From: Patrick Steinhardt <ps@pks.im>
Date: Thu, 01 Oct 2026 13:15:51 GMT
Subject: Re: [PATCH v2 4/7] xdiff: NUL-terminate buffers read by read_mmfile()
Message-ID: <ar5dB4p6pQITEUq6@pks.im>
In-Reply-To: <20260930234413.GD1347555@coredump.intra.peff.net>

On Wed, Sep 30, 2026 at 07:44:13PM -0400, Jeff King wrote:
> Since an mmfile_t is a ptr/len pair, our read_mmfile() allocates exactly
> the number of bytes we claim to store. But in many other places in Git,
> we add an extra NUL "just in case", which can help avoid read overruns
> due to off-by-ones or the use of string functions.
> 
> I don't know of any path that would benefit from this, but I noticed it
> while converting ll_ext_merge() to use read_mmfile(), since its original
> code did add a NUL byte (even though I cannot find any case where it
> would have mattered). Let's add the same defensive NUL in read_mmfile()
> by using xmallocz() instead of xmalloc().

Nit: I guess this is an artifact from the reorder, but this sounds as if
`ll_ext_merge()` wouldn't append the NUL byte anymore. But at this step
it still does, as the change to `read_mmfile()` now happens before the
change to `ll_ext_merge()`.

Patrick