On 8/15/19 8:41 AM, Gregor Riepl wrote: >> Testing for "( defined(__sun) && defined(__SVR4))" actually assumes that >> Solaris runs on SPARC only which is not the case. Solaris runs on x86 >> as well and x86 doesn't have these alignment issues. Hence, the proper >> way is to detect the architecture and not the opering system. > > Doesn't that contradict with what the patch does? > > (...) > This merely adds a test for the SPARC architecture, but still keeps the > special case for Solaris. Hence it will still enable alignment on Solaris/x86.
Well, you're quoting part of my mail only which is where the confusion comes from. In the text, I'm actually saying that the current patch is certainly not the best approach. I am providing a patch to fix this issue in Debian and usually those patches are minimally invasive so that the maintainer easily understands what they are doing. At the same time, I'm suggesting a more generic approach for the fix that should be used upstream instead of the patch I'm providing for Debian. So, yes, the attached patch doesn't fix Solaris x86, but it's a Debian-specific patch anyway and will never be relevant for Solaris x86. It's up to upstream whether they are using the more generic approach where they replace __sun with __sparc which is what I suggested upstream [1]. Adrian > [1] https://tug.org/pipermail/tlbuild/2019q3/004551.html -- .''`. John Paul Adrian Glaubitz : :' : Debian Developer - [email protected] `. `' Freie Universitaet Berlin - [email protected] `- GPG: 62FF 8A75 84E0 2956 9546 0006 7426 3B37 F5B5 F913
