… OMP formats should be revised, but there are so many :-/ The optimum may differ between different CPUs so I started do document what it was tuned for. This also states that is *is* tuned at all, so don't you dare just copying a format and leave that comment...
Signed-off-by: Dhiru Kholia <email@example.com> Signed-off-by: Narendra Kangralkar <firstname.lastname@example.org>
…ous fix did not catch all cases.
…is practically" This reverts commit 3b53da3. Did not work for me. Segfaulted in unstable, hung in bleeding.
…ically the same but this is cleaner.
mem_alloc_tiny() is changed to a normal alloc. But it was supposed to still maintain requested alignment!