| 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
| ||
|
Message-ID: <CAC-pXoP8jXhFiV2g4gPNvCPs41Y=k21tmJdSzF35SVvR1SR9hw@mail.gmail.com>
Date: Mon, 13 Dec 2021 15:49:00 -0800
From: Steve Cho <stevecho@...omium.org>
To: "yunfei.dong@...iatek.com" <yunfei.dong@...iatek.com>
Cc: Alexandre Courbot <acourbot@...omium.org>,
Hans Verkuil <hverkuil-cisco@...all.nl>,
Tzung-Bi Shih <tzungbi@...omium.org>,
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>,
Hsin-Yi Wang <hsinyi@...omium.org>,
Fritz Koenig <frkoenig@...omium.org>,
Dafna Hirschfeld <dafna.hirschfeld@...labora.com>,
Benjamin Gaignard <benjamin.gaignard@...labora.com>,
Daniel Vetter <daniel@...ll.ch>,
dri-devel <dri-devel@...ts.freedesktop.org>,
Irui Wang <irui.wang@...iatek.com>,
AngeloGioacchino Del Regno
<angelogioacchino.delregno@...labora.com>,
linux-media@...r.kernel.org, devicetree@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
srv_heupstream@...iatek.com, linux-mediatek@...ts.infradead.org,
Project_Global_Chrome_Upstream_Group@...iatek.com
Subject: Re: [PATCH v12, 13/19] media: mtk-vcodec: Add work queue for core
hardware decode
Hi Yunfei,
Thank you for addressing my questions and comments.
Note that my comments for formatting are usually based on standards
from chromium & chrome os.
So feel free to reflect only if you think it is relevant.
Reviewed-by: Steve Cho <stevecho@...omium.org>
On Mon, Dec 13, 2021 at 12:52 AM yunfei.dong@...iatek.com
<yunfei.dong@...iatek.com> wrote:
>
> Hi steve,
>
> Thanks for your suggestion.
>
> On Thu, 2021-12-09 at 15:44 -0800, Steve Cho wrote:
> > On Wed, Dec 1, 2021 at 7:46 PM Yunfei Dong <yunfei.dong@...iatek.com>
> > wrote:
> > >
> > > Add work queue to process core hardware information.
> > > First, get lat_buf from message queue, then call core
> > > hardware of each codec(H264/VP9/AV1) to decode, finally
> > > puts lat_buf back to the message.
> > >
> > > Signed-off-by: Yunfei Dong <yunfei.dong@...iatek.com>
> > > ---
> > > .../platform/mtk-vcodec/mtk_vcodec_dec_drv.c | 16 +++++++-
> > > .../platform/mtk-vcodec/mtk_vcodec_drv.h | 3 ++
> > > .../platform/mtk-vcodec/vdec_msg_queue.c | 41
> > > ++++++++++++++++---
> > > .../platform/mtk-vcodec/vdec_msg_queue.h | 8 ++--
> > > 4 files changed, 57 insertions(+), 11 deletions(-)
> > >
> > > diff --git a/drivers/media/platform/mtk-vcodec/mtk_vcodec_dec_drv.c
> > > b/drivers/media/platform/mtk-vcodec/mtk_vcodec_dec_drv.c
> > > index d460703f335d..4fbff61d2334 100644
> > > --- a/drivers/media/platform/mtk-vcodec/mtk_vcodec_dec_drv.c
> > > +++ b/drivers/media/platform/mtk-vcodec/mtk_vcodec_dec_drv.c
> > > @@ -341,6 +341,17 @@ static int mtk_vcodec_probe(struct
> > > platform_device *pdev)
> > > goto err_dec_pm;
> > > }
> > >
> > > + if (IS_VDEC_LAT_ARCH(dev->vdec_pdata->hw_arch)) {
> > > + vdec_msg_queue_init_ctx(&dev->msg_queue_core_ctx,
> > > MTK_VDEC_CORE);
> > > + dev->core_workqueue =
> > > alloc_ordered_workqueue("core-decoder",
> > > + WQ_MEM_RECLAIM | WQ_FREEZABLE);
> > > + if (!dev->core_workqueue) {
> > > + mtk_v4l2_err("Failed to create core
> > > workqueue");
> > > + ret = -EINVAL;
> > > + goto err_res;
> > > + }
> > > + }
> > > +
> > > for (i = 0; i < MTK_VDEC_HW_MAX; i++)
> > > mutex_init(&dev->dec_mutex[i]);
> > > spin_lock_init(&dev->irqlock);
> > > @@ -351,7 +362,7 @@ static int mtk_vcodec_probe(struct
> > > platform_device *pdev)
> > > ret = v4l2_device_register(&pdev->dev, &dev->v4l2_dev);
> > > if (ret) {
> > > mtk_v4l2_err("v4l2_device_register err=%d", ret);
> > > - goto err_res;
> > > + goto err_core_workq;
> > > }
> > >
> > > init_waitqueue_head(&dev->queue);
> > > @@ -450,6 +461,9 @@ static int mtk_vcodec_probe(struct
> > > platform_device *pdev)
> > > video_unregister_device(vfd_dec);
> > > err_dec_alloc:
> > > v4l2_device_unregister(&dev->v4l2_dev);
> > > +err_core_workq:
> > > + if (IS_VDEC_LAT_ARCH(dev->vdec_pdata->hw_arch))
> > > + destroy_workqueue(dev->core_workqueue);
> > > err_res:
> > > mtk_vcodec_release_dec_pm(&dev->pm);
> > > err_dec_pm:
> > > diff --git a/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h
> > > b/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h
> > > index cbaed96dcfa2..a558cc16026d 100644
> > > --- a/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h
> > > +++ b/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h
> > > @@ -27,6 +27,7 @@
> > > #define MTK_VCODEC_MAX_PLANES 3
> > > #define MTK_V4L2_BENCHMARK 0
> > > #define WAIT_INTR_TIMEOUT_MS 1000
> > > +#define IS_VDEC_LAT_ARCH(hw_arch) ((hw_arch) >=
> > > MTK_VDEC_LAT_SINGLE_CORE)
> >
> > Basic question: What is practical meaning of this? What architectures
> > are supported?
> >
> This definition is used to separate different architectures.
> Pure single core/lat single core at current period is supported.
> > >
> > > /*
> > > * enum mtk_hw_reg_idx - MTK hw register base index
> > > @@ -464,6 +465,7 @@ struct mtk_vcodec_enc_pdata {
> > > * @dec_capability: used to identify decode capability, ex: 4k
> > > * @enc_capability: used to identify encode capability
> > > *
> > > + * @core_workqueue: queue used for core hardware decode
> > > * @msg_queue_core_ctx: msg queue context used for core workqueue
> > > *
> > > * @subdev_dev: subdev hardware device
> > > @@ -506,6 +508,7 @@ struct mtk_vcodec_dev {
> > > unsigned int dec_capability;
> > > unsigned int enc_capability;
> > >
> > > + struct workqueue_struct *core_workqueue;
> > > struct vdec_msg_queue_ctx msg_queue_core_ctx;
> > >
> > > void *subdev_dev[MTK_VDEC_HW_MAX];
> > > diff --git a/drivers/media/platform/mtk-vcodec/vdec_msg_queue.c
> > > b/drivers/media/platform/mtk-vcodec/vdec_msg_queue.c
> > > index 913aefa67618..24f1d03df9f1 100644
> > > --- a/drivers/media/platform/mtk-vcodec/vdec_msg_queue.c
> > > +++ b/drivers/media/platform/mtk-vcodec/vdec_msg_queue.c
> > > @@ -68,6 +68,9 @@ int vdec_msg_queue_qbuf(struct vdec_msg_queue_ctx
> > > *msg_ctx, struct vdec_lat_buf
> > >
> > > if (msg_ctx->hardware_index != MTK_VDEC_CORE)
> > > wake_up_all(&msg_ctx->ready_to_use);
> > > + else
> > > + queue_work(buf->ctx->dev->core_workqueue,
> > > + &buf->ctx->msg_queue.core_work);
> >
> > need {} for else here?
> >
> If condition not add "{}", else need not to add "{}" ?
I was suggesting as it had 2 lines.
> > >
> > > mtk_v4l2_debug(3, "enqueue buf type: %d addr: 0x%p num:
> > > %d",
> > > msg_ctx->hardware_index, buf, msg_ctx->ready_num);
> > > @@ -169,8 +172,7 @@ bool vdec_msg_queue_wait_lat_buf_full(struct
> > > vdec_msg_queue *msg_queue)
> > > return false;
> > > }
> > >
> > > -void vdec_msg_queue_deinit(
> > > - struct vdec_msg_queue *msg_queue,
> > > +void vdec_msg_queue_deinit(struct vdec_msg_queue *msg_queue,
> > > struct mtk_vcodec_ctx *ctx)
> > > {
> > > struct vdec_lat_buf *lat_buf;
> > > @@ -196,10 +198,36 @@ void vdec_msg_queue_deinit(
> > > }
> > > }
> > >
> > > -int vdec_msg_queue_init(
> > > - struct vdec_msg_queue *msg_queue,
> > > - struct mtk_vcodec_ctx *ctx,
> > > - core_decode_cb_t core_decode,
> > > +static void vdec_msg_queue_core_work(struct work_struct *work)
> > > +{
> > > + struct vdec_msg_queue *msg_queue =
> > > + container_of(work, struct vdec_msg_queue,
> > > core_work);
> > > + struct mtk_vcodec_ctx *ctx =
> > > + container_of(msg_queue, struct mtk_vcodec_ctx,
> > > msg_queue);
> > > + struct mtk_vcodec_dev *dev = ctx->dev;
> > > + struct vdec_lat_buf *lat_buf;
> > > +
> > > + lat_buf = vdec_msg_queue_dqbuf(&dev->msg_queue_core_ctx);
> > > + if (!lat_buf)
> > > + return;
> >
> > If we were to return in this error condition,
> > isn't it better to also differentiate this error with return code and
> > change void return type?
> >
> vdec_msg_queue_core_work function is callback for param "func" in
> struct work_struct, need not to add return value.
>
> > > +
> > > + ctx = lat_buf->ctx;
> > > + mtk_vcodec_set_curr_ctx(dev, ctx, MTK_VDEC_CORE);
> > > +
> > > + lat_buf->core_decode(lat_buf);
> > > +
> > > + mtk_vcodec_set_curr_ctx(dev, NULL, MTK_VDEC_CORE);
> > > + vdec_msg_queue_qbuf(&ctx->msg_queue.lat_ctx, lat_buf);
> > > +
> > > + if (!list_empty(&ctx->msg_queue.lat_ctx.ready_queue)) {
> > > + mtk_v4l2_debug(3, "re-schedule to decode for core",
> > > + dev->msg_queue_core_ctx.ready_num);
> > > + queue_work(dev->core_workqueue, &msg_queue-
> > > >core_work);
> > > + }
> > > +}
> > > +
> > > +int vdec_msg_queue_init(struct vdec_msg_queue *msg_queue,
> > > + struct mtk_vcodec_ctx *ctx, core_decode_cb_t
> > > core_decode,
> > > int private_size)
> > > {
> > > struct vdec_lat_buf *lat_buf;
> > > @@ -210,6 +238,7 @@ int vdec_msg_queue_init(
> > > return 0;
> > >
> > > vdec_msg_queue_init_ctx(&msg_queue->lat_ctx,
> > > MTK_VDEC_LAT0);
> > > + INIT_WORK(&msg_queue->core_work, vdec_msg_queue_core_work);
> > > msg_queue->wdma_addr.size = vde_msg_queue_get_trans_size(
> > > ctx->picinfo.buf_w, ctx->picinfo.buf_h);
> > >
> > > diff --git a/drivers/media/platform/mtk-vcodec/vdec_msg_queue.h
> > > b/drivers/media/platform/mtk-vcodec/vdec_msg_queue.h
> > > index 21a9c0aeb1b4..43eae638a2a8 100644
> > > --- a/drivers/media/platform/mtk-vcodec/vdec_msg_queue.h
> > > +++ b/drivers/media/platform/mtk-vcodec/vdec_msg_queue.h
> > > @@ -67,6 +67,7 @@ struct vdec_lat_buf {
> > > * @wdma_addr: wdma address used for ube
> > > * @wdma_rptr_addr: ube read point
> > > * @wdma_wptr_addr: ube write point
> > > + * @core_work: core hardware work
> > > * @lat_ctx: used to store lat buffer list
> > > */
> > > struct vdec_msg_queue {
> > > @@ -76,6 +77,7 @@ struct vdec_msg_queue {
> > > uint64_t wdma_rptr_addr;
> > > uint64_t wdma_wptr_addr;
> > >
> > > + struct work_struct core_work;
> > > struct vdec_msg_queue_ctx lat_ctx;
> > > };
> > >
> > > @@ -86,10 +88,8 @@ struct vdec_msg_queue {
> > > * @core_decode: core decode callback for each codec
> > > * @private_size: the private data size used to share with core
> > > */
> > > -int vdec_msg_queue_init(
> > > - struct vdec_msg_queue *msg_queue,
> > > - struct mtk_vcodec_ctx *ctx,
> > > - core_decode_cb_t core_decode,
> > > +int vdec_msg_queue_init(struct vdec_msg_queue *msg_queue,
> > > + struct mtk_vcodec_ctx *ctx, core_decode_cb_t
> > > core_decode,
> >
> > Not sure about the formatting rule, but is it supposed to be one
> > param per line?
> > If so, this comment also applied to function definition part.
> >
> > > int private_size);
> > >
> I try to review other files, it looks that two or more parameters per
> line is ok. Whether you mean that we'd better to write one parameter
> per line?
> int vdec_msg_queue_init(struct vdec_msg_queue *msg_queue,
> struct mtk_vcodec_ctx *ctx,
> core_decode_cb_t
Yes, that's what I meant, but you are welcome to keep current change.
>
> Thanks,
> Yunfei Dong
> > > /**
> > > --
> > > 2.25.1
> > >
Powered by blists - more mailing lists