Revert using MCP510 in bootstrap #118810
Merged
Revert using MCP510 in bootstrap #118810
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This reverts commit 40c3d35. The option was dogfooded for using lld with MCP510 , but it broke testing with LLD, because we don't pass `-Zunstable-options` on enough places.
rustbot
added
S-waiting-on-review
|
It fixes the failures for me. @bors r+ p=1 Priority bump because this breaks tests for compiler devs who set |
bors
added
S-waiting-on-bors
bors
added a commit
to rust-lang-ci/rust
that referenced
this pull request
Dec 10, 2023
…nethercote Revert using MCP510 in bootstrap This reverts commit 40c3d35. The commit started dogfooding MCP510 to enable `lld` when building the compiler , but it broke tests, because we don't pass `-Zunstable-options` on enough places. This PR hotfixes that, and temporarily makes the "self-contained" option.. not very self-contained. I'll send a proper fix later, but I want to unblock rustc developres that use `lld` locally. r? `@nnethercote` (who discovered the problem)
|
The job Click to see the possible cause of the failure (guessed by this bot) |
|
💔 Test failed - checks-actions |
bors
added
S-waiting-on-review
|
@bors retry |
bors
added
S-waiting-on-bors
|
☀️ Test successful - checks-actions |
|
Finished benchmarking commit (1cb200c): comparison URL. Overall result: no relevant changes - no action needed@rustbot label: -perf-regression Instruction countThis benchmark run did not return any relevant results for this metric. Max RSS (memory usage)ResultsThis is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.
CyclesResultsThis is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.
Binary sizeThis benchmark run did not return any relevant results for this metric. Bootstrap: 669.249s -> 668.219s (-0.15%) |
matthiaskrgr
added a commit
to matthiaskrgr/rust
that referenced
this pull request
Jan 4, 2025
…zkan Allow using self-contained LLD in bootstrap In rust-lang#116278, I added a `"self-contained"` mode to the `rust.use-lld` bootstrap option, which was designed for using the built-in LLD for linking compiler artifacts. However, this was later reverted in rust-lang#118810. This PR brings the old logic back, which switches LLD in bootstrap from `-fuse-ld=lld` to [MCP510](rust-lang/compiler-team#510 way of passing linker flags to enable LLD (both external and self-contained). So this does two changes: 1) Goes from `-fuse-ld=lld` to MCP510 2) Actually makes it possible to use the self-contained LLD to compile compiler artifacts Regarding the second commit: Since rust-lang#86113, we have been passing `-fuse-ld=lld` as a target flag to all tests when `use-lld = true` is enabled. This kind of worked for all tests, since it was just a linker argument, which has bypassed any compiler checks, and probably resulted only in some warning if the given target linker didn't actually support LLD. However, after the first commit, some tests actually start failing with this approach: ``` error: linker flavor `gnu-lld-cc` is incompatible with the current target | = note: compatible flavors are: llbc, ptx ``` So the second commit removes the passing of LLD flags as target flags to tests. I don't think that it's a good idea to pass specific compiler flags to all tests unconditionally, tbh. The doctest command from rust-lang#86113 doesn't go through compiletest anymore, and doctests should be quite a lot faster since rust-lang#126245 in general. CC `@the8472` If someone has a beefy machine, it would be nice to test whether this doesn't regress test execution speed. How to do that: 1) Enable `rust.use-lld = true` and `rust.lld = true` in `config.toml` 2) Benchmark `./x test tests/ui --force-rerun` between `master` and this PR Once this is tested in the wild, I would like to make the self-contained LLD the default in CI, hopefully to make CI builds faster. r? `@onur-ozkan`
rust-timer
added a commit
to rust-lang-ci/rust
that referenced
this pull request
Jan 4, 2025
Rollup merge of rust-lang#135001 - Kobzol:bootstrap-mcp-510, r=onur-ozkan Allow using self-contained LLD in bootstrap In rust-lang#116278, I added a `"self-contained"` mode to the `rust.use-lld` bootstrap option, which was designed for using the built-in LLD for linking compiler artifacts. However, this was later reverted in rust-lang#118810. This PR brings the old logic back, which switches LLD in bootstrap from `-fuse-ld=lld` to [MCP510](rust-lang/compiler-team#510 way of passing linker flags to enable LLD (both external and self-contained). So this does two changes: 1) Goes from `-fuse-ld=lld` to MCP510 2) Actually makes it possible to use the self-contained LLD to compile compiler artifacts Regarding the second commit: Since rust-lang#86113, we have been passing `-fuse-ld=lld` as a target flag to all tests when `use-lld = true` is enabled. This kind of worked for all tests, since it was just a linker argument, which has bypassed any compiler checks, and probably resulted only in some warning if the given target linker didn't actually support LLD. However, after the first commit, some tests actually start failing with this approach: ``` error: linker flavor `gnu-lld-cc` is incompatible with the current target | = note: compatible flavors are: llbc, ptx ``` So the second commit removes the passing of LLD flags as target flags to tests. I don't think that it's a good idea to pass specific compiler flags to all tests unconditionally, tbh. The doctest command from rust-lang#86113 doesn't go through compiletest anymore, and doctests should be quite a lot faster since rust-lang#126245 in general. CC `@the8472` If someone has a beefy machine, it would be nice to test whether this doesn't regress test execution speed. How to do that: 1) Enable `rust.use-lld = true` and `rust.lld = true` in `config.toml` 2) Benchmark `./x test tests/ui --force-rerun` between `master` and this PR Once this is tested in the wild, I would like to make the self-contained LLD the default in CI, hopefully to make CI builds faster. r? `@onur-ozkan`
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This reverts commit 40c3d35. The commit started dogfooding MCP510 to enable
lldwhen building the compiler , but it broke tests, because we don't pass-Zunstable-optionson enough places. This PR hotfixes that, and temporarily makes the "self-contained" option.. not very self-contained. I'll send a proper fix later, but I want to unblock rustc developres that uselldlocally.r? @nnethercote (who discovered the problem)