[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240327090835.3232629-1-zhangpeng362@huawei.com>
Date: Wed, 27 Mar 2024 17:08:35 +0800
From: Peng Zhang <zhangpeng362@...wei.com>
To: <linux-mm@...ck.org>, <linux-kernel@...r.kernel.org>
CC: <akpm@...ux-foundation.org>, <peterx@...hat.com>, <david@...hat.com>,
<axelrasmussen@...gle.com>, <surenb@...gle.com>, <lokeshgidra@...gle.com>,
<Liam.Howlett@...cle.com>, <wangkefeng.wang@...wei.com>,
<sunnanyong@...wei.com>
Subject: [PATCH] userfaultfd: early return in dup_userfaultfd()
From: ZhangPeng <zhangpeng362@...wei.com>
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
Signed-off-by: ZhangPeng <zhangpeng362@...wei.com>
---
fs/userfaultfd.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
index 60dcfafdc11a..3e6ddda6f159 100644
--- a/fs/userfaultfd.c
+++ b/fs/userfaultfd.c
@@ -657,7 +657,10 @@ int dup_userfaultfd(struct vm_area_struct *vma, struct list_head *fcs)
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);
--
2.25.1
Powered by blists - more mailing lists