Re: [PATCH v2] ASoC: Intel: avs: Fix potential buffer overflow by snprintf()

[Greg KH]

On Thu, Nov 13, 2025 at 09:46:12AM +0100, Cezary Rojewski wrote:
> On 2025-11-12 8:20 PM, Greg KH wrote:
> > On Wed, Nov 12, 2025 at 11:48:51PM +0530, hariconscious@xxxxxxxxx wrote:
> > > From: HariKrishna Sagala <hariconscious@xxxxxxxxx>
> > > 
> > > snprintf() returns the would-be-filled size when the string overflows
> > > the given buffer size, hence using this value may result in a buffer
> > > overflow (although it's unrealistic).
> > 
> > unrealistic == impossible
> > 
> > So why make this change at all?
> 
> The problem will never occur in production-scenario given the AudioDSP
> firmware limitation - max ~10 probe-point entries so, the built string will
> be far away from 4K_SZ bytes.
> 
> If the verdict is: ignore the recommendation as the problem is unrealistic,
> I'm OK with that. Typically though I'd prefer to stick to the
> recommendations.

That's fine, but don't claim that it fixes a buffer overflow when that
is NOT what this is doing at all.

> > > This patch replaces it with a safer version, scnprintf() for papering
> > > over such a potential issue.
> > 
> > Don't "paper over", actually fix real things.
> > 
> > 
> > > Link: https://github.com/KSPP/linux/issues/105
> > > 'Fixes: 5a565ba23abe ("ASoC: Intel: avs: Probing and firmware tracing
> > > over debugfs")'
> > 
> > No, this is not a "fix".
> 
> The patch isn't worded well, that's clear.
> While the patch is an outcome of static-analysis, isn't it good to have
> 'Fixes:' to point out the offending commit regardless?

No, it is not "fixing" anything.  Please don't claim that it does.  It
is "just" a code transformation to get rid of an api that some people do
not like.

thanks,

greg k-h