->msg_queue, pfc-
> > >vsi.trans.dma_addr_end);
> > +
> > + if (IS_VDEC_INNER_RACING(instance->ctx->dev->dec_capability))
> > + vdec_msg_queue_update_ube_rptr(&ctx->msg_queue, pfc-
> > >vsi.trans.dma_addr);
> > + else
> > + vdec_msg_queue_update_ube_rptr(&ctx->msg_queue, pfc-
> > >vsi.trans.dma_addr_end);
> > +
> > ctx->dev->vdec_pdata->cap_to_disp(ctx, 0, lat_buf-
> > >src_buf_req);
> >
> > return 0;
> > @@ -2204,7 +2215,12 @@ static int vdec_vp9_slice_core_decode(struct
> > vdec_lat_buf *lat_buf)
> > err:
> > if (ctx && pfc) {
> > /* always update read pointer */
> > - vdec_msg_queue_update_ube_rptr(&ctx->msg_queue, pfc-
> > >vsi.trans.dma_addr_end);
> > + if (IS_VDEC_INNER_RACING(instance->ctx->dev-
> > >dec_capability))
> > + vdec_msg_queue_update_ube_rptr(&ctx->msg_queue,
> > + pfc-
> > >vsi.trans.dma_addr);
> > + else
> > + vdec_msg_queue_update_ube_rptr(&ctx->msg_queue,
> > + pfc-
> > >vsi.trans.dma_addr_end);
> >
> > if (fb)
> > ctx->dev->vdec_pdata->cap_to_disp(ctx, 1,
> > lat_buf->src_buf_req);
Powered by blists - more mailing lists