iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

powerpc/64s/radix: add warning and comments in mm_cpumask trim

Browse Source 
Add a comment explaining part of the logic for mm_cpumask trimming, and
add a (hopefully graceful) check and warning in case something gets it
wrong.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20201217134731.488135-2-npiggin@gmail.com
This commit is contained in:
Nicholas Piggin 2020-12-17 23:47:25 +10:00 committed by Michael Ellerman
parent e79b76e03b
commit a2496049f1
1 changed files with 21 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
arch/powerpc/mm/book3s64/radix_tlb.c
View File

@ -653,13 +653,14 @@ static void do_exit_flush_lazy_tlb(void *arg)
{ {
struct mm_struct *mm = arg; struct mm_struct *mm = arg;
unsigned long pid = mm->context.id; unsigned long pid = mm->context.id;
int cpu = smp_processor_id();
/* /*
* A kthread could have done a mmget_not_zero() after the flushing CPU * A kthread could have done a mmget_not_zero() after the flushing CPU
* checked mm_is_singlethreaded, and be in the process of * checked mm_cpumask, and be in the process of kthread_use_mm when
* kthread_use_mm when interrupted here. In that case, current->mm will * interrupted here. In that case, current->mm will be set to mm,
* be set to mm, because kthread_use_mm() setting ->mm and switching to * because kthread_use_mm() setting ->mm and switching to the mm is
* the mm is done with interrupts off. * done with interrupts off.
*/ */
if (current->mm == mm) if (current->mm == mm)
goto out_flush; goto out_flush;
@ -673,8 +674,22 @@ static void do_exit_flush_lazy_tlb(void *arg)
mmdrop(mm); mmdrop(mm);
} }
atomic_dec(&mm->context.active_cpus); /*
cpumask_clear_cpu(smp_processor_id(), mm_cpumask(mm)); * This IPI is only initiated from a CPU which is running mm which
* is a single-threaded process, so there will not be another racing
* IPI coming in where we would find our cpumask already clear.
*
* Nothing else clears our bit in the cpumask except CPU offlining,
* in which case we should not be taking IPIs here. However check
* this just in case the logic is wrong somewhere, and don't underflow
* the active_cpus count.
*/
if (cpumask_test_cpu(cpu, mm_cpumask(mm))) {
atomic_dec(&mm->context.active_cpus);
cpumask_clear_cpu(cpu, mm_cpumask(mm));
} else {
WARN_ON_ONCE(1);
}
out_flush: out_flush:
_tlbiel_pid(pid, RIC_FLUSH_ALL); _tlbiel_pid(pid, RIC_FLUSH_ALL);