public inbox for isar-users@googlegroups.com
 help / color / mirror / Atom feed
From: Uladzimir Bely <ubely@ilbers.de>
To: Henning Schild <henning.schild@siemens.com>
Cc: isar-users@googlegroups.com, Jan Kiszka <jan.kiszka@siemens.com>
Subject: Re: [PATCH] expand-on-first-boot: Fix early exit in helper script
Date: Fri, 10 Jun 2022 13:03:36 +0300	[thread overview]
Message-ID: <4916772.Qq0lBPeGtt@home> (raw)
In-Reply-To: <20220610113253.123f3f75@md1za8fc.ad001.siemens.net>

In the email from Friday, 10 June 2022 12:32:53 +03 user Henning Schild wrote:
> Am Fri, 10 Jun 2022 10:13:54 +0300
> 
> schrieb Uladzimir Bely <ubely@ilbers.de>:
> > In the email from Friday, 10 June 2022 08:14:19 +03 user Jan Kiszka
> > 
> > wrote:
> > > On 10.06.22 05:19, Uladzimir Bely wrote:
> > > > Changes in 342a751e introduced a regression when error from 'find'
> > > > makes helper script early exit (with -1 error code) just after an
> > > > attempt to assign ROOT_DEV_SLAVE variable. So, the last partition
> > > > remained not resized.
> > > > 
> > > > Passing 'find' output to 'head -1' masks possible error from
> > > > 'find' without breaking the logic and continues the script
> > > > execution.
> > > > 
> > > > Signed-off-by: Uladzimir Bely <ubely@ilbers.de>
> > > > ---
> > > > 
> > > >  .../expand-on-first-boot/files/expand-last-partition.sh
> > > >  
> > > > | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
> > > > 
> > > > diff --git
> > > > a/meta/recipes-support/expand-on-first-boot/files/expand-last-partitio
> > > > n.s
> > > > h
> > > > b/meta/recipes-support/expand-on-first-boot/files/expand-last-partitio
> > > > n.s
> > > > h index f3e29a1c..934a014f 100755
> > > > ---
> > > > a/meta/recipes-support/expand-on-first-boot/files/expand-last-partitio
> > > > n.s
> > > > h +++
> > > > b/meta/recipes-support/expand-on-first-boot/files/expand-last-partitio
> > > > n.s
> > > > h @@ -11,7 +11,7 @@ set -e
> > > > 
> > > >  ROOT_DEV="$(findmnt / -o source -n)"
> > > >  ROOT_DEV_NAME=${ROOT_DEV##*/}
> > > > 
> > > > -ROOT_DEV_SLAVE=$(find /sys/block/"${ROOT_DEV_NAME}"/slaves
> > > > -mindepth 1 -print -quit 2>/dev/null) +ROOT_DEV_SLAVE=$(find
> > > > /sys/block/"${ROOT_DEV_NAME}"/slaves -mindepth 1 -print -quit
> > > > 2>/dev/null | head -1)>
> > > > 
> > > >  if [ -n "${ROOT_DEV_SLAVE}" ]; then
> > > >  
> > > >  	ROOT_DEV=/dev/${ROOT_DEV_SLAVE##*/}
> > > >  
> > > >  fi
> > > 
> > > "-quit" is supposed to obsolete "head -1". Feel free to prove me
> > > wrong, but that was the reason for dropping it.
> > > 
> > > Jan
> > 
> > Yes, "-quit" makes 'find' to exit after first match. But if nothing
> > found, the return code of 'find' is "-1" and due to script's "set -e"
> > it just exits.
> > 
> > An example that demonstrates the behaviour (on raspberry):
> > 
> > root@isar:~# cat /boot/test.sh
> > #!/bin/sh
> > 
> > set -e
> > 
> > echo "1: start"
> > ROOT_DEV="$(findmnt / -o source -n)"
> > echo "2: ROOT_DEV=${ROOT_DEV}"
> > ROOT_DEV_NAME=${ROOT_DEV##*/}
> > echo "3: ROOT_DEV_NAME=${ROOT_DEV_NAME}"
> > ROOT_DEV_SLAVE=$(find /sys/block/"${ROOT_DEV_NAME}"/slaves -mindepth
> > 1 -print -quit
> > 2>/dev/null)
> > echo "4: ROOT_DEV_SLAVE=${ROOT_DEV_SLAVE}"
> > root@isar:~# sh /boot/test.sh
> > 1: start
> > 2: ROOT_DEV=/dev/mmcblk0p2
> > 3: ROOT_DEV_NAME=mmcblk0p2
> > root@isar:~# echo $?
> > 1
> > root@isar:~# ls /sys/block/ | grep mmc
> > mmcblk0
> > root@isar:~#
> > So, the easiest way to mask error returned by 'find' is to use
> > something like 'head -1' or 'tee / dev/null'. In this case script
> > just continues with empty ROOT_DEV_SLAVE.
> 
> i think something other than head would be better to catch the case
> where find does not find anything.
> 
> i.e.
> 
> rc=0
> find || rc=$?
> if [ $rc -eq 0 -a -n "${ROOT_DEV_SLAVE}" ]; then
> 
> or simply use a "find || true"
> 
> Henning

Yes, '|| true' also works. And it looks better then 'tee' or 'head'.

I'd also like to make this more clear for me... On my machine I have 
ROOT_DEV_NAME=nvme0n1p3 (partition #3), but in
"/sys/block/" I have only "nvme0n1" (whole device) directory with empty 
"slaves" inside.

Similar thing I see on Raspberry - ROOT_DEV_NAME is a partition on sdcard, 
while only whole (mmcblk0) device is mentioned in "/sys/block/"

In which cases nodes for the partition could exist in "/sys/block", so that 
'find' would not produce errors?

-- 
Uladzimir Bely




      reply	other threads:[~2022-06-10 10:03 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-10  3:19 Uladzimir Bely
2022-06-10  5:14 ` Jan Kiszka
2022-06-10  7:13   ` Uladzimir Bely
2022-06-10  9:32     ` Henning Schild
2022-06-10 10:03       ` Uladzimir Bely [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=4916772.Qq0lBPeGtt@home \
    --to=ubely@ilbers.de \
    --cc=henning.schild@siemens.com \
    --cc=isar-users@googlegroups.com \
    --cc=jan.kiszka@siemens.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