This bug has been fixed on Ubuntu 14.04 ESM,
linux kernel package version 3.13.0-175.226.
Marking the LP bug as Fix Released / Trusty.
** Changed in: linux (Ubuntu Trusty)
Status: Fix Committed => Fix Released
--
You received this bug notification because you are a member of Ubuntu
Verification on Trusty ESM updates.
$ uname -rv
3.13.0-175-generic #226-Ubuntu SMP Fri Nov 8 15:26:34 UTC 2019
The problem does not happen with the test-case.
(kmod.c updated for the new kernel uname strings, same instruction addresses
from test build apply.)
Test case snippets:
Migrated from
** Changed in: linux (Ubuntu Trusty)
Status: In Progress => Fix Committed
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1813018
Title:
Kernel Oops - unable to handle kernel paging request;
** Tags added: sts
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1813018
Title:
Kernel Oops - unable to handle kernel paging request; RIP is at
wait_migrate_huge_page+0x51/0x70
To manage
** Description changed:
[Impact]
* Users on NUMA systems (mostly servers) with
NUMA balancing enabled (which is by default)
might hit a crash/BUG() on a race condition
if two simultaneous page faults of the same
transparent hugepage go into the path for
migration
Test-case snippet on original kernel on ESM (3.13.0-174.225)
[ 1387.632017] cpu 4/pid 1920/task TWO :: change_prot_numa() :: address =
0x7f4560e0, end = 0x7f456100
[ 1389.640322] cpu 5/pid 1921/task ONE :: do_huge_pmd_numa_page() :: addr/mask
= 0x7f4560e0, addr = 0x7f4560e0, pmd
Test-case snippet on original kernel (3.13.0-170.220)
[ 330.980173] cpu 5/pid 2126/task ONE :: Stage 4. T1 wake up T2... it may BUG!
<...>
[ 331.975122] cpu 6/pid 2125/task TWO :: Stage 4. T2 sleep for 1s... BUG
afterward?
[ 332.980237] BUG: unable to handle kernel paging request at
Not quite -- using a pastebin (expire 1y):
https://pastebin.ubuntu.com/p/pjMNxkJfJy/
** Description changed:
[Impact]
* Users on NUMA systems (mostly servers) with
NUMA balancing enabled (which is by default)
might hit a crash/BUG() on a race condition
if two simultaneous
** Attachment added: "kmod.c"
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1813018/+attachment/5301231/+files/kmod.c
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1813018
Title:
Kernel
Test-case snippet on modified kernel on ESM (3.13.0-174.225 + patch)
The problem doesn't happen anymore.
[ 169.972017] cpu 4/pid 1765/task TWO :: change_prot_numa() :: address =
0x7f534f80, end = 0x7f534fa0
[ 171.976313] cpu 5/pid 1766/task ONE :: do_huge_pmd_numa_page() ::
** Attachment added: "test.c"
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1813018/+attachment/5301230/+files/test.c
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1813018
Title:
Kernel
Checking if this snippet is properly spaced in a LP comment text
(instead of Description.)
Task 1 / CPU 1Task 2 / CPU 2
do_huge_pmd_numa_page() do_huge_pmd_numa_page()
- pmd_lock().
- trylock_page()
** Description changed:
- Kernel oops occurs randomly every now and then, seemingly when running
- memory-intensive processes (so far, it happened to me when using bowtie2
- or STAR).
+ [Impact]
+
+ * Users on NUMA systems (mostly servers) with
+NUMA balancing enabled (which is by default)
** Also affects: linux (Ubuntu Trusty)
Importance: Undecided
Status: New
** Changed in: linux (Ubuntu)
Status: Confirmed => Invalid
** Changed in: linux (Ubuntu Trusty)
Status: New => In Progress
** Changed in: linux (Ubuntu Trusty)
Importance: Undecided => Medium
**
@mluypaert
Thanks for the details.
Do you know whether the problem reproduces with any bowtie2 example that
I could run for myself? I'm not familiar w/ it.
Apparently there's a workaround for it, if you're willing to test: to
disable NUMA balancing.
This _might_ impact performance on some
Hi @mfo
I still hit this problem on instances running the 3.13.0 Kernel, but on
other instances where I upgraded the Kernel to the HWE Kernel (derived
from Ubuntu 16.04) I no longer have the issue.
It's hard to reproduce, since the problems occur only occasionally and
randomly. Out of 626
Hi @mluypaert,
Do you still hit this problem?
Would you be able to provide steps of how you run bowtie2/star that
reproudces the problem?
And/or be able to try a workaround that might prevent it from happening?
Thank you,
Mauricio
--
You received this bug notification because you are a
apport information
** Tags added: apport-collected ec2-images trusty
** Description changed:
Kernel oops occurs randomly every now and then, seemingly when running
memory-intensive processes (so far, it happened to me when using bowtie2
or STAR).
Running Ubuntu 14.04 LTS on AWS EC2
18 matches
Mail list logo