On 08/05/2016 19:40, Douglas R. Reno wrote: > > > On Sun, May 8, 2016 at 1:34 AM, Pierre Labastie <[email protected] > <mailto:[email protected]>> wrote: > > On 06/05/2016 16:44, Pierre Labastie wrote: > > Hi, > > > > It looks like GCC 6 adds a level of rigor to C/C++ compiling, and that > code > > which used to work, does not anymore because it relied on not rigorously > > defined behavior. > > > > I guess this is the case for a lot of packages in BLFS. I already have > found > > (sorry if some of these have been mentionned already, I do not have a > way to > > search the mail archives ATM): > > harbuzz: may be built with: CC="gcc -std=c99" CXX="g++ -std=c++98" > ./configure... > > php: may be built with: CC="gcc -std=c99" ./configure... > > libsoup: same > > xorg-evdev-driver: same > > libdrm: same > > gstreamer: same > > gstreamer-plgins-base: same (not tried the other plugins, but I guess > they > > need the same) > > openjdk: fix not ready yet > > > > Update: > > When using GCC 6 from an LFS build, only openjdk needs a fix. All the > other > packages mentioned above work fine. > > Pierre > > I have some theories as to why this is and how to fix it. I am probably going > to fire up one of my old LFS-7.9 GCC5 builds and see what I can do. > > I remember reading that libtool should be recompiled when upgrading the > compiler. Anyone sure if I am correct? > > Douglas R. Reno >
What I can tell is that some of the failures were coming from undefined types (like uint or u_char), which are standard glibc types. I guess the reason is that header files were not input in the necessary order. Why it is so and whether it has something to do with libtool, I cannot tell. Pierre -- http://lists.linuxfromscratch.org/listinfo/blfs-dev FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page
