DMA Engine development
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Bartosz Golaszewski" <bartosz.golaszewski@oss.qualcomm.com>
Cc: vkoul@kernel.org, Frank.Li@kernel.org, dmaengine@vger.kernel.org
Subject: Re: [PATCH v17 07/14] crypto: qce - Cancel work on device detach
Date: Tue, 19 May 2026 13:47:07 +0000	[thread overview]
Message-ID: <20260519134708.09803C2BCB3@smtp.kernel.org> (raw)
In-Reply-To: <20260519-qcom-qce-cmd-descr-v17-7-53a595414b79@oss.qualcomm.com>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Incorrect devres registration order allows DMA callbacks to schedule work after it has been cancelled, causing a use-after-destroy and use-after-free on driver unbind.
--

commit 0da71e23ef53eb17f76a32aa2d25f7d24ce0dd3a
Author: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>

crypto: qce - Cancel work on device detach

This commit adds a devres action to cancel done_work on device removal or
probe error, ensuring the workqueue is properly cleaned up.

Fixes: eb7986e5e14d ("crypto: qce - convert tasklet to workqueue")
Closes: https://sashiko.dev/#/patchset/20260427-qcom-qce-cmd-descr-v16-0-945fd1cafbbc%40oss.qualcomm.com?part=7

> diff --git a/drivers/crypto/qce/core.c b/drivers/crypto/qce/core.c
> index b966f3365b7de..e82fc862c74b2 100644
> --- a/drivers/crypto/qce/core.c
> +++ b/drivers/crypto/qce/core.c
[ ... ]
> @@ -240,6 +247,11 @@ static int qce_crypto_probe(struct platform_device *pdev)
>  		return ret;
>  
>  	INIT_WORK(&qce->done_work, qce_req_done_work);
> +
> +	ret = devm_add_action_or_reset(dev, qce_cancel_work, &qce->done_work);
> +	if (ret)
> +		return ret;
> +

Since devres cleanup actions execute in reverse order of their registration,
won't qce_cancel_work() be executed before the DMA channels are terminated?

If an active hardware DMA transfer completes immediately after
cancel_work_sync() returns, the DMA completion interrupt could fire. Could
this invoke the DMA callback and schedule_work(&qce->done_work), re-queuing
the work that was just cancelled?

Because the devres teardown will then proceed to destroy the mutex, release
the DMA channels, and free the qce structure, won't this result in a
use-after-free when qce_req_done_work() eventually executes?

Would it be safer to register the cancel action before requesting the DMA
channels via devm_qce_dma_request(), ensuring that the source of
asynchronous events is disabled before the background work is cancelled?

>  	crypto_init_queue(&qce->queue, QCE_QUEUE_LENGTH);
>  
>  	qce->async_req_enqueue = qce_async_request_enqueue;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260519-qcom-qce-cmd-descr-v17-0-53a595414b79@oss.qualcomm.com?part=7

  reply	other threads:[~2026-05-19 13:47 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-19 13:17 [PATCH v17 00/14] crypto/dmaengine: qce: introduce BAM locking and use DMA for register I/O Bartosz Golaszewski
2026-05-19 13:17 ` [PATCH v17 01/14] dmaengine: constify struct dma_descriptor_metadata_ops Bartosz Golaszewski
2026-05-19 13:17 ` [PATCH v17 02/14] dmaengine: qcom: bam_dma: free interrupt before the clock in error path Bartosz Golaszewski
2026-05-19 13:48   ` sashiko-bot
2026-05-19 13:17 ` [PATCH v17 03/14] dmaengine: qcom: bam_dma: convert tasklet to a BH workqueue Bartosz Golaszewski
2026-05-19 13:44   ` sashiko-bot
2026-05-19 13:17 ` [PATCH v17 04/14] dmaengine: qcom: bam_dma: Extend the driver's device match data Bartosz Golaszewski
2026-05-19 13:17 ` [PATCH v17 05/14] dmaengine: qcom: bam_dma: Add pipe_lock_supported flag support Bartosz Golaszewski
2026-05-19 13:17 ` [PATCH v17 06/14] dmaengine: qcom: bam_dma: add support for BAM locking Bartosz Golaszewski
2026-05-19 13:54   ` sashiko-bot
2026-05-19 13:17 ` [PATCH v17 07/14] crypto: qce - Cancel work on device detach Bartosz Golaszewski
2026-05-19 13:47   ` sashiko-bot [this message]
2026-05-19 13:17 ` [PATCH v17 08/14] crypto: qce - Include algapi.h in the core.h header Bartosz Golaszewski
2026-05-19 13:17 ` [PATCH v17 09/14] crypto: qce - Remove unused ignore_buf Bartosz Golaszewski
2026-05-19 13:48   ` sashiko-bot
2026-05-19 13:17 ` [PATCH v17 10/14] crypto: qce - Simplify arguments of devm_qce_dma_request() Bartosz Golaszewski
2026-05-19 13:17 ` [PATCH v17 11/14] crypto: qce - Use existing devres APIs in devm_qce_dma_request() Bartosz Golaszewski
2026-05-19 14:00   ` sashiko-bot
2026-05-19 13:17 ` [PATCH v17 12/14] crypto: qce - Map crypto memory for DMA Bartosz Golaszewski
2026-05-19 13:52   ` sashiko-bot
2026-05-19 13:17 ` [PATCH v17 13/14] crypto: qce - Add BAM DMA support for crypto register I/O Bartosz Golaszewski
2026-05-19 13:56   ` sashiko-bot
2026-05-19 13:17 ` [PATCH v17 14/14] crypto: qce - Communicate the base physical address to the dmaengine Bartosz Golaszewski
2026-05-19 14:17   ` sashiko-bot

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260519134708.09803C2BCB3@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=Frank.Li@kernel.org \
    --cc=bartosz.golaszewski@oss.qualcomm.com \
    --cc=dmaengine@vger.kernel.org \
    --cc=sashiko-reviews@lists.linux.dev \
    --cc=vkoul@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox