New profile protocol

- when kernel profiles a process for the first time it saves an entry
  describing the process [endpoint|name]

- every profile sample is only [endpoint|pc]

- profile utility creates a table of endpoint <-> name relations and
  translates endpoints of samples into names and writing out the
  results to comply with the processing tools

- "task" endpoints like KERNEL are negative thus we must cast it to
  unsigned when hashing

kernel/proc.h

@@ -241,6 +241,7 @@ struct proc {
 					 * regs are significant (initialized)*/
 #define MF_SENDING_FROM_KERNEL	0x2000 /* message of this process is from kernel */
 #define MF_CONTEXT_SET	0x4000 /* don't touch context */
+#define MF_SPROF_SEEN	0x8000 /* profiling has seen this process */
 
 /* Magic process table addresses. */
 #define BEG_PROC_ADDR (&proc[0])

kernel/profile.c

@@ -60,6 +60,36 @@ PUBLIC void stop_profile_clock()
   rm_irq_handler(&profile_clock_hook);
 }
 
+PRIVATE sprof_save_sample(struct proc * p)
+{
+	struct sprof_sample s;
+
+	s.proc = p->p_endpoint;
+	s.pc = (void *) p->p_reg.pc;
+
+	/* Store sample (process name and program counter). */
+	data_copy(KERNEL, (vir_bytes) &s,
+			sprof_ep, sprof_data_addr_vir + sprof_info.mem_used,
+			sizeof(s));
+
+	sprof_info.mem_used += sizeof(s);
+}
+
+PRIVATE sprof_save_proc(struct proc * p)
+{
+	struct sprof_proc s;
+
+	s.proc = p->p_endpoint;
+	memcpy(s.name, p->p_name, P_NAME_LEN);
+
+	/* Store sample (process name and program counter). */
+	data_copy(KERNEL, (vir_bytes) &s,
+			sprof_ep, sprof_data_addr_vir + sprof_info.mem_used,
+			sizeof(s));
+
+	sprof_info.mem_used += sizeof(s);
+}
+
 /*===========================================================================*
  *			profile_clock_handler                           *
  *===========================================================================*/
@@ -79,29 +109,17 @@ PRIVATE int profile_clock_handler(irq_hook_t *hook)
 
   p = get_cpulocal_var(proc_ptr);
 
-  /* All is OK */
+  if (!(p->p_misc_flags & MF_SPROF_SEEN)) {
+	  p->p_misc_flags |= MF_SPROF_SEEN;
+	  sprof_save_proc(p);
+  }
 
-  /* Idle process? */
-  if (priv(p)->s_proc_nr == IDLE) {
-	sprof_info.idle_samples++;
-  } else
   /* Runnable system process? */
-  if (priv(p)->s_flags & SYS_PROC && proc_is_runnable(p)) {
-	/* Note: k_reenter is always 0 here. */
-
-	/* Store sample (process name and program counter). */
-	data_copy(KERNEL, (vir_bytes) p->p_name,
-		sprof_ep, sprof_data_addr_vir + sprof_info.mem_used,
-		strlen(p->p_name));
-
-	data_copy(KERNEL, (vir_bytes) &p->p_reg.pc, sprof_ep,
-		(vir_bytes) (sprof_data_addr_vir + sprof_info.mem_used +
-					sizeof(p->p_name)),
-		(vir_bytes) sizeof(p->p_reg.pc));
-
-	sprof_info.mem_used += sizeof(sprof_sample);
-
-	sprof_info.system_samples++;
+  if (p->p_endpoint == IDLE)
+	  sprof_info.idle_samples++;
+  else if (priv(p)->s_flags & SYS_PROC && proc_is_runnable(p)) {
+	  sprof_save_sample(p);
+	  sprof_info.system_samples++;
   } else {
 	/* User process. */
 	sprof_info.user_samples++;

kernel/system/do_sprofile.c

@@ -20,6 +20,14 @@
 /* user address to write info struct */
 PRIVATE vir_bytes sprof_info_addr_vir;
 
+PRIVATE clean_seen_flag(void)
+{
+	int i;
+
+	for (i = 0; i < NR_TASKS + NR_PROCS; i++)
+		proc[i].p_misc_flags &= ~MF_SPROF_SEEN;
+}
+
 /*===========================================================================*
  *				do_sprofile				     *
  *===========================================================================*/
@@ -62,6 +70,8 @@ PUBLIC int do_sprofile(struct proc * caller, message * m_ptr)
 	
 	sprofiling = 1;
 
+	clean_seen_flag();
+
   	return OK;
 
   case PROF_STOP:
@@ -82,6 +92,8 @@ PUBLIC int do_sprofile(struct proc * caller, message * m_ptr)
 	data_copy(KERNEL, (vir_bytes) &sprof_info,
 		sprof_ep, sprof_info_addr_vir, sizeof(sprof_info));
 
+	clean_seen_flag();
+
   	return OK;
 
   default: