[Bug] (Purley) x86/hpet: Reduce HPET counter read contention
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
intel |
Fix Released
|
Undecided
|
Unassigned | ||
linux (Ubuntu) |
Fix Released
|
Undecided
|
Unassigned | ||
Yakkety |
Fix Released
|
Undecided
|
Tim Gardner | ||
Zesty |
Fix Released
|
Undecided
|
Unassigned |
Bug Description
Issue description:
Boot ubuntu 16.10 and upstream v4.8 kernel on Purley, sometimes we saw following call trace:
[ 8.740679] hpet0: at MMIO 0xfed00000, IRQs 2, 8, 0, 0, 0, 0, 0, 0^M
[ 8.747701] hpet0: 8 comparators, 64-bit 24.000000 MHz counter^M
[ 32.765648] NMI watchdog: BUG: soft lockup - CPU#129 stuck for 22s! [watchdog/
[ 32.779913] NMI watchdog: BUG: soft lockup - CPU#130 stuck for 22s! [swapper/130:0]^M
[ 32.779915] Modules linked in:^M
[ 32.779918] CPU: 130 PID: 0 Comm: swapper/130 Not tainted 3.10.0-
[ 32.779919] Hardware name: Intel Corporation PURLEY/PURLEY, BIOS PLYDCRB1.
[ 32.779921] task: ffff8830b3fb3ec0 ti: ffff8801738e0000 task.ti: ffff8801738e0000^M
[ 32.779930] RIP: 0010:[<
[ 32.779932] RSP: 0000:ffff880173
<snip>
(All cpus are stuck and NMI dump resulting in a large amount of output.)
Issue solution:
v4.9 kernel has a commit to fix this issue, please backport it to 16.10:
f99fd22 x86/hpet: Reduce HPET counter read contention
Changed in linux (Ubuntu Yakkety): | |
status: | In Progress → Fix Committed |
tags: |
added: verification-done-yakkety removed: verification-needed-yakkety |
Changed in intel: | |
status: | New → Fix Released |
https:/ /lists. ubuntu. com/archives/ kernel- team/2016- November/ 081204. html