On 08/05/2016 11:07 AM, Guillaume Munch wrote: > Le 05/08/2016 à 08:39, Stephan Witt a écrit : >> Am 04.08.2016 um 14:55 schrieb Guillaume Munch <g...@lyx.org>: >>> >>> Le 04/08/2016 à 08:48, Stephan Witt a écrit : >>>> >>>> On my Mac with clang I cannot compile after this commit :( >>>> >>>> src/frontends/qt4/GuiFontLoader.cpp:380:2: error: thread-local >>>> storage is not supported for the current target >>>> thread_local vector<bool> cache_set(NUM_FAMILIES, false); >>>> src/frontends/qt4/GuiFontLoader.cpp:381:2: error: thread-local >>>> storage is not supported for the current target >>>> thread_local vector<bool> cache(NUM_FAMILIES, false); >>>> >>>> http://stackoverflow.com/questions/28094794/why-does-apple-clang-disallow-c11-thread-local-when-official-clang-supports >>>> >>>> tries to explain why. >>>> >>> >>> Thanks for the pointer. I reverted the patch. It looks to me that it is >>> not possible to use thread_local unless the support for Xcode < 8 is >>> dropped. For now one has to do without it. >> > > But, this is very unfortunate. Do you have an idea of a time frame when > it will become reasonable to require Xcode >= 8 ? I am not familiar with > the Apple environment, but I expect months if not years.
I've suggested it before, but: Why can't we just test for this in configure? Granted, it is a hassle, but we keep running into problems very much like this one. See also the issues with mingw. Richard
