The quilt patch titled
Subject: userfaultfd: early return in dup_userfaultfd()
has been removed from the -mm tree. Its filename was
userfaultfd-early-return-in-dup_userfaultfd.patch
This patch was dropped because it was merged into the mm-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
------------------------------------------------------
From: ZhangPeng <zhangpeng362@xxxxxxxxxx>
Subject: userfaultfd: early return in dup_userfaultfd()
Date: Wed, 27 Mar 2024 17:08:35 +0800
When vma->vm_userfaultfd_ctx.ctx is NULL, vma->vm_flags should have
cleared __VM_UFFD_FLAGS. Therefore, there is no need to down_write or
clear the flag, which will affect fork performance. Fix this by
returning early if octx is NULL in dup_userfaultfd().
By applying this patch we can get a 1.3% performance improvement for
lmbench fork_prot. Results are as follows:
base early return
Process fork+exit: 419.1106 413.4804
Link: https://lkml.kernel.org/r/20240327090835.3232629-1-zhangpeng362@xxxxxxxxxx
Signed-off-by: ZhangPeng <zhangpeng362@xxxxxxxxxx>
Reviewed-by: Peter Xu <peterx@xxxxxxxxxx>
Cc: Axel Rasmussen <axelrasmussen@xxxxxxxxxx>
Cc: David Hildenbrand <david@xxxxxxxxxx>
Cc: Kefeng Wang <wangkefeng.wang@xxxxxxxxxx>
Cc: Liam R. Howlett <Liam.Howlett@xxxxxxxxxx>
Cc: Lokesh Gidra <lokeshgidra@xxxxxxxxxx>
Cc: Nanyong Sun <sunnanyong@xxxxxxxxxx>
Cc: Suren Baghdasaryan <surenb@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/userfaultfd.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
--- a/fs/userfaultfd.c~userfaultfd-early-return-in-dup_userfaultfd
+++ a/fs/userfaultfd.c
@@ -657,7 +657,10 @@ int dup_userfaultfd(struct vm_area_struc
struct userfaultfd_fork_ctx *fctx;
octx = vma->vm_userfaultfd_ctx.ctx;
- if (!octx || !(octx->features & UFFD_FEATURE_EVENT_FORK)) {
+ if (!octx)
+ return 0;
+
+ if (!(octx->features & UFFD_FEATURE_EVENT_FORK)) {
vma_start_write(vma);
vma->vm_userfaultfd_ctx = NULL_VM_UFFD_CTX;
userfaultfd_set_vm_flags(vma, vma->vm_flags & ~__VM_UFFD_FLAGS);
_
Patches currently in -mm which might be from zhangpeng362@xxxxxxxxxx are
filemap-replace-pte_offset_map-with-pte_offset_map_nolock.patch
