Re: [PATCH] wifi: mwifiex: bound SDIO fw dump count by memory table size

[Brian Norris <briannorris@chromium.org>]

On Fri, Sep 04, 2026 at 12:57:55PM +0000, Tianchu Chen wrote:
> From: Tianchu Chen <flynnnchen@tencent.com>
> 
> mwifiex_sdio_fw_dump() reads the number of memory regions (dump_num) from 
> a device register and uses it to index the static 15-entry
> mem_type_mapping_tbl[] without any bounds check. A bogus device can
> report a dump_num larger than the table, making each loop iteration
> store a vmalloc pointer and a device-controlled size past the end
> of the table into adjacent .data (up to ~5.7KB when dump_num is 255).
> 
> The dump path runs after command timeouts or firmware crashes, so a
> bogus device can first force an error to trigger the dump and
> then report a bogus dump_num.
> 
> Aborting the dump only loses the firmware snapshot; the card reset 
> that follows is unaffected, so error recovery still completes.
> 
> Discovered by Atuin - Automated Vulnerability Discovery Engine.
> 
> Fixes: 54881c6b37c8d ("mwifiex: add firmware dump feature for SDIO")
> Cc: stable@vger.kernel.org
> Assisted-by: LLM
> Signed-off-by: Tianchu Chen <flynnnchen@tencent.com>

I'm not sure if we're in the business of distrusting SDIO registers, but
sure?

Acked-by: Brian Norris <briannorris@chromium.org>

We also read this same register a bit later and use it for computing
potentially-enormous buffers to allocate and fill. Dunno if we consider
OOM a potential vulnerability (or infinite loop / DoS, or ...).