Re: [PATCH] staging: vme_user: validate slave window size against buffer size

["gregkh@linuxfoundation.org" <gregkh@linuxfoundation.org>]

On Sat, May 09, 2026 at 03:58:45PM +0900, 木口璃音 wrote:
> This patch addresses the OOB read/write reported earlier
> in the security@kernel.org thread (now handled publicly per
> Greg's and Willy's guidance).
> 
> Tested on Linux 7.1.0-rc2 with KASAN; all three reproducers
> fail with -EINVAL after applying this patch and produce no
> KASAN splat.
> 
> >From 506ecfc9b8608fb3a56477b8fd205238a1bf66ff Mon Sep 17 00:00:00 2001
> From: Rion Kiguchi <kiguchi.r.sec@gmail.com>
> Date: Sat, 9 May 2026 15:38:33 +0900
> Subject: [PATCH] staging: vme_user: validate slave window size against buffer
>  size

This all doesn't belong in the body of the email, please just use git
send-email to send the patch.

> 
> The VME_SET_SLAVE ioctl in drivers/staging/vme_user/vme_user.c accepts
> a user-controlled slave.size and forwards it to vme_slave_set() without
> comparing it against image[minor].size_buf. The slave-image kernel
> buffer is allocated at probe time with a fixed size of PCI_BUF_SIZE
> (0x20000 / 128 KiB), but the configured VME window size can be made
> much larger via the ioctl.
> 
> The subsequent read() / write() handlers (vme_user_read /
> vme_user_write) clamp the I/O range against vme_get_size() (the
> configured window size, attacker-controlled) but never consult
> size_buf. The slave I/O paths buffer_to_user() and buffer_from_user()
> then index image[minor].kern_buf with *ppos values up to
> image_size - 1, well beyond the actual allocation.
> 
> Result: a local user with read/write access to /dev/bus/vme/s* can
> trigger out-of-bounds read and write of the kernel slab adjacent to
> the slave-image buffer.
> 
> Fix: reject slave.size > size_buf in the VME_SET_SLAVE handler. Also
> add defensive bounds checks against size_buf in buffer_to_user() and
> buffer_from_user() so that the I/O paths cannot exceed the
> allocation even if a future ioctl path forgets to validate.
> 
> Reported-by: Pochix1103 <kiguchi.r.sec@gmail.com>

Ok, but:

> Cc: stable@vger.kernel.org
> Signed-off-by: Rion Kiguchi <kiguchi.r.sec@gmail.com>

You don't have a reported-by and a signed-off-by for the same thing, if
you author and sign off, it's implied you are reporting the issue :)

Also, you have to document the AI tool you used to find and fix this as
per out documentation.

thanks,

greg k-h