First push to mozilla-inbound showing this is https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&revision=8266c4f82c1a. This is breaking Android artifact builds and will likely have changed the constants sent to our metrics and Telemetry collection, so we either need to rollback to earlier definitions or prepare to change what we expect.

So the reason this is happening is that config.sub does some more normalization, and we don't run config.sub in pyconfigure. I guess until we stop using TARGET_CPU and TARGET_OS, we should use config.sub to normalize them.

Until we stop relying on the raw_cpu and raw_os values from target and host, we need to keep normalizing them like old-configure.in did, which involves filtering them through config.sub. Review commit: https://reviewboard.mozilla.org/r/40905/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/40905/

Comment on attachment 8731884 [details] MozReview Request: Bug 1257648 - Always filter target/host triplets with config.sub https://reviewboard.mozilla.org/r/40905/#review37447 Output of `./mach configure ; ./mach environment -v | grep TARGET_` before: TARGET_CPU = arm TARGET_OS = androideabi ANDROID_TARGET_SDK: 23 MOZ_LIBSTDCXX_TARGET_VERSION: TARGET_CPU: arm TARGET_OS: androideabi TARGET_XPCOM_ABI: ANDROID_TARGET_SDK after: TARGET_CPU = arm TARGET_OS = android ANDROID_TARGET_SDK: 23 MOZ_LIBSTDCXX_TARGET_VERSION: TARGET_CPU: arm TARGET_OS: android TARGET_XPCOM_ABI: ANDROID_TARGET_SDK I can't verify that |mach artifact install| works since the actual upstream files are badly named, but this should start generating new ones correctly.