Re: [PATCH v4] mm: Avoid unnecessary page fault retires on shared memory types
From: Heiko Carstens <hca@linux.ibm.com>
Date: 2022-05-29 20:37:05
Also in:
linux-alpha, linux-m68k, linux-mips, linux-mm, linux-riscv, linux-s390, linux-sh, linux-um, lkml, sparclinux
Subsystem:
s390 architecture, the rest · Maintainers:
Heiko Carstens, Vasily Gorbik, Alexander Gordeev, Linus Torvalds
onstr@monstr.eu>, Thomas Bogendoerfer [off-list ref], linux-parisc@vger.kernel.org, Max Filippov [off-list ref], linux-kernel@vger.kernel.org, Johannes Berg [off-list ref], Dinh Nguyen [off-list ref], linux-riscv@lists.infradead.org, Palmer Dabbelt [off-list ref], Sven Schnelle [off-list ref], linux-alpha@vger.kernel.org, Ivan Kokshaysky [off-list ref], Andrew Morton [off-list ref], linuxppc-dev@lists.ozlabs.org, "David S . Miller" [off-list ref] Errors-To: linuxppc-dev-bounces+linuxppc-dev=archiver.kernel.org@lists.ozlabs.org Sender: "Linuxppc-dev" [off-list ref] On Fri, May 27, 2022 at 03:39:36PM -0400, Peter Xu wrote:
quoted hunk ↗ jump to hunk
diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index e173b6187ad5..4608cc962ecf 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c@@ -433,6 +433,17 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access) goto out_up; goto out; } + + /* The fault is fully completed (including releasing mmap lock) */ + if (fault & VM_FAULT_COMPLETED) { + /* + * Gmap will need the mmap lock again, so retake it. TODO: + * only conditionally take the lock when CONFIG_PGSTE set. + */ + mmap_read_lock(mm); + goto out_gmap; + } + if (unlikely(fault & VM_FAULT_ERROR)) goto out_up;
Guess the patch below on top of your patch is what we want. Just for clarification: if gmap is not NULL then the process is a kvm process. So, depending on the workload, this optimization makes sense.
diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
index 4608cc962ecf..e1d40ca341b7 100644
--- a/arch/s390/mm/fault.c
+++ b/arch/s390/mm/fault.c@@ -436,12 +436,11 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access) /* The fault is fully completed (including releasing mmap lock) */ if (fault & VM_FAULT_COMPLETED) { - /* - * Gmap will need the mmap lock again, so retake it. TODO: - * only conditionally take the lock when CONFIG_PGSTE set. - */ - mmap_read_lock(mm); - goto out_gmap; + if (gmap) { + mmap_read_lock(mm); + goto out_gmap; + } + goto out; } if (unlikely(fault & VM_FAULT_ERROR))