All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
To: linux-block@vger.kernel.org
Cc: hch@lst.de, Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Subject: [PATCH V2 0/6] null_blk: simplify null_handle_cmd()
Date: Tue,  9 Jul 2019 12:21:26 -0700	[thread overview]
Message-ID: <20190709192132.24723-1-chaitanya.kulkarni@wdc.com> (raw)

Hi,

The core function where we handle the request null_handle_cmd() in the
null_blk does various things based on how null_blk is configured :-

1. Handle throttling.
2. Handle badblocks.
3. Handle Memory backed device operations.
4. Handle Zoned Block device operations.
5. Completion of the requests.

With all the above functionality present in the one function,
null_handle_cmd() is growing and becoming unreasonably lengthy when
we want to add more features such as new Zone requests which is being
worked on (See [1], [2]).

This is a cleanup patch-series which refactors the code in the
null_handle_cmd(). We create a clear interface for each of the above
mentioned functionality which leads to having null_handle_cmd() more
clear and easy to manage with future changes. Please have a look at
NVMe PCIe Driver (nvme_queue_rq()) (See [3]) which has a similar code
structure and nicely structured code for doing various things such as
setting up commands, mapping of the block layer requests, mapping
PRPs/SGLs, handling integrity requests and finally submitting the
NVMe Command etc.

With this patch-series we fix the following issues with the current 
code :-

1. Get rid of the multiple nesting levels (> 2).
2. Easy to read, debug and extend the code for specific feature.
3. Get rid of the various line foldings which are there in the code
   due to multiple level of nesting under if conditions.
4. Precise definition for the null_handle_cmd() and clear error
   handling as helpers are responsible for handling errors.

Please consider this for 5.3.

Cc: hch@lst.de

Regards,
Chaitanya

[1] https://www.spinics.net/lists/linux-block/msg41884.html
[2] https://www.spinics.net/lists/linux-block/msg41883.html
[3] https://github.com/torvalds/linux/blob/master/drivers/nvme/host/pci.c

* Changes from V1:-
1. Move bio vs req code into the callers for the null_handle_cmd() and
   add required arguments to simplify the code in the same function.
2. Get rid of the extra braces for the null_handle_zoned().
3. Get rid of the multiple returns style and the goto.
4. For throttling, code keep the check in the caller.
5. Add uniform code format for setting the cmd->error in the
   null_handle_cmd() and make required code changes so that each
   feature specific function will return blk_status_t value.

Chaitanya Kulkarni (6):
  null_blk: move duplicate code to callers
  null_blk: create a helper for throttling
  null_blk: create a helper for badblocks
  null_blk: create a helper for mem-backed ops
  null_blk: create a helper for zoned devices
  null_blk: create a helper for req completion

 drivers/block/null_blk.h       |  20 ++--
 drivers/block/null_blk_main.c  | 178 +++++++++++++++++++--------------
 drivers/block/null_blk_zoned.c |  29 +++---
 3 files changed, 128 insertions(+), 99 deletions(-)

-- 
2.17.0


             reply	other threads:[~2019-07-09 19:21 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-07-09 19:21 Chaitanya Kulkarni [this message]
2019-07-09 19:21 ` [PATCH V2 1/6] null_blk: move duplicate code to callers Chaitanya Kulkarni
2019-07-09 19:21 ` [PATCH V2 2/6] null_blk: create a helper for throttling Chaitanya Kulkarni
2019-07-09 19:21 ` [PATCH V2 3/6] null_blk: create a helper for badblocks Chaitanya Kulkarni
2019-07-09 19:21 ` [PATCH V2 4/6] null_blk: create a helper for mem-backed ops Chaitanya Kulkarni
2019-07-09 19:21 ` [PATCH V2 5/6] null_blk: create a helper for zoned devices Chaitanya Kulkarni
2019-07-09 19:21 ` [PATCH V2 6/6] null_blk: create a helper for req completion Chaitanya Kulkarni

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=20190709192132.24723-1-chaitanya.kulkarni@wdc.com \
    --to=chaitanya.kulkarni@wdc.com \
    --cc=hch@lst.de \
    --cc=linux-block@vger.kernel.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.