#1016526 nodejs test failures on mips64el/riscv64 with gcc 12

Package:
gcc-12
Source:
gcc-12
Description:
GNU C compiler
Submitter:
Jérémy Lal
Date:
2026-07-27 17:35:09 UTC
Severity:
normal
Tags:
#1016526#5
Date:
2022-08-02 10:50:05 UTC
From:
To:
Hi,

while building nodejs 18.6.0+dfsg-4 on mips64el, some SSL tests did regress,
https://buildd.debian.org/status/logs.php?pkg=nodejs&arch=mips64el
(look for "not ok" tests).

In particular, test-crypto-sign-verify test failed with an unexpected error:
'error:1E08010C:DECODER routines::unsupported'

I suppose the root cause is similar to the one for
[EC code appears to be broken on s390x](https://bugs.debian.org/1016290),
but it's not the same bug.

Also note that nodejs tests run with
CipherString = DEFAULT:@SECLEVEL=1

Also the same tests don't fail on mipsel, or any other arch,
including s390x.

Jérémy
#1016526#12
Date:
2022-08-02 14:35:47 UTC
From:
To:
The same test also fails when building nodejs on riscv64.
#1016526#19
Date:
2022-08-02 16:40:15 UTC
From:
To:
For some reason it seems to expect to fail with some other error
instead.

It's currently unclear that this is an openssl bug. Can you show
what that code is trying to do, and what it expects openssl to return?

There seem to be various other test suite failures that don't seem to be
related to openssl at all.

It will be totally unrelated. That one is very specific to s390x.


Kurt
#1016526#24
Date:
2022-08-02 18:18:32 UTC
From:
To:
Le mar. 2 août 2022 à 18:40, Kurt Roeckx <kurt@roeckx.be> a écrit :

Indeed, something has changed somewhere else.

Jérémy
#1016526#51
Date:
2022-08-12 19:10:15 UTC
From:
To:
I've been looking at this because I assumed that something may be
different in the init path on mips vs everywhere else leading to a
similar kind of an error like the last time where the error strings were
missing. That said, I haven't seen much difference between mips64el and
amd64 here. Therefore I just wiped my home on eller.
Let me know if you want me to look at this again.

Sebastian
#1016526#56
Date:
2022-08-12 19:24:39 UTC
From:
To:
Le ven. 12 août 2022 à 21:10, Sebastian Andrzej Siewior <
sebastian@breakpoint.cc> a écrit :

Thanks a lot. I assumed the same, that's why I notified it here.

The actual issue is somewhere else between gcc11 and gcc12.

Jérémy