Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

hdr_recl calls zthr_wakeup() on destroyed zthr #9047

Merged
merged 1 commit into from
Jul 18, 2019
Merged

hdr_recl calls zthr_wakeup() on destroyed zthr #9047

merged 1 commit into from
Jul 18, 2019

Conversation

sdimitro
Copy link
Contributor

There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
that flag, destroying the zthr, and goes into
buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Signed-off-by: Serapheim Dimitropoulos serapheim@delphix.com

How Has This Been Tested?

Has been internally soaking in DelphixOS for ~6 months now

Types of changes

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Performance enhancement (non-breaking change which improves efficiency)
  • Code cleanup (non-breaking change which makes code smaller or more readable)
  • Breaking change (fix or feature that would cause existing functionality to change)
  • Documentation (a change to man pages or other documentation)

Checklist:

@sdimitro sdimitro requested review from ahrens and brad-lewis July 16, 2019 23:04
@sdimitro sdimitro added the Status: Code Review Needed Ready for review and testing label Jul 16, 2019
@sdimitro
hdr_recl calls zthr_wakeup() on destroyed zthr
7b90af2 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
@behlendorf behlendorf added Status: Accepted Ready to integrate (reviewed, tested) and removed Status: Code Review Needed Ready for review and testing labels Jul 18, 2019
@behlendorf behlendorf merged commit 1c44a5c into openzfs:master Jul 18, 2019
@codecov
Copy link

codecov bot commented Jul 18, 2019

Codecov Report

Merging #9047 into master will decrease coverage by 0.04%.
The diff coverage is 100%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master    #9047      +/-   ##
==========================================
- Coverage   78.75%   78.71%   -0.05%     
==========================================
  Files         402      402              
  Lines      121005   121007       +2     
==========================================
- Hits        95300    95249      -51     
- Misses      25705    25758      +53
Flag Coverage Δ
#kernel 79.59% <100%> (+0.02%) ⬆️
#user 66.29% <100%> (-0.24%) ⬇️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a649768...7b90af2. Read the comment docs.

TulsiJain pushed a commit to TulsiJain/zfs that referenced this pull request Jul 20, 2019
@sdimitro
hdr_recl calls zthr_wakeup() on destroyed zthr
2ba964e 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
TulsiJain pushed a commit to TulsiJain/zfs that referenced this pull request Jul 20, 2019
@sdimitro
hdr_recl calls zthr_wakeup() on destroyed zthr
069b042 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit to tonyhutter/zfs that referenced this pull request Aug 13, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
c81b40f 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit to tonyhutter/zfs that referenced this pull request Aug 21, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
a935f44 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit to tonyhutter/zfs that referenced this pull request Aug 22, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
a5ed6ad 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit to tonyhutter/zfs that referenced this pull request Aug 23, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
30c49a6 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit to tonyhutter/zfs that referenced this pull request Sep 17, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
b740a79 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit to tonyhutter/zfs that referenced this pull request Sep 18, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
a77c16b 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit to tonyhutter/zfs that referenced this pull request Sep 23, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
c3e2d4f 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes openzfs#9047
tonyhutter pushed a commit that referenced this pull request Sep 26, 2019
@sdimitro @tonyhutter
hdr_recl calls zthr_wakeup() on destroyed zthr
bbbe4b0 
There exists a race condition were hdr_recl() calls
zthr_wakeup() on a destroyed zthr. The timeline is the
following:

[1] hdr_recl() runs first and goes intro zthr_wakeup()
    because arc_initialized is set.
[2] arc_fini() is called by another thread, zeroes
    that flag, destroying the zthr, and goes into
    buf_init().
[3] hdr_recl() tries to enter the destroyed mutex
    and we blow up.

This patch ensures that the ARC's zthrs are not offloaded
any new work once arc_initialized is set and then destroys
them after all of the ARC state has been deleted.

Reviewed by: Matt Ahrens <matt@delphix.com>
Reviewed by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Serapheim Dimitropoulos <serapheim@delphix.com>
Closes #9047
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@ahrens ahrens ahrens approved these changes

@behlendorf behlendorf behlendorf approved these changes

@brad-lewis brad-lewis Awaiting requested review from brad-lewis

Assignees
No one assigned
Labels
Status: Accepted Ready to integrate (reviewed, tested)
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@sdimitro @behlendorf @ahrens