[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: tricore-debian-cross-container CI job failing
From: |
Peter Maydell |
Subject: |
Re: tricore-debian-cross-container CI job failing |
Date: |
Thu, 12 Jan 2023 14:30:45 +0000 |
On Thu, 12 Jan 2023 at 14:13, Peter Maydell <peter.maydell@linaro.org> wrote:
>
> The tricore-debian-cross-container CI job has started failing for
> no obvious reason:
>
> https://gitlab.com/qemu-project/qemu/-/jobs/3593581274
>
> gcc -DHAVE_CONFIG_H -I. -I. -I. -D_GNU_SOURCE -I. -I. -I../bfd
> -I./../bfd -I./../include -I./../intl -I../intl -w
> -DLOCALEDIR="\"/usr/local/share/locale\"" -W -Wall -Wstrict-prototypes
> -Wmissing-prototypes -w -c `test -f 'ldgram.c' || echo './'`ldgram.c
> `test -f ldlex.l || echo './'`ldlex.l
> /bin/sh: 1: ldlex.l: not found
For comparison, here's another run of the same job, which
succeeded:
https://gitlab.com/qemu-project/qemu/-/jobs/3594096306
The difference is that in the success case Make doesn't think
it needs to rebuild ldlex.c or ldgram.c, but in the failure
case it does (and falls over because it can't). I don't know
why this is, maybe a random timestamp thing?
Anyway, to make the job reliable I think we need to have
flex and bison in the container.
thanks
-- PMM