[ofa-general] Re: [PATCH] Opensm: main exit codes
Sasha Khapyorsky
sashak at voltaire.com
Wed Nov 26 02:13:05 PST 2008
Hi Tim,
On 14:03 Mon 24 Nov , Timothy A. Meier wrote:
> >
> > And are there any of such tools? Or any *real* use?
> >
>
> Chicken/Egg? Currently, we depend on only ZERO or non-zero. Although OpenSM returns "other" values
> on exit, they aren't really formalized or documented. Hence the patch. ;^)
And after this patch it is still be not formalized - there are another
places in OpenSM where exit(N) is called. For example what could you do
with exit(YY_EXIT_FAILURE)?
> Personally, I have (and create) several different versions of opensm with small customizations,
> and test them on our cluster testbeds. I often will start/stop them in a variety of configurations
> (with and without plugins, more than one sm on a node, etc.) and if and when opensm doesn't
> startup normally, it would be nice to have a meaningful exit code.
>
> Perhaps others might find it useful as well, or for some future use.
Maybe, but for this clear rules should be defined and applied, not just
several exit codes. Ideally such work could be done in parallel - OpenSM
and analyzing tool (not a Chicken/Egg :)).
> But again, I originally considered this more as code cleanup. Converting the exits, returns, and aborts
> to provide a more consistent interface to the system.
Ok, if it is only the purpose we can do something like this (assuming
all exit(), abort(), etc. and not only in main.c are converted), but in
this case I would suggest to start with very limited error codes set, and
to not add OSM_EXIT_TYPE_NORMAL - "0" looks better and it is fine for the
system too. And in any case I don't see this as OFED materials.
Sasha
More information about the general
mailing list