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, dmaengine@vger.kernel.org, Frank.Li@kernel.org
Subject: Re: [PATCH v23 10/14] crypto: qce - Simplify arguments of devm_qce_dma_request()
Date: Tue, 21 Jul 2026 13:44:35 +0000	[thread overview]
Message-ID: <20260721134436.694E71F00A3A@smtp.kernel.org> (raw)
In-Reply-To: <20260721-qcom-qce-cmd-descr-v23-10-1be02deffda9@oss.qualcomm.com>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Device unbind sequence incorrectly releases DMA channels and frees memory before the background work item is cancelled, allowing the work item to trigger a Use-After-Free on these resources.
--

--- Patch [10]: [PATCH v23 10/14] crypto: qce - Simplify arguments of devm_qce_dma_request() ---
commit 6680b197a5f4cdb8f7389167e1c7991cd29ab241
Author: Bartosz Golaszewski <brgl@kernel.org>
Subject: crypto: qce - Simplify arguments of devm_qce_dma_request()

This function can extract all the information it needs from struct
qce_device alone so simplify its arguments. This is done in preparation
for adding support for register I/O over DMA which will require
accessing even more fields from struct qce_device.

> diff --git a/drivers/crypto/qce/core.c b/drivers/crypto/qce/core.c
> index dd860435d2c47..aa4a0b1774908 100644
> --- a/drivers/crypto/qce/core.c
> +++ b/drivers/crypto/qce/core.c
> @@ -237,7 +237,7 @@ static int qce_crypto_probe(struct platform_device *pdev)
>  	if (ret)
>  		return ret;
>  
> -	ret = devm_qce_dma_request(qce->dev, &qce->dma);
> +	ret = devm_qce_dma_request(qce);

[Severity: High]
Does this devres teardown order create a use-after-free during device unbind?

In qce_crypto_probe(), devm_add_action_or_reset() is called for qce_cancel_work
before calling devm_qce_dma_request():

drivers/crypto/qce/core.c:qce_crypto_probe() {
    ...
	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;
    ...
	ret = devm_qce_dma_request(qce);
    ...
}

Since devres actions execute in reverse order of registration, the DMA resources
freed in qce_dma_release() (which is registered inside devm_qce_dma_request)
will be released before the background work is cancelled.

If the device is unbound while a crypto request is pending, qce_dma_release()
will free the DMA channels and the heap-allocated dma->result_buf.

The background work qce->done_work could then run, dequeue the next request,
and submit it to the freed DMA channels via qce_dma_prep_sgs(), writing to the
freed dma->result_buf.

A later fix in commit "crypto: qce - Cancel work on device detach" resolves
this inverted teardown order by grouping termination and freeing together.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260721-qcom-qce-cmd-descr-v23-0-1be02deffda9@oss.qualcomm.com?part=10

  reply	other threads:[~2026-07-21 13:44 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-21 13:33 [PATCH v23 00/14] crypto/dmaengine: qce: introduce BAM locking and use DMA for register I/O Bartosz Golaszewski
2026-07-21 13:33 ` [PATCH v23 01/14] dmaengine: constify struct dma_descriptor_metadata_ops Bartosz Golaszewski
2026-07-21 13:33 ` [PATCH v23 02/14] dmaengine: qcom: bam_dma: free interrupt before the clock in error path Bartosz Golaszewski
2026-07-21 13:48   ` sashiko-bot
2026-07-21 13:33 ` [PATCH v23 03/14] dmaengine: qcom: bam_dma: convert tasklet to a BH workqueue Bartosz Golaszewski
2026-07-21 13:52   ` sashiko-bot
2026-07-21 13:33 ` [PATCH v23 04/14] dmaengine: qcom: bam_dma: Extend the driver's device match data Bartosz Golaszewski
2026-07-21 13:33 ` [PATCH v23 05/14] dmaengine: qcom: bam_dma: Add pipe_lock_supported flag support Bartosz Golaszewski
2026-07-21 13:45   ` sashiko-bot
2026-07-21 13:33 ` [PATCH v23 06/14] dmaengine: qcom: bam_dma: add support for BAM locking Bartosz Golaszewski
2026-07-21 13:48   ` sashiko-bot
2026-07-21 18:07   ` Mukesh Savaliya
2026-07-21 13:33 ` [PATCH v23 07/14] crypto: qce - Cancel work on device detach Bartosz Golaszewski
2026-07-21 13:47   ` sashiko-bot
2026-07-21 13:33 ` [PATCH v23 08/14] crypto: qce - Include algapi.h in the core.h header Bartosz Golaszewski
2026-07-21 13:33 ` [PATCH v23 09/14] crypto: qce - Remove unused ignore_buf Bartosz Golaszewski
2026-07-21 13:46   ` sashiko-bot
2026-07-21 13:33 ` [PATCH v23 10/14] crypto: qce - Simplify arguments of devm_qce_dma_request() Bartosz Golaszewski
2026-07-21 13:44   ` sashiko-bot [this message]
2026-07-21 13:33 ` [PATCH v23 11/14] crypto: qce - Use existing devres APIs in devm_qce_dma_request() Bartosz Golaszewski
2026-07-21 13:47   ` sashiko-bot
2026-07-21 18:16   ` Mukesh Savaliya
2026-07-21 13:33 ` [PATCH v23 12/14] crypto: qce - Map crypto memory for DMA Bartosz Golaszewski
2026-07-21 13:48   ` sashiko-bot
2026-07-21 18:20   ` Mukesh Savaliya
2026-07-21 13:33 ` [PATCH v23 13/14] crypto: qce - Add BAM DMA support for crypto register I/O Bartosz Golaszewski
2026-07-21 13:50   ` sashiko-bot
2026-07-21 13:33 ` [PATCH v23 14/14] crypto: qce - Communicate the base physical address to the dmaengine Bartosz Golaszewski
2026-07-21 13:55   ` 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=20260721134436.694E71F00A3A@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