[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

--- Comment #10 from Ngie Cooper  ---
(In reply to Konstantin Belousov from comment #9)

Yes, it's been fixed for a while (apologies for not following up on the bug):
https://ci.freebsd.org/job/FreeBSD-head-amd64-test/3646/testReport/lib.libthr/cond_test/cond_timedwait_race/
.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

Konstantin Belousov  changed:

   What|Removed |Added

 Status|In Progress |Closed
 Resolution|--- |FIXED

--- Comment #9 from Konstantin Belousov  ---
I believe that this is fixed but I never received the confirmation.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

Ed Maste  changed:

   What|Removed |Added

 Status|New |In Progress
 CC||ema...@freebsd.org

--- Comment #8 from Ed Maste  ---
Does this still need MFC?

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

Jan Beich  changed:

   What|Removed |Added

 Attachment #178503|text/x-csrc |text/plain
  mime type||

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

Mark Linimon  changed:

   What|Removed |Added

   Keywords||patch

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

--- Comment #7 from commit-h...@freebsd.org ---
A commit references this bug:

Author: kib
Date: Wed Jan  4 16:10:52 UTC 2017
New revision: 311287
URL: https://svnweb.freebsd.org/changeset/base/311287

Log:
  __vdso_gettc(): be extra careful with /dev/hpet mappings, never unmap
  the mapping which might be accessed by other threads.

  If a pointer to the /dev/hpet register page mapping was stored into
  the hpet_dev_map, other threads might access the page at any time.
  Never unmap it, instead, keep track of mappings for all hpet units in
  smal array.  Store pointer to the newly mapped registers page using
  CAS, to detect parallel mappings.

  It appeared relatively easy to demonstrate the problem by arranging
  two threads which perform gettimeofday(2) concurently, first time in
  the process address space, when HPET is used for timecounter.

  PR:   215715
  Sponsored by: The FreeBSD Foundation
  MFC after:1 week

Changes:
  head/lib/libc/x86/sys/__vdso_gettc.c

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

--- Comment #6 from Konstantin Belousov  ---
Created attachment 178503
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=178503=edit
A test program illustrating bug with two threads executing gettimeofday()
concurrently

The program makes an easy reproduction case for the bug I fixed in the patch. 
The timecounter should be set to HPET, then the program run.  It forks children
which try to execute gettimeofday() concurrently, and at the end print stats of
normally exited and signalled children.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

--- Comment #4 from Konstantin Belousov  ---
I do not see how the backtrace you posted has anything to do with eventtimers. 
But it might be relevant for timecounter HPET hardware.  Please show sysctl
kern.timecounter output on your test machine.

In fact, I do see an issue in multithreaded program which happens to execute
e.g. gettimeofday(2) in two threads in parallel, regardless of your test
program.  If your machine uses HPET for timecounter, please test the patch
attached.

In fact, using LAPIC for eventtimers and HPET for timecounter is very weird
configuration.  Also weird is that there is no HPET in the eventtimers hw list.
Is it VM ? Show verbose dmesg.

BTW, core file does not have any use outside of your machine.

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

--- Comment #3 from Ngie Cooper  ---
Created attachment 178455
  --> https://bugs.freebsd.org/bugzilla/attachment.cgi?id=178455=edit
test program, corefile, and debug symbols

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"

[Bug 215715] cond_test:cond_timedwait_race crashes with SIGSEGV periodically when calling binuptime

[bugzilla-noreply]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=215715

Ngie Cooper  changed:

   What|Removed |Added

Summary|cond_test:cond_timedwait_ra |cond_test:cond_timedwait_ra
   |ce crashes with SIGSEGV |ce crashes with SIGSEGV
   |periodically|periodically when calling
   ||binuptime
  Component|tests   |bin
   Assignee|freebsd-test...@freebsd.org |freebsd-bugs@FreeBSD.org

-- 
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"