Re: [PATCH v2] media: mtk-jpeg: drain hardware completion before freeing context

[Nicolas Dufresne <nicolas.dufresne@collabora.com>]

[-- Attachment #1: Type: text/plain, Size: 12066 bytes --]

Le samedi 18 juillet 2026 à 19:24 +0800, Guangshuo Li a écrit :
> The change referenced by the Fixes tag cancels ctx->jpeg_work before
> freeing the JPEG context.
> 
> That prevents the worker itself from accessing the context after it has
> been freed. However, the multi-core workers return after programming a
> hardware instance. The IRQ handler or the per-hardware timeout work can
> then continue to access the context through hw_param.curr_ctx after the
> worker has completed.
> 
> If userspace closes the file while a hardware job is still pending,
> cancel_work_sync() can return and mtk_jpeg_release() can free the
> context before the IRQ or timeout path has finished using it.
> 
> Track the number of in-flight hardware jobs for each context and wait
> for them to complete in mtk_jpeg_release().
> 
> Use a per-hardware active flag to ensure that only the IRQ handler or
> the timeout work completes each job. Clear hw_param.curr_ctx and drop
> the in-flight count only after the completion path has finished all
> accesses to the context.
> 
> Fixes: 34c519feef3e ("media: mtk-jpeg: fix use-after-free in release path due to uncancelled work")
> Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com>
> ---
> v2:
>   - Replace the undefined mtk_jpeg_release_hw() call with per-context
>     in-flight hardware job tracking.
>   - Ensure that only the IRQ or timeout path completes each hardware job.
>   - Clear hw_param.curr_ctx after the completion path stops using the
>     context.
>   - Keep hardware instances busy until the timeout path has finished
>     accessing their saved context and buffers.
> 
>  .../platform/mediatek/jpeg/mtk_jpeg_core.c    | 12 ++++++++-
>  .../platform/mediatek/jpeg/mtk_jpeg_core.h    | 14 +++++++++++
>  .../platform/mediatek/jpeg/mtk_jpeg_dec_hw.c  | 25 +++++++++++++------
>  .../platform/mediatek/jpeg/mtk_jpeg_enc_hw.c  | 25 +++++++++++++------
>  4 files changed, 61 insertions(+), 15 deletions(-)
> 
> diff --git a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.c b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.c
> index d147ec483081..d41c0e0516b9 100644
> --- a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.c
> +++ b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.c
> @@ -1161,6 +1161,8 @@ static int mtk_jpeg_open(struct file *file)
>  	}
>  
>  	INIT_WORK(&ctx->jpeg_work, jpeg->variant->jpeg_worker);
> +	atomic_set(&ctx->hw_jobs, 0);
> +	init_waitqueue_head(&ctx->hw_jobs_wq);
>  	INIT_LIST_HEAD(&ctx->dst_done_queue);
>  	spin_lock_init(&ctx->done_queue_lock);
>  	v4l2_fh_init(&ctx->fh, vfd);
> @@ -1202,8 +1204,12 @@ static int mtk_jpeg_release(struct file *file)
>  	struct mtk_jpeg_dev *jpeg = video_drvdata(file);
>  	struct mtk_jpeg_ctx *ctx = mtk_jpeg_file_to_ctx(file);
>  
> -	if (jpeg->variant->jpeg_worker)
> +	if (jpeg->variant->jpeg_worker) {
>  		cancel_work_sync(&ctx->jpeg_work);
> +		wait_event(ctx->hw_jobs_wq,
> +			   atomic_read(&ctx->hw_jobs) == 0);
> +	}
> +
>  	mutex_lock(&jpeg->lock);
>  	v4l2_m2m_ctx_release(ctx->fh.m2m_ctx);

The things is that we already paid the price of this complexity in the m2m
framework by synchronously waiting for the last v4l2_m2m_job_finish() to be
called from the driver. I think we should not add a mechanism on top, but rather
figure-out why this is not working for your usage of this driver.

Nicolas

>  	v4l2_ctrl_handler_free(&ctx->ctrl_hdl);
> @@ -1640,6 +1646,8 @@ static void mtk_jpegenc_worker(struct work_struct *work)
>  	v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
>  	v4l2_m2m_dst_buf_remove(ctx->fh.m2m_ctx);
>  
> +	atomic_inc(&ctx->hw_jobs);
> +	atomic_set_release(&comp_jpeg[hw_id]->hw_param.job_active, 1);
>  	schedule_delayed_work(&comp_jpeg[hw_id]->job_timeout_work,
>  			      msecs_to_jiffies(MTK_JPEG_HW_TIMEOUT_MSEC));
>  
> @@ -1759,6 +1767,8 @@ static void mtk_jpegdec_worker(struct work_struct *work)
>  		goto setdst_end;
>  	}
>  
> +	atomic_inc(&ctx->hw_jobs);
> +	atomic_set_release(&comp_jpeg[hw_id]->hw_param.job_active, 1);
>  	schedule_delayed_work(&comp_jpeg[hw_id]->job_timeout_work,
>  			      msecs_to_jiffies(MTK_JPEG_HW_TIMEOUT_MSEC));
>  
> diff --git a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.h b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.h
> index 02ed0ed5b736..be8ada60782a 100644
> --- a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.h
> +++ b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_core.h
> @@ -103,6 +103,7 @@ struct mtk_jpeg_hw_param {
>  	struct vb2_v4l2_buffer *src_buffer;
>  	struct vb2_v4l2_buffer *dst_buffer;
>  	struct mtk_jpeg_ctx *curr_ctx;
> +	atomic_t job_active;
>  };
>  
>  enum mtk_jpegenc_hw_id {
> @@ -282,6 +283,8 @@ struct mtk_jpeg_q_data {
>   * @restart_interval:		jpeg encoder restart interval
>   * @ctrl_hdl:			controls handler
>   * @jpeg_work:			jpeg encoder workqueue
> + * @hw_jobs:			number of hardware jobs still referencing this context
> + * @hw_jobs_wq:			wait queue for hardware job completion
>   * @total_frame_num:		encoded frame number
>   * @dst_done_queue:		encoded frame buffer queue
>   * @done_queue_lock:		encoded frame operation spinlock
> @@ -299,6 +302,8 @@ struct mtk_jpeg_ctx {
>  	struct v4l2_ctrl_handler ctrl_hdl;
>  
>  	struct work_struct jpeg_work;
> +	atomic_t hw_jobs;
> +	wait_queue_head_t hw_jobs_wq;
>  	u32 total_frame_num;
>  	struct list_head dst_done_queue;
>  	/* spinlock protecting the encode done buffer */
> @@ -306,4 +311,13 @@ struct mtk_jpeg_ctx {
>  	u32 last_done_frame_num;
>  };
>  
> +static inline void
> +mtk_jpeg_hw_job_done(struct mtk_jpeg_hw_param *hw_param,
> +		     struct mtk_jpeg_ctx *ctx)
> +{
> +	WRITE_ONCE(hw_param->curr_ctx, NULL);
> +	if (atomic_dec_and_test(&ctx->hw_jobs))
> +		wake_up(&ctx->hw_jobs_wq);
> +}
> +
>  #endif /* _MTK_JPEG_CORE_H */
> diff --git a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_dec_hw.c b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_dec_hw.c
> index 32372781daf5..c2d40653891a 100644
> --- a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_dec_hw.c
> +++ b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_dec_hw.c
> @@ -527,7 +527,12 @@ static void mtk_jpegdec_timeout_work(struct work_struct *work)
>  			     job_timeout_work.work);
>  	struct mtk_jpeg_dev *master_jpeg = cjpeg->master_dev;
>  	struct vb2_v4l2_buffer *src_buf, *dst_buf;
> +	struct mtk_jpeg_ctx *ctx;
> +
> +	if (atomic_cmpxchg(&cjpeg->hw_param.job_active, 1, 0) != 1)
> +		return;
>  
> +	ctx = cjpeg->hw_param.curr_ctx;
>  	src_buf = cjpeg->hw_param.src_buffer;
>  	dst_buf = cjpeg->hw_param.dst_buffer;
>  	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf);
> @@ -535,11 +540,13 @@ static void mtk_jpegdec_timeout_work(struct work_struct *work)
>  	mtk_jpeg_dec_reset(cjpeg->reg_base);
>  	clk_disable_unprepare(cjpeg->jdec_clk.clks->clk);
>  	pm_runtime_put(cjpeg->dev);
> +	v4l2_m2m_buf_done(src_buf, buf_state);
> +	mtk_jpegdec_put_buf(cjpeg);
> +	mtk_jpeg_hw_job_done(&cjpeg->hw_param, ctx);
> +
>  	cjpeg->hw_state = MTK_JPEG_HW_IDLE;
>  	atomic_inc(&master_jpeg->hw_rdy);
>  	wake_up(&master_jpeg->hw_wq);
> -	v4l2_m2m_buf_done(src_buf, buf_state);
> -	mtk_jpegdec_put_buf(cjpeg);
>  }
>  
>  static irqreturn_t mtk_jpegdec_hw_irq_handler(int irq, void *priv)
> @@ -555,12 +562,10 @@ static irqreturn_t mtk_jpegdec_hw_irq_handler(int irq, void *priv)
>  	struct mtk_jpegdec_comp_dev *jpeg = priv;
>  	struct mtk_jpeg_dev *master_jpeg = jpeg->master_dev;
>  
> -	cancel_delayed_work(&jpeg->job_timeout_work);
> +	if (atomic_cmpxchg(&jpeg->hw_param.job_active, 1, 0) != 1)
> +		return IRQ_HANDLED;
>  
> -	ctx = jpeg->hw_param.curr_ctx;
> -	src_buf = jpeg->hw_param.src_buffer;
> -	dst_buf = jpeg->hw_param.dst_buffer;
> -	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf);
> +	cancel_delayed_work(&jpeg->job_timeout_work);
>  
>  	irq_status = mtk_jpeg_dec_get_int_status(jpeg->reg_base);
>  	dec_irq_ret = mtk_jpeg_dec_enum_result(irq_status);
> @@ -570,6 +575,11 @@ static irqreturn_t mtk_jpegdec_hw_irq_handler(int irq, void *priv)
>  	if (dec_irq_ret != MTK_JPEG_DEC_RESULT_EOF_DONE)
>  		dev_warn(jpeg->dev, "Jpg Dec occurs unknown Err.");
>  
> +	ctx = jpeg->hw_param.curr_ctx;
> +	src_buf = jpeg->hw_param.src_buffer;
> +	dst_buf = jpeg->hw_param.dst_buffer;
> +	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf);
> +
>  	jpeg_src_buf =
>  		container_of(src_buf, struct mtk_jpeg_src_buf, b);
>  
> @@ -582,6 +592,7 @@ static irqreturn_t mtk_jpegdec_hw_irq_handler(int irq, void *priv)
>  	mtk_jpegdec_put_buf(jpeg);
>  	pm_runtime_put(ctx->jpeg->dev);
>  	clk_disable_unprepare(jpeg->jdec_clk.clks->clk);
> +	mtk_jpeg_hw_job_done(&jpeg->hw_param, ctx);
>  
>  	jpeg->hw_state = MTK_JPEG_HW_IDLE;
>  	wake_up(&master_jpeg->hw_wq);
> diff --git a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_enc_hw.c b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_enc_hw.c
> index b312a15d707b..b08f94845613 100644
> --- a/drivers/media/platform/mediatek/jpeg/mtk_jpeg_enc_hw.c
> +++ b/drivers/media/platform/mediatek/jpeg/mtk_jpeg_enc_hw.c
> @@ -257,7 +257,12 @@ static void mtk_jpegenc_timeout_work(struct work_struct *work)
>  	struct mtk_jpeg_dev *master_jpeg = cjpeg->master_dev;
>  	enum vb2_buffer_state buf_state = VB2_BUF_STATE_ERROR;
>  	struct vb2_v4l2_buffer *src_buf, *dst_buf;
> +	struct mtk_jpeg_ctx *ctx;
> +
> +	if (atomic_cmpxchg(&cjpeg->hw_param.job_active, 1, 0) != 1)
> +		return;
>  
> +	ctx = cjpeg->hw_param.curr_ctx;
>  	src_buf = cjpeg->hw_param.src_buffer;
>  	dst_buf = cjpeg->hw_param.dst_buffer;
>  	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf);
> @@ -265,11 +270,13 @@ static void mtk_jpegenc_timeout_work(struct work_struct *work)
>  	mtk_jpeg_enc_reset(cjpeg->reg_base);
>  	clk_disable_unprepare(cjpeg->venc_clk.clks->clk);
>  	pm_runtime_put(cjpeg->dev);
> +	v4l2_m2m_buf_done(src_buf, buf_state);
> +	mtk_jpegenc_put_buf(cjpeg);
> +	mtk_jpeg_hw_job_done(&cjpeg->hw_param, ctx);
> +
>  	cjpeg->hw_state = MTK_JPEG_HW_IDLE;
>  	atomic_inc(&master_jpeg->hw_rdy);
>  	wake_up(&master_jpeg->hw_wq);
> -	v4l2_m2m_buf_done(src_buf, buf_state);
> -	mtk_jpegenc_put_buf(cjpeg);
>  }
>  
>  static irqreturn_t mtk_jpegenc_hw_irq_handler(int irq, void *priv)
> @@ -283,12 +290,10 @@ static irqreturn_t mtk_jpegenc_hw_irq_handler(int irq, void *priv)
>  	struct mtk_jpegenc_comp_dev *jpeg = priv;
>  	struct mtk_jpeg_dev *master_jpeg = jpeg->master_dev;
>  
> -	cancel_delayed_work(&jpeg->job_timeout_work);
> +	if (atomic_cmpxchg(&jpeg->hw_param.job_active, 1, 0) != 1)
> +		return IRQ_HANDLED;
>  
> -	ctx = jpeg->hw_param.curr_ctx;
> -	src_buf = jpeg->hw_param.src_buffer;
> -	dst_buf = jpeg->hw_param.dst_buffer;
> -	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf);
> +	cancel_delayed_work(&jpeg->job_timeout_work);
>  
>  	irq_status = readl(jpeg->reg_base + JPEG_ENC_INT_STS) &
>  		JPEG_ENC_INT_STATUS_MASK_ALLIRQ;
> @@ -297,6 +302,11 @@ static irqreturn_t mtk_jpegenc_hw_irq_handler(int irq, void *priv)
>  	if (!(irq_status & JPEG_ENC_INT_STATUS_DONE))
>  		dev_warn(jpeg->dev, "Jpg Enc occurs unknown Err.");
>  
> +	ctx = jpeg->hw_param.curr_ctx;
> +	src_buf = jpeg->hw_param.src_buffer;
> +	dst_buf = jpeg->hw_param.dst_buffer;
> +	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf);
> +
>  	result_size = mtk_jpeg_enc_get_file_size(jpeg->reg_base,
>  						 ctx->jpeg->variant->support_34bit);
>  	vb2_set_plane_payload(&dst_buf->vb2_buf, 0, result_size);
> @@ -305,6 +315,7 @@ static irqreturn_t mtk_jpegenc_hw_irq_handler(int irq, void *priv)
>  	mtk_jpegenc_put_buf(jpeg);
>  	pm_runtime_put(ctx->jpeg->dev);
>  	clk_disable_unprepare(jpeg->venc_clk.clks->clk);
> +	mtk_jpeg_hw_job_done(&jpeg->hw_param, ctx);
>  
>  	jpeg->hw_state = MTK_JPEG_HW_IDLE;
>  	wake_up(&master_jpeg->hw_wq);

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 228 bytes --]