Title: RE: [openib-general] Re: [PATCH] [2/2] change QP state to SQE
> -Original Message-
> From: Roland Dreier [mailto:[EMAIL PROTECTED]]
> Sent: Friday, November 12, 2004 7:42 PM
>
>
> I thought about this a little, and it seems that having the
> CQ poll
Hal> For SQD2SQD, path migration state is missing as is remote
Hal> node address vector, .
Good catch on the IB_QP_AV (actually IB_QP_PATH_MIG_STATE was there).
Hal> Is IB_QP_TIMEOUT local ACK timeout ?
Yes.
Hal> Also, does MAX_QP_RD_ATOMIC handle both local and destination?
On Fri, 2004-11-12 at 23:21, Roland Dreier wrote:
> OK, here's a patch that adds support for "Current QP state" in the
> modify QP verb. Does this look OK?
Looks good to me.
A few comments/questions relative to IBA 1.2 vol 1 table 91 (p.569-572):
For SQD2SQD, path migration state is missing as
OK, here's a patch that adds support for "Current QP state" in the
modify QP verb. Does this look OK?
Thanks,
Roland
Index: infiniband/include/ib_verbs.h
===
--- infiniband/include/ib_verbs.h (revision 1223)
+++ infiniband/i
Sean> That would work fine, and be only a minor update to the MAD
Sean> code. Will you be generated a patch for mthca?
Yes, eventually. (ib_verbs.h will also need an update to add the
field to ib_qp_attr)
- R.
___
openib-general mailing list
Roland Dreier wrote:
I thought about this a little, and it seems that having the CQ poll
operation update the QP state is not the right solution. It seems it
would be better to add support for the "Current QP state" modifier for
the modify QP operation and expect the consumer to use that to
indica
I thought about this a little, and it seems that having the CQ poll
operation update the QP state is not the right solution. It seems it
would be better to add support for the "Current QP state" modifier for
the modify QP operation and expect the consumer to use that to
indicate that the QP is in
Roland Dreier wrote:
Sean> This should transition the QP state to SQE when encountering
Sean> a send error on the CQ. There may be a better way of doing
Sean> this; I didn't spend a lot of time studying the code.
Thanks for the patch... let me look at how I want to do this (and
probabl
Sean> This should transition the QP state to SQE when encountering
Sean> a send error on the CQ. There may be a better way of doing
Sean> this; I didn't spend a lot of time studying the code.
Thanks for the patch... let me look at how I want to do this (and
probably handle transitions
