--- /dev/null
+diff -Nurb linux-2.6.22-580/drivers/oprofile/cpu_buffer.c linux-2.6.22-590/drivers/oprofile/cpu_buffer.c
+--- linux-2.6.22-580/drivers/oprofile/cpu_buffer.c 2007-07-08 19:32:17.000000000 -0400
++++ linux-2.6.22-590/drivers/oprofile/cpu_buffer.c 2008-02-27 13:53:47.000000000 -0500
+@@ -21,6 +21,7 @@
+ #include <linux/oprofile.h>
+ #include <linux/vmalloc.h>
+ #include <linux/errno.h>
++#include <linux/arrays.h>
+
+ #include "event_buffer.h"
+ #include "cpu_buffer.h"
+@@ -143,6 +144,14 @@
+ b->head_pos = 0;
+ }
+
++struct event_spec {
++ unsigned int pc;
++ unsigned long dcookie;
++ unsigned count;
++};
++
++extern void (*rec_event)(void *,unsigned int);
++
+ static inline void
+ add_sample(struct oprofile_cpu_buffer * cpu_buf,
+ unsigned long pc, unsigned long event)
+@@ -151,6 +160,7 @@
+ entry->eip = pc;
+ entry->event = event;
+ increment_head(cpu_buf);
++
+ }
+
+ static inline void
+@@ -237,12 +247,66 @@
+ oprofile_end_trace(cpu_buf);
+ }
+
++static int proc_pid_cmdline(struct task_struct *task, char * buffer)
++{
++ int res = 0;
++ unsigned int len;
++ struct mm_struct *mm = get_task_mm(task);
++ if (!mm)
++ goto out;
++ if (!mm->arg_end)
++ goto out_mm; /* Shh! No looking before we're done */
++
++ len = mm->arg_end - mm->arg_start;
++
++ if (len > PAGE_SIZE)
++ len = PAGE_SIZE;
++
++ res = access_process_vm(task, mm->arg_start, buffer, len, 0);
++
++ // If the nul at the end of args has been overwritten, then
++ // assume application is using setproctitle(3).
++ if (res > 0 && buffer[res-1] != '\0' && len < PAGE_SIZE) {
++ len = strnlen(buffer, res);
++ if (len < res) {
++ res = len;
++ } else {
++ len = mm->env_end - mm->env_start;
++ if (len > PAGE_SIZE - res)
++ len = PAGE_SIZE - res;
++ res += access_process_vm(task, mm->env_start, buffer+res, len, 0);
++ res = strnlen(buffer, res);
++ }
++ }
++out_mm:
++ mmput(mm);
++out:
++ return res;
++}
++
++
++
+ void oprofile_add_sample(struct pt_regs * const regs, unsigned long event)
+ {
+ int is_kernel = !user_mode(regs);
+ unsigned long pc = profile_pc(regs);
++ int res=0;
+
++ if (rec_event) {
++ struct event esig;
++ struct event_spec espec;
++ /*res = proc_pid_cmdline(current, espec->appname);*/
++ esig.task = current;
++ espec.pc=pc;
++ espec.count=1;
++ esig.event_data=&espec;
++ esig.event_type=event; /* index in the event array currently set up */
++ /* make sure the counters are loaded in the order we want them to show up*/
++ (*rec_event)(&esig, 1);
++ }
++ else {
+ oprofile_add_ext_sample(pc, regs, event, is_kernel);
++ }
+ }
+
+ void oprofile_add_pc(unsigned long pc, int is_kernel, unsigned long event)
+diff -Nurb linux-2.6.22-580/drivers/oprofile/oprof.c linux-2.6.22-590/drivers/oprofile/oprof.c
+--- linux-2.6.22-580/drivers/oprofile/oprof.c 2007-07-08 19:32:17.000000000 -0400
++++ linux-2.6.22-590/drivers/oprofile/oprof.c 2008-02-27 13:48:29.000000000 -0500
+@@ -33,6 +33,8 @@
+ */
+ static int timer = 0;
+
++extern void *rec_event;
++
+ int oprofile_setup(void)
+ {
+ int err;
+@@ -53,7 +55,7 @@
+ * us missing task deaths and eventually oopsing
+ * when trying to process the event buffer.
+ */
+- if ((err = sync_start()))
++ if (err = sync_start())
+ goto out3;
+
+ is_setup = 1;
+diff -Nurb linux-2.6.22-580/fs/exec.c linux-2.6.22-590/fs/exec.c
+--- linux-2.6.22-580/fs/exec.c 2008-02-27 13:46:38.000000000 -0500
++++ linux-2.6.22-590/fs/exec.c 2008-02-27 13:49:58.000000000 -0500
+@@ -52,6 +52,7 @@
+ #include <linux/audit.h>
+ #include <linux/signalfd.h>
+ #include <linux/vs_memory.h>
++#include <linux/dcookies.h>
+
+ #include <asm/uaccess.h>
+ #include <asm/mmu_context.h>
+@@ -488,6 +489,10 @@
+
+ if (!err) {
+ struct inode *inode = nd.dentry->d_inode;
++ unsigned long cookie;
++ if (!nd.dentry->d_cookie)
++ get_dcookie(nd.dentry, nd.mnt, &cookie);
++
+ file = ERR_PTR(-EACCES);
+ if (!(nd.mnt->mnt_flags & MNT_NOEXEC) &&
+ S_ISREG(inode->i_mode)) {
+diff -Nurb linux-2.6.22-580/include/linux/arrays.h linux-2.6.22-590/include/linux/arrays.h
+--- linux-2.6.22-580/include/linux/arrays.h 1969-12-31 19:00:00.000000000 -0500
++++ linux-2.6.22-590/include/linux/arrays.h 2008-02-27 13:48:29.000000000 -0500
+@@ -0,0 +1,35 @@
++#ifndef __ARRAYS_H__
++#define __ARRAYS_H__
++#include <linux/list.h>
++
++#define SAMPLING_METHOD_DEFAULT 0
++#define SAMPLING_METHOD_LOG 1
++
++/* Every probe has an array handler */
++
++/* XXX - Optimize this structure */
++
++struct array_handler {
++ struct list_head link;
++ unsigned int (*hash_func)(void *);
++ unsigned int (*sampling_func)(void *,int,void *);
++ unsigned short size;
++ unsigned int threshold;
++ unsigned char **expcount;
++ unsigned int sampling_method;
++ unsigned int **arrays;
++ unsigned int arraysize;
++ unsigned int num_samples[2];
++ void **epoch_samples; /* size-sized lists of samples */
++ unsigned int (*serialize)(void *, void *);
++ unsigned char code[5];
++};
++
++struct event {
++ struct list_head link;
++ void *event_data;
++ unsigned int count;
++ unsigned int event_type;
++ struct task_struct *task;
++};
++#endif
+diff -Nurb linux-2.6.22-580/include/linux/sched.h linux-2.6.22-590/include/linux/sched.h
+--- linux-2.6.22-580/include/linux/sched.h 2008-02-27 13:46:40.000000000 -0500
++++ linux-2.6.22-590/include/linux/sched.h 2008-02-27 13:48:29.000000000 -0500
+@@ -849,7 +849,7 @@
+ unsigned int btrace_seq;
+ #endif
+ unsigned long sleep_avg;
+- unsigned long long timestamp, last_ran;
++ unsigned long long timestamp, last_ran, last_interrupted, last_ran_j;
+ unsigned long long sched_time; /* sched_clock time spent running */
+ enum sleep_type sleep_type;
+
+diff -Nurb linux-2.6.22-580/kernel/fork.c linux-2.6.22-590/kernel/fork.c
+--- linux-2.6.22-580/kernel/fork.c 2008-02-27 13:46:40.000000000 -0500
++++ linux-2.6.22-590/kernel/fork.c 2008-02-27 13:48:29.000000000 -0500
+@@ -197,6 +197,8 @@
+ tsk->btrace_seq = 0;
+ #endif
+ tsk->splice_pipe = NULL;
++ //tsk->cmdline[0]='\0';
++ tsk->last_interrupted = 0;
+ return tsk;
+ }
+
+diff -Nurb linux-2.6.22-580/kernel/sched.c linux-2.6.22-590/kernel/sched.c
+--- linux-2.6.22-580/kernel/sched.c 2008-02-27 13:46:40.000000000 -0500
++++ linux-2.6.22-590/kernel/sched.c 2008-02-27 14:08:26.000000000 -0500
+@@ -10,7 +10,7 @@
+ * 1998-11-19 Implemented schedule_timeout() and related stuff
+ * by Andrea Arcangeli
+ * 2002-01-04 New ultra-scalable O(1) scheduler by Ingo Molnar:
+- * hybrid priority-list and round-robin design with
++ * hybrid priority-list and round-robin deventn with
+ * an array-switch method of distributing timeslices
+ * and per-CPU runqueues. Cleanups and useful suggestions
+ * by Davide Libenzi, preemptible kernel bits by Robert Love.
+@@ -56,6 +56,7 @@
+
+ #include <asm/tlb.h>
+ #include <asm/unistd.h>
++#include <linux/arrays.h>
+ #include <linux/vs_sched.h>
+ #include <linux/vs_cvirt.h>
+
+@@ -3608,6 +3609,8 @@
+
+ #endif
+
++extern void (*rec_event)(void *,unsigned int);
++
+ static inline int interactive_sleep(enum sleep_type sleep_type)
+ {
+ return (sleep_type == SLEEP_INTERACTIVE ||
+@@ -3617,16 +3620,51 @@
+ /*
+ * schedule() is the main scheduler function.
+ */
++
++struct event_spec {
++ unsigned long pc;
++ unsigned long dcookie;
++ unsigned count;
++ unsigned char reason;
++};
++
++#define top_esp (THREAD_SIZE - sizeof(unsigned long))
++#define top_ebp (THREAD_SIZE - 2*sizeof(unsigned long))
++
++static inline unsigned long my_get_wchan(struct task_struct *p)
++{
++ unsigned long ebp, esp, eip;
++ unsigned long stack_page;
++ int count = 0;
++ stack_page = (unsigned long)task_stack_page(p);
++ esp = p->thread.esp;
++ if (!stack_page || esp < stack_page || esp > top_esp+stack_page)
++ return 0;
++ /* include/asm-i386/system.h:switch_to() pushes ebp last. */
++ ebp = *(unsigned long *) esp;
++ do {
++ if (ebp < stack_page || ebp > top_ebp+stack_page)
++ return 0;
++ eip = *(unsigned long *) (ebp+4);
++ if (!in_sched_functions(eip))
++ return eip;
++ ebp = *(unsigned long *) ebp;
++ } while (count++ < 16);
++ return 0;
++}
++/* CHOPSTIX */
++
+ asmlinkage void __sched schedule(void)
+ {
+ struct task_struct *prev, *next;
+ struct prio_array *array;
+ struct list_head *queue;
+ unsigned long long now;
+- unsigned long run_time;
++ unsigned long run_time, diff;
+ int cpu, idx, new_prio;
+ long *switch_count;
+ struct rq *rq;
++ int sampling_reason;
+
+ /*
+ * Test if we are atomic. Since do_exit() needs to call into
+@@ -3680,6 +3718,7 @@
+ switch_count = &prev->nivcsw;
+ if (prev->state && !(preempt_count() & PREEMPT_ACTIVE)) {
+ switch_count = &prev->nvcsw;
++
+ if (unlikely((prev->state & TASK_INTERRUPTIBLE) &&
+ unlikely(signal_pending(prev))))
+ prev->state = TASK_RUNNING;
+@@ -3689,6 +3728,9 @@
+ vx_uninterruptible_inc(prev);
+ }
+ deactivate_task(prev, rq);
++ if (prev->state & TASK_INTERRUPTIBLE) {
++ prev->last_interrupted=jiffies;
++ }
+ }
+ }
+
+@@ -3763,8 +3805,44 @@
+ prev->sleep_avg -= run_time;
+ if ((long)prev->sleep_avg <= 0)
+ prev->sleep_avg = 0;
++
+ prev->timestamp = prev->last_ran = now;
+
++ /* CHOPSTIX */
++
++ prev->last_ran_j = jiffies;
++ if (next->last_interrupted) {
++ diff = (jiffies-next->last_interrupted);
++ next->last_interrupted = 0;
++ sampling_reason = 0;
++ }
++ else {
++ diff = jiffies-next->last_ran_j;
++ sampling_reason = 1;
++ }
++
++ if (rec_event && (diff>HZ/5)) {
++ struct event event;
++ struct event_spec espec;
++ unsigned long eip;
++ unsigned int state = next->state;
++
++ espec.reason = sampling_reason;
++
++ next->state = 0;
++ eip = next->thread.esp;
++ next->state = state;
++
++ next->last_interrupted = 0;
++ event.event_data=&espec;
++ event.task=next;
++ espec.pc=eip;
++ event.event_type=2;
++ /* index in the event array currently set up */
++ /* make sure the counters are loaded in the order we want them to show up*/
++ (*rec_event)(&event, diff);
++ }
++
+ sched_info_switch(prev, next);
+ if (likely(prev != next)) {
+ next->timestamp = next->last_ran = now;
+@@ -7275,3 +7353,7 @@
+ }
+
+ #endif
++
++void (*rec_event)(void *,unsigned int);
++EXPORT_SYMBOL(rec_event);
++EXPORT_SYMBOL(in_sched_functions);
+diff -Nurb linux-2.6.22-580/kernel/sched.c.rej linux-2.6.22-590/kernel/sched.c.rej
+--- linux-2.6.22-580/kernel/sched.c.rej 1969-12-31 19:00:00.000000000 -0500
++++ linux-2.6.22-590/kernel/sched.c.rej 2008-02-27 13:48:29.000000000 -0500
+@@ -0,0 +1,18 @@
++***************
++*** 56,61 ****
++
++ #include <asm/tlb.h>
++ #include <asm/unistd.h>
++
++ /*
++ * Scheduler clock - returns current time in nanosec units.
++--- 56,64 ----
++
++ #include <asm/tlb.h>
++ #include <asm/unistd.h>
+++ #include <linux/arrays.h>
+++
+++
++
++ /*
++ * Scheduler clock - returns current time in nanosec units.
git://git.onelab.eu / linux-2.6.git / commitdiff