Created attachment 423674 [details] Build.log.gz Hi! Chromium build failed in my system during regular world update: [code] <command-line>:0:0: warning: "_FORTIFY_SOURCE" redefined <built-in>: note: this is the location of the previous definition In file included from ../../base/stl_util.h:11:0, from ../../base/memory/scoped_vector.h:14, from ../../extensions/renderer/worker_script_context_set.h:10, from ../../extensions/renderer/worker_script_context_set.cc:5: /usr/lib/gcc/x86_64-pc-linux-gnu/4.9.3/include/g++-v4/functional:2505:13: error: expected type-specifier before ')' token operator==(const function<_Res(_Args...)>& __f, nullptr_t) noexcept ^ ninja: build stopped: subcommand failed. [/code] The reason is unclear to me. Previous update was Ok. "emerge --info" and build logs are attached. Regards, Alex
Created attachment 423676 [details] emerge --info output
I can confirm the failure. I bumped dev-libs/re2 to 20151101 in my overlay and merged =app-arch/snappy-1.1.3. chromium compiled successfully.
Pretty sure this is a dupe of bug #571156
Strangely, but it successfully compiled with "MAKEOPTS=-j1 emerge chromium". And even more strange is that single-threaded compilation for chromium (and some other packages) has been set in my system for years, in /etc/portage/env and /etc/portage/package.env, and it's worked out Ok untill the above chromium build.
I couldn't repro this and at this point the bug is rather old. I think re2 is a red herring. For now we're using the bundled version. I also don't see how serial or parallel compilation could influence a compile error in a .cc file not involving anything generated. This is really mysterious. If it's still happening with new versions, just let me know - either reopen or file new bug.
