public inbox for linux-scsi@vger.kernel.org
 help / color / mirror / Atom feed
From: asutoshd@codeaurora.org
To: Steffen Maier <maier@linux.ibm.com>
Cc: linux-scsi@vger.kernel.org, linux-scsi-owner@vger.kernel.org
Subject: Re: Query: SCSI Device node creation when UFS is loaded as a module
Date: Thu, 24 Oct 2019 10:05:57 -0700	[thread overview]
Message-ID: <e1d00a73efd3b73adae441ff5e00c4b4@codeaurora.org> (raw)
In-Reply-To: <d2804026-7908-4601-3216-e60d51131984@linux.ibm.com>

On 2019-10-24 02:50, Steffen Maier wrote:
> On 10/24/19 1:51 AM, asutoshd@codeaurora.org wrote:
>> Hi
>> I'm loading the ufs-qcom driver as a module but am not seeing the 
>> /dev/sda* device nodes.
>> Looks like it's not being created.
>> 
>> I find the sda nodes in other paths being enumerated though:
>> 
>> / # find /sys -name sda
>> /sys/kernel/debug/block/sda
>> /sys/class/block/sda
>> /sys/devices/platform/<...>/<xxx>.ufshc/host0/target0:0:0/0:0:0:0/block/sda
>> /sys/block/sda
>> 
>> All Luns are detected and I see sda is detected and prints for all the 
>> Luns as below -:
>> sd 0:0:0:0: [sda] .... ....-byte logical blocks:
>> 
>> ... so on ...
>> 
>> But if I link it statically instead of a module, it works fine. All 
>> device nodes are created.
>> 
>> I'm trying to figure out where/how in SCSI does it create these device 
>> nodes - /dev/sd<a/b/c/d> ?
> 
> That's from (systemd-)udevd user space based on uevents from the 
> kernel.
> 
>> I've looked into sd.c but I couldn't figure out the exact place yet.
> 
> Yeah, based on the SCSI device probe and add lun, the high level
> driver sd would emit udev events for block devices.

Thanks Steffen and Avri.

I understood that sd.c creates the devices based on scsi_device and 
establishes a relation between scsi_device (sdp) and scsi_disk (sdkp).
But I couldn't figure out why it creates the device fine when the 
ufs-qcom is linked statically, but is unable to create the devices when 
inserted as a module.

Does it follow a different method to create the device nodes when 
statically linked and needs (systemd-)udevd to handle the events during 
dynamic insertion?

Thanks
asd

      reply	other threads:[~2019-10-24 17:06 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-23 23:51 Query: SCSI Device node creation when UFS is loaded as a module asutoshd
2019-10-24  6:14 ` Avri Altman
2019-10-24  9:50 ` Steffen Maier
2019-10-24 17:05   ` asutoshd [this message]

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=e1d00a73efd3b73adae441ff5e00c4b4@codeaurora.org \
    --to=asutoshd@codeaurora.org \
    --cc=linux-scsi-owner@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=maier@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