All of lore.kernel.org
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Dave Jones <davej@redhat.com>,
	Steven Rostedt <rostedt@goodmis.org>
Subject: [ 24/60] tracing: Use flag buffer_disabled for irqsoff tracer
Date: Mon, 12 Aug 2013 23:33:51 -0700	[thread overview]
Message-ID: <20130813063335.268798611@linuxfoundation.org> (raw)
In-Reply-To: <20130813063331.950321483@linuxfoundation.org>

3.10-stable review patch.  If anyone has any objections, please let me know.

------------------

From: "Steven Rostedt (Red Hat)" <rostedt@goodmis.org>

commit 10246fa35d4ffdfe472185d4cbf9c2dfd9a9f023 upstream.

If the ring buffer is disabled and the irqsoff tracer records a trace it
will clear out its buffer and lose the data it had previously recorded.

Currently there's a callback when writing to the tracing_of file, but if
tracing is disabled via the function tracer trigger, it will not inform
the irqsoff tracer to stop recording.

By using the "mirror" flag (buffer_disabled) in the trace_array, that keeps
track of the status of the trace_array's buffer, it gives the irqsoff
tracer a fast way to know if it should record a new trace or not.
The flag may be a little behind the real state of the buffer, but it
should not affect the trace too much. It's more important for the irqsoff
tracer to be fast.

Reported-by: Dave Jones <davej@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

---
 kernel/trace/trace.c         |  101 +++++++++++++++++++++++++++++--------------
 kernel/trace/trace_irqsoff.c |    4 -
 2 files changed, 72 insertions(+), 33 deletions(-)

--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -246,9 +246,24 @@ cycle_t ftrace_now(int cpu)
 	return ts;
 }
 
+/**
+ * tracing_is_enabled - Show if global_trace has been disabled
+ *
+ * Shows if the global trace has been enabled or not. It uses the
+ * mirror flag "buffer_disabled" to be used in fast paths such as for
+ * the irqsoff tracer. But it may be inaccurate due to races. If you
+ * need to know the accurate state, use tracing_is_on() which is a little
+ * slower, but accurate.
+ */
 int tracing_is_enabled(void)
 {
-	return tracing_is_on();
+	/*
+	 * For quick access (irqsoff uses this in fast path), just
+	 * return the mirror variable of the state of the ring buffer.
+	 * It's a little racy, but we don't really care.
+	 */
+	smp_rmb();
+	return !global_trace.buffer_disabled;
 }
 
 /*
@@ -361,6 +376,23 @@ unsigned long trace_flags = TRACE_ITER_P
 	TRACE_ITER_GRAPH_TIME | TRACE_ITER_RECORD_CMD | TRACE_ITER_OVERWRITE |
 	TRACE_ITER_IRQ_INFO | TRACE_ITER_MARKERS | TRACE_ITER_FUNCTION;
 
+void tracer_tracing_on(struct trace_array *tr)
+{
+	if (tr->trace_buffer.buffer)
+		ring_buffer_record_on(tr->trace_buffer.buffer);
+	/*
+	 * This flag is looked at when buffers haven't been allocated
+	 * yet, or by some tracers (like irqsoff), that just want to
+	 * know if the ring buffer has been disabled, but it can handle
+	 * races of where it gets disabled but we still do a record.
+	 * As the check is in the fast path of the tracers, it is more
+	 * important to be fast than accurate.
+	 */
+	tr->buffer_disabled = 0;
+	/* Make the flag seen by readers */
+	smp_wmb();
+}
+
 /**
  * tracing_on - enable tracing buffers
  *
@@ -369,15 +401,7 @@ unsigned long trace_flags = TRACE_ITER_P
  */
 void tracing_on(void)
 {
-	if (global_trace.trace_buffer.buffer)
-		ring_buffer_record_on(global_trace.trace_buffer.buffer);
-	/*
-	 * This flag is only looked at when buffers haven't been
-	 * allocated yet. We don't really care about the race
-	 * between setting this flag and actually turning
-	 * on the buffer.
-	 */
-	global_trace.buffer_disabled = 0;
+	tracer_tracing_on(&global_trace);
 }
 EXPORT_SYMBOL_GPL(tracing_on);
 
@@ -571,6 +595,23 @@ void tracing_snapshot_alloc(void)
 EXPORT_SYMBOL_GPL(tracing_snapshot_alloc);
 #endif /* CONFIG_TRACER_SNAPSHOT */
 
+void tracer_tracing_off(struct trace_array *tr)
+{
+	if (tr->trace_buffer.buffer)
+		ring_buffer_record_off(tr->trace_buffer.buffer);
+	/*
+	 * This flag is looked at when buffers haven't been allocated
+	 * yet, or by some tracers (like irqsoff), that just want to
+	 * know if the ring buffer has been disabled, but it can handle
+	 * races of where it gets disabled but we still do a record.
+	 * As the check is in the fast path of the tracers, it is more
+	 * important to be fast than accurate.
+	 */
+	tr->buffer_disabled = 1;
+	/* Make the flag seen by readers */
+	smp_wmb();
+}
+
 /**
  * tracing_off - turn off tracing buffers
  *
@@ -581,26 +622,29 @@ EXPORT_SYMBOL_GPL(tracing_snapshot_alloc
  */
 void tracing_off(void)
 {
-	if (global_trace.trace_buffer.buffer)
-		ring_buffer_record_off(global_trace.trace_buffer.buffer);
-	/*
-	 * This flag is only looked at when buffers haven't been
-	 * allocated yet. We don't really care about the race
-	 * between setting this flag and actually turning
-	 * on the buffer.
-	 */
-	global_trace.buffer_disabled = 1;
+	tracer_tracing_off(&global_trace);
 }
 EXPORT_SYMBOL_GPL(tracing_off);
 
 /**
+ * tracer_tracing_is_on - show real state of ring buffer enabled
+ * @tr : the trace array to know if ring buffer is enabled
+ *
+ * Shows real state of the ring buffer if it is enabled or not.
+ */
+int tracer_tracing_is_on(struct trace_array *tr)
+{
+	if (tr->trace_buffer.buffer)
+		return ring_buffer_record_is_on(tr->trace_buffer.buffer);
+	return !tr->buffer_disabled;
+}
+
+/**
  * tracing_is_on - show state of ring buffers enabled
  */
 int tracing_is_on(void)
 {
-	if (global_trace.trace_buffer.buffer)
-		return ring_buffer_record_is_on(global_trace.trace_buffer.buffer);
-	return !global_trace.buffer_disabled;
+	return tracer_tracing_is_on(&global_trace);
 }
 EXPORT_SYMBOL_GPL(tracing_is_on);
 
@@ -4060,7 +4104,7 @@ static int tracing_wait_pipe(struct file
 		 *
 		 * iter->pos will be 0 if we haven't read anything.
 		 */
-		if (!tracing_is_enabled() && iter->pos)
+		if (!tracing_is_on() && iter->pos)
 			break;
 	}
 
@@ -5772,15 +5816,10 @@ rb_simple_read(struct file *filp, char _
 	       size_t cnt, loff_t *ppos)
 {
 	struct trace_array *tr = filp->private_data;
-	struct ring_buffer *buffer = tr->trace_buffer.buffer;
 	char buf[64];
 	int r;
 
-	if (buffer)
-		r = ring_buffer_record_is_on(buffer);
-	else
-		r = 0;
-
+	r = tracer_tracing_is_on(tr);
 	r = sprintf(buf, "%d\n", r);
 
 	return simple_read_from_buffer(ubuf, cnt, ppos, buf, r);
@@ -5802,11 +5841,11 @@ rb_simple_write(struct file *filp, const
 	if (buffer) {
 		mutex_lock(&trace_types_lock);
 		if (val) {
-			ring_buffer_record_on(buffer);
+			tracer_tracing_on(tr);
 			if (tr->current_trace->start)
 				tr->current_trace->start(tr);
 		} else {
-			ring_buffer_record_off(buffer);
+			tracer_tracing_off(tr);
 			if (tr->current_trace->stop)
 				tr->current_trace->stop(tr);
 		}
--- a/kernel/trace/trace_irqsoff.c
+++ b/kernel/trace/trace_irqsoff.c
@@ -373,7 +373,7 @@ start_critical_timing(unsigned long ip,
 	struct trace_array_cpu *data;
 	unsigned long flags;
 
-	if (likely(!tracer_enabled))
+	if (!tracer_enabled || !tracing_is_enabled())
 		return;
 
 	cpu = raw_smp_processor_id();
@@ -416,7 +416,7 @@ stop_critical_timing(unsigned long ip, u
 	else
 		return;
 
-	if (!tracer_enabled)
+	if (!tracer_enabled || !tracing_is_enabled())
 		return;
 
 	data = per_cpu_ptr(tr->trace_buffer.data, cpu);



  parent reply	other threads:[~2013-08-13  6:54 UTC|newest]

Thread overview: 65+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-08-13  6:33 [ 00/60] 3.10.7-stable review Greg Kroah-Hartman
2013-08-13  6:33 ` [ 01/60] SCSI: Dont attempt to send extended INQUIRY command if skip_vpd_pages is set Greg Kroah-Hartman
2013-08-13  6:33 ` [ 02/60] SCSI: megaraid_sas: megaraid_sas driver init fails in kdump kernel Greg Kroah-Hartman
2013-08-13  6:33 ` [ 03/60] virtio-scsi: Fix virtqueue affinity setup Greg Kroah-Hartman
2013-08-13  6:33 ` [ 04/60] powerpc: On POWERNV enable PPC_DENORMALISATION by default Greg Kroah-Hartman
2013-08-13  6:33 ` [ 05/60] powerpc: Fix hypervisor facility unavaliable vector number Greg Kroah-Hartman
2013-08-13  6:33 ` [ 06/60] powerpc: Rework setting up H/FSCR bit definitions Greg Kroah-Hartman
2013-08-13  6:33 ` [ 07/60] powerpc: Fix context switch DSCR on POWER8 Greg Kroah-Hartman
2013-08-13  6:33 ` [ 08/60] powerpc: Save the TAR register earlier Greg Kroah-Hartman
2013-08-13  6:33 ` [ 09/60] powerpc/tm: Fix context switching TAR, PPR and DSCR SPRs Greg Kroah-Hartman
2013-08-13  6:33 ` [ 10/60] ext4: destroy ext4_es_cachep on module unload Greg Kroah-Hartman
2013-08-13  6:33 ` [ 11/60] ext4: make sure group number is bumped after a inode allocation race Greg Kroah-Hartman
2013-08-13  6:33 ` [ 12/60] ext4: fix retry handling in ext4_ext_truncate() Greg Kroah-Hartman
2013-08-13  6:33 ` [ 13/60] regmap: cache: Make sure to sync the last register in a block Greg Kroah-Hartman
2013-08-13  6:33 ` [ 14/60] regmap: Add missing header for !CONFIG_REGMAP stubs Greg Kroah-Hartman
2013-08-13  6:33 ` [ 15/60] hwmon: (adt7470) Fix incorrect return code check Greg Kroah-Hartman
2013-08-13  6:33 ` [ 16/60] staging: zcache: fix "zcache=" kernel parameter Greg Kroah-Hartman
2013-08-13  6:33 ` [ 17/60] media: em28xx: fix assignment of the eeprom data Greg Kroah-Hartman
2013-08-13  6:33 ` [ 18/60] i2c: i2c-mxs: Use DMA mode even for small transfers Greg Kroah-Hartman
2013-08-13  6:33 ` [ 19/60] cpufreq: loongson2: fix regression related to clock management Greg Kroah-Hartman
2013-08-13  6:33 ` [ 20/60] cpufreq: rename ignore_nice as ignore_nice_load Greg Kroah-Hartman
2013-08-13  6:33 ` [ 21/60] ACPI / PM: Walk physical_node_list under physical_node_lock Greg Kroah-Hartman
2013-08-13  6:33 ` [ 22/60] tracing: Fix fields of struct trace_iterator that are zeroed by mistake Greg Kroah-Hartman
2013-08-13  6:33 ` [ 23/60] tracing: Make TRACE_ITER_STOP_ON_FREE stop the correct buffer Greg Kroah-Hartman
2013-08-13  6:33 ` Greg Kroah-Hartman [this message]
2013-08-13  6:33 ` [ 25/60] tracing: Fix reset of time stamps during trace_clock changes Greg Kroah-Hartman
2013-08-13  6:33 ` [ 26/60] ALSA: usb-audio: do not trust too-big wMaxPacketSize values Greg Kroah-Hartman
2013-08-13  6:33 ` [ 27/60] ALSA: 6fire: fix DMA issues with URB transfer_buffer usage Greg Kroah-Hartman
2013-08-13  6:33 ` [ 28/60] Btrfs: release both paths before logging dir/changed extents Greg Kroah-Hartman
2013-08-13  6:33 ` [ 29/60] LOCKD: Dont call utsname()->nodename from nlmclnt_setlockargs Greg Kroah-Hartman
2013-08-13  6:33 ` [ 30/60] SUNRPC: Dont auto-disconnect from the local rpcbind socket Greg Kroah-Hartman
2013-08-13  6:33 ` [ 31/60] SUNRPC: If the rpcbind channel is disconnected, fail the call to unregister Greg Kroah-Hartman
2013-08-13  6:33 ` [ 32/60] virtio/console: Quit from splice_write if pipe->nrbufs is 0 Greg Kroah-Hartman
2013-08-13  6:34 ` [ 33/60] virtio/console: Add pipe_lock/unlock for splice_write Greg Kroah-Hartman
2013-08-13  6:34 ` [ 34/60] virtio: console: fix race with port unplug and open/close Greg Kroah-Hartman
2013-08-13  6:34 ` [ 35/60] virtio: console: fix race in port_fops_open() and port unplug Greg Kroah-Hartman
2013-08-13  6:34 ` [ 36/60] virtio: console: clean up port data immediately at time of unplug Greg Kroah-Hartman
2013-08-13  6:34 ` [ 37/60] virtio: console: fix raising SIGIO after port unplug Greg Kroah-Hartman
2013-08-13  6:34 ` [ 38/60] virtio: console: return -ENODEV on all read operations after unplug Greg Kroah-Hartman
2013-08-13  6:34 ` [ 39/60] drm/cirrus: Invalidate page tables when pinning a BO Greg Kroah-Hartman
2013-08-13  6:34 ` [ 40/60] drm/mgag200: " Greg Kroah-Hartman
2013-08-13  6:34 ` [ 41/60] drm/ast: invalidate " Greg Kroah-Hartman
2013-08-13  6:34 ` [ 42/60] drm: Dont pass negative delta to ktime_sub_ns() Greg Kroah-Hartman
2013-08-13  6:34 ` [ 43/60] drm/radeon: select audio dto based on encoder id for DCE3 Greg Kroah-Hartman
2013-08-13  6:34 ` [ 44/60] drm/radeon: stop sending invalid UVD destroy msg Greg Kroah-Hartman
2013-08-13  6:34 ` [ 45/60] ext4: allow the mount options nodelalloc and data=journal Greg Kroah-Hartman
2013-08-13  6:34 ` [ 46/60] ext4: fix mount/remount error messages for incompatible mount options Greg Kroah-Hartman
2013-08-13  6:34 ` [ 47/60] ext4: flush the extent status cache during EXT4_IOC_SWAP_BOOT Greg Kroah-Hartman
2013-08-13  6:34 ` [ 48/60] cifs: extend the buffer length enought for sprintf() using Greg Kroah-Hartman
2013-08-13  6:34 ` [ 49/60] cifs: dont instantiate new dentries in readdir for inodes that need to be revalidated immediately Greg Kroah-Hartman
2013-08-13  6:34 ` [ 50/60] zram: allow request end to coincide with disksize Greg Kroah-Hartman
2013-08-13  6:34 ` [ 51/60] usb: core: dont try to reset_device() a port that got just disconnected Greg Kroah-Hartman
2013-08-13  6:34 ` [ 52/60] debugfs: debugfs_remove_recursive() must not rely on list_empty(d_subdirs) Greg Kroah-Hartman
2013-08-13  6:34 ` [ 53/60] reiserfs: fix deadlock in umount Greg Kroah-Hartman
2013-08-13  6:34 ` [ 54/60] drm/i915: initialize gt_lock early with other spin locks Greg Kroah-Hartman
2013-08-13  6:34 ` [ 55/60] drm/radeon: fix halting UVD Greg Kroah-Hartman
2013-08-13  6:34 ` [ 56/60] drm/radeon: only save UVD bo when we have open handles Greg Kroah-Hartman
2013-08-13  6:34 ` [ 57/60] drm/radeon: always program the MC on startup Greg Kroah-Hartman
2013-08-13  6:34 ` [ 58/60] SCSI: nsp32: use mdelay instead of large udelay constants Greg Kroah-Hartman
2013-08-13  6:34 ` [ 59/60] mtd: omap2: allow bulding as a module Greg Kroah-Hartman
2013-08-13  6:34 ` [ 60/60] MIPS: Expose missing pci_io{map,unmap} declarations Greg Kroah-Hartman
2013-08-13 11:57 ` [ 00/60] 3.10.7-stable review Guenter Roeck
2013-08-13 17:57   ` Greg Kroah-Hartman
2013-08-13 17:18 ` Shuah Khan
2013-08-13 17:57   ` Greg Kroah-Hartman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20130813063335.268798611@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=davej@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=stable@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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.