* [PATCH] rxrpc: Call state should be read with READ_ONCE() under some circumstances
@ 2021-01-07 5:14 Baptiste Lepers
2021-01-07 11:05 ` David Howells
0 siblings, 1 reply; 3+ messages in thread
From: Baptiste Lepers @ 2021-01-07 5:14 UTC (permalink / raw)
Cc: davem, dhowells, kuba, linux-afs, netdev, Baptiste Lepers
The call state may be changed at any time by the data-ready routine in
response to received packets, so if the call state is to be read and acted
upon several times in a function, READ_ONCE() must be used unless the call
state lock is held.
Signed-off-by: Baptiste Lepers <baptiste.lepers@gmail.com>
---
net/rxrpc/input.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c
index 667c44aa5a63..dc201363f2c4 100644
--- a/net/rxrpc/input.c
+++ b/net/rxrpc/input.c
@@ -430,7 +430,7 @@ static void rxrpc_input_data(struct rxrpc_call *call, struct sk_buff *skb)
return;
}
- if (call->state == RXRPC_CALL_SERVER_RECV_REQUEST) {
+ if (state == RXRPC_CALL_SERVER_RECV_REQUEST) {
unsigned long timo = READ_ONCE(call->next_req_timo);
unsigned long now, expect_req_by;
--
2.17.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] rxrpc: Call state should be read with READ_ONCE() under some circumstances
2021-01-07 5:14 [PATCH] rxrpc: Call state should be read with READ_ONCE() under some circumstances Baptiste Lepers
@ 2021-01-07 11:05 ` David Howells
2021-01-07 23:23 ` Baptiste Lepers
0 siblings, 1 reply; 3+ messages in thread
From: David Howells @ 2021-01-07 11:05 UTC (permalink / raw)
To: Baptiste Lepers; +Cc: dhowells, davem, kuba, linux-afs, netdev
Baptiste Lepers <baptiste.lepers@gmail.com> wrote:
> The call state may be changed at any time by the data-ready routine in
> response to received packets, so if the call state is to be read and acted
> upon several times in a function, READ_ONCE() must be used unless the call
> state lock is held.
I'm going to add:
As it happens, we used READ_ONCE() to read the state a few lines above the
unmarked read in rxrpc_input_data(), so use that value rather than
re-reading it.
to the commit message, if that's okay by you.
David
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] rxrpc: Call state should be read with READ_ONCE() under some circumstances
2021-01-07 11:05 ` David Howells
@ 2021-01-07 23:23 ` Baptiste Lepers
0 siblings, 0 replies; 3+ messages in thread
From: Baptiste Lepers @ 2021-01-07 23:23 UTC (permalink / raw)
To: David Howells
Cc: David Miller, Jakub Kicinski, linux-afs, Network Development
Yes, that's fine with me. Thanks for the clarification of the commit message.
(Sorry for the double send, I forgot to enable plain text on my
previous answer.)
Baptiste.
On Thu, Jan 7, 2021 at 10:05 PM David Howells <dhowells@redhat.com> wrote:
>
> Baptiste Lepers <baptiste.lepers@gmail.com> wrote:
>
> > The call state may be changed at any time by the data-ready routine in
> > response to received packets, so if the call state is to be read and acted
> > upon several times in a function, READ_ONCE() must be used unless the call
> > state lock is held.
>
> I'm going to add:
>
> As it happens, we used READ_ONCE() to read the state a few lines above the
> unmarked read in rxrpc_input_data(), so use that value rather than
> re-reading it.
>
> to the commit message, if that's okay by you.
>
> David
>
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2021-01-07 23:24 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-01-07 5:14 [PATCH] rxrpc: Call state should be read with READ_ONCE() under some circumstances Baptiste Lepers
2021-01-07 11:05 ` David Howells
2021-01-07 23:23 ` Baptiste Lepers
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.