Bootstrap x86_64 musl by itself

[mati865]

It should slightly reduce build time and prepares ground for musl native tests.

NOTE: I haven't tested artifacts yet (only the build).

Can I have double try?

r? @alexcrichton

[alexcrichton]

@bors: try

Nice!

[bors]

⌛ Trying commit edb396d with merge a90f8373dc60c0a5f657a8c48011deb357a90671...

[rust-highfive]

The job dist-x86_64-musl of your PR failed on Travis (raw log). Through arcane magic we have determined that the following fragments from the build log may contain information about the problem.

Click to expand the log.

travis_fold:end:services

travis_fold:start:git.checkout
travis_time:start:027cce68
$ git clone --depth=2 --branch=try https://github.com/rust-lang/rust.git rust-lang/rust
---
[00:13:49] Step 8/11 : ENV HOSTS=x86_64-unknown-linux-musl
[00:13:49]  ---> Running in a26d952b2f61
[00:13:50] Removing intermediate container a26d952b2f61
[00:13:50]  ---> d87b354f036f
[00:13:50] Step 9/11 : ENV RUST_CONFIGURE_ARGS       --musl-root-x86_64=/usr/local/x86_64-linux-musl       --enable-extended       --disable-docs       --set target.x86_64-unknown-linux-musl.crt-static=false       --build $HOSTS       --set target.x86_64-unknown-linux-musl.cc=x86_64-linux-musl-gcc       --set target.x86_64-unknown-linux-musl.cxx=x86_64-linux-musl-g++       --set target.x86_64-unknown-linux-musl.linker=x86_64-linux-musl-gcc
[00:13:50] Removing intermediate container 363d3d999733
[00:13:50]  ---> 608a6747e2a6
[00:13:50] Step 10/11 : ENV CFLAGS_x86_64_unknown_linux_musl="-Wa,-mrelax-relocations=no -Wa,--compress-debug-sections=none     -Wl,--compress-debug-sections=none"
[00:13:50]  ---> Running in 276c37783adf

I'm a bot! I can only do what humans tell me to, so if this was not helpful or you have suggestions for improvements, please ping or otherwise contact @TimNN. (Feature Requests)

[bors]

💔 Test failed - checks-travis

[alexcrichton]

@bors: retry

[bors]

⌛ Trying commit edb396d with merge 9c10e7a...

[bors]

☀️ Try build successful - checks-travis
Build commit: 9c10e7a

[mati865]

Not really an improvement, I'll check artifacts and log tomorrow.

[mati865]

I tested it in Alpine and Ubuntu:14.04 containers and found no issues.

Build time compared to normal (r+) jobs decreased only a little. There are less compilers build but everything takes longer. I don't know whether try builds are generally slower, musl toolchain is slower or sccache has to fill.

Recent musl releases gained various optimisations, I can upgrade and see how it performs.

[alexcrichton]

That's still relatively close to the 2h limit and it makes sense to me in that it compiles less so it'll eventually be faster. Would you want to go ahead and land this or should we wait for more performance investigations of glibc vs musl?

[mati865]

I think newer musl version could cut off few more minute but it's not that big deal and I'll bump musl anyway if I get guard pages to work.
Using clang+lld could be more beneficial but I think binutils caused enough headache so far.
Maybe somebody skilled will debug it but it's out of scope

Dropped travis commit, it's ready to go.

[alexcrichton]

@bors: r+

[bors]

📌 Commit d402415 has been approved by alexcrichton

[bors]

⌛ Testing commit d402415 with merge 3ee9363...

[bors]

☀️ Test successful - checks-travis, status-appveyor
Approved by: alexcrichton
Pushing 3ee9363 to master...

[mati865]

Looking at build time on r+ job and previous PRs it saves 10-20 minutes. That's reasonable saving.