Linux-SCSI Archive mirror
 help / color / mirror / Atom feed
From: Allen Pais <apais@linux.microsoft.com>
To: linux-scsi@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
	target-devel@vger.kernel.org, megaraidlinux.pdl@broadcom.com,
	jejb@linux.ibm.com, hare@suse.com, martin.petersen@oracle.com,
	linuxdrivers@attotech.com, tyreld@linux.ibm.com,
	mpe@ellerman.id.au, npiggin@gmail.com,
	christophe.leroy@csgroup.eu, aneesh.kumar@kernel.org,
	naveen.n.rao@linux.ibm.com, artur.paszkiewicz@intel.co,
	kashyap.desai@broadcom.com, sumit.saxena@broadcom.com,
	shivasharan.srikanteshwara@broadcom.com,
	chandrakanth.patil@broadcom.com, jinpu.wang@cloud.ionos.com
Subject: [PATCH 0/1] Convert tasklets to bottom half workqueues
Date: Thu,  2 May 2024 20:34:32 +0000	[thread overview]
Message-ID: <20240502203433.15811-1-apais@linux.microsoft.com> (raw)

I am submitting this patch which converts instances of tasklets
in drivers/scsi/* to bottom half workqueues. I appreciate your
feedback and suggestion on the changes.

Note: The patch is only compile tested.

In the patcheset, you will notice *FIXME* in two places:
1. pm8001/pm8001_init.c @ pm8001_work(struct work_struct *t)
2. pmcraid.c @ pmcraid_work_function(struct work_struct *t)

The current implementation limits context-aware processing
within work functions due to the lack of a mechanism to identify
the source work_struct in the array. The proposed solution wraps
each work_struct with a struct work_wrapper, adding crucial context
like the array index and a reference to the parent data structure.

Ex:

#define SOME_CONSTANT 10
struct xxx_data {

.....
struct work_struct work[SOME_CONSTANT]:
.....
};

The xxx_data module currently uses an array of work_structs
for scheduling work, but it lacks the ability to identify which
array element is associated with a specific invocation of the work
function. This limitation prevents the execution of context-specific
actions based on the source of the work request.

The proposed solution is to introduce a struct work_wrapper that
encapsulates each work_struct along with additional metadata,
including an index and a pointer to the parent xxx_data structure.
This enhancement allows the work function to access necessary
context information.

Changes:

1. Definition of struct work_wrapper:

struct work_wrapper {
    struct work_struct work;
    struct xxx_data *data;
    int index;
};

struct xxx_data {
    struct work_wrapper work[SOME_CONSTANT];
};

During initialization:

for (int i = 0; i < SOME_CONSTANT; i++) {
    p->work[i].data = p;
    p->work[i].index = i;
    INIT_WORK(&p->work[i].work, work_func);
}

And it's usage in the handler:

void work_func(struct work_struct *t)
{
    struct work_wrapper *wrapper = from_work(wrapper, t, work);
    struct xxx_data *a = wrapper->data;
    int index = wrapper->index;

    ....
}

If the above is solution is acceptable, I can have the same
incorporated in version 2.

Thanks.

Allen Pais (1):
  [RFC] scsi: Convert from tasklet to BH workqueue

 drivers/scsi/aic7xxx/aic7xxx_osm.c          |  2 +-
 drivers/scsi/aic94xx/aic94xx_hwi.c          | 14 ++--
 drivers/scsi/aic94xx/aic94xx_hwi.h          |  5 +-
 drivers/scsi/aic94xx/aic94xx_scb.c          | 36 +++++-----
 drivers/scsi/aic94xx/aic94xx_task.c         | 14 ++--
 drivers/scsi/aic94xx/aic94xx_tmf.c          | 34 +++++-----
 drivers/scsi/esas2r/esas2r.h                | 12 ++--
 drivers/scsi/esas2r/esas2r_init.c           | 14 ++--
 drivers/scsi/esas2r/esas2r_int.c            | 18 ++---
 drivers/scsi/esas2r/esas2r_io.c             |  2 +-
 drivers/scsi/esas2r/esas2r_main.c           | 16 ++---
 drivers/scsi/ibmvscsi/ibmvfc.c              | 16 ++---
 drivers/scsi/ibmvscsi/ibmvfc.h              |  3 +-
 drivers/scsi/ibmvscsi/ibmvscsi.c            | 16 ++---
 drivers/scsi/ibmvscsi/ibmvscsi.h            |  3 +-
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c    | 15 ++---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.h    |  3 +-
 drivers/scsi/isci/host.c                    | 12 ++--
 drivers/scsi/isci/host.h                    |  8 +--
 drivers/scsi/isci/init.c                    |  4 +-
 drivers/scsi/megaraid/mega_common.h         |  5 +-
 drivers/scsi/megaraid/megaraid_mbox.c       | 21 +++---
 drivers/scsi/megaraid/megaraid_sas.h        |  4 +-
 drivers/scsi/megaraid/megaraid_sas_base.c   | 32 +++++----
 drivers/scsi/megaraid/megaraid_sas_fusion.c | 16 ++---
 drivers/scsi/mvsas/mv_init.c                | 27 ++++----
 drivers/scsi/mvsas/mv_sas.h                 |  9 +--
 drivers/scsi/pm8001/pm8001_init.c           | 57 ++++++++--------
 drivers/scsi/pm8001/pm8001_sas.h            |  2 +-
 drivers/scsi/pmcraid.c                      | 75 ++++++++++-----------
 drivers/scsi/pmcraid.h                      |  5 +-
 31 files changed, 249 insertions(+), 251 deletions(-)

-- 
2.17.1


             reply	other threads:[~2024-05-02 20:34 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-02 20:34 Allen Pais [this message]
2024-05-02 20:34 ` [PATCH] [RFC] scsi: Convert from tasklet to BH workqueue Allen Pais
2024-05-03  2:03   ` Michael Ellerman
2024-05-03 15:32     ` Allen Pais

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240502203433.15811-1-apais@linux.microsoft.com \
    --to=apais@linux.microsoft.com \
    --cc=aneesh.kumar@kernel.org \
    --cc=artur.paszkiewicz@intel.co \
    --cc=chandrakanth.patil@broadcom.com \
    --cc=christophe.leroy@csgroup.eu \
    --cc=hare@suse.com \
    --cc=jejb@linux.ibm.com \
    --cc=jinpu.wang@cloud.ionos.com \
    --cc=kashyap.desai@broadcom.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=linuxdrivers@attotech.com \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=martin.petersen@oracle.com \
    --cc=megaraidlinux.pdl@broadcom.com \
    --cc=mpe@ellerman.id.au \
    --cc=naveen.n.rao@linux.ibm.com \
    --cc=npiggin@gmail.com \
    --cc=shivasharan.srikanteshwara@broadcom.com \
    --cc=sumit.saxena@broadcom.com \
    --cc=target-devel@vger.kernel.org \
    --cc=tyreld@linux.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).