1759213 - Build broken on Fedora 35 x86_64

Status: RESOLVED DUPLICATE of bug 1759125

(Firefox Build System :: Mach Core, defect)

Description

[Jim Blandy :jimb]

The following changeset breaks the build on Fedora:

changeset:   682052:78117d649b5d
user:        Mike Hommey <[e-mail]>
date:        Fri Mar 11 04:10:48 2022 +0000
summary:     Bug 1758781 - Use the WinMsvc.cmake toolchain file to build compiler-rt. r=firefox-build-system-reviewers,mhentges

I'll attach the error output.

Comment 1

[Jim Blandy :jimb]

Attachment: mach error output

Comment 2

[Mitchell Hentges [:mhentges] 🦀]

Can you share the results of your hg status and hg id?

Comment 3

[Jim Blandy :jimb]

[jimb@akatsuki central]$ hg status
[jimb@akatsuki central]$ hg parent
changeset:   682052:78117d649b5d
user:        Mike Hommey <[e-mail]>
date:        Fri Mar 11 04:10:48 2022 +0000
summary:     Bug 1758781 - Use the WinMsvc.cmake toolchain file to build compiler-rt. r=firefox-build-system-reviewers,mhentges

[jimb@akatsuki central]$ hg id
78117d649b5d
[jimb@akatsuki central]$ 

Comment 4

[Mitchell Hentges [:mhentges] 🦀]

Aha, I've go a repro, but I've also got a workaround: hg pull && hg up central. My new hg id is 81f71425dfe3, and bootstrapping clang works OK there.
I'll leave the NI on glandium, because I'm not sure why 78117d649b5d doesn't want to play ball - my best theory is that it's one revision in a larger push, and a later revision in that push also changed the clang toolchain input files. So, when the artifacts were built, it was only for the "newest" revision in the push.

Comment 5

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1758781

Comment 7

[Mitchell Hentges [:mhentges] 🦀]

This isn't a duplicate of 1759125.
One is about zstandard, the other is ERROR!!!!!! Could not find artifacts for a toolchain build namedtoolchain-linux64-clang``.

Comment 8

[Jim Blandy :jimb]

If I build from 81f71425dfe3, I get a different failure:

 0:03.69 Setting up artifact sysroot-x86_64-linux-gnu.tar.zst
 0:03.69 Using artifact from local cache: /home/jimb/.mozbuild/toolchains/c48bdbc2e4ee67f6-sysroot-x86_64-linux-gnu.tar.zst
 0:05.41 Error running mach:
 0:05.41     ['--log-no-times', 'artifact', 'toolchain', '--from-task', 'TQ-MzrjPSx6R24m3A80cSw:public/build/sysroot-x86_64-linux-gnu.tar.zst']
 0:05.41 The error occurred in code that was called by the mach command. This is either
 0:05.41 a bug in the called code itself or in the way that mach is calling it.
 0:05.41 You can invoke |./mach busted| to check if this issue is already on file. If it
 0:05.41 isn't, please use |./mach busted file artifact| to report it. If |./mach busted| is
 0:05.41 misbehaving, you can also inspect the dependencies of bug 1543241.
 0:05.41 If filing a bug, please include the full output of mach, including this error
 0:05.41 message.
 0:05.41 The details of the failure are as follows:
 0:05.41 ModuleNotFoundError: No module named 'zstandard'
 0:05.41   File "/home/jimb/moz/central/python/mozbuild/mozbuild/artifact_commands.py", line 576, in artifact_toolchain
 0:05.41     unpack_file(local)
 0:05.41   File "/home/jimb/moz/central/python/mozbuild/mozbuild/action/tooltool.py", line 993, in unpack_file
 0:05.41     import zstandard
 0:05.41 Sentry event ID: 48be00e810314f0b907e0a2b9b63786a
 0:05.51 Sentry is attempting to send 0 pending error messages
 0:05.51 Waiting up to 2 seconds
 0:05.51 Press Ctrl-C to quit
 0:05.60 ERROR: Command '['/home/jimb/moz/central/obj-x86_64-pc-linux-gnu/_virtualenvs/build/bin/python', '/home/jimb/moz/central/mach', '--log-no-times', 'artifact', 'toolchain', '--from-task', 'TQ-MzrjPSx6R24m3A80cSw:public/build/sysroot-x86_64-linux-gnu.tar.zst']' returned non-zero exit status 1.

This failure was introduced in 4365905c4c1a:

changeset:   682054:4365905c4c1a
user:        Mike Hommey <[e-mail]>
date:        Fri Mar 11 04:10:49 2022 +0000
summary:     Bug 1758781 - Remove now unnecessary patch to link llvm-symbolizer statically. r=firefox-build-system-reviewers,mhentges

Comment 9

[Mitchell Hentges [:mhentges] 🦀]

Now that issue, with zstandard, is from the ticket glandium linked.

Comment 10

[Jim Blandy :jimb]

Indeed. That's the bug that's blocking me at the moment, so I'm more interested in that than this.

If this bug no longer describes a problem you think we should fix, I'm happy to close this.

Comment 11

[Mitchell Hentges [:mhentges] 🦀]

Indeed. That's the bug that's blocking me at the moment, so I'm more interested in that than this.

True, I've got a patch up and I'm standing by for review. I'd recommend either applying that patch locally, backing out the offending patch (1337faa66745), or working on a previous head of central to workaround the issue.

If this bug no longer describes a problem you think we should fix, I'm happy to close this.

There's still an underlying issue of you updating to central (78117d649b5d was central to you, when you originally updated to it, right?), and being unable to bootstrap, which I'm interested in hearing from glandium about.

Comment 12

[Mike Hommey [:glandium]]

The different error is from trying to bootstrap off of either 78117d649b5d or 68b936acd34c, which aren't individual pushes.