[ofiwg] RFC on error handling in fi_getinfo call

[Coulter, Susan K]

On Jan 15, 2015, at 4:46 PM, "Hefty, Sean" <sean.hefty at intel.com>
 wrote:

>> Thank you.  And I guess the operative word here is "silently".
>> Is that the bit that would require an API change?  ( not being silent
>> about the sockets choice ? )
> 
> The API returns a list of fi_info structures on success or an error.  This could be changed to:
> 
> - Return fi_info structures even if an error is returned
> - Return multiple error codes + fi_info list

> - Carry error information as part of fi_info

This last option seems nice.
How much work is it to modify the fi_info structure?

> ?
> 
> It's a matter of complexity being exposed to the user.  


> We've also discussed adding runtime options to disable built-in providers as a work-around for buggy providers.

This is also a nice option - but again I don't have a feel for the amount of work this involves.

> 
> - Sean

====================================

Susan Coulter
HPC-3 Network/Infrastructure
505-667-8425
Increase the Peace...
An eye for an eye leaves the whole world blind
====================================