It would seem that Chromium actually builds and installs correctly (as it is the version showing up in the "About" menu of the application), but there is an odd die message at the end of the compilation. >>> Original instance of package unmerged safely. * Messages for package www-client/chromium-5.0.307.7: * ERROR: www-client/chromium-5.0.307.7 failed: * compilation failed * * Call stack: * ebuild.sh, line 54: Called src_compile * environment, line 2607: Called die * The specific snippet of code: * emake -r V=1 chrome chrome_sandbox BUILDTYPE=Release rootdir="${S}" CC=$(tc-getCC) CXX=$(tc-getCXX) AR=$(tc-getAR) RANLIB=$(tc-getRANLIB) || die "compilation failed" * * If you need support, post the output of 'emerge --info =www-client/chromium-5.0.307.7', * the complete build log and the output of 'emerge -pqv =www-client/chromium-5.0.307.7'. * The complete build log is located at '/var/tmp/portage/www-client/chromium-5.0.307.7/temp/build.log'. * The ebuild environment file is located at '/var/tmp/portage/www-client/chromium-5.0.307.7/temp/environment'. * S: '/var/tmp/portage/www-client/chromium-5.0.307.7/work/chromium-5.0.307.7' >>> Auto-cleaning packages... >>> No outdated packages were found on your system. * Regenerating GNU info directory index... * Processed 90 info files. Reproducible: Always Steps to Reproduce: 1. Keyword Chromium (if running stable) 2. emerge chromium 3. Builds successfully (or so it would seem), but the die message appears. Actual Results: Die message appears after the original instance was unmerged successfully. Expected Results: Should install correctly, and produce no error message.
You are confused by --keep-going and the echo ELOG features which makes the messages re-appear upon emerge has finished the rest of its job. chrome has clearly failed to recompile for you. The "Original instance of package unmerged safely." referrs to last successful package in the list, not to www-client/chromium-5.0.307.7 ;)
Nathan, are you able to obtain the build log? Does chromium really fail to build, or is it some unstable portage weirdness?
After having just updated to 5.0.307.11, there were no more odd die messages at the end of the compilation. It would seem that this was either a fluke, or has been fixed with this newer version of the ebuild. I will close this bug for now, unless someone else has further information.