PPC smbfs history (circa 2.1.24) question

[Urban Widmark]

On Thu, 29 Mar 2001, Jim McDonough wrote:

> I have verified that smbmount works fine on the following big-endian
> setups:
> LinuxPPC on an RS/6000 model 43P-133, kernel 2.2.17 (with CHRP patches)
> Suse 7.0 on a System/390 G6 VM guest, kernel 2.2.16
> Suse 7.0 on a System/390 Multiprise 2000 VM guest, kernel 2.2.16 (should be
> the same as the G6, but you might be surprised)
> 
> I tested connecting both to a Win2K pro system, and several samba 2.2
> current CVS (3/29) systems.
> 
> Is there some test suite I should run to verify this?  I'm just mounting
> and casually browsing...

No, that's fine. Thanks a lot.

Even a small test is a verification that most things work. There may of
course be a few bugs in specific functions, but most follow the same
pattern.


There is LTP (http://oss.sgi.com/projects/ltp/) for generic kernel tests.
Most of those tests can be run with smbfs as the working directory. Some
test are expected to fail, however. I have a list somewhere of ltp tests
that I thought made sense.

I have also been using ltp to write/run some small regression tests for
smbfs. Not really anything for publication yet, and a while back one of my
tests deleted all the other tests ...

There are no coverage tests to verify that all paths lead to endianess
heaven. Or even a testsuite that "tickles" all the SMB calls. So there is
no easy way to run something and then say that smbfs seems ok ...

[/me stamps the next smbfs patch with "PPC certified by Jim McDonough"
 anyway :)]

/Urban