#14880 closed enhancement (fixed)
gcc-11.1.0 (wait for LFS)
Reported by: | Douglas R. Reno | Owned by: | Douglas R. Reno |
---|---|---|---|
Priority: | normal | Milestone: | 11.0 |
Component: | BOOK | Version: | SVN |
Severity: | normal | Keywords: | |
Cc: |
Description (last modified by ) ¶
New major version
Change History (13)
comment:1 by , 4 years ago
Description: | modified (diff) |
---|---|
Summary: | gcc-10.3.0 (wait for LFS) → gcc-11.1.0 (wait for LFS) |
follow-up: 4 comment:2 by , 4 years ago
fuse-3.10.2 and 3.10.3 FTBFS. Workaround:
sed '/iostream/a #include<limits>' example/cxxopts.hpp
comment:3 by , 4 years ago
Workaround for llvm-11.1.0 (not needed for 12.0 which already has this):
sed -i '/#include <vector>/i #include <limits>' utils/benchmark/src/benchmark_register.h
comment:4 by , 4 years ago
Replying to Xi Ruoyao:
fuse-3.10.2 and 3.10.3 FTBFS. Workaround:
sed '/iostream/a #include<limits>' example/cxxopts.hpp
I guess that should be sed -i ...
comment:5 by , 4 years ago
For rxvt-unicode-0.9.22, remove a dynamic exception specification (deprecated since C++11)
sed -i 's/\(default_loop .*AUTO.\).*/\1/' libev/ev++.h
comment:6 by , 4 years ago
Fluxbox reports: error: ordered comparison of pointer with integer zero ('unsigned char*' and 'int') Gentoo have a patch to fix that, we can use a sed:
sed -i '/text_prop.value > 0/s/>/!=/' util/fluxbox-remote.cc
comment:7 by , 4 years ago
For qt-5.15.2 I've uploaded qt-everywhere-src-5.15.2-gcc11-1.patch.
qtwebengine is not yet fixed.
comment:8 by , 4 years ago
I'm uploading qtwebengine-20210401-build_fixes-2.patch (adds various headers for gcc11).
comment:9 by , 4 years ago
I've pushed a patch for vlc-3.0.12. It's only two files so can probably be done with seds, but I lack motivation after fighting g++17 all day.
comment:10 by , 4 years ago
Owner: | changed from | to
---|---|
Status: | new → assigned |
Grab stuff that needs to be synced to LFS and some other stuff to work on today.
comment:11 by , 4 years ago
One small problem, hopefully it is only temporary.
I build random versions of firefox beta on this machine, to keep abreast of what is changing. I had not tried to do that on gcc-11.1 until yesterday. It failed abysmally, after 45.53 seconds it failed:
0:45.53 /scratch/working/firefox-89.0/firefox-build-dir/dist/include/nsTHashtable.h:317:27: error: no matching function for call to 'nsTHashtable<detail::VoidPtrHashKey>::WithEntryHandle(const void*&, const fallible_t&, nsTHashtable<detail::VoidPtrHashKey>::PutEntry(nsTHashtable<detail::VoidPtrHashKey>::KeyType, const fallible_t&)::<lambda(auto:7)>)' 0:45.53 317 | return WithEntryHandle(aKey, aFallible, [](auto maybeEntryHandle) { 0:45.53 | ~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0:45.53 318 | return maybeEntryHandle ? maybeEntryHandle->OrInsert() : nullptr; 0:45.53 | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0:45.53 319 | }); 0:45.53 | ~~ [...] 0:45.54 /scratch/working/firefox-89.0/firefox-build-dir/dist/include/nsTHashtable.h:443:8: note: template argument deduction/substitution failed: [...] 0:45.54 make[4]: Leaving directory '/scratch/working/firefox-89.0/firefox-build-dir/accessible/aom' 0:45.54 make[3]: *** [/scratch/working/firefox-89.0/config/recurse.mk:72: accessible/aom/target-objects] Error 2 0:45.54 make[3]: *** Waiting for unfinished jobs....
and then rust carried on burning cycles for another 16m47 (8 cores).
I'm totally at a loss for this and I guess most people are using clang++ for firefox. With clang-12.0.0 it does build, and also with gcc-10 on an 10.1 system. Not a showstopper for the book, but we might eventually have to revert to clang for the next esr.
comment:12 by , 4 years ago
Resolution: | → fixed |
---|---|
Status: | assigned → closed |
Fixed at @229d06924a16001b642c88a78e297eff97f8697f
Bump to 11.1.0. Jakub says it's scheduled to be released on Apr 27.