Hi Vishal, hi Jeff,
good morning from Germany ...
> Oops! My fault for not paying attention to the fact that Vishal
> committed some IB changes and forgetting to roll a new beta. 7.1b14 is
> now out (http://www.lam-mpi.org/beta/) that has these changes. It's
> identical to last night's snapshot (except for the version number, of
> course), so if you already grabbed that, you don't need to get b14.
>
> We would love to get some external feedback on our ib RPI module.
> Please let us know how it goes.
I got the snapshot and compiled it but the problem is the same.
I had a look at the lam sources, the problem seems to occur in
function 'int poll_cq(struct _proc *p)',
file ssi_rpi_ib_actions.c:
if ((ret != LAM_IB_OK) || (cdesc.status != LAM_IB_SUCCESS))
ret is LAM_IB_OK in my case but cdesc.status is 10 (in one
case 5) which should be a VAPI_REM_ACCESS_ERR.
I have no experience with Infiniband at all, so I have no
idea what the problem is. Maybe our Infiniband installation
is not okay. Is there an easy way to check this out?
Can you recommend a kind of 'Infiniband tutorial'?
Bye and thank you,
Charlie
|
