[EXT] NACK: [SRU] [Focal] [PATCH v3 0/4] qla2xxx: bug fixes

[Nilesh Javali]

Tim,

The patches 1 and 2 can be cherry-picked.
However patches 3 and 4 diverge from the upstream due to a dependent commit,
84ed362ac40c scsi: qla2xxx: Dual FCP-NVMe target port support

We wanted to keep the bug fixes and code changes limited and port only the upstream commits
that fixes the issue seen in Bug 1860724. Hence the commit 84ed362ac40c is skipped and that is the reason
for code divergence.

For the verification part I would let Saurav K / Quinn T comment who are involved in triaging
this particular bug and identified the fixes.

Thanks,
Nilesh

> -----Original Message-----
> From: Tim Gardner <tim.gardner at canonical.com>
> Sent: Thursday, June 3, 2021 5:34 PM
> To: Nilesh Javali <njavali at marvell.com>; kernel-team at lists.ubuntu.com
> Cc: GR-QLogic-Storage-Upstream <GR-QLogic-Storage-
> Upstream at marvell.com>
> Subject: [EXT] NACK: [SRU] [Focal] [PATCH v3 0/4] qla2xxx: bug fixes
> 
> External Email
> 
> ----------------------------------------------------------------------
> Hi Nilesh - I hate to be tedious, but I couldn't get patch 2 or 3 to
> apply to Focal tip. I was able to cherry-pick 2, but 3 still wouldn't apply.
> 
> Patch 4 backport is quite a bit different then the upstream commit. Have
> you run this kernel on any other hosts with this storage adapter to make
> sure there have been no regressions ? There is no indication in the bug
> report. We really want to be careful since (I think) this is a common
> device in the server world, i.e., cloud.
> 
> rtg
> 
> On 6/2/21 3:25 AM, Nilesh Javali wrote:
> > BugLink: https://urldefense.proofpoint.com/v2/url?u=https-
> 3A__bugs.launchpad.net_bugs_1860724&d=DwICaQ&c=nKjWec2b6R0mOyP
> az7xtfQ&r=FAW9wuzbtHIZL7SV63sr8rG59Hctu-
> eGu0G9pxwOXgQ&m=S9f8VMVydi92_urTmLA60MnouLP-
> G0QZCAtKwiv6u5A&s=vfBsH2fbmCjoi1qpH0mcFDxV8_ol341Oe_XqpMjPQ08&
> e=
> >
> > SRU Justification:
> >
> > [Impact]
> >
> > * On Ubuntu 20.04, NVMe Target will not get discovered in N2N
> > * configuration. The discovery was failing because Initiator was
> > * trying FC PRLI instead of NVMe PRLI.
> >
> > [Fix]
> >
> > * The driver code changes to handle NVMe N2N configuration and send
> PRLI
> > * based on the target.
> >
> > [Test Plan]
> >
> > * Load driver in NVMe N2N configuration and NVMe devices should be
> > * discovered by an Initiator.
> >
> > [Where problems could occur]
> >
> > * NVMe N2N configuration.
> >
> > [Other Info]
> >
> > * The bug is with inbox qla2xxx driver and needs update.
> > * All the qla2xxx driver bug fixes are part of upstream kernel.
> > * The repo used is,
> > * https://urldefense.proofpoint.com/v2/url?u=https-
> 3A__git.launchpad.net_-7Eubuntu-2Dkernel_ubuntu_-2Bsource_linux_-
> 2Bgit_focal&d=DwICaQ&c=nKjWec2b6R0mOyPaz7xtfQ&r=FAW9wuzbtHIZL7S
> V63sr8rG59Hctu-eGu0G9pxwOXgQ&m=S9f8VMVydi92_urTmLA60MnouLP-
> G0QZCAtKwiv6u5A&s=jA7n5gQeoEjumbmjvWWjN5iqrSSfB8KsC14r_03N3cU&
> e=
> > * The base-commit used to apply these bug fixes is at the bottom.
> >
> > v3:
> > - Updated cherry picked and backported commit details for each patch
> > - Updated subject line to remove pull request
> > - Add v3 and v2 change list to cover letter
> >
> > v2:
> > - Added SRU justification to cover letter
> > - Updated subject line of each patch
> > - Added mainline commit details for each patch
> >
> > Arun Easi (1):
> >    scsi: qla2xxx: Fix point-to-point (N2N) device discovery issue
> >
> > Quinn Tran (2):
> >    scsi: qla2xxx: Serialize fc_port alloc in N2N
> >    scsi: qla2xxx: Set Nport ID for N2N
> >
> > Quinn Trann (1):
> >    scsi: qla2xxx: Fix N2N and NVMe connect retry failure
> >
> >   drivers/scsi/qla2xxx/qla_def.h  |  2 +
> >   drivers/scsi/qla2xxx/qla_init.c | 68 ++++++++++++++++++++++++++-------
> >   drivers/scsi/qla2xxx/qla_mbx.c  | 47 ++++++++++++-----------
> >   drivers/scsi/qla2xxx/qla_os.c   | 10 +++--
> >   4 files changed, 88 insertions(+), 39 deletions(-)
> >
> >
> > base-commit: b2b21184870d7b313573f1ad27d259adb27dc3af
> >
> 
> --
> -----------
> Tim Gardner
> Canonical, Inc