public inbox for isar-users@googlegroups.com
 help / color / mirror / Atom feed
From: Henning Schild <henning.schild@siemens.com>
To: Anton Mikanovich <amikan@ilbers.de>
Cc: isar-users@googlegroups.com
Subject: Re: [PATCH v3 0/5] Rebuild mount logic
Date: Tue, 8 Jun 2021 09:58:32 +0200	[thread overview]
Message-ID: <20210608095832.0f25116c@md1za8fc.ad001.siemens.net> (raw)
In-Reply-To: <20210508062518.83852-1-amikan@ilbers.de>

This series makes things not build in containers anymore, which we
heavily rely on for CI. Exceptions in the mount logic.

I would suggest to revert and investigate.

Henning

Am Sat,  8 May 2021 09:25:13 +0300
schrieb Anton Mikanovich <amikan@ilbers.de>:

> Mounting is perfomed very inconsistent currently in Isar. A lot of
> mount points are left after task execution and stays mounted until
> cleanup in build_completed hook.
> 
> We already experienced issues with stuck forever on unmounting at
> "Couldn't unmount, retrying..." error.
> It is also quite easy to face with a lot of ramaining mounts after the
> failed or interrupted manual builds.
> 
> This patchset implements unmounting paths after usage. In the cases of
> mount sharing corresponding request counting is performed, so real
> mount will be done only at first request, real unmount will be done
> only after the last task will try to unmount the path.
> 
> Meantime build_completed will use emergency only, so following warning
> added to show something goes wrong.
> 
> ---
> Changes since v2:
> - Prevent race conditions.
> - Fix building qemuarm64-stretch.
> - Minimize locks.
> - Cleanup counter files.
> - Unmount /proc/sys/fs/binfmt_misc also.
> Changes since v1:
> - Make all mounts be unmounted after use.
> - Warn about lost mounts after finish.
> 
> Anton Mikanovich (5):
>   dpkg: Make mount buildroot reliable
>   buildchroot: Unmount buildchroot mounts if not needed
>   rootfs: Unmount rootfs mounts if not needed
>   wic: Unmount dirs after usage
>   events: Warn if mounted paths left
> 
>  meta/classes/buildchroot.bbclass              |  47 +++++++
>  meta/classes/dpkg-base.bbclass                | 127
> +++++++++++------- meta/classes/dpkg-gbp.bbclass                 |
> 8 +- meta/classes/dpkg.bbclass                     |  14 +-
>  meta/classes/ext4-img.bbclass                 |   2 +
>  meta/classes/fit-img.bbclass                  |   2 +
>  meta/classes/image-tools-extension.bbclass    |   2 +
>  meta/classes/image.bbclass                    |  24 ++--
>  meta/classes/initramfs.bbclass                |   2 +
>  meta/classes/isar-events.bbclass              |   1 +
>  meta/classes/rootfs.bbclass                   |  62 ++++++++-
>  meta/classes/ubi-img.bbclass                  |   2 +
>  meta/classes/ubifs-img.bbclass                |   2 +
>  meta/classes/wic-img.bbclass                  |  10 ++
>  .../buildchroot/buildchroot.inc               |   8 ++
>  15 files changed, 232 insertions(+), 81 deletions(-)
> 


  parent reply	other threads:[~2021-06-08  8:02 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-08  6:25 Anton Mikanovich
2021-05-08  6:25 ` [PATCH v3 1/5] dpkg: Make mount buildroot reliable Anton Mikanovich
2021-06-08  8:09   ` Henning Schild
2021-06-08  8:31     ` Anton Mikanovich
2021-06-08  8:49       ` Henning Schild
2021-07-01 11:55   ` Jan Kiszka
2021-05-08  6:25 ` [PATCH v3 2/5] buildchroot: Unmount buildchroot mounts if not needed Anton Mikanovich
2021-05-08  6:25 ` [PATCH v3 3/5] rootfs: Unmount rootfs " Anton Mikanovich
2021-05-08  6:25 ` [PATCH v3 4/5] wic: Unmount dirs after usage Anton Mikanovich
2021-05-08  6:25 ` [PATCH v3 5/5] events: Warn if mounted paths left Anton Mikanovich
2021-07-01 11:51   ` Jan Kiszka
2021-06-08  7:58 ` Henning Schild [this message]
2021-06-08  8:15   ` [PATCH v3 0/5] Rebuild mount logic Anton Mikanovich
2021-06-08  8:42     ` Henning Schild
2021-06-08  8:54     ` Henning Schild
2021-06-10  7:12       ` Anton Mikanovich
2021-06-10 15:02         ` Henning Schild
2021-06-21 15:54 ` Jan Kiszka
2021-06-22 14:40   ` Anton Mikanovich

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=20210608095832.0f25116c@md1za8fc.ad001.siemens.net \
    --to=henning.schild@siemens.com \
    --cc=amikan@ilbers.de \
    --cc=isar-users@googlegroups.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