Techwizz/kernel-fxtec-pro1x
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

perf: Optimize ring-buffer write by depending on control dependencies

Browse source 
Remove a full barrier from the ring-buffer write path by relying on
a control dependency to order a LOAD -> STORE scenario.

Cc: "Paul E. McKenney" <paulmck@us.ibm.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/n/tip-8alv40z6ikk57jzbaobnxrjl@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This commit is contained in:
Peter Zijlstra 2013-11-25 11:49:10 +01:00 committed by Ingo Molnar
parent 813932149e
commit c7f2e3cd6c
1 changed files with 26 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

42
kernel/events/ring_buffer.c
View file

@ -61,19 +61,20 @@ static void perf_output_put_handle(struct perf_output_handle *handle)
*
* kernel user
*
* READ ->data_tail READ ->data_head
* smp_mb() (A) smp_rmb() (C)
* WRITE $data READ $data
* smp_wmb() (B) smp_mb() (D)
* STORE ->data_head WRITE ->data_tail
* if (LOAD ->data_tail) { LOAD ->data_head
* (A) smp_rmb() (C)
* STORE $data LOAD $data
* smp_wmb() (B) smp_mb() (D)
* STORE ->data_head STORE ->data_tail
* }
*
* Where A pairs with D, and B pairs with C.
*
* I don't think A needs to be a full barrier because we won't in fact
* write data until we see the store from userspace. So we simply don't
* issue the data WRITE until we observe it. Be conservative for now.
* In our case (A) is a control dependency that separates the load of
* the ->data_tail and the stores of $data. In case ->data_tail
* indicates there is no room in the buffer to store $data we do not.
*
* OTOH, D needs to be a full barrier since it separates the data READ
* D needs to be a full barrier since it separates the data READ
* from the tail WRITE.
*
* For B a WMB is sufficient since it separates two WRITEs, and for C
@ -81,7 +82,7 @@ static void perf_output_put_handle(struct perf_output_handle *handle)
*
* See perf_output_begin().
*/
smp_wmb();
smp_wmb(); /* B, matches C */
rb->user_page->data_head = head;
/*
@ -144,17 +145,26 @@ int perf_output_begin(struct perf_output_handle *handle,
if (!rb->overwrite &&
unlikely(CIRC_SPACE(head, tail, perf_data_size(rb)) < size))
goto fail;
/*
* The above forms a control dependency barrier separating the
* @tail load above from the data stores below. Since the @tail
* load is required to compute the branch to fail below.
*
* A, matches D; the full memory barrier userspace SHOULD issue
* after reading the data and before storing the new tail
* position.
*
* See perf_output_put_handle().
*/
head += size;
} while (local_cmpxchg(&rb->head, offset, head) != offset);
/*
* Separate the userpage->tail read from the data stores below.
* Matches the MB userspace SHOULD issue after reading the data
* and before storing the new tail position.
*
* See perf_output_put_handle().
* We rely on the implied barrier() by local_cmpxchg() to ensure
* none of the data stores below can be lifted up by the compiler.
*/
smp_mb();
if (unlikely(head - local_read(&rb->wakeup) > rb->watermark))
local_add(rb->watermark, &rb->wakeup);