Re: [vfs-next] RCU call-traces (task: modprobe)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Aug 8, 2022 at 3:43 PM Paul E. McKenney <paulmck@xxxxxxxxxx> wrote:
>
> On Sun, Aug 07, 2022 at 08:56:25AM +0200, Greg KH wrote:
> > On Sat, Aug 06, 2022 at 09:32:47AM -0700, Paul E. McKenney wrote:
> > > On Sat, Aug 06, 2022 at 08:31:26AM +0200, Sedat Dilek wrote:
> > > > Hi Al and Paul,
> > > >
> > > > I pulled latest vfs-next on top of Linux v5.19 and see this in dmesg-logs:
> > > >
> > > > root# LC_ALL=C dmesg -T | egrep -B5 -A15 'Call Trace:'
> > > > [Sat Aug  6 06:51:15 2022] Bluetooth: SCO socket layer initialized
> > > > [Sat Aug  6 06:51:15 2022] rcu: INFO: rcu_preempt detected expedited
> > > > stalls on CPUs/tasks: { 0-... } 6 jiffies s: 101 root: 0x1/.
> > >
> > > This is because of the CONFIG_RCU_EXP_CPU_STALL_TIMEOUT=20 in your
> > > .config file.  This is selected in kernels built with CONFIG_ANDROID=y,
> > > and results in an ultra-short 20-millisecond RCU CPU stall timeout for
> > > expedited grace periods.  Yes, the Android guys want and are getting
> > > milliseconds-scale response times from synchronize_rcu_expedited().
> > >
> > > The last time someone ran into this, the upshot was that CONFIG_ANDROID
> > > was to be removed: https://lwn.net/Articles/899743/
> > >
> > > My guess is that the CONFIG_ANDROID-removal patches are still on their way
> > > into mainline (they are in -next), but I added Greg KH and Christoph on CC
> > > just in case some other issue arose.
> >
> > Not that I know of, those changes are all now in Linus's tree.
>
> And now that I do another pull, I do see them, thank you!
>
> So Sedat and Ben can just merge current mainline and be fine.
>

For the records:

commit 1045a06724f322ed61f1ffb994427c7bdbe64647
"remove CONFIG_ANDROID"

diff --git a/kernel/rcu/Kconfig.debug b/kernel/rcu/Kconfig.debug
index 9b64e55d4f615..e875f4f889656 100644
--- a/kernel/rcu/Kconfig.debug
+++ b/kernel/rcu/Kconfig.debug
@@ -86,8 +86,7 @@ config RCU_EXP_CPU_STALL_TIMEOUT
  int "Expedited RCU CPU stall timeout in milliseconds"
  depends on RCU_STALL_COMMON
  range 0 21000
- default 20 if ANDROID
- default 0 if !ANDROID
+ default 0
  help
    If a given expedited RCU grace period extends more than the
    specified number of milliseconds, a CPU stall warning is printed.

-Sedat-

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=1045a06724f322ed61f1ffb994427c7bdbe64647



[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux