linux_old1/kernel/marker.c

859 lines
24 KiB
C
Raw Normal View History

/*
* Copyright (C) 2007 Mathieu Desnoyers
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*/
#include <linux/module.h>
#include <linux/mutex.h>
#include <linux/types.h>
#include <linux/jhash.h>
#include <linux/list.h>
#include <linux/rcupdate.h>
#include <linux/marker.h>
#include <linux/err.h>
extern struct marker __start___markers[];
extern struct marker __stop___markers[];
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/* Set to 1 to enable marker debug output */
const int marker_debug;
/*
* markers_mutex nests inside module_mutex. Markers mutex protects the builtin
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* and module markers and the hash table.
*/
static DEFINE_MUTEX(markers_mutex);
/*
* Marker hash table, containing the active markers.
* Protected by module_mutex.
*/
#define MARKER_HASH_BITS 6
#define MARKER_TABLE_SIZE (1 << MARKER_HASH_BITS)
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/*
* Note about RCU :
* It is used to make sure every handler has finished using its private data
* between two consecutive operation (add or remove) on a given marker. It is
* also used to delay the free of multiple probes array until a quiescent state
* is reached.
* marker entries modifications are protected by the markers_mutex.
*/
struct marker_entry {
struct hlist_node hlist;
char *format;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
void (*call)(const struct marker *mdata, /* Probe wrapper */
void *call_private, const char *fmt, ...);
struct marker_probe_closure single;
struct marker_probe_closure *multi;
int refcount; /* Number of times armed. 0 if disarmed. */
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
struct rcu_head rcu;
void *oldptr;
unsigned char rcu_pending:1;
unsigned char ptype:1;
char name[0]; /* Contains name'\0'format'\0' */
};
static struct hlist_head marker_table[MARKER_TABLE_SIZE];
/**
* __mark_empty_function - Empty probe callback
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* @probe_private: probe private data
* @call_private: call site private data
* @fmt: format string
* @...: variable argument list
*
* Empty callback provided as a probe to the markers. By providing this to a
* disabled marker, we make sure the execution flow is always valid even
* though the function pointer change and the marker enabling are two distinct
* operations that modifies the execution flow of preemptible code.
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
void __mark_empty_function(void *probe_private, void *call_private,
const char *fmt, va_list *args)
{
}
EXPORT_SYMBOL_GPL(__mark_empty_function);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/*
* marker_probe_cb Callback that prepares the variable argument list for probes.
* @mdata: pointer of type struct marker
* @call_private: caller site private data
* @fmt: format string
* @...: Variable argument list.
*
* Since we do not use "typical" pointer based RCU in the 1 argument case, we
* need to put a full smp_rmb() in this branch. This is why we do not use
* rcu_dereference() for the pointer read.
*/
void marker_probe_cb(const struct marker *mdata, void *call_private,
const char *fmt, ...)
{
va_list args;
char ptype;
/*
* preempt_disable does two things : disabling preemption to make sure
* the teardown of the callbacks can be done correctly when they are in
* modules and they insure RCU read coherency.
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
*/
preempt_disable();
ptype = mdata->ptype;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (likely(!ptype)) {
marker_probe_func *func;
/* Must read the ptype before ptr. They are not data dependant,
* so we put an explicit smp_rmb() here. */
smp_rmb();
func = mdata->single.func;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/* Must read the ptr before private data. They are not data
* dependant, so we put an explicit smp_rmb() here. */
smp_rmb();
va_start(args, fmt);
func(mdata->single.probe_private, call_private, fmt, &args);
va_end(args);
} else {
struct marker_probe_closure *multi;
int i;
/*
* multi points to an array, therefore accessing the array
* depends on reading multi. However, even in this case,
* we must insure that the pointer is read _before_ the array
* data. Same as rcu_dereference, but we need a full smp_rmb()
* in the fast path, so put the explicit barrier here.
*/
smp_read_barrier_depends();
multi = mdata->multi;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
for (i = 0; multi[i].func; i++) {
va_start(args, fmt);
multi[i].func(multi[i].probe_private, call_private, fmt,
&args);
va_end(args);
}
}
preempt_enable();
}
EXPORT_SYMBOL_GPL(marker_probe_cb);
/*
* marker_probe_cb Callback that does not prepare the variable argument list.
* @mdata: pointer of type struct marker
* @call_private: caller site private data
* @fmt: format string
* @...: Variable argument list.
*
* Should be connected to markers "MARK_NOARGS".
*/
void marker_probe_cb_noarg(const struct marker *mdata,
void *call_private, const char *fmt, ...)
{
va_list args; /* not initialized */
char ptype;
preempt_disable();
ptype = mdata->ptype;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (likely(!ptype)) {
marker_probe_func *func;
/* Must read the ptype before ptr. They are not data dependant,
* so we put an explicit smp_rmb() here. */
smp_rmb();
func = mdata->single.func;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/* Must read the ptr before private data. They are not data
* dependant, so we put an explicit smp_rmb() here. */
smp_rmb();
func(mdata->single.probe_private, call_private, fmt, &args);
} else {
struct marker_probe_closure *multi;
int i;
/*
* multi points to an array, therefore accessing the array
* depends on reading multi. However, even in this case,
* we must insure that the pointer is read _before_ the array
* data. Same as rcu_dereference, but we need a full smp_rmb()
* in the fast path, so put the explicit barrier here.
*/
smp_read_barrier_depends();
multi = mdata->multi;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
for (i = 0; multi[i].func; i++)
multi[i].func(multi[i].probe_private, call_private, fmt,
&args);
}
preempt_enable();
}
EXPORT_SYMBOL_GPL(marker_probe_cb_noarg);
static void free_old_closure(struct rcu_head *head)
{
struct marker_entry *entry = container_of(head,
struct marker_entry, rcu);
kfree(entry->oldptr);
/* Make sure we free the data before setting the pending flag to 0 */
smp_wmb();
entry->rcu_pending = 0;
}
static void debug_print_probes(struct marker_entry *entry)
{
int i;
if (!marker_debug)
return;
if (!entry->ptype) {
printk(KERN_DEBUG "Single probe : %p %p\n",
entry->single.func,
entry->single.probe_private);
} else {
for (i = 0; entry->multi[i].func; i++)
printk(KERN_DEBUG "Multi probe %d : %p %p\n", i,
entry->multi[i].func,
entry->multi[i].probe_private);
}
}
static struct marker_probe_closure *
marker_entry_add_probe(struct marker_entry *entry,
marker_probe_func *probe, void *probe_private)
{
int nr_probes = 0;
struct marker_probe_closure *old, *new;
WARN_ON(!probe);
debug_print_probes(entry);
old = entry->multi;
if (!entry->ptype) {
if (entry->single.func == probe &&
entry->single.probe_private == probe_private)
return ERR_PTR(-EBUSY);
if (entry->single.func == __mark_empty_function) {
/* 0 -> 1 probes */
entry->single.func = probe;
entry->single.probe_private = probe_private;
entry->refcount = 1;
entry->ptype = 0;
debug_print_probes(entry);
return NULL;
} else {
/* 1 -> 2 probes */
nr_probes = 1;
old = NULL;
}
} else {
/* (N -> N+1), (N != 0, 1) probes */
for (nr_probes = 0; old[nr_probes].func; nr_probes++)
if (old[nr_probes].func == probe
&& old[nr_probes].probe_private
== probe_private)
return ERR_PTR(-EBUSY);
}
/* + 2 : one for new probe, one for NULL func */
new = kzalloc((nr_probes + 2) * sizeof(struct marker_probe_closure),
GFP_KERNEL);
if (new == NULL)
return ERR_PTR(-ENOMEM);
if (!old)
new[0] = entry->single;
else
memcpy(new, old,
nr_probes * sizeof(struct marker_probe_closure));
new[nr_probes].func = probe;
new[nr_probes].probe_private = probe_private;
entry->refcount = nr_probes + 1;
entry->multi = new;
entry->ptype = 1;
debug_print_probes(entry);
return old;
}
static struct marker_probe_closure *
marker_entry_remove_probe(struct marker_entry *entry,
marker_probe_func *probe, void *probe_private)
{
int nr_probes = 0, nr_del = 0, i;
struct marker_probe_closure *old, *new;
old = entry->multi;
debug_print_probes(entry);
if (!entry->ptype) {
/* 0 -> N is an error */
WARN_ON(entry->single.func == __mark_empty_function);
/* 1 -> 0 probes */
WARN_ON(probe && entry->single.func != probe);
WARN_ON(entry->single.probe_private != probe_private);
entry->single.func = __mark_empty_function;
entry->refcount = 0;
entry->ptype = 0;
debug_print_probes(entry);
return NULL;
} else {
/* (N -> M), (N > 1, M >= 0) probes */
for (nr_probes = 0; old[nr_probes].func; nr_probes++) {
if ((!probe || old[nr_probes].func == probe)
&& old[nr_probes].probe_private
== probe_private)
nr_del++;
}
}
if (nr_probes - nr_del == 0) {
/* N -> 0, (N > 1) */
entry->single.func = __mark_empty_function;
entry->refcount = 0;
entry->ptype = 0;
} else if (nr_probes - nr_del == 1) {
/* N -> 1, (N > 1) */
for (i = 0; old[i].func; i++)
if ((probe && old[i].func != probe) ||
old[i].probe_private != probe_private)
entry->single = old[i];
entry->refcount = 1;
entry->ptype = 0;
} else {
int j = 0;
/* N -> M, (N > 1, M > 1) */
/* + 1 for NULL */
new = kzalloc((nr_probes - nr_del + 1)
* sizeof(struct marker_probe_closure), GFP_KERNEL);
if (new == NULL)
return ERR_PTR(-ENOMEM);
for (i = 0; old[i].func; i++)
if ((probe && old[i].func != probe) ||
old[i].probe_private != probe_private)
new[j++] = old[i];
entry->refcount = nr_probes - nr_del;
entry->ptype = 1;
entry->multi = new;
}
debug_print_probes(entry);
return old;
}
/*
* Get marker if the marker is present in the marker hash table.
* Must be called with markers_mutex held.
* Returns NULL if not present.
*/
static struct marker_entry *get_marker(const char *name)
{
struct hlist_head *head;
struct hlist_node *node;
struct marker_entry *e;
u32 hash = jhash(name, strlen(name), 0);
head = &marker_table[hash & ((1 << MARKER_HASH_BITS)-1)];
hlist_for_each_entry(e, node, head, hlist) {
if (!strcmp(name, e->name))
return e;
}
return NULL;
}
/*
* Add the marker to the marker hash table. Must be called with markers_mutex
* held.
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
static struct marker_entry *add_marker(const char *name, const char *format)
{
struct hlist_head *head;
struct hlist_node *node;
struct marker_entry *e;
size_t name_len = strlen(name) + 1;
size_t format_len = 0;
u32 hash = jhash(name, name_len-1, 0);
if (format)
format_len = strlen(format) + 1;
head = &marker_table[hash & ((1 << MARKER_HASH_BITS)-1)];
hlist_for_each_entry(e, node, head, hlist) {
if (!strcmp(name, e->name)) {
printk(KERN_NOTICE
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
"Marker %s busy\n", name);
return ERR_PTR(-EBUSY); /* Already there */
}
}
/*
* Using kmalloc here to allocate a variable length element. Could
* cause some memory fragmentation if overused.
*/
e = kmalloc(sizeof(struct marker_entry) + name_len + format_len,
GFP_KERNEL);
if (!e)
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
return ERR_PTR(-ENOMEM);
memcpy(&e->name[0], name, name_len);
if (format) {
e->format = &e->name[name_len];
memcpy(e->format, format, format_len);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (strcmp(e->format, MARK_NOARGS) == 0)
e->call = marker_probe_cb_noarg;
else
e->call = marker_probe_cb;
trace_mark(core_marker_format, "name %s format %s",
e->name, e->format);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
} else {
e->format = NULL;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
e->call = marker_probe_cb;
}
e->single.func = __mark_empty_function;
e->single.probe_private = NULL;
e->multi = NULL;
e->ptype = 0;
e->refcount = 0;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
e->rcu_pending = 0;
hlist_add_head(&e->hlist, head);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
return e;
}
/*
* Remove the marker from the marker hash table. Must be called with mutex_lock
* held.
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
static int remove_marker(const char *name)
{
struct hlist_head *head;
struct hlist_node *node;
struct marker_entry *e;
int found = 0;
size_t len = strlen(name) + 1;
u32 hash = jhash(name, len-1, 0);
head = &marker_table[hash & ((1 << MARKER_HASH_BITS)-1)];
hlist_for_each_entry(e, node, head, hlist) {
if (!strcmp(name, e->name)) {
found = 1;
break;
}
}
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (!found)
return -ENOENT;
if (e->single.func != __mark_empty_function)
return -EBUSY;
hlist_del(&e->hlist);
/* Make sure the call_rcu has been executed */
if (e->rcu_pending)
rcu_barrier();
kfree(e);
return 0;
}
/*
* Set the mark_entry format to the format found in the element.
*/
static int marker_set_format(struct marker_entry **entry, const char *format)
{
struct marker_entry *e;
size_t name_len = strlen((*entry)->name) + 1;
size_t format_len = strlen(format) + 1;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
e = kmalloc(sizeof(struct marker_entry) + name_len + format_len,
GFP_KERNEL);
if (!e)
return -ENOMEM;
memcpy(&e->name[0], (*entry)->name, name_len);
e->format = &e->name[name_len];
memcpy(e->format, format, format_len);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (strcmp(e->format, MARK_NOARGS) == 0)
e->call = marker_probe_cb_noarg;
else
e->call = marker_probe_cb;
e->single = (*entry)->single;
e->multi = (*entry)->multi;
e->ptype = (*entry)->ptype;
e->refcount = (*entry)->refcount;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
e->rcu_pending = 0;
hlist_add_before(&e->hlist, &(*entry)->hlist);
hlist_del(&(*entry)->hlist);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/* Make sure the call_rcu has been executed */
if ((*entry)->rcu_pending)
rcu_barrier();
kfree(*entry);
*entry = e;
trace_mark(core_marker_format, "name %s format %s",
e->name, e->format);
return 0;
}
/*
* Sets the probe callback corresponding to one marker.
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
static int set_marker(struct marker_entry **entry, struct marker *elem,
int active)
{
int ret;
WARN_ON(strcmp((*entry)->name, elem->name) != 0);
if ((*entry)->format) {
if (strcmp((*entry)->format, elem->format) != 0) {
printk(KERN_NOTICE
"Format mismatch for probe %s "
"(%s), marker (%s)\n",
(*entry)->name,
(*entry)->format,
elem->format);
return -EPERM;
}
} else {
ret = marker_set_format(entry, elem->format);
if (ret)
return ret;
}
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/*
* probe_cb setup (statically known) is done here. It is
* asynchronous with the rest of execution, therefore we only
* pass from a "safe" callback (with argument) to an "unsafe"
* callback (does not set arguments).
*/
elem->call = (*entry)->call;
/*
* Sanity check :
* We only update the single probe private data when the ptr is
* set to a _non_ single probe! (0 -> 1 and N -> 1, N != 1)
*/
WARN_ON(elem->single.func != __mark_empty_function
&& elem->single.probe_private
!= (*entry)->single.probe_private &&
!elem->ptype);
elem->single.probe_private = (*entry)->single.probe_private;
/*
* Make sure the private data is valid when we update the
* single probe ptr.
*/
smp_wmb();
elem->single.func = (*entry)->single.func;
/*
* We also make sure that the new probe callbacks array is consistent
* before setting a pointer to it.
*/
rcu_assign_pointer(elem->multi, (*entry)->multi);
/*
* Update the function or multi probe array pointer before setting the
* ptype.
*/
smp_wmb();
elem->ptype = (*entry)->ptype;
elem->state = active;
return 0;
}
/*
* Disable a marker and its probe callback.
* Note: only waiting an RCU period after setting elem->call to the empty
* function insures that the original callback is not used anymore. This insured
* by preempt_disable around the call site.
*/
static void disable_marker(struct marker *elem)
{
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/* leave "call" as is. It is known statically. */
elem->state = 0;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
elem->single.func = __mark_empty_function;
/* Update the function before setting the ptype */
smp_wmb();
elem->ptype = 0; /* single probe */
/*
* Leave the private data and id there, because removal is racy and
* should be done only after an RCU period. These are never used until
* the next initialization anyway.
*/
}
/**
* marker_update_probe_range - Update a probe range
* @begin: beginning of the range
* @end: end of the range
*
* Updates the probe callback corresponding to a range of markers.
*/
void marker_update_probe_range(struct marker *begin,
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
struct marker *end)
{
struct marker *iter;
struct marker_entry *mark_entry;
mutex_lock(&markers_mutex);
for (iter = begin; iter < end; iter++) {
mark_entry = get_marker(iter->name);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (mark_entry) {
set_marker(&mark_entry, iter,
!!mark_entry->refcount);
/*
* ignore error, continue
*/
} else {
disable_marker(iter);
}
}
mutex_unlock(&markers_mutex);
}
/*
* Update probes, removing the faulty probes.
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
*
* Internal callback only changed before the first probe is connected to it.
* Single probe private data can only be changed on 0 -> 1 and 2 -> 1
* transitions. All other transitions will leave the old private data valid.
* This makes the non-atomicity of the callback/private data updates valid.
*
* "special case" updates :
* 0 -> 1 callback
* 1 -> 0 callback
* 1 -> 2 callbacks
* 2 -> 1 callbacks
* Other updates all behave the same, just like the 2 -> 3 or 3 -> 2 updates.
* Site effect : marker_set_format may delete the marker entry (creating a
* replacement).
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
static void marker_update_probes(void)
{
/* Core kernel markers */
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
marker_update_probe_range(__start___markers, __stop___markers);
/* Markers in modules. */
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
module_update_markers();
}
/**
* marker_probe_register - Connect a probe to a marker
* @name: marker name
* @format: format string
* @probe: probe handler
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* @probe_private: probe private data
*
* private data must be a valid allocated memory address, or NULL.
* Returns 0 if ok, error value on error.
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* The probe address must at least be aligned on the architecture pointer size.
*/
int marker_probe_register(const char *name, const char *format,
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
marker_probe_func *probe, void *probe_private)
{
struct marker_entry *entry;
int ret = 0;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
struct marker_probe_closure *old;
mutex_lock(&markers_mutex);
entry = get_marker(name);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (!entry) {
entry = add_marker(name, format);
if (IS_ERR(entry)) {
ret = PTR_ERR(entry);
goto end;
}
}
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
/*
* If we detect that a call_rcu is pending for this marker,
* make sure it's executed now.
*/
if (entry->rcu_pending)
rcu_barrier();
old = marker_entry_add_probe(entry, probe, probe_private);
if (IS_ERR(old)) {
ret = PTR_ERR(old);
goto end;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
}
mutex_unlock(&markers_mutex);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
marker_update_probes(); /* may update entry */
mutex_lock(&markers_mutex);
entry = get_marker(name);
WARN_ON(!entry);
entry->oldptr = old;
entry->rcu_pending = 1;
/* write rcu_pending before calling the RCU callback */
smp_wmb();
markers: use synchronize_sched() Markers do not mix well with CONFIG_PREEMPT_RCU because it uses preempt_disable/enable() and not rcu_read_lock/unlock for minimal intrusiveness. We would need call_sched and sched_barrier primitives. Currently, the modification (connection and disconnection) of probes from markers requires changes to the data structure done in RCU-style : a new data structure is created, the pointer is changed atomically, a quiescent state is reached and then the old data structure is freed. The quiescent state is reached once all the currently running preempt_disable regions are done running. We use the call_rcu mechanism to execute kfree() after such quiescent state has been reached. However, the new CONFIG_PREEMPT_RCU version of call_rcu and rcu_barrier does not guarantee that all preempt_disable code regions have finished, hence the race. The "proper" way to do this is to use rcu_read_lock/unlock, but we don't want to use it to minimize intrusiveness on the traced system. (we do not want the marker code to call into much of the OS code, because it would quickly restrict what can and cannot be instrumented, such as the scheduler). The temporary fix, until we get call_rcu_sched and rcu_barrier_sched in mainline, is to use synchronize_sched before each call_rcu calls, so we wait for the quiescent state in the system call code path. It will slow down batch marker enable/disable, but will make sure the race is gone. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-03 04:04:51 +08:00
#ifdef CONFIG_PREEMPT_RCU
synchronize_sched(); /* Until we have the call_rcu_sched() */
#endif
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
call_rcu(&entry->rcu, free_old_closure);
end:
mutex_unlock(&markers_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(marker_probe_register);
/**
* marker_probe_unregister - Disconnect a probe from a marker
* @name: marker name
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* @probe: probe function pointer
* @probe_private: probe private data
*
* Returns the private data given to marker_probe_register, or an ERR_PTR().
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* We do not need to call a synchronize_sched to make sure the probes have
* finished running before doing a module unload, because the module unload
* itself uses stop_machine(), which insures that every preempt disabled section
* have finished.
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
int marker_probe_unregister(const char *name,
marker_probe_func *probe, void *probe_private)
{
struct marker_entry *entry;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
struct marker_probe_closure *old;
int ret = -ENOENT;
mutex_lock(&markers_mutex);
entry = get_marker(name);
if (!entry)
goto end;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (entry->rcu_pending)
rcu_barrier();
old = marker_entry_remove_probe(entry, probe, probe_private);
mutex_unlock(&markers_mutex);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
marker_update_probes(); /* may update entry */
mutex_lock(&markers_mutex);
entry = get_marker(name);
if (!entry)
goto end;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
entry->oldptr = old;
entry->rcu_pending = 1;
/* write rcu_pending before calling the RCU callback */
smp_wmb();
markers: use synchronize_sched() Markers do not mix well with CONFIG_PREEMPT_RCU because it uses preempt_disable/enable() and not rcu_read_lock/unlock for minimal intrusiveness. We would need call_sched and sched_barrier primitives. Currently, the modification (connection and disconnection) of probes from markers requires changes to the data structure done in RCU-style : a new data structure is created, the pointer is changed atomically, a quiescent state is reached and then the old data structure is freed. The quiescent state is reached once all the currently running preempt_disable regions are done running. We use the call_rcu mechanism to execute kfree() after such quiescent state has been reached. However, the new CONFIG_PREEMPT_RCU version of call_rcu and rcu_barrier does not guarantee that all preempt_disable code regions have finished, hence the race. The "proper" way to do this is to use rcu_read_lock/unlock, but we don't want to use it to minimize intrusiveness on the traced system. (we do not want the marker code to call into much of the OS code, because it would quickly restrict what can and cannot be instrumented, such as the scheduler). The temporary fix, until we get call_rcu_sched and rcu_barrier_sched in mainline, is to use synchronize_sched before each call_rcu calls, so we wait for the quiescent state in the system call code path. It will slow down batch marker enable/disable, but will make sure the race is gone. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-03 04:04:51 +08:00
#ifdef CONFIG_PREEMPT_RCU
synchronize_sched(); /* Until we have the call_rcu_sched() */
#endif
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
call_rcu(&entry->rcu, free_old_closure);
remove_marker(name); /* Ignore busy error message */
ret = 0;
end:
mutex_unlock(&markers_mutex);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
return ret;
}
EXPORT_SYMBOL_GPL(marker_probe_unregister);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
static struct marker_entry *
get_marker_from_private_data(marker_probe_func *probe, void *probe_private)
{
struct marker_entry *entry;
unsigned int i;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
struct hlist_head *head;
struct hlist_node *node;
for (i = 0; i < MARKER_TABLE_SIZE; i++) {
head = &marker_table[i];
hlist_for_each_entry(entry, node, head, hlist) {
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (!entry->ptype) {
if (entry->single.func == probe
&& entry->single.probe_private
== probe_private)
return entry;
} else {
struct marker_probe_closure *closure;
closure = entry->multi;
for (i = 0; closure[i].func; i++) {
if (closure[i].func == probe &&
closure[i].probe_private
== probe_private)
return entry;
}
}
}
}
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
return NULL;
}
/**
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* marker_probe_unregister_private_data - Disconnect a probe from a marker
* @probe: probe function
* @probe_private: probe private data
*
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* Unregister a probe by providing the registered private data.
* Only removes the first marker found in hash table.
* Return 0 on success or error value.
* We do not need to call a synchronize_sched to make sure the probes have
* finished running before doing a module unload, because the module unload
* itself uses stop_machine(), which insures that every preempt disabled section
* have finished.
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
int marker_probe_unregister_private_data(marker_probe_func *probe,
void *probe_private)
{
struct marker_entry *entry;
int ret = 0;
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
struct marker_probe_closure *old;
mutex_lock(&markers_mutex);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
entry = get_marker_from_private_data(probe, probe_private);
if (!entry) {
ret = -ENOENT;
goto end;
}
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (entry->rcu_pending)
rcu_barrier();
old = marker_entry_remove_probe(entry, NULL, probe_private);
mutex_unlock(&markers_mutex);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
marker_update_probes(); /* may update entry */
mutex_lock(&markers_mutex);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
entry = get_marker_from_private_data(probe, probe_private);
WARN_ON(!entry);
entry->oldptr = old;
entry->rcu_pending = 1;
/* write rcu_pending before calling the RCU callback */
smp_wmb();
markers: use synchronize_sched() Markers do not mix well with CONFIG_PREEMPT_RCU because it uses preempt_disable/enable() and not rcu_read_lock/unlock for minimal intrusiveness. We would need call_sched and sched_barrier primitives. Currently, the modification (connection and disconnection) of probes from markers requires changes to the data structure done in RCU-style : a new data structure is created, the pointer is changed atomically, a quiescent state is reached and then the old data structure is freed. The quiescent state is reached once all the currently running preempt_disable regions are done running. We use the call_rcu mechanism to execute kfree() after such quiescent state has been reached. However, the new CONFIG_PREEMPT_RCU version of call_rcu and rcu_barrier does not guarantee that all preempt_disable code regions have finished, hence the race. The "proper" way to do this is to use rcu_read_lock/unlock, but we don't want to use it to minimize intrusiveness on the traced system. (we do not want the marker code to call into much of the OS code, because it would quickly restrict what can and cannot be instrumented, such as the scheduler). The temporary fix, until we get call_rcu_sched and rcu_barrier_sched in mainline, is to use synchronize_sched before each call_rcu calls, so we wait for the quiescent state in the system call code path. It will slow down batch marker enable/disable, but will make sure the race is gone. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-03 04:04:51 +08:00
#ifdef CONFIG_PREEMPT_RCU
synchronize_sched(); /* Until we have the call_rcu_sched() */
#endif
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
call_rcu(&entry->rcu, free_old_closure);
remove_marker(entry->name); /* Ignore busy error message */
end:
mutex_unlock(&markers_mutex);
return ret;
}
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
EXPORT_SYMBOL_GPL(marker_probe_unregister_private_data);
/**
* marker_get_private_data - Get a marker's probe private data
* @name: marker name
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* @probe: probe to match
* @num: get the nth matching probe's private data
*
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
* Returns the nth private data pointer (starting from 0) matching, or an
* ERR_PTR.
* Returns the private data pointer, or an ERR_PTR.
* The private data pointer should _only_ be dereferenced if the caller is the
* owner of the data, or its content could vanish. This is mostly used to
* confirm that a caller is the owner of a registered probe.
*/
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
void *marker_get_private_data(const char *name, marker_probe_func *probe,
int num)
{
struct hlist_head *head;
struct hlist_node *node;
struct marker_entry *e;
size_t name_len = strlen(name) + 1;
u32 hash = jhash(name, name_len-1, 0);
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
int i;
head = &marker_table[hash & ((1 << MARKER_HASH_BITS)-1)];
hlist_for_each_entry(e, node, head, hlist) {
if (!strcmp(name, e->name)) {
Linux Kernel Markers: support multiple probes RCU style multiple probes support for the Linux Kernel Markers. Common case (one probe) is still fast and does not require dynamic allocation or a supplementary pointer dereference on the fast path. - Move preempt disable from the marker site to the callback. Since we now have an internal callback, move the preempt disable/enable to the callback instead of the marker site. Since the callback change is done asynchronously (passing from a handler that supports arguments to a handler that does not setup the arguments is no arguments are passed), we can safely update it even if it is outside the preempt disable section. - Move probe arm to probe connection. Now, a connected probe is automatically armed. Remove MARK_MAX_FORMAT_LEN, unused. This patch modifies the Linux Kernel Markers API : it removes the probe "arm/disarm" and changes the probe function prototype : it now expects a va_list * instead of a "...". If we want to have more than one probe connected to a marker at a given time (LTTng, or blktrace, ssytemtap) then we need this patch. Without it, connecting a second probe handler to a marker will fail. It allow us, for instance, to do interesting combinations : Do standard tracing with LTTng and, eventually, to compute statistics with SystemTAP, or to have a special trigger on an event that would call a systemtap script which would stop flight recorder tracing. Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> Cc: Christoph Hellwig <hch@infradead.org> Cc: Mike Mason <mmlnx@us.ibm.com> Cc: Dipankar Sarma <dipankar@in.ibm.com> Cc: David Smith <dsmith@redhat.com> Cc: "Paul E. McKenney" <paulmck@us.ibm.com> Cc: "Frank Ch. Eigler" <fche@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-14 07:03:37 +08:00
if (!e->ptype) {
if (num == 0 && e->single.func == probe)
return e->single.probe_private;
else
break;
} else {
struct marker_probe_closure *closure;
int match = 0;
closure = e->multi;
for (i = 0; closure[i].func; i++) {
if (closure[i].func != probe)
continue;
if (match++ == num)
return closure[i].probe_private;
}
}
}
}
return ERR_PTR(-ENOENT);
}
EXPORT_SYMBOL_GPL(marker_get_private_data);