Re: CVE-2024-39471: drm/amdgpu: add error handle to avoid out-of-bounds

Re: CVE-2024-39471: drm/amdgpu: add error handle to avoid out-of-bounds

[Siddh Raman Pant]

[-- Attachment #1: Type: text/plain, Size: 605 bytes --]

On Tue, 25 Jun 2024 16:29:04 +0200, Greg Kroah-Hartman wrote:
> In the Linux kernel, the following vulnerability has been resolved:
> 
> drm/amdgpu: add error handle to avoid out-of-bounds
> 
> if the sdma_v4_0_irq_id_to_seq return -EINVAL, the process should
> be stop to avoid out-of-bounds read, so directly return -EINVAL.
> 
> The Linux kernel CVE team has assigned CVE-2024-39471 to this issue.

This commit has a bug which was fixed by 6769a23697f1. It should be
immediately backported, otherwise this "fix" doesn't do anything since
gcc will optimise out the check.

Thanks,
Siddh

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: CVE-2024-39471: drm/amdgpu: add error handle to avoid out-of-bounds

[gregkh]

On Wed, Jul 24, 2024 at 10:05:53AM +0000, Siddh Raman Pant wrote:
> On Tue, 25 Jun 2024 16:29:04 +0200, Greg Kroah-Hartman wrote:
> > In the Linux kernel, the following vulnerability has been resolved:
> > 
> > drm/amdgpu: add error handle to avoid out-of-bounds
> > 
> > if the sdma_v4_0_irq_id_to_seq return -EINVAL, the process should
> > be stop to avoid out-of-bounds read, so directly return -EINVAL.
> > 
> > The Linux kernel CVE team has assigned CVE-2024-39471 to this issue.
> 
> This commit has a bug which was fixed by 6769a23697f1. It should be
> immediately backported, otherwise this "fix" doesn't do anything since
> gcc will optimise out the check.

<formletter>

This is not the correct way to submit patches for inclusion in the
stable kernel tree.  Please read:
    https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
for how to do this properly.

</formletter>

(hint, send the request to stable@vger.kernel.org, not
cve@kernel.org...)

Also, any specific reason why that commit was NOT marked for stable
inclusion by the developers/maintainers involved with it and you didn't
cc any of them here either :(

thanks,

greg k-h

Re: CVE-2024-39471: drm/amdgpu: add error handle to avoid out-of-bounds

[Siddh Raman Pant]

[-- Attachment #1: Type: text/plain, Size: 1346 bytes --]

On Wed, Jul 24 2024 at 15:44:12 +0530, gregkh@linuxfoundation.org
wrote:
> On Wed, Jul 24, 2024 at 10:05:53AM +0000, Siddh Raman Pant wrote:
> > On Tue, 25 Jun 2024 16:29:04 +0200, Greg Kroah-Hartman wrote:
> > > In the Linux kernel, the following vulnerability has been resolved:
> > > 
> > > drm/amdgpu: add error handle to avoid out-of-bounds
> > > 
> > > if the sdma_v4_0_irq_id_to_seq return -EINVAL, the process should
> > > be stop to avoid out-of-bounds read, so directly return -EINVAL.
> > > 
> > > The Linux kernel CVE team has assigned CVE-2024-39471 to this issue.
> > 
> > This commit has a bug which was fixed by 6769a23697f1. It should be
> > immediately backported, otherwise this "fix" doesn't do anything since
> > gcc will optimise out the check.
> 
> <formletter>
> 
> This is not the correct way to submit patches for inclusion in the
> stable kernel tree.  Please read:
>     https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html
> for how to do this properly.
> 
> </formletter>
> 
> (hint, send the request to stable@vger.kernel.org, not
> cve@kernel.org...)
> 
> Also, any specific reason why that commit was NOT marked for stable
> inclusion by the developers/maintainers involved with it and you didn't
> cc any of them here either :(

Oops, sorry, my bad.

Siddh

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: CVE-2024-39471: drm/amdgpu: add error handle to avoid out-of-bounds

[Siddh Raman Pant]

[-- Attachment #1: Type: text/plain, Size: 675 bytes --]

(Mail V2: Send to correct mailing list and CCing relevant people.)

On Tue, 25 Jun 2024 16:29:04 +0200, Greg Kroah-Hartman wrote:
> In the Linux kernel, the following vulnerability has been resolved:
> 
> drm/amdgpu: add error handle to avoid out-of-bounds
> 
> if the sdma_v4_0_irq_id_to_seq return -EINVAL, the process should
> be stop to avoid out-of-bounds read, so directly return -EINVAL.
> 
> The Linux kernel CVE team has assigned CVE-2024-39471 to this issue.

This commit has a bug which was fixed by 6769a23697f1. It should be
immediately backported, otherwise this "fix" doesn't do anything since
gcc will optimise out the check.

Thanks,
Siddh

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]