When built on linux 3.x, python used to report os.platform == 'linux3', thus confusing many exiusting scripts, including some used to build chromium. While bug 376219 tracked that confusion, and eventually announced that chromium would now in fact compile on linux 3, bug 375257 tackled the issue at the root and included a patch from python upstream that will now cause that string to be 'linux2' again for the sake of compatibility. So now we need to somehow revert that chromium "fix", or hope for upstream to do so. Grepping the chromium sources, it appers that they simply dropped all linux2 support, and now support linux3 exclusively. But maybe I'm missing something. >>> Configuring source in /var/tmp/portage/www-client/chromium-14.0.835.163/work/chromium-14.0.835.163 ... build/gyp_chromium --depth=. -Ddisable_sse2=1 -Ddisable_nacl=1 -Denable_webrtc=0 -Duse_system_bzip2=1 -Duse_system_flac=1 -Duse_system_icu=1 -Duse_system_libevent=1 -Duse_system_libjpeg=1 -Duse_system_libpng=1 -Duse_system_libwebp=1 -Duse_system_libxml=1 -Duse_system_speex=1 -Duse_system_xdg_utils=1 -Duse_system_zlib=1 -Duse_cups=1 -Duse_gconf=1 -Duse_gnome_keyring=1 -Dlinux_link_gnome_keyring=1 -Dlinux_sandbox_path=/usr/lib64/chromium-browser/chrome_sandbox -Dlinux_sandbox_chrome_path=/usr/lib64/chromium-browser/chrome -Dproprietary_codecs=1 -Dffmpeg_branding=Chrome -Dtarget_arch=x64 -Dwerror= Updating projects from gyp files... Traceback (most recent call last): File "build/gyp_chromium", line 170, in <module> sys.exit(gyp.main(args)) File "/var/tmp/portage/www-client/chromium-14.0.835.163/work/chromium-14.0.835.163/tools/gyp/pylib/gyp/__init__.py", line 338, in main 'sunos5': 'make',}[sys.platform] ] KeyError: 'linux2' * ERROR: www-client/chromium-14.0.835.163 failed (configure phase): * (no error message) * * Call stack: * ebuild.sh, line 91: Called src_configure * environment, line 6468: Called die * The specific snippet of code: * egyp ${myconf} || die
Do you still have that bashrc hack in place from bug 376219 comment 25?
(In reply to comment #1) > Do you still have that bashrc hack in place from bug 376219 comment 25? Yes, stupid me. Sorry for the noise, and thanks for finding this! Would only have had to look up a few lines in order to find the elog message I specifically added for this purpose. It even was in the elog-generated email which informed me of the build failure. I'm really blind today! Thanks for catching this. I see this applies to bug 376219 comment 30 as well.