Re: [PATCH 0/2] vtime: Remove pair of seqcount on context switch

From: Frederic Weisbecker
Date: Mon Oct 07 2019 - 12:51:46 EST


On Mon, Oct 07, 2019 at 06:20:31PM +0200, Ingo Molnar wrote:
>
> * Frederic Weisbecker <frederic@xxxxxxxxxx> wrote:
>
> > Extracted from a larger queue that fixes kcpustat on nohz_full, these
> > two patches have value on their own as they remove two write barriers
> > on nohz_full context switch.
> >
> > Frederic Weisbecker (2):
> > vtime: Rename vtime_account_system() to vtime_account_kernel()
> > vtime: Spare a seqcount lock/unlock cycle on context switch
> >
> > arch/ia64/kernel/time.c | 4 +--
> > arch/powerpc/kernel/time.c | 6 ++--
> > arch/s390/kernel/vtime.c | 4 +--
> > include/linux/context_tracking.h | 4 +--
> > include/linux/vtime.h | 38 ++++++++++++------------
> > kernel/sched/cputime.c | 50 ++++++++++++++++++--------------
> > 6 files changed, 57 insertions(+), 49 deletions(-)
>
> Which tree is this against? Doesn't apply cleanly to v5.4-rc2 nor v5.3.
> Does it have any prereqs perhaps?

Indeed, you need to apply this fix first: https://lore.kernel.org/lkml/20190925214242.21873-1-frederic@xxxxxxxxxx/

"[PATCH] sched/vtime: Fix guest/system mis-accounting on task switch"

Thanks!