lkml.org 
[lkml]   [2015]   [May]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
Subject[RFC PATCH] perf: enable fsync to flush buffered samples
Date
Instead of having a PERF_EVENT_IOC_FLUSH ioctl this instead allows
userspace to use fsync for flushing pmu samples, as suggested by Ingo
Molnar - thanks.

For reference I've also pushed a patch to my Mesa branch to test
this: https://github.com/rib/mesa wip/rib/oa-hsw-4.0.0

- Robert

--- >8 ---

To allow for pmus that may have internal buffering (e.g. the hardware
writes out data to a circular buffer which is only periodically
forwarded to userspace via perf) this enables userspace to explicitly
ensure it has received all samples before a point in time.

Signed-off-by: Robert Bragg <robert@sixbynine.org>
---
include/linux/perf_event.h | 7 +++++++
kernel/events/core.c | 23 +++++++++++++++++++++++
2 files changed, 30 insertions(+)

diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 04e98c8..d7fac05 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -305,6 +305,13 @@ struct pmu {
* Free pmu-private AUX data structures
*/
void (*free_aux) (void *aux); /* optional */
+
+ /*
+ * Flush buffered samples (E.g. for pmu hardware that writes samples to
+ * some intermediate buffer) userspace may need to explicitly ensure
+ * such samples have been forwarded to perf.
+ */
+ int (*flush) (struct perf_event *event); /*optional */
};

/**
diff --git a/kernel/events/core.c b/kernel/events/core.c
index 2ba89a1..a604e0c 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -4728,6 +4728,28 @@ static int perf_fasync(int fd, struct file *filp, int on)
return 0;
}

+static int perf_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
+{
+ struct perf_event *event = filp->private_data;
+ struct perf_event_context *ctx;
+ int ret;
+
+ /* We don't have a use for synchonizing a specific range, or datasync
+ * but lets not silently ignore them in case we think of uses later...
+ */
+ if (start != 0 || end != LLONG_MAX || datasync != 0)
+ return -EINVAL;
+
+ if (!event->pmu->flush)
+ return 0;
+
+ ctx = perf_event_ctx_lock(event);
+ ret = event->pmu->flush(event);
+ perf_event_ctx_unlock(event, ctx);
+
+ return ret;
+}
+
static const struct file_operations perf_fops = {
.llseek = no_llseek,
.release = perf_release,
@@ -4737,6 +4759,7 @@ static const struct file_operations perf_fops = {
.compat_ioctl = perf_compat_ioctl,
.mmap = perf_mmap,
.fasync = perf_fasync,
+ .fsync = perf_fsync,
};

/*
--
2.4.1


\
 
 \ /
  Last update: 2015-05-21 20:21    [W:0.051 / U:1.104 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site