Bug 1467723 - regression: perf commit cc1582c231ea breaks debugging with "rr"
Summary: regression: perf commit cc1582c231ea breaks debugging with "rr"
Keywords:
Status: CLOSED CURRENTRELEASE
Alias: None
Product: Fedora
Classification: Fedora
Component: kernel
Version: 25
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Kernel Maintainer List
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2017-07-04 21:06 UTC by Michael Stahl
Modified: 2017-09-29 18:51 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2017-09-29 18:51:46 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)

Description Michael Stahl 2017-07-04 21:06:51 UTC
Description of problem:
replay debugging with rr doesn't work

Version-Release number of selected component (if applicable):
kernel-4.11.7-200.fc25.x86_64

How reproducible:
Sometimes, but not always

Steps to Reproduce:
1. rr record instdir/program/soffice
2. trigger some hard to reproduce crash
3. rr replay ~/.local/share/rr/latest

Actual results:

[ERROR /home/roc/rr/rr/src/ReplaySession.cc:545:guard_overshoot() errno: SUCCESS] Replay diverged; target registers mismatched:
[FATAL /home/roc/rr/rr/src/ReplaySession.cc:549:guard_overshoot() errno: SUCCESS]
 (task 14923 (rec:14535) at time 42867)
 -> Assertion `false' failed to hold. overshot target ticks=2734735504 by 33386


Expected results:

working replay debugging

Additional info:

4.11.3 does not have this problem: kernel-core-4.11.3-202.fc25.x86_64

rr developers claim this is a regression of commit cc1582c231ea
"perf/core: Drop kernel samples even though :u is specified"

LKML thread:
http://www.spinics.net/lists/kernel/msg2544195.html

Comment 1 Michael Stahl 2017-09-29 18:51:46 UTC
works for me now, was fixed reportedly in 4.12.4 and 4.13


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