* [PATCH] xen/time: use READ_ONCE
@ 2015-11-20 15:02 ` Stefano Stabellini
0 siblings, 0 replies; 4+ messages in thread
From: Stefano Stabellini @ 2015-11-20 15:02 UTC (permalink / raw)
To: xen-devel
Cc: linux-kernel, Stefano.Stabellini, Stefano Stabellini,
Mark Rutland, konrad.wilk, david.vrabel, boris.ostrovsky
Use READ_ONCE through the code, rather than explicit barriers.
Suggested-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
CC: Mark Rutland <mark.rutland@arm.com>
CC: konrad.wilk@oracle.com
CC: david.vrabel@citrix.com
CC: boris.ostrovsky@oracle.com
---
drivers/xen/time.c | 17 +++++++----------
1 file changed, 7 insertions(+), 10 deletions(-)
diff --git a/drivers/xen/time.c b/drivers/xen/time.c
index 433fe24..7107842 100644
--- a/drivers/xen/time.c
+++ b/drivers/xen/time.c
@@ -25,7 +25,7 @@ static u64 get64(const u64 *p)
if (BITS_PER_LONG < 64) {
u32 *p32 = (u32 *)p;
- u32 h, l;
+ u32 h, l, h2;
/*
* Read high then low, and then make sure high is
@@ -34,15 +34,14 @@ static u64 get64(const u64 *p)
* XXX some clean way to make this endian-proof?
*/
do {
- h = p32[1];
- barrier();
- l = p32[0];
- barrier();
- } while (p32[1] != h);
+ h = READ_ONCE(p32[1]);
+ l = READ_ONCE(p32[0]);
+ h2 = READ_ONCE(p32[1]);
+ } while(h2 != h);
ret = (((u64)h) << 32) | l;
} else
- ret = *p;
+ ret = READ_ONCE(*p);
return ret;
}
@@ -66,9 +65,7 @@ void xen_get_runstate_snapshot(struct vcpu_runstate_info *res)
*/
do {
state_time = get64(&state->state_entry_time);
- barrier();
- *res = *state;
- barrier();
+ *res = READ_ONCE(*state);
} while (get64(&state->state_entry_time) != state_time);
}
--
1.7.10.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH] xen/time: use READ_ONCE
@ 2015-11-20 15:02 ` Stefano Stabellini
0 siblings, 0 replies; 4+ messages in thread
From: Stefano Stabellini @ 2015-11-20 15:02 UTC (permalink / raw)
To: xen-devel
Cc: linux-kernel, Stefano.Stabellini, Stefano Stabellini,
Mark Rutland, konrad.wilk, david.vrabel, boris.ostrovsky
Use READ_ONCE through the code, rather than explicit barriers.
Suggested-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
CC: Mark Rutland <mark.rutland@arm.com>
CC: konrad.wilk@oracle.com
CC: david.vrabel@citrix.com
CC: boris.ostrovsky@oracle.com
---
drivers/xen/time.c | 17 +++++++----------
1 file changed, 7 insertions(+), 10 deletions(-)
diff --git a/drivers/xen/time.c b/drivers/xen/time.c
index 433fe24..7107842 100644
--- a/drivers/xen/time.c
+++ b/drivers/xen/time.c
@@ -25,7 +25,7 @@ static u64 get64(const u64 *p)
if (BITS_PER_LONG < 64) {
u32 *p32 = (u32 *)p;
- u32 h, l;
+ u32 h, l, h2;
/*
* Read high then low, and then make sure high is
@@ -34,15 +34,14 @@ static u64 get64(const u64 *p)
* XXX some clean way to make this endian-proof?
*/
do {
- h = p32[1];
- barrier();
- l = p32[0];
- barrier();
- } while (p32[1] != h);
+ h = READ_ONCE(p32[1]);
+ l = READ_ONCE(p32[0]);
+ h2 = READ_ONCE(p32[1]);
+ } while(h2 != h);
ret = (((u64)h) << 32) | l;
} else
- ret = *p;
+ ret = READ_ONCE(*p);
return ret;
}
@@ -66,9 +65,7 @@ void xen_get_runstate_snapshot(struct vcpu_runstate_info *res)
*/
do {
state_time = get64(&state->state_entry_time);
- barrier();
- *res = *state;
- barrier();
+ *res = READ_ONCE(*state);
} while (get64(&state->state_entry_time) != state_time);
}
--
1.7.10.4
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [Xen-devel] [PATCH] xen/time: use READ_ONCE
2015-11-20 15:02 ` Stefano Stabellini
@ 2015-11-26 17:45 ` David Vrabel
-1 siblings, 0 replies; 4+ messages in thread
From: David Vrabel @ 2015-11-26 17:45 UTC (permalink / raw)
To: Stefano Stabellini, xen-devel
Cc: Mark Rutland, linux-kernel, david.vrabel, boris.ostrovsky
On 20/11/15 15:02, Stefano Stabellini wrote:
> Use READ_ONCE through the code, rather than explicit barriers.
Applied to for-linus-4.5, thanks.
David
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] xen/time: use READ_ONCE
@ 2015-11-26 17:45 ` David Vrabel
0 siblings, 0 replies; 4+ messages in thread
From: David Vrabel @ 2015-11-26 17:45 UTC (permalink / raw)
To: Stefano Stabellini, xen-devel
Cc: Mark Rutland, boris.ostrovsky, linux-kernel, david.vrabel
On 20/11/15 15:02, Stefano Stabellini wrote:
> Use READ_ONCE through the code, rather than explicit barriers.
Applied to for-linus-4.5, thanks.
David
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2015-11-26 17:46 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-11-20 15:02 [PATCH] xen/time: use READ_ONCE Stefano Stabellini
2015-11-20 15:02 ` Stefano Stabellini
2015-11-26 17:45 ` [Xen-devel] " David Vrabel
2015-11-26 17:45 ` David Vrabel
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.