All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dan Carpenter <error27@gmail.com>
To: Antti Laakso <antti.laakso@linux.intel.com>
Cc: linux-media@vger.kernel.org
Subject: [bug report] media: ipu6: Move isys fw mapping to pci_probe
Date: Tue, 15 Sep 2026 17:32:33 +0300	[thread overview]
Message-ID: <aqlXAXcXvMM3HWtJ@stanley.mountain> (raw)

Hello Antti Laakso,

Commit 69ebb1bfc865 ("media: ipu6: Move isys fw mapping to
pci_probe") from Aug 27, 2026 (linux-next), leads to the following
Smatch static checker warning:

	drivers/media/pci/intel/ipu6/ipu6.c:865 ipu6_pci_remove()
	warn: variable dereferenced before check 'isp->psys' (see line 844)

drivers/media/pci/intel/ipu6/ipu6.c
    836 static void ipu6_pci_remove(struct pci_dev *pdev)
    837 {
    838         struct ipu6_device *isp = pci_get_drvdata(pdev);
    839         unsigned long dir;
    840 
    841         devm_free_irq(&pdev->dev, pdev->irq, isp);
    842 
    843         dir = IS_IPU7(isp) ? DMA_BIDIRECTIONAL : DMA_TO_DEVICE;
    844         ipu6_cpd_free_pkg_dir(isp->psys);

ipu6_cpd_free_pkg_dir() dereferences isp->psys so

    845         ipu6_unmap_fw_region(isp->psys, dir);
    846 
    847         if (isp->isys) {

is too late.  Fortunately, it's also unnecessary.  Just delete it.

    848                 ipu6_cpd_free_pkg_dir(isp->isys);
    849                 if (isp->isys->fw_sgt.nents)
    850                         ipu6_unmap_fw_region(isp->isys, dir);
    851         }
    852 
    853         vfree(isp->fw_code_region);
    854         isp->fw_code_region = NULL;
    855 
    856         ipu6_buttress_exit(isp);
    857 
    858         ipu6_bus_del_devices(pdev);

I'm using ChatGPT to review Smatch warnings these days and it
gives a bonus bug that ipu6_bus_del_devices() frees isp->psys.

"ipu6_bus_del_devices() calls:

	auxiliary_device_delete(&adev->auxdev);
	auxiliary_device_uninit(&adev->auxdev);

The uninit drops the device reference, and ipu6_bus_release() frees
the containing ipu6_bus_device. Consequently, accessing isp->psys->mmu
or isp->isys->mmu afterward can use freed memory."

    859 
    860         pm_runtime_forbid(&pdev->dev);
    861         pm_runtime_get_noresume(&pdev->dev);
    862 
    863         release_firmware(isp->cpd_fw);
    864 
--> 865         if (isp->psys)
    866                 ipu6_mmu_cleanup(isp->psys->mmu);
                                         ^^^^^^^^^

    867         if (isp->isys)
    868                 ipu6_mmu_cleanup(isp->isys->mmu);
                                         ^^^^^^^^^
So these dereferences are a use after free.  Originally, this code saved
the pointer the mmus before calling ipu6_bus_del_devices().

    869 }

This email is a free service from the Smatch-CI project [smatch.sf.net].

regards,
dan carpenter

                 reply	other threads:[~2026-09-15 14:32 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=aqlXAXcXvMM3HWtJ@stanley.mountain \
    --to=error27@gmail.com \
    --cc=antti.laakso@linux.intel.com \
    --cc=linux-media@vger.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.