IB/srp: merge fixes from MLNX_OFED

IB/srp: merge fixes from MLNX_OFED

[Sebastian Riemer]

Hi Sagi,

is that "/mswg/git/mlnx_ofed/mlnx-ofed-2.x-kernel.git" tree from the
MLNX_OFED public by any chance?

There are fixes included relevant for the mainline. Would be strange if
I would send the patches as somebody at Mellanox discovered and fixed
the issues.

I've hit a kernel panic today during testing caused by the loop around
ib_fmr_pool_unmap(). The loop has been fixed in MLNX_OFED. So there
should be a patch sent for it to the linux-rdma mailing list.

I've also noticed the added target locking around target->free_tx
handling in srp_rport_reconnect(). There are cases e.g. in
srp_queuecommand() where holding the rport mutex isn't enough to protect
it. So for me this looks right.

Then, in srp_create_target() I've noticed the check of the return value
of ib_query_gid(). Makes completely sense to check it.

Please send patches for so obvious fixes to the mailing list! There is a
very good chance that they get accepted.

Cheers,
Sebastian
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: IB/srp: merge fixes from MLNX_OFED

[Bart Van Assche]

On 02/18/14 17:47, Sebastian Riemer wrote:
> I've also noticed the added target locking around target->free_tx
> handling in srp_rport_reconnect(). There are cases e.g. in
> srp_queuecommand() where holding the rport mutex isn't enough to protect
> it. So for me this looks right.
> 
> Then, in srp_create_target() I've noticed the check of the return value
> of ib_query_gid(). Makes completely sense to check it.
> 
> Please send patches for so obvious fixes to the mailing list! There is a
> very good chance that they get accepted.

Hello Sebastian,

If I hadn't been busy testing scsi-mq patches recently, the SRP
initiator patches you referred to would already have been posted for
review on the linux-rdma mailing list. I hope to find some time one of
the next days to post a few SRP initiator patches on the linux-rdma
mailing list - not only the ones referred to above but also a few other
patches.

Bart.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: IB/srp: merge fixes from MLNX_OFED

[Sagi Grimberg]

On 2/18/2014 7:05 PM, Bart Van Assche wrote:
> On 02/18/14 17:47, Sebastian Riemer wrote:
>> I've also noticed the added target locking around target->free_tx
>> handling in srp_rport_reconnect(). There are cases e.g. in
>> srp_queuecommand() where holding the rport mutex isn't enough to protect
>> it. So for me this looks right.
>>
>> Then, in srp_create_target() I've noticed the check of the return value
>> of ib_query_gid(). Makes completely sense to check it.
>>
>> Please send patches for so obvious fixes to the mailing list! There is a
>> very good chance that they get accepted.
> Hello Sebastian,
>
> If I hadn't been busy testing scsi-mq patches recently, the SRP
> initiator patches you referred to would already have been posted for
> review on the linux-rdma mailing list. I hope to find some time one of
> the next days to post a few SRP initiator patches on the linux-rdma
> mailing list - not only the ones referred to above but also a few other
> patches.

Hey Sebastian & Bart,

Yes, I did plan to submit a few patches for SRP soon (I'm currently busy 
playing with scsi-mq and cleaning-up iSER DIF patches).
Thanks for the reminder, you are absolutely right!

Regarding the FMR unmap crash, I experienced it when running our 
distro-backported MLNX_OFED
and hadn't got a chance to see if it reproduces in upstream yet. Thanks 
for confirming this reproduces here as well.
Bart, Are you familiar with this issue? it should reproduce over mlx5 
driver or SRIOV (where FMRs are not supported).

Regarding the rest,
Let's let Bart propose his fixes and I'll make sure to see if I have 
anything to add there.

Sagi.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: IB/srp: merge fixes from MLNX_OFED

[Bart Van Assche]

On 02/18/14 18:25, Sagi Grimberg wrote:
> Regarding the FMR unmap crash, I experienced it when running our
> distro-backported MLNX_OFED
> and hadn't got a chance to see if it reproduces in upstream yet. Thanks
> for confirming this reproduces here as well.
> Bart, Are you familiar with this issue? it should reproduce over mlx5
> driver or SRIOV (where FMRs are not supported).

The only issue I am aware of with regard to FMR and mlx4 was specific to
a single vendor kernel in which the FMR code had been modified. mlx4 +
FMR works fine AFAIK with RHEL, SLES and with upstream kernels.
Unfortunately I have not yet had the chance to give the mlx5 driver a try.

Bart.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html