[PATCH v1 2/2] epoll: Use __user_write_access_begin() and unsafe_put_user() in epoll_put_uevent().

[Kuniyuki Iwashima]

epoll_put_uevent() calls __put_user() twice, which are inlined
to two calls of out-of-line functions, __put_user_nocheck_4()
and __put_user_nocheck_8().

Both functions wrap mov with a stac/clac pair, which is expensive
on an AMD EPYC 7B12 64-Core Processor platform.

  __put_user_nocheck_4  /proc/kcore [Percent: local period]
  Percent │
    89.91 │      stac
     0.19 │      mov  %eax,(%rcx)
     0.15 │      xor  %ecx,%ecx
     9.69 │      clac
     0.06 │    ← retq

This was remarkable while testing neper/tcp_rr with 1000 flows per
thread.

  Overhead  Shared O  Symbol
    10.08%  [kernel]  [k] _copy_to_iter
     7.12%  [kernel]  [k] ip6_output
     6.40%  [kernel]  [k] sock_poll
     5.71%  [kernel]  [k] move_addr_to_user
     4.39%  [kernel]  [k] __put_user_nocheck_4
     ...
     1.06%  [kernel]  [k] ep_try_send_events
     ...                  ^- epoll_put_uevent() was inlined
     0.78%  [kernel]  [k] __put_user_nocheck_8

Use __user_write_access_begin() and unsafe_put_user() in
epoll_put_uevent().

We see 1% improvement on tcp_rr throughput by just saving a single
stac/clac pair.

Another option would be to use can_do_masked_user_access()
and masked_user_access_begin(), but we saw ~5% regression with
unnecessary 3 operations for address masking, which is already
checked by ep_check_params().

Suggested-by: Eric Dumazet <edumazet at google.com>
Signed-off-by: Kuniyuki Iwashima <kuniyu at google.com>
---
 include/linux/eventpoll.h | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/include/linux/eventpoll.h b/include/linux/eventpoll.h
index ccb478eb174b..efc0aa2d496f 100644
--- a/include/linux/eventpoll.h
+++ b/include/linux/eventpoll.h
@@ -82,11 +82,14 @@ static inline struct epoll_event __user *
 epoll_put_uevent(__poll_t revents, __u64 data,
 		 struct epoll_event __user *uevent)
 {
-	if (__put_user(revents, &uevent->events) ||
-	    __put_user(data, &uevent->data))
-		return NULL;
-
-	return uevent+1;
+	__user_write_access_begin(uevent, sizeof(*uevent));
+	unsafe_put_user(revents, &uevent->events, efault);
+	unsafe_put_user(data, &uevent->data, efault);
+	user_access_end();
+	return uevent + 1;
+efault:
+	user_access_end();
+	return NULL;
 }
 #endif
 
-- 
2.51.1.814.gb8fa24458f-goog