This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: RISC-V test results
- From: "Maciej W. Rozycki" <macro at wdc dot com>
- To: DJ Delorie <dj at redhat dot com>
- Cc: libc-alpha at sourceware dot org, alistair dot francis at wdc dot com
- Date: Thu, 1 Aug 2019 14:45:05 +0100 (BST)
- Subject: Re: RISC-V test results
- Ironport-sdr: 5zKTSyHN3laup9wPkQ4XHotjhUbX92eN1Dk8eZ8gSJzisegSextvJSrtOUZNbxX0UJRxGIFcyf 6d997iEQB0pvtoX1hvnbqkwOOyKHpDYarsIqtJ6ADOce8elkun3pB9FpXXS819Hb4dDVFUadNa RbUZMwM6JsXu4ERKI1Kc9V8s0lDVCXGzqlaF9xLmnehb2dzx7eQ5By7BfKq6ncZourm1GP+2VW pp7K74NtVbJKLDT62FR1GLPDfQMYdFzyQXoHK8lWKjmMI5C4/GRks5qrqCKu+XhJbAE3TuoBLb ZzQ=
- Ironport-sdr: niQBd0BWjTdV6IydPBO1JV96/TT1Jg6BkRE/EA19PwmSoM6Mi0n1LPREWsO6SVh80WwgVWSE47 C66tdbBLSBB+x1sTdzrkCIfuAhxN4BpZqZGDfLbPXUHblndZQPp7EGQHj8a7rsmiqfK7pFWfzK V+lrYk7I7nLxl3VC292Fu4uROhkt6QT96usBzeNT6tKsSundwfxt0m7bdw1U3dlH/uTGs3ymmm sIHZONsp58oXRYDETQ2r/89kV/piT9MU6xCVOcaaiDFS5qvlODI+XhzmIfx3EK4+eiZrCszWm9 +kdX3wo5u6oVzfgZ+XQcrflS
- Ironport-sdr: TE73OK6RdgqS7QvbTxLBpn50sLtH9CROKkfO3KIFl+BDQWjc44qNueUYB4aKVJA237l977MVNu oTwyyVeW4H4NKARpvxw8agjKpiDmEOmuhfs/CeizsHUwVA1DoEVQ4nCJGQmkiQ9PBHfOlHndUA DtnGvgCara4IUEvyApjmNt/CBy9RuWX8Hu8DJMp8SCIXurY84BIKi9KOV1JcwTy8lZDtm57wO9 4iCj2rkhVwc0hvRx5ghwGrcmEK/5lWeawawNqzLd7rYpH72IgjlCx2WsilvPeWvIWdOwqfGHHG /oY=
- References: <xny30dg4s2.fsf@greed.delorie.com>
On Wed, 31 Jul 2019, DJ Delorie wrote:
> My environment is a HiFiveU board with a local SSD (no nfs involved),
> running Fedora 28 but with some upgrades (versions noted in wiki). It's
> one of the Fedora/RISC-V build machines so I have faith in it :-)
>
> The build was a native build, no cross-anything.
For the record mine is Fedora 29, however in a cross-testing environment
it hardly matters as long as it doesn't crash. I think such a setup has
an advantage in that it relies less on native features, although on the
other hand native testing puts more stress on the system under test, and
in any case collecting results from different configurations provides
better coverage.
> > I had to bump up the timeout factor to avoid excessive timeouts and
>
> I set TIMEOUTFACTOR to 20 for hifive builds, even with a local SSD.
Sure, I think the vast majority of the test cases is not
filesystem-bound.
Maciej