On 5 September 2018 at 10:54, Gert Wollny <gert.wol...@collabora.com> wrote:
> I see, my "ralloc" was just a typo then. What I meant was that when > realloc succeeds then the old pointer is either invalid or its > allocation was directly extended. In both cases your proposal to free > the new pointer if the second reallocation fails is not correct, > because the memory the original pointer pointed to is no longer > available like before the realloc call. > You're spot on - I withdraw my buggy suggestion :-\ Skimming through Mesa - there are very few cases were we care about realloc (regardless of case) failure. So my suggestion all together is a bit of meh. Pardon for the noise -Emil _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
