Re: make check

[Marcus Brinkmann]

On Sun, Mar 24, 2002 at 07:52:08AM -0800, James Morrison wrote:
>   I'm curious what types of checks would be in a make check target for the
> hurd.
> Would have to be run in a sub-hurd to test all the hurd translators and libc
> functions?  Should the test programs be compiled within that test sub-hurd?

No, you can attach a second exec translator to any node and test it's
low level primitives (RPCs) manually, of course.  Same goes for any
translator.

Checks for glibc should be in glibc.  The Hurd should have checks for
Hurd servers.  This includes a check for each bug found, especially
for border conditions (like symlinks to empty names etc, although this
would probably be more an example for glibc).

I think it would be a very good idea to run a POSIX conformance test
suite on the Hurd.  There are several out there, AFAIK.  They test a lot
of stuff that should work for us.

Equally important is to stress the actually running servers.  There are
stress tests out there, for example file systems tests which make a lot
of weird filesystem operations.  Running these against the Hurd servers
and debugging crashes, memory leaks etc would be another very good idea.

Thanks,
Marcus