Bug 148076 - [linux] linux_sys_futex: unknown op 265
Summary: [linux] linux_sys_futex: unknown op 265
Status: Closed FIXED
Alias: None
Product: Base System
Classification: Unclassified
Component: kern (show other bugs)
Version: Unspecified
Hardware: Any Any
: Normal Affects Only Me
Assignee: Dmitry Chagin
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-06-23 00:50 UTC by Andrea Magliano
Modified: 2014-10-29 09:27 UTC (History)
0 users

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Andrea Magliano 2010-06-23 00:50:01 UTC
'emerge' stalls indefinitely, no network activity.

'top' shows stalled python:
  PID USERNAME  THR PRI NICE   SIZE    RES STATE    TIME   WCPU COMMAND
 1124 root        1  76    0 18852K 17764K ttyout   0:01  6.69% python2.6

'dmesg' (new line after emerge try):
linux_sys_futex: unknown op 265

Strangely 'emerge --sync' succeeds, but ping fails as well:
ping: recvmsg: Invalid argument

How-To-Repeat: install linux_dist-gentoo-stage3-20100601_1

mount following:
devfs           /usr/local/gentoo-stage3/dev devfs      rw 0 0
linproc         /usr/local/gentoo-stage3/proc     linprocfs       rw 0 0
linsys          /usr/local/gentoo-stage3/sys      linsysfs        rw 0 0

sudo /usr/sbin/chroot /usr/local/gentoo-stage3 su -

emerge <whatever>

'emerge' stalls indefinitely
Comment 1 Mark Linimon freebsd_committer freebsd_triage 2010-06-23 02:17:20 UTC
Responsible Changed
From-To: freebsd-ports-bugs->gabor

Fix synopsis and assign.
Comment 2 Gabor Kovesdan freebsd_committer freebsd_triage 2010-07-12 22:15:16 UTC
Responsible Changed
From-To: gabor->dchagin

This is not a problem in the port but in the underlying Linux emulation 
layer, thus it cannot be fixed in the port. Recategorize and change 
synopsis accordingly. Reassign to dchagin@, who agreed to take a look at it.
Comment 3 Jung-uk Kim freebsd_committer freebsd_triage 2010-10-06 21:12:02 UTC
> linux_sys_futex: unknown op 265

This is (FUTEX_WAIT_BITSET | FUTEX_CLOCK_REALTIME) and it is available 
from Linux 2.6.28:

http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6-stable.git;a=shortlog;h=1acdac104668a0834cfa267de9946fac7764d486

In turn, FUTEX_WAIT_BITSET itself is available from Linux 2.6.25:

http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6-stable.git;a=shortlog;h=cd689985cf49f6ff5c8eddc48d98b9d581d9475d

I think we should suspend this PR until Linuxulator adds syscalls 
beyond 2.6.16.

FYI...

Jung-uk Kim
Comment 4 Alexander Best freebsd_committer freebsd_triage 2010-11-19 13:43:36 UTC
please note that r215338 should reduce the output storm of such warnings,
however the futex implementation is still missing support for FUTEX_WAIT_BITSET
and a few others.

cheers.
alex

-- 
a13x
Comment 5 Alexander Best freebsd_committer freebsd_triage 2011-01-30 07:49:39 UTC
State Changed
From-To: open->patched

For now we simply made FUTEX_CLOCK_REALTIME a no op. The linuxulator should no 
longer complain about it. A real implementation however is neded to fully 
support futex operators FUTEX_WAIT_BITSET and FUTEX_WAIT_REQUEUE_PI, which 
support switching between CLOCK_MONOTONIC and FUTEX_CLOCK_REALTIME.
Comment 6 Dmitry Chagin freebsd_committer freebsd_triage 2014-10-29 09:27:42 UTC
merge to stable/8 by r219167