The patch titled
Subject: ipc,msg: provide barrier pairings for lockless receive
has been removed from the -mm tree. Its filename was
ipcmsg-provide-barrier-pairings-for-lockless-receive.patch
This patch was dropped because an updated version will be merged
------------------------------------------------------
From: Davidlohr Bueso <dave@xxxxxxxxxxxx>
Subject: ipc,msg: provide barrier pairings for lockless receive
We currently use a full barrier on the sender side to to avoid receiver
tasks disappearing on us while still performing on the sender side wakeup.
We lack however, the proper CPU-CPU interactions pairing on the receiver
side which busy-waits for the message. Similarly, we do not need a full
smp_mb, and can relax the semantics for the writer and reader sides of the
message. This is safe as we are only ordering loads and stores to r_msg.
And in both smp_wmb and smp_rmb, there are no stores after the calls
_anyway_.
This obviously applies for pipelined_send and expunge_all, for EIRDM when
destroying a queue.
Signed-off-by: Davidlohr Bueso <dbueso@xxxxxxx>
Cc: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
ipc/msg.c | 31 ++++++++++++++++++++++---------
1 file changed, 22 insertions(+), 9 deletions(-)
diff -puN ipc/msg.c~ipcmsg-provide-barrier-pairings-for-lockless-receive ipc/msg.c
--- a/ipc/msg.c~ipcmsg-provide-barrier-pairings-for-lockless-receive
+++ a/ipc/msg.c
@@ -196,7 +196,7 @@ static void expunge_all(struct msg_queue
* or dealing with -EAGAIN cases. See lockless receive part 1
* and 2 in do_msgrcv().
*/
- smp_mb();
+ smp_wmb();
msr->r_msg = ERR_PTR(res);
}
}
@@ -580,7 +580,7 @@ static inline int pipelined_send(struct
/* initialize pipelined send ordering */
msr->r_msg = NULL;
wake_up_process(msr->r_tsk);
- smp_mb(); /* see barrier comment below */
+ smp_wmb(); /* see barrier comment below */
msr->r_msg = ERR_PTR(-E2BIG);
} else {
msr->r_msg = NULL;
@@ -589,11 +589,12 @@ static inline int pipelined_send(struct
wake_up_process(msr->r_tsk);
/*
* Ensure that the wakeup is visible before
- * setting r_msg, as the receiving end depends
- * on it. See lockless receive part 1 and 2 in
- * do_msgrcv().
+ * setting r_msg, as the receiving can otherwise
+ * exit - once r_msg is set, the receiver can
+ * continue. See lockless receive part 1 and 2
+ * in do_msgrcv().
*/
- smp_mb();
+ smp_wmb();
msr->r_msg = msg;
return 1;
@@ -934,10 +935,22 @@ long do_msgrcv(int msqid, void __user *b
* wake_up_process(). There is a race with exit(), see
* ipc/mqueue.c for the details.
*/
- msg = (struct msg_msg *)msr_d.r_msg;
- while (msg == NULL) {
- cpu_relax();
+ for (;;) {
+ /*
+ * Pairs with writer barrier in pipelined_send
+ * or expunge_all
+ */
+ smp_rmb();
msg = (struct msg_msg *)msr_d.r_msg;
+ if (msg)
+ break;
+
+ /*
+ * The cpu_relax() call is a compiler barrier
+ * which forces everything in this loop to be
+ * re-loaded.
+ */
+ cpu_relax();
}
/* Lockless receive, part 3:
_
Patches currently in -mm which might be from dave@xxxxxxxxxxxx are
mm-hugetlb-document-the-reserve-map-region-tracking-routines.patch
mm-hugetlb-compute-return-the-number-of-regions-added-by-region_add.patch
mm-hugetlb-handle-races-in-alloc_huge_page-and-hugetlb_reserve_pages.patch
mm-hugetlb-handle-races-in-alloc_huge_page-and-hugetlb_reserve_pages-v4.patch
linux-next.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
