[PATCH 3/3] x86,tlb: put cpumask_test_cpu in prev == next under CONFIG_DEBUG_VM

Rik van Riel posted 3 patches 2 weeks ago
There is a newer version of this series
[PATCH 3/3] x86,tlb: put cpumask_test_cpu in prev == next under CONFIG_DEBUG_VM
Posted by Rik van Riel 2 weeks ago
On a web server workload, the cpumask_test_cpu inside the
WARN_ON_ONCE in the prev == next branch takes about 17% of
all the CPU time of switch_mm_irqs_off.

On a large fleet, this WARN_ON_ONCE has not fired in at least
a month, possibly never.

Move this test under CONFIG_DEBUG_VM so it does not get compiled
in production kernels.

Signed-off-by: Rik van Riel <riel@surriel.com>
---
 arch/x86/mm/tlb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 9d0d34576928..1aac4fa90d3d 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -568,7 +568,7 @@ void switch_mm_irqs_off(struct mm_struct *unused, struct mm_struct *next,
 		 * mm_cpumask. The TLB shootdown code can figure out from
 		 * cpu_tlbstate_shared.is_lazy whether or not to send an IPI.
 		 */
-		if (WARN_ON_ONCE(prev != &init_mm &&
+		if (IS_ENABLED(CONFIG_DEBUG_VM) && WARN_ON_ONCE(prev != &init_mm &&
 				 !cpumask_test_cpu(cpu, mm_cpumask(next))))
 			cpumask_set_cpu(cpu, mm_cpumask(next));
 
-- 
2.45.2
[tip: x86/mm] x86/mm/tlb: Put cpumask_test_cpu() check in switch_mm_irqs_off() under CONFIG_DEBUG_VM
Posted by tip-bot2 for Rik van Riel 1 week, 3 days ago
The following commit has been merged into the x86/mm branch of tip:

Commit-ID:     7e33001b8b9a78062679e0fdf5b0842a49063135
Gitweb:        https://git.kernel.org/tip/7e33001b8b9a78062679e0fdf5b0842a49063135
Author:        Rik van Riel <riel@surriel.com>
AuthorDate:    Fri, 08 Nov 2024 19:27:50 -05:00
Committer:     Ingo Molnar <mingo@kernel.org>
CommitterDate: Wed, 13 Nov 2024 03:42:41 +01:00

x86/mm/tlb: Put cpumask_test_cpu() check in switch_mm_irqs_off() under CONFIG_DEBUG_VM

On a web server workload, the cpumask_test_cpu() inside the
WARN_ON_ONCE() in the 'prev == next branch' takes about 17% of
all the CPU time of switch_mm_irqs_off().

On a large fleet, this WARN_ON_ONCE() has not fired in at least
a month, possibly never.

Move this test under CONFIG_DEBUG_VM so it does not get compiled
in production kernels.

Signed-off-by: Rik van Riel <riel@surriel.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/r/20241109003727.3958374-4-riel@surriel.com
---
 arch/x86/mm/tlb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index 86593d1..b0d5a64 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -568,7 +568,7 @@ void switch_mm_irqs_off(struct mm_struct *unused, struct mm_struct *next,
 		 * mm_cpumask. The TLB shootdown code can figure out from
 		 * cpu_tlbstate_shared.is_lazy whether or not to send an IPI.
 		 */
-		if (WARN_ON_ONCE(prev != &init_mm &&
+		if (IS_ENABLED(CONFIG_DEBUG_VM) && WARN_ON_ONCE(prev != &init_mm &&
 				 !cpumask_test_cpu(cpu, mm_cpumask(next))))
 			cpumask_set_cpu(cpu, mm_cpumask(next));