New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
emscripten: 1.37.16 -> 1.37.34 #35636
emscripten: 1.37.16 -> 1.37.34 #35636
Conversation
Semi-automatic update. These checks were performed: - built on NixOS - found 1.37.34 with grep in /nix/store/mdr47v3wdjmzic5c2nvdx5krdwl6bcxf-emscripten-1.37.34
@GrahamcOfBorg build emscripten |
Success on x86_64-darwin (full log) Partial log (click to expand)
|
Failure on aarch64-linux (full log) Partial log (click to expand)
|
Success on x86_64-linux (full log) Partial log (click to expand)
|
Is the compiler tested as part of building itself, or should that be done as well? Never poked at the emscripten bits... |
You meant @matthewbauer |
@matthiasbeyer it worked out in the end :) |
While debugging my follow-up PR, I found that this breaks emscriptenPackages? @GrahamcOfBorg build emscriptenPackages.zlib Like I said before, does the resulting compiler still work? Apparently not... (at least not the way we used it) :(. |
No attempt on aarch64-linux The following builds were skipped because they don't evaluate on aarch64-linux: emscriptenPackages.zlib No log is available. |
Failure on x86_64-darwin (full log) Attempted: emscriptenPackages.zlib Partial log (click to expand)
|
…1.37.16-to-1.37.34" This reverts commit 5c02b9a, reversing changes made to 2f2ad74. #35636 (comment)
Success on x86_64-linux (full log) Attempted: emscriptenPackages.zlib Partial log (click to expand)
|
@qknight It was already reverted. |
@ryantm thanks for your effort! see this: emscripten-core/emscripten-fastcomp#224
well, it didn't warn IIRC
i found that out the hard way. in general we should change nixpkgs expressions so that one defines the version number as our roadmap should be:
i've been playing with 1.37.36 and implemented the unit testing:
altough there are lots of errors, this is not looking so bad. i'd love to select some meaningful tests and get these into nixpkgs. we can't do all, this requires at least 4220.752s to complete, which is too long! building emscriptenPackage.zlibstill this won't compile with the new toolchain. no clue why, anyone?
|
Just a guess but maybe try setting 'hardeningDisable = [ "format" ]' |
@matthewbauer where should i try hardeningDisable? seems the python executable was missing and therefore i'm no facing:
|
LONG_BIT is no longer a problem, i've added this:
to libxml2 |
@matthewbauer maybe you could check emscripten-core/emscripten#6341 for an obvious error i've made. |
@qknight "hardeningDisable" was just a guess based on the "Compiler error reporting is too harsh" line. But it sounds like python was the real culprit. |
@qknight Looks like this might be the culprit:
This was introduced in @Ericson2314's cross compiling patches. |
PR #26007 used these to avoid causing a mass rebuild. Now that we know things work, we do that to clean up.
@matthewbauer you are amazing! setting: |
status update:
see https://github.com/qknight/nixpkgs/tree/emscripten-1.37.16-to-1.37.36 |
if you want to test my changes:
|
so everything from my perspective is now working. someone interested in helping me to write the documentation? |
\o/ done! |
@ryantm @matthewbauer please have a look at #37291 |
Semi-automatic update. These checks were performed: