Don't depend on llvm-nm to detect archive indexes #10300
Merged
Conversation
This file contains hidden or 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
sbc100
force-pushed
the
fix_auto_index
branch
from
3b10530 to
19e3da4
Compare
kripken
approved these changes
Jan 30, 2020
tests/test_other.py
Outdated
| self.assertContained('libfoo.a: archive has no index; run ranlib to add one', stderr) | ||
| # The default behavior is to add archive indexes automatically. | ||
| run_process([PYTHON, EMCC, 'libfoo.a', 'hello_world.o']) | ||
| run_process([PYTHON, EMCC, 'libfoo.a', 'hello_world.o']) |
There was a problem hiding this comment.
why is this line duplicating the one before it?
There is an llvm but that mean that certain binary files get mistaken for COFF object which means llvm-nm will error out: https://bugs.llvm.org/show_bug.cgi?id=44683 Instead write out own trivial ar parser to detect the presence of that index. Fixes: #10195
sbc100
force-pushed
the
fix_auto_index
branch
from
413ec7f to
6b5bced
Compare
aheejin
added a commit
to aheejin/emscripten
that referenced
this pull request
Oct 2, 2020
This test was added in emscripten-core#10300 to check if we can handle non-objects in archives, especially rust metadata files that start with two leading zeroes. LLVM's file magic identifier thinks files with two leading zeroes are COFF files. But Rust metadata files used to start with two leading zeroes too, resulting in an error in LLVM tools. So emscripten-core#10300 bypassed use of `llvm-nm` to avoid that. We were still using `llvm-ranlib`, which also ran the LLVM magic identifier when trying to create a symbol table for an object, but `llvm-ranlib` so far has ignored those errors. But recently llvm/llvm-project@a20168d made them explicit errors, so we couldn't run `llvm-ranlib` anymore with archives containing objects that start with two leading zeroes. But this is not relevant anymore because Rust fixed their object file format in rust-lang/rust#66235, so their files are not mistaken by LLVM for COFF files anymore. So this PR fixes this test to not include a binary with two leading zeros, while still testing archival of non-object files.
aheejin
added a commit
that referenced
this pull request
Oct 2, 2020
This test was added in #10300 to check if we can handle non-objects in archives, especially rust metadata files that start with two leading zeroes. LLVM's file magic identifier thinks files with two leading zeroes are COFF files. But Rust metadata files used to start with two leading zeroes too, resulting in an error in LLVM tools. So #10300 bypassed use of `llvm-nm` to avoid that. We were still using `llvm-ranlib`, which also ran the LLVM magic identifier when trying to create a symbol table for an object, but `llvm-ranlib` so far has ignored those errors. But recently llvm/llvm-project@a20168d made them explicit errors, so we couldn't run `llvm-ranlib` anymore with archives containing objects that start with two leading zeroes. But this is not relevant anymore because Rust fixed their object file format in rust-lang/rust#66235, so their files are not mistaken by LLVM for COFF files anymore. So this PR fixes this test to not include a binary with two leading zeros, while still testing archival of non-object files.
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.
There is an llvm but that mean that certain binary files get mistaken
for COFF object which means llvm-nm will error out:
https://bugs.llvm.org/show_bug.cgi?id=44683
Instead write out own trivial ar parser to detect the presence of that
index.
Fixes: #10195