diff mbox series

[v4,23/26] mm/x86: Use general page fault accounting

Message ID 20200630204601.39591-1-peterx@redhat.com (mailing list archive)
State New, archived
Headers show
Series mm: Page fault accounting cleanups | expand

Commit Message

Peter Xu June 30, 2020, 8:46 p.m. UTC
Use the general page fault accounting by passing regs into handle_mm_fault().

CC: Dave Hansen <dave.hansen@linux.intel.com>
CC: Andy Lutomirski <luto@kernel.org>
CC: Peter Zijlstra <peterz@infradead.org>
CC: Thomas Gleixner <tglx@linutronix.de>
CC: Ingo Molnar <mingo@redhat.com>
CC: Borislav Petkov <bp@alien8.de>
CC: x86@kernel.org
CC: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
---
 arch/x86/mm/fault.c | 17 ++---------------
 1 file changed, 2 insertions(+), 15 deletions(-)

Comments

Dave Hansen July 1, 2020, 3:35 p.m. UTC | #1
On 6/30/20 1:46 PM, Peter Xu wrote:
> Use the general page fault accounting by passing regs into handle_mm_fault().
...
> -	/*
> -	 * Major/minor page fault accounting. If any of the events
> -	 * returned VM_FAULT_MAJOR, we account it as a major fault.
> -	 */
> -	if (major) {
> -		tsk->maj_flt++;
> -		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, regs, address);
> -	} else {
> -		tsk->min_flt++;
> -		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, regs, address);
> -	}
> -
>  	check_v8086_mode(regs, address, tsk);
>  }

I did a quick grep and it wasn't obvious to me how
PERF_COUNT_SW_PAGE_FAULTS_MIN/MAJ get bumped in the handle_mm_fault() path.

Are you sure they get set?
Peter Xu July 3, 2020, 12:46 a.m. UTC | #2
Hi, Dave,

On Wed, Jul 01, 2020 at 08:35:40AM -0700, Dave Hansen wrote:
> On 6/30/20 1:46 PM, Peter Xu wrote:
> > Use the general page fault accounting by passing regs into handle_mm_fault().
> ...
> > -	/*
> > -	 * Major/minor page fault accounting. If any of the events
> > -	 * returned VM_FAULT_MAJOR, we account it as a major fault.
> > -	 */
> > -	if (major) {
> > -		tsk->maj_flt++;
> > -		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, regs, address);
> > -	} else {
> > -		tsk->min_flt++;
> > -		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, regs, address);
> > -	}
> > -
> >  	check_v8086_mode(regs, address, tsk);
> >  }
> 
> I did a quick grep and it wasn't obvious to me how
> PERF_COUNT_SW_PAGE_FAULTS_MIN/MAJ get bumped in the handle_mm_fault() path.
> 
> Are you sure they get set?

Sorry for missing the context.  This patch is based on the 1st patch of the
same series:

https://lore.kernel.org/lkml/20200630204504.38516-1-peterx@redhat.com/

Both of them are set there.

Thanks,
diff mbox series

Patch

diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c
index fe3ca00eb121..9ac80bb87781 100644
--- a/arch/x86/mm/fault.c
+++ b/arch/x86/mm/fault.c
@@ -1140,7 +1140,7 @@  void do_user_addr_fault(struct pt_regs *regs,
 	struct vm_area_struct *vma;
 	struct task_struct *tsk;
 	struct mm_struct *mm;
-	vm_fault_t fault, major = 0;
+	vm_fault_t fault;
 	unsigned int flags = FAULT_FLAG_DEFAULT;
 
 	tsk = current;
@@ -1292,8 +1292,7 @@  void do_user_addr_fault(struct pt_regs *regs,
 	 * userland). The return to userland is identified whenever
 	 * FAULT_FLAG_USER|FAULT_FLAG_KILLABLE are both set in flags.
 	 */
-	fault = handle_mm_fault(vma, address, flags, NULL);
-	major |= fault & VM_FAULT_MAJOR;
+	fault = handle_mm_fault(vma, address, flags, regs);
 
 	/* Quick path to respond to signals */
 	if (fault_signal_pending(fault, regs)) {
@@ -1320,18 +1319,6 @@  void do_user_addr_fault(struct pt_regs *regs,
 		return;
 	}
 
-	/*
-	 * Major/minor page fault accounting. If any of the events
-	 * returned VM_FAULT_MAJOR, we account it as a major fault.
-	 */
-	if (major) {
-		tsk->maj_flt++;
-		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, regs, address);
-	} else {
-		tsk->min_flt++;
-		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, regs, address);
-	}
-
 	check_v8086_mode(regs, address, tsk);
 }
 NOKPROBE_SYMBOL(do_user_addr_fault);