RE: [PATCH v3 0/4] io_uring iopoll in scsi layer

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



> Hi,
> I don't understand how this patchset works. My testing shows scsi_debug is
> broken and I will be sending a correcting patch shortly (similar to the
> one I
> sent you on 20210108).

Hi Doug -

scsi_debug patch from this series works on my setup. I was under impression
that you want this patch to be available in tree and on top of current
patchset, you want to have further incremental update.
What do you suggest ? Do you want me to wait for your updated patch OR We
can ask Martin to pick all the patches except scsi_debug ? You can post
scsi_debug changes as another series or separate patch.

I have few more megaraid_sas patches in pipeline, so I am looking for this
series to be available as baseline.

Kashyap

>
> The scsi_debug driver is a simplified LLD that needs to know in advance
> whether a request/command issued to it will be using the .mq_poll
> callback.
> Perhaps you have found another way but one simple way to find that out is
> this test:
>     if (request->cmd_flags & REQ_HIPRI)
>

Agree. I am not very much familiar with scsi_debug code so used current code
change as starting point and from there things can be improved.

> In the case of scsi_debug (after my patch) the delay associated with the
> command is not wired up to generate an event which leads to completion.
> Instead, callbacks through .mq_poll are expected and they will check if
> that
> delay has expired, if not the callback returns 0. When the delay has
> expired
> and a .mq_poll is received then completion occurs.
>
> Doug Gilbert
>
> > v3 ->
> > - added reviewed-by tag
> > - Fix comment provided by Hannes for below patch.
> > https://patchwork.kernel.org/project/linux-scsi/patch/20201203034100.2
> > 9716-3-kashyap.desai@xxxxxxxxxxxx/
> > - Fix Functional issue of poll_queues settings not working in v2.
> >
> > v2 ->
> > - updated feedback from v1.
> > - added reviewed-by & tested-by tag
> > - remove flood of prints in scsi_debug driver during iopoll
> >    reported by Douglas Gilbert.
> > - added new patch to support to get shost from hctx.
> >    added new helper function "scsi_init_hctx"
> >
> > v1 ->
> > Fixed warnings in scsi_debug driver.
> > Reported-by: kernel test robot <lkp@xxxxxxxxx>
> >
> > Kashyap Desai (4):
> >    add io_uring with IOPOLL support in scsi layer
> >    megaraid_sas: iouring iopoll support
> >    scsi_debug : iouring iopoll support
> >    scsi: set shost as hctx driver_data
> >
> >   drivers/scsi/megaraid/megaraid_sas.h        |   3 +
> >   drivers/scsi/megaraid/megaraid_sas_base.c   |  87 +++++++++++--
> >   drivers/scsi/megaraid/megaraid_sas_fusion.c |  42 ++++++-
> >   drivers/scsi/megaraid/megaraid_sas_fusion.h |   2 +
> >   drivers/scsi/scsi_debug.c                   | 130 ++++++++++++++++++++
> >   drivers/scsi/scsi_lib.c                     |  29 ++++-
> >   include/scsi/scsi_cmnd.h                    |   1 +
> >   include/scsi/scsi_host.h                    |  11 ++
> >   8 files changed, 291 insertions(+), 14 deletions(-)
> >
> >
> > base-commit: a927ec3995427e9c47752900ad2df0755d02aba5
> >

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Index of Archives]     [SCSI Target Devel]     [Linux SCSI Target Infrastructure]     [Kernel Newbies]     [IDE]     [Security]     [Git]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux ATA RAID]     [Linux IIO]     [Samba]     [Device Mapper]

  Powered by Linux