Skip to contentSkip to conversation

arm64/mm: define ARCH_SUPPORTS_SPECULATIVE_PAGE_FAULT

PATCH New Ganesh Mahendran 1 participant 2 messages May 4, 2018

Subsystems (2)
Memory management arm64
lore
Share & tools
lore

Ganesh Mahendran

author
#0 lore raw

PATCH v2 1/2 arm64/mm: define ARCH_SUPPORTS_SPECULATIVE_PAGE_FAULT

Set ARCH_SUPPORTS_SPECULATIVE_PAGE_FAULT for arm64. This enables Speculative Page Fault handler.

Signed-off-by Ganesh Mahendran <[email protected]>

---
v2: remove "if SMP"

---
 arch/arm64/Kconfig | 1 +
 1 file changed, 1 insertion(+)
arch/arm64/Kconfig +1 -0
index eb2cf49..b3ca29d 100644
@@ -144,6 +144,7 @@ config ARM64
 	select SPARSE_IRQ
 	select SYSCTL_EXCEPTION_TRACE
 	select THREAD_INFO_IN_TASK
+	select ARCH_SUPPORTS_SPECULATIVE_PAGE_FAULT
 	help
 	  ARM 64-bit (AArch64) Linux support.
Signature · 1 line
1.9.1
author
#1 lore raw

PATCH v2 2/2 arm64/mm: add speculative page fault

This patch enables the speculative page fault on the arm64 architecture.

I completed spf porting in 4.9. From the test result, we can see app launching time improved by about 10% in average. For the apps which have more than 50 threads, 15% or even more improvement can be got.

Signed-off-by Ganesh Mahendran <[email protected]>

---
v2:
move find_vma() to do_page_fault()
remove IS_ENABLED()
remove fault != VM_FAULT_SIGSEGV check
initilize vma = NULL

---
 arch/arm64/mm/fault.c | 29 +++++++++++++++++++++++++----
 1 file changed, 25 insertions(+), 4 deletions(-)
arch/arm64/mm/fault.c +25 -4
index 4165485..efd5956 100644
@@ -320,14 +320,12 @@ static void do_bad_area(unsigned long addr, unsigned int esr, struct pt_regs *re
 #define VM_FAULT_BADMAP		0x010000
 #define VM_FAULT_BADACCESS	0x020000
 
-static int __do_page_fault(struct mm_struct *mm, unsigned long addr,
+static int __do_page_fault(struct vm_area_struct *vma, unsigned long addr,
 			   unsigned int mm_flags, unsigned long vm_flags,
 			   struct task_struct *tsk)
 {
-	struct vm_area_struct *vma;
 	int fault;
 
-	vma = find_vma(mm, addr);
 	fault = VM_FAULT_BADMAP;
 	if (unlikely(!vma))
 		goto out;

@@ -371,6 +369,7 @@ static int __kprobes do_page_fault(unsigned long addr, unsigned int esr,
 	int fault, major = 0;
 	unsigned long vm_flags = VM_READ | VM_WRITE;
 	unsigned int mm_flags = FAULT_FLAG_ALLOW_RETRY | FAULT_FLAG_KILLABLE;
+	struct vm_area_struct *vma = NULL;
 
 	if (notify_page_fault(regs, esr))
 		return 0;

@@ -410,6 +409,16 @@ static int __kprobes do_page_fault(unsigned long addr, unsigned int esr,
 	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, addr);
 
 	/*
+	 * let's try a speculative page fault without grabbing the
+	 * mmap_sem.
+	 */
+	fault = handle_speculative_fault(mm, addr, mm_flags, &vma);
+	if (fault != VM_FAULT_RETRY) {
+		perf_sw_event(PERF_COUNT_SW_SPF, 1, regs, addr);
+		goto done;
+	}
+
+	/*
 	 * As per x86, we may deadlock here. However, since the kernel only
 	 * validly references user space from well defined areas of the code,
 	 * we can bug out early if this is from code which shouldn't.

@@ -431,7 +440,10 @@ static int __kprobes do_page_fault(unsigned long addr, unsigned int esr,
 #endif
 	}
 
-	fault = __do_page_fault(mm, addr, mm_flags, vm_flags, tsk);
+	if (!vma || !can_reuse_spf_vma(vma, addr))
+		vma = find_vma(mm, addr);
+
+	fault = __do_page_fault(vma, addr, mm_flags, vm_flags, tsk);
 	major |= fault & VM_FAULT_MAJOR;
 
 	if (fault & VM_FAULT_RETRY) {

@@ -454,11 +466,20 @@ static int __kprobes do_page_fault(unsigned long addr, unsigned int esr,
 		if (mm_flags & FAULT_FLAG_ALLOW_RETRY) {
 			mm_flags &= ~FAULT_FLAG_ALLOW_RETRY;
 			mm_flags |= FAULT_FLAG_TRIED;
+
+			/*
+			 * Do not try to reuse this vma and fetch it
+			 * again since we will release the mmap_sem.
+			 */
+			vma = NULL;
+
 			goto retry;
 		}
 	}
 	up_read(&mm->mmap_sem);
 
+done:
+
 	/*
 	 * Handle the "normal" (no error) case first.
 	 */
Signature · 1 line
1.9.1