Re: [PATCH 2/3] read-cache: use get_be32 instead of hand-rolled ntoh_l

On Thu, Jan 23, 2014 at 03:34:16PM -0800, Jonathan Nieder wrote:

Line 1484 looks more problematic:

		disk_ce = (struct ondisk_cache_entry *)((char *)mmap + src_offset);

In v4 indexes, src_offset doesn't have any particular alignment so
this conversion has undefined behavior.

Do you know if any tests exercise this code with paths that don't
have convenient length?

My impression was that we are not testing v4 index at all (and grepping
for `--index-version`, which I think is the only way to write it,
supports that).

-Peff