[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-mm
Subject:    [PATCH 25/29] mm: disable speculative faults for single threaded user space
From:       Michel Lespinasse <michel () lespinasse ! org>
Date:       2021-04-30 19:52:26
Message-ID: 20210430195232.30491-26-michel () lespinasse ! org
[Download RAW message or body]

Performance tuning: single threaded userspace does not benefit from
speculative page faults, so we turn them off to avoid any related
(small) extra overheads.

Signed-off-by: Michel Lespinasse <michel@lespinasse.org>
---
 arch/x86/mm/fault.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c
index 463061186827..b5c21585e35f 100644
--- a/arch/x86/mm/fault.c
+++ b/arch/x86/mm/fault.c
@@ -1321,6 +1321,14 @@ void do_user_addr_fault(struct pt_regs *regs,
 #endif
 
 #ifdef CONFIG_SPECULATIVE_PAGE_FAULT
+
+	/*
+	 * No need to try speculative faults for kernel or
+	 * single threaded user space.
+	 */
+	if (!(flags & FAULT_FLAG_USER) || atomic_read(&mm->mm_users) == 1)
+		goto no_spf;
+
 	count_vm_event(SPF_ATTEMPT);
 	seq = mmap_seq_read_start(mm);
 	if (seq & 1)
@@ -1353,7 +1361,9 @@ void do_user_addr_fault(struct pt_regs *regs,
 
 spf_abort:
 	count_vm_event(SPF_ABORT);
-#endif
+no_spf:
+
+#endif	/* CONFIG_SPECULATIVE_PAGE_FAULT */
 
 	/*
 	 * Kernel-mode access to the user address space should only occur
-- 
2.20.1


[prev in list] [next in list] [prev in thread] [next in thread]