Enable more tests for Emscripten targets

[tlively]

Specifically, enable SIMD intrinsic tests, target-independent asm!
tests, and tests that were ignored due to old fastcomp bugs that have
long since become irrelevant.

Also re-enables asmjs tests that pass -g by reducing the corresponding
Emscripten debug level so that it does not error out with a message
about source maps being unsupported. This required fixing the asmjs
target's arch string to be "asmjs" rather than "wasm32", and
that in turn required auditing many cfg attributes and fixing them to
include both asmjs and wasm32.

For many tests that could not be re-enabled, also adds an explanatory
comment to make future auditing easier. In particular, tests requiring
threads are left to be re-enabled in future work.

r? @alexcrichton

[rust-highfive]

The job x86_64-gnu-llvm-7 of your PR failed (pretty log, 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.

2020-01-10T06:03:19.4780981Z ##[command]git remote add origin https://github.com/rust-lang/rust
2020-01-10T06:03:19.4792474Z ##[command]git config gc.auto 0
2020-01-10T06:03:19.4795074Z ##[command]git config --get-all http.https://github.com/rust-lang/rust.extraheader
2020-01-10T06:03:19.4797234Z ##[command]git config --get-all http.proxy
2020-01-10T06:03:19.4800483Z ##[command]git -c http.extraheader="AUTHORIZATION: basic ***" fetch --force --tags --prune --progress --no-recurse-submodules --depth=2 origin +refs/heads/*:refs/remotes/origin/* +refs/pull/68082/merge:refs/remotes/pull/68082/merge

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)

[alexcrichton]

Nice! Looks like there's a rustfmt thing and otherwise just one small nit, but otherwise r=me

Thanks!

[tlively]

@alexcrichton I took a look at the timings, and the wasm32 bot here takes about 25 minutes longer than the wasm32 bot on the latest bors jobs (2h44m intead of 2h18m). The JS bot (which is disabled) is only a couple minutes shy of the 3 hour mark. How much of a problem is this?

[alexcrichton]

Hm ok that's a little excessive in terms of time, mind keeping the wasm32 tests as-is and we can reevaluate when we get some more powerful builders?

[tlively]

Actually, looking at some of the other builders, 2h44m doesn't seem so bad. Plenty of other builders are in the 2h40m or even 2h50m range. On the latest run a lot of the Windows builders have just passed the 3h mark: https://dev.azure.com/rust-lang/rust/_build/results?buildId=18111.

I would be happy to disable some of the longer-running wasm tests to save some time, but I would want to specifically disable long-running tests rather than just leaving historically buggy tests disabled. If you have an idea of what a reasonable runtime goal would be, I can selectively disable tests to try to meet that goal.

On a related note, is there a way to avoid rebuilding LLVM on every run of the builders?

[bors]

☔ The latest upstream changes (presumably #67000) made this pull request unmergeable. Please resolve the merge conflicts.

[tlively]

@alexcrichton The merge conflict was just a deleted test, so it didn't seem worth going through re-enabling and disabling the wasm/asmjs CI for the PR. This should be good to go unless you have thoughts on the run time.

[alexcrichton]

Hm I'd personally prefer at least a good headroom with our slowest tests, I'd rather not try to get up next to them in terms of execution time. 3h is already far too long and we've been looking to reduce it. I think it's probably best to leave the set of tests as-is for now and we can investigate expanding coverage in a separate PR?

On a related note, is there a way to avoid rebuilding LLVM on every run of the builders?

Not currently, but LLVM is cached with sccache, so it's not supposed to be as slow as a full build.

[bors]

☔ The latest upstream changes (presumably #68311) made this pull request unmergeable. Please resolve the merge conflicts.

[JohnCSimon]

Ping from triage:
@tlively this PR has sat idle for over a week. Thanks.

[tlively]

Thanks, @JohnCSimon! The current status of this PR is that I need to disable a large number of long-running tests to get the bot time under control. If that doesn’t work, I’ll probably want to remove most of the test changes from the PR. I am not yet actively working on these changes yet.

[JohnCSimon]

Ping from triage -
@tlively - you have some merge conflicts,
can you post your status on this PR? Thanks.

[tlively]

Still no progress here, but I’ll put this on my list of things to do soon.

[JohnCSimon]

Ping from triage
@tlively - just letting you know, this has sat idle for two weeks.

[tlively]

I'll close this for now since I am not actively working on it. I'll get back to it at some point, but I don't know when.