Re: Bug#1122332: mariadb: FTBFS on sh4: builds unstable starting from 2025

Hello Otto,

On Tue, 2025-12-09 at 18:18 -0800, Otto Kek=C3=A4l=C3=A4inen wrote:
> The MariaDB builds on sh4 have started to fail frequently in 2025. See
> trend at https://buildd.debian.org/status/logs.php?pkg=3Dmariadb&arch=3Ds=
h4
> with most builds passing in 2023-2024 but then starting from early
> 2025 builds frequently fail. However, they seem to pass on rebuilds.
>=20
> The build failures of MariaDB however are not simple "out of disk
> space" or "out of memory" issues, but crashes. I wonder if sh4 porters
> could help test MariaDB builds to debug what is going on and if the
> frequent build failures are a sign of a latent real code error?

The latest failure seen on sh4 in sid is because of cmake crashing which
could be because of libuv1 using io_uring. However, I had io_uring disabled
in libuv1 in the past [1], so I'll have to dig into the crash to see what
the problem is.

Adrian

> [1] http://bugs.debian.org/1111616

--=20
 .''`.  John Paul Adrian Glaubitz
: :' :  Debian Developer
`. `'   Physicist
  `-    GPG: 62FF 8A75 84E0 2956 9546  0006 7426 3B37 F5B5 F913