[openib-general] libehca causes segfault when not physically present..

[Troy Benjegerdes]

On Thu, Nov 03, 2005 at 11:13:58AM -0800, Roland Dreier wrote:
>     Heiko> this bug should be fixed in OpenIB trunk 3960.
> 
> It's good to see this fixed and all the other cleanups in this
> checkin.  I'll have to go back to my ehca code reviewing....
> 
> However, when this code moves upstream, you'll have to make your
> changes in smaller digestible chunks.  The diff between r3959 and
> r3960 is rather gigantic:
> 
>  33 files changed, 945 insertions(+), 1163 deletions(-)
> 
> And this piece:
> 
>     > -MODULE_VERSION("EHCA2_0035");
>     > +MODULE_VERSION("EHCA2_0037");
> 
> indicates that there was a 0036 that you never let anyone see.

I'll second the comment about smaller digestible chunks. A second thing
I don't completely understand is the vast size difference between the
ehca and mthca drivers. Is the ehca really that much more complex?

I also want to comment that EHCA is the only thing that's versioned that
is easy to tell what version of the module is actually loaded at the
moment. I'd rather have versions I don't see float by than see every 
file in mthca get updated, but no version rev.

I tried adding some code to generate a version string from the outupt of
svnversion but it didn't work too well.

The same goes for OpenSM as well.. the only version string you get when
starting it is 'Opensm-1.1.0', which isn't very usefull.

And once that's figured out, maybe we can start thinking about how to
make sure kernel module versions match userspace versions. Personally,
I'd like to see the ehca functions exported as a VDSO.