Bug 1096185 - Do not expand mountpoint path, since it might cause deadlock
Summary: Do not expand mountpoint path, since it might cause deadlock
Keywords:
Status: CLOSED INSUFFICIENT_DATA
Alias: None
Product: Fedora
Classification: Fedora
Component: mock
Version: rawhide
Hardware: Unspecified
OS: Unspecified
low
low
Target Milestone: ---
Assignee: Miroslav Suchý
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2014-05-09 11:37 UTC by Vít Ondruch
Modified: 2014-11-03 17:24 UTC (History)
3 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2014-11-03 17:24:00 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)
Do not expand the mountpoint path (1.18 KB, patch)
2014-05-09 11:37 UTC, Vít Ondruch
no flags Details | Diff

Description Vít Ondruch 2014-05-09 11:37:13 UTC
Created attachment 894017 [details]
Do not expand the mountpoint path

Description of problem:
Probing mountpoints might cause problems, when NFS mount is not working as expected. This is unnecessary issue.


Version-Release number of selected component (if applicable):
$ rpm -q mock
mock-1.1.38-1.fc20.noarch


Additional info:
Anyway, trying to unmount something which was not mounted by mock itself seems fishy and it causes another pain, when I am using mock shell with mounted development directory during development. In case I need to install some additional package, I have to remount the directory again. So I would suggest to drop it.

Comment 1 Miroslav Suchý 2014-07-17 13:04:40 UTC
I tried really hard to reproduce. But no luck.
No luck with pure NFS mounts.
I got some weird behaviour with 
  foo.bar.com:/exports/archive         /mnt/archive       nfs ro,noauto,x-systemd.automount,x-systemd.device-timeout=10 0 0
mounts (which Vít is using). But that behave strangly even without mock (I suspect something smelly in systemd).
But I was not able to reproduce anything deterministic.

Without real reproducer, I doubt I can do anything. Unless somebody provide some reproducer, I will close it in few months.

Comment 2 Miroslav Suchý 2014-11-03 17:24:00 UTC
I tried again (BTW the code is now in _mount_is_ours() and is refactored) and no luck to reproduce.
Feel free to reopen when there is better reproducer.


Note You need to log in before you can comment on or make changes to this bug.