зеркало из https://github.com/mozilla/pjs.git
850 строки
23 KiB
C++
850 строки
23 KiB
C++
/* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
|
|
/* ***** BEGIN LICENSE BLOCK *****
|
|
* Version: MPL 1.1/GPL 2.0/LGPL 2.1
|
|
*
|
|
* The contents of this file are subject to the Mozilla Public License Version
|
|
* 1.1 (the "License"); you may not use this file except in compliance with
|
|
* the License. You may obtain a copy of the License at
|
|
* http://www.mozilla.org/MPL/
|
|
*
|
|
* Software distributed under the License is distributed on an "AS IS" basis,
|
|
* WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
|
|
* for the specific language governing rights and limitations under the
|
|
* License.
|
|
*
|
|
* The Original Code is mozilla.org code.
|
|
*
|
|
* The Initial Developer of the Original Code is Netscape Communications Corp.
|
|
* Portions created by the Initial Developer are Copyright (C) 1998
|
|
* the Initial Developer. All Rights Reserved.
|
|
*
|
|
* Contributor(s):
|
|
* Randell Jesup (recent improvements, threads, etc)
|
|
*
|
|
* Alternatively, the contents of this file may be used under the terms of
|
|
* either the GNU General Public License Version 2 or later (the "GPL"), or
|
|
* the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
|
|
* in which case the provisions of the GPL or the LGPL are applicable instead
|
|
* of those above. If you wish to allow use of your version of this file only
|
|
* under the terms of either the GPL or the LGPL, and not to allow others to
|
|
* use your version of this file under the terms of the MPL, indicate your
|
|
* decision by deleting the provisions above and replace them with the notice
|
|
* and other provisions required by the GPL or the LGPL. If you do not delete
|
|
* the provisions above, a recipient may use your version of this file under
|
|
* the terms of any one of the MPL, the GPL or the LGPL.
|
|
*
|
|
* ***** END LICENSE BLOCK ***** */
|
|
|
|
#include "leaky.h"
|
|
#include "intcnt.h"
|
|
|
|
#include <sys/types.h>
|
|
#include <sys/mman.h>
|
|
#include <sys/stat.h>
|
|
#include <fcntl.h>
|
|
#include <unistd.h>
|
|
#include <string.h>
|
|
#ifndef NTO
|
|
#include <getopt.h>
|
|
#endif
|
|
#include <assert.h>
|
|
#include <stdlib.h>
|
|
#include <stdio.h>
|
|
|
|
#ifdef NTO
|
|
#include <mem.h>
|
|
#endif
|
|
|
|
#ifndef FALSE
|
|
#define FALSE 0
|
|
#endif
|
|
#ifndef TRUE
|
|
#define TRUE 1
|
|
#endif
|
|
|
|
static const u_int DefaultBuckets = 10007; // arbitrary, but prime
|
|
static const u_int MaxBuckets = 1000003; // arbitrary, but prime
|
|
|
|
//----------------------------------------------------------------------
|
|
|
|
int main(int argc, char** argv)
|
|
{
|
|
leaky* l = new leaky;
|
|
|
|
l->initialize(argc, argv);
|
|
l->outputfd = stdout;
|
|
|
|
for (int i = 0; i < l->numLogFiles; i++) {
|
|
if (l->output_dir || l->numLogFiles > 1) {
|
|
char name[2048]; // XXX fix
|
|
if (l->output_dir)
|
|
snprintf(name,sizeof(name),"%s/%s.html",l->output_dir,argv[l->logFileIndex + i]);
|
|
else
|
|
snprintf(name,sizeof(name),"%s.html",argv[l->logFileIndex + i]);
|
|
|
|
fprintf(stderr,"opening %s\n",name);
|
|
l->outputfd = fopen(name,"w");
|
|
// if an error we won't process the file
|
|
}
|
|
if (l->outputfd) { // paranoia
|
|
l->open(argv[l->logFileIndex + i]);
|
|
|
|
if (l->outputfd != stderr) {
|
|
fclose(l->outputfd);
|
|
l->outputfd = NULL;
|
|
}
|
|
}
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
char *
|
|
htmlify(const char *in)
|
|
{
|
|
const char *p = in;
|
|
char *out, *q;
|
|
int n = 0;
|
|
size_t newlen;
|
|
|
|
// Count the number of '<' and '>' in the input.
|
|
while ((p = strpbrk(p, "<>")))
|
|
{
|
|
++n;
|
|
++p;
|
|
}
|
|
|
|
// Knowing the number of '<' and '>', we can calculate the space
|
|
// needed for the output string.
|
|
newlen = strlen(in) + n * 3 + 1;
|
|
out = new char[newlen];
|
|
|
|
// Copy the input to the output, with substitutions.
|
|
p = in;
|
|
q = out;
|
|
do
|
|
{
|
|
if (*p == '<')
|
|
{
|
|
strcpy(q, "<");
|
|
q += 4;
|
|
}
|
|
else if (*p == '>')
|
|
{
|
|
strcpy(q, ">");
|
|
q += 4;
|
|
}
|
|
else
|
|
{
|
|
*q++ = *p;
|
|
}
|
|
p++;
|
|
} while (*p);
|
|
*q = '\0';
|
|
|
|
return out;
|
|
}
|
|
|
|
leaky::leaky()
|
|
{
|
|
applicationName = NULL;
|
|
progFile = NULL;
|
|
|
|
quiet = true;
|
|
showAddress = false;
|
|
showThreads = false;
|
|
stackDepth = 100000;
|
|
onlyThread = 0;
|
|
|
|
mappedLogFile = -1;
|
|
firstLogEntry = lastLogEntry = 0;
|
|
|
|
sfd = -1;
|
|
externalSymbols = 0;
|
|
usefulSymbols = 0;
|
|
numExternalSymbols = 0;
|
|
lowestSymbolAddr = 0;
|
|
highestSymbolAddr = 0;
|
|
|
|
loadMap = NULL;
|
|
|
|
collect_last = false;
|
|
collect_start = -1;
|
|
collect_end = -1;
|
|
}
|
|
|
|
leaky::~leaky()
|
|
{
|
|
}
|
|
|
|
void leaky::usageError()
|
|
{
|
|
fprintf(stderr, "Usage: %s [-v] [-t] [-e exclude] [-i include] [-s stackdepth] [--last] [--all] [--start n [--end m]] [--output-dir dir] prog log [log2 ...]\n", (char*) applicationName);
|
|
fprintf(stderr,
|
|
"\t-v: verbose\n"
|
|
"\t-t | --threads: split threads\n"
|
|
"\t--only-thread n: only profile thread N\n"
|
|
"\t-i include-id: stack must include specified id\n"
|
|
"\t-e exclude-id: stack must NOT include specified id\n"
|
|
"\t-s stackdepth: Limit depth looked at from captured stack frames\n"
|
|
"\t--last: only profile the last capture section\n"
|
|
"\t--start n [--end m]: profile n to m (or end) capture sections\n"
|
|
"\t--output-dir dir: write output files to dir\n"
|
|
"\tIf there's one log, output goes to stdout unless --output-dir is set\n"
|
|
"\tIf there are more than one log, output files will be named with .html added\n"
|
|
);
|
|
exit(-1);
|
|
}
|
|
|
|
static struct option longopts[] = {
|
|
{ "threads", 0, NULL, 't' },
|
|
{ "only-thread", 1, NULL, 'T' },
|
|
{ "last", 0, NULL, 'l' },
|
|
{ "start", 1, NULL, 'x' },
|
|
{ "end", 1, NULL, 'n' },
|
|
{ "output-dir", 1, NULL, 'd' },
|
|
{ NULL, 0, NULL, 0 },
|
|
};
|
|
|
|
void leaky::initialize(int argc, char** argv)
|
|
{
|
|
applicationName = argv[0];
|
|
applicationName = strrchr(applicationName, '/');
|
|
if (!applicationName) {
|
|
applicationName = argv[0];
|
|
} else {
|
|
applicationName++;
|
|
}
|
|
|
|
int arg;
|
|
int errflg = 0;
|
|
int longindex = 0;
|
|
|
|
onlyThread = 0;
|
|
output_dir = NULL;
|
|
|
|
// XXX tons of cruft here left over from tracemalloc
|
|
// XXX The -- options shouldn't need short versions, or they should be documented
|
|
while (((arg = getopt_long(argc, argv, "adEe:gh:i:r:Rs:tT:qvx:ln:",longopts,&longindex)) != -1)) {
|
|
switch (arg) {
|
|
case '?':
|
|
default:
|
|
fprintf(stderr,"error: unknown option %c\n",optopt);
|
|
errflg++;
|
|
break;
|
|
case 'a':
|
|
break;
|
|
case 'A': // not implemented
|
|
showAddress = true;
|
|
break;
|
|
case 'd':
|
|
output_dir = optarg; // reference to an argv pointer
|
|
break;
|
|
case 'R':
|
|
break;
|
|
case 'e':
|
|
exclusions.add(optarg);
|
|
break;
|
|
case 'g':
|
|
break;
|
|
case 'r': // not implemented
|
|
roots.add(optarg);
|
|
if (!includes.IsEmpty()) {
|
|
errflg++;
|
|
}
|
|
break;
|
|
case 'i':
|
|
includes.add(optarg);
|
|
if (!roots.IsEmpty()) {
|
|
errflg++;
|
|
}
|
|
break;
|
|
case 'h':
|
|
break;
|
|
case 's':
|
|
stackDepth = atoi(optarg);
|
|
if (stackDepth < 2) {
|
|
stackDepth = 2;
|
|
}
|
|
break;
|
|
case 'x':
|
|
// --start
|
|
collect_start = atoi(optarg);
|
|
break;
|
|
case 'n':
|
|
// --end
|
|
collect_end = atoi(optarg);
|
|
break;
|
|
case 'l':
|
|
// --last
|
|
collect_last = true;
|
|
break;
|
|
case 'q':
|
|
break;
|
|
case 'v':
|
|
quiet = !quiet;
|
|
break;
|
|
case 't':
|
|
showThreads = true;
|
|
break;
|
|
case 'T':
|
|
showThreads = true;
|
|
onlyThread = atoi(optarg);
|
|
break;
|
|
}
|
|
}
|
|
if (errflg || ((argc - optind) < 2)) {
|
|
usageError();
|
|
}
|
|
progFile = argv[optind++];
|
|
logFileIndex = optind;
|
|
numLogFiles = argc - optind;
|
|
if (!quiet)
|
|
fprintf(stderr,"numlogfiles = %d\n",numLogFiles);
|
|
}
|
|
|
|
static void* mapFile(int fd, u_int flags, off_t* sz)
|
|
{
|
|
struct stat sb;
|
|
if (fstat(fd, &sb) < 0) {
|
|
perror("fstat");
|
|
exit(-1);
|
|
}
|
|
void* base = mmap(0, (int)sb.st_size, flags, MAP_PRIVATE, fd, 0);
|
|
if (!base) {
|
|
perror("mmap");
|
|
exit(-1);
|
|
}
|
|
*sz = sb.st_size;
|
|
return base;
|
|
}
|
|
|
|
void leaky::LoadMap()
|
|
{
|
|
malloc_map_entry mme;
|
|
char name[1000];
|
|
|
|
if (!loadMap) {
|
|
// all files use the same map
|
|
int fd = ::open(M_MAPFILE, O_RDONLY);
|
|
if (fd < 0) {
|
|
perror("open: " M_MAPFILE);
|
|
exit(-1);
|
|
}
|
|
for (;;) {
|
|
int nb = read(fd, &mme, sizeof(mme));
|
|
if (nb != sizeof(mme)) break;
|
|
nb = read(fd, name, mme.nameLen);
|
|
if (nb != (int)mme.nameLen) break;
|
|
name[mme.nameLen] = 0;
|
|
if (!quiet) {
|
|
fprintf(stderr,"%s @ %lx\n", name, mme.address);
|
|
}
|
|
|
|
LoadMapEntry* lme = new LoadMapEntry;
|
|
lme->address = mme.address;
|
|
lme->name = strdup(name);
|
|
lme->next = loadMap;
|
|
loadMap = lme;
|
|
}
|
|
close(fd);
|
|
}
|
|
}
|
|
|
|
void leaky::open(char *logFile)
|
|
{
|
|
int threadArray[100]; // should auto-expand
|
|
int last_thread = -1;
|
|
int numThreads = 0;
|
|
int section = -1;
|
|
bool collecting = false;
|
|
|
|
LoadMap();
|
|
|
|
setupSymbols(progFile);
|
|
|
|
// open up the log file
|
|
if (mappedLogFile)
|
|
::close(mappedLogFile);
|
|
|
|
mappedLogFile = ::open(logFile, O_RDONLY);
|
|
if (mappedLogFile < 0) {
|
|
perror("open");
|
|
exit(-1);
|
|
}
|
|
off_t size;
|
|
firstLogEntry = (malloc_log_entry*) mapFile(mappedLogFile, PROT_READ, &size);
|
|
lastLogEntry = (malloc_log_entry*)((char*)firstLogEntry + size);
|
|
|
|
if (!collect_last || collect_start < 0) {
|
|
collecting = true;
|
|
}
|
|
|
|
// First, restrict it to the capture sections specified (all, last, start/end)
|
|
// This loop walks through all the call stacks we recorded
|
|
for (malloc_log_entry* lep=firstLogEntry;
|
|
lep < lastLogEntry;
|
|
lep = reinterpret_cast<malloc_log_entry*>(&lep->pcs[lep->numpcs])) {
|
|
|
|
if (lep->flags & JP_FIRST_AFTER_PAUSE) {
|
|
section++;
|
|
if (collect_last) {
|
|
firstLogEntry = lep;
|
|
numThreads = 0;
|
|
collecting = true;
|
|
}
|
|
if (collect_start == section) {
|
|
collecting = true;
|
|
firstLogEntry = lep;
|
|
}
|
|
if (collect_end == section) {
|
|
collecting = false;
|
|
lastLogEntry = lep;
|
|
}
|
|
if (!quiet)
|
|
fprintf(stderr,"New section %d: first=%p, last=%p, collecting=%d\n",
|
|
section,(void*)firstLogEntry,(void*)lastLogEntry,collecting);
|
|
}
|
|
|
|
// Capture thread info at the same time
|
|
|
|
// Find all the threads captured
|
|
|
|
// pthread/linux docs say the signal can be delivered to any thread in
|
|
// the process. In practice, it appears in Linux that it's always
|
|
// delivered to the thread that called setitimer(), and each thread can
|
|
// have a separate itimer. There's a support library for gprof that
|
|
// overlays pthread_create() to set timers in any threads you spawn.
|
|
if (showThreads && collecting) {
|
|
if (lep->thread != last_thread)
|
|
{
|
|
int i;
|
|
for (i=0; i<numThreads; i++)
|
|
{
|
|
if (lep->thread == threadArray[i])
|
|
break;
|
|
}
|
|
if (i == numThreads &&
|
|
i < (int) (sizeof(threadArray)/sizeof(threadArray[0])))
|
|
{
|
|
threadArray[i] = lep->thread;
|
|
numThreads++;
|
|
if (!quiet)
|
|
fprintf(stderr,"new thread %d\n",lep->thread);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
if (!quiet)
|
|
fprintf(stderr,"Done collecting: sections %d: first=%p, last=%p, numThreads=%d\n",
|
|
section,(void*)firstLogEntry,(void*)lastLogEntry,numThreads);
|
|
|
|
fprintf(outputfd,"<html><head><title>Jprof Profile Report</title></head><body>\n");
|
|
fprintf(outputfd,"<h1><center>Jprof Profile Report</center></h1>\n");
|
|
|
|
if (showThreads)
|
|
{
|
|
fprintf(stderr,"Num threads %d\n",numThreads);
|
|
|
|
fprintf(outputfd,"<hr>Threads:<p><pre>\n");
|
|
for (int i=0; i<numThreads; i++)
|
|
{
|
|
fprintf(outputfd," <a href=\"#thread_%d\">%d</a> ",
|
|
threadArray[i],threadArray[i]);
|
|
}
|
|
fprintf(outputfd,"</pre>");
|
|
|
|
for (int i=0; i<numThreads; i++)
|
|
{
|
|
if (!onlyThread || onlyThread == threadArray[i])
|
|
analyze(threadArray[i]);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
analyze(0);
|
|
}
|
|
|
|
fprintf(outputfd,"</pre></body></html>\n");
|
|
}
|
|
|
|
//----------------------------------------------------------------------
|
|
|
|
|
|
static int symbolOrder(void const* a, void const* b)
|
|
{
|
|
Symbol const* ap = (Symbol const *)a;
|
|
Symbol const* bp = (Symbol const *)b;
|
|
return ap->address == bp->address ? 0 :
|
|
(ap->address > bp->address ? 1 : -1);
|
|
}
|
|
|
|
void leaky::ReadSharedLibrarySymbols()
|
|
{
|
|
LoadMapEntry* lme = loadMap;
|
|
while (NULL != lme) {
|
|
ReadSymbols(lme->name, lme->address);
|
|
lme = lme->next;
|
|
}
|
|
}
|
|
|
|
void leaky::setupSymbols(const char *fileName)
|
|
{
|
|
if (usefulSymbols == 0) {
|
|
// only read once!
|
|
|
|
// Read in symbols from the program
|
|
ReadSymbols(fileName, 0);
|
|
|
|
// Read in symbols from the .so's
|
|
ReadSharedLibrarySymbols();
|
|
|
|
if (!quiet) {
|
|
fprintf(stderr,"A total of %d symbols were loaded\n", usefulSymbols);
|
|
}
|
|
|
|
// Now sort them
|
|
qsort(externalSymbols, usefulSymbols, sizeof(Symbol), symbolOrder);
|
|
lowestSymbolAddr = externalSymbols[0].address;
|
|
highestSymbolAddr = externalSymbols[usefulSymbols-1].address;
|
|
}
|
|
}
|
|
|
|
// Binary search the table, looking for a symbol that covers this
|
|
// address.
|
|
int leaky::findSymbolIndex(u_long addr)
|
|
{
|
|
u_int base = 0;
|
|
u_int limit = usefulSymbols - 1;
|
|
Symbol* end = &externalSymbols[limit];
|
|
while (base <= limit) {
|
|
u_int midPoint = (base + limit)>>1;
|
|
Symbol* sp = &externalSymbols[midPoint];
|
|
if (addr < sp->address) {
|
|
if (midPoint == 0) {
|
|
return -1;
|
|
}
|
|
limit = midPoint - 1;
|
|
} else {
|
|
if (sp+1 < end) {
|
|
if (addr < (sp+1)->address) {
|
|
return midPoint;
|
|
}
|
|
} else {
|
|
return midPoint;
|
|
}
|
|
base = midPoint + 1;
|
|
}
|
|
}
|
|
return -1;
|
|
}
|
|
|
|
Symbol* leaky::findSymbol(u_long addr)
|
|
{
|
|
int idx = findSymbolIndex(addr);
|
|
|
|
if(idx<0) {
|
|
return NULL;
|
|
} else {
|
|
return &externalSymbols[idx];
|
|
}
|
|
}
|
|
|
|
//----------------------------------------------------------------------
|
|
|
|
bool leaky::excluded(malloc_log_entry* lep)
|
|
{
|
|
if (exclusions.IsEmpty()) {
|
|
return false;
|
|
}
|
|
|
|
char** pcp = &lep->pcs[0];
|
|
u_int n = lep->numpcs;
|
|
for (u_int i = 0; i < n; i++, pcp++) {
|
|
Symbol* sp = findSymbol((u_long) *pcp);
|
|
if (sp && exclusions.contains(sp->name)) {
|
|
return true;
|
|
}
|
|
}
|
|
return false;
|
|
}
|
|
|
|
bool leaky::included(malloc_log_entry* lep)
|
|
{
|
|
if (includes.IsEmpty()) {
|
|
return true;
|
|
}
|
|
|
|
char** pcp = &lep->pcs[0];
|
|
u_int n = lep->numpcs;
|
|
for (u_int i = 0; i < n; i++, pcp++) {
|
|
Symbol* sp = findSymbol((u_long) *pcp);
|
|
if (sp && includes.contains(sp->name)) {
|
|
return true;
|
|
}
|
|
}
|
|
return false;
|
|
}
|
|
|
|
//----------------------------------------------------------------------
|
|
|
|
void leaky::displayStackTrace(FILE* out, malloc_log_entry* lep)
|
|
{
|
|
char** pcp = &lep->pcs[0];
|
|
u_int n = (lep->numpcs < stackDepth) ? lep->numpcs : stackDepth;
|
|
for (u_int i = 0; i < n; i++, pcp++) {
|
|
u_long addr = (u_long) *pcp;
|
|
Symbol* sp = findSymbol(addr);
|
|
if (sp) {
|
|
fputs(sp->name, out);
|
|
if (showAddress) {
|
|
fprintf(out, "[%p]", (char*)addr);
|
|
}
|
|
}
|
|
else {
|
|
fprintf(out, "<%p>", (char*)addr);
|
|
}
|
|
fputc(' ', out);
|
|
}
|
|
fputc('\n', out);
|
|
}
|
|
|
|
void leaky::dumpEntryToLog(malloc_log_entry* lep)
|
|
{
|
|
printf("%ld\t", lep->delTime);
|
|
printf(" --> ");
|
|
displayStackTrace(outputfd, lep);
|
|
}
|
|
|
|
void leaky::generateReportHTML(FILE *fp, int *countArray, int count, int thread)
|
|
{
|
|
fprintf(fp,"<center>");
|
|
if (showThreads)
|
|
{
|
|
fprintf(fp,"<hr><A NAME=thread_%d><b>Thread: %d</b></A><p>",
|
|
thread,thread);
|
|
}
|
|
fprintf(fp,"<A href=#flat_%d>flat</A><b> | </b><A href=#hier_%d>hierarchical</A>",
|
|
thread,thread);
|
|
fprintf(fp,"</center><P><P><P>\n");
|
|
|
|
int totalTimerHits = count;
|
|
int *rankingTable = new int[usefulSymbols];
|
|
|
|
for(int cnt=usefulSymbols; --cnt>=0; rankingTable[cnt]=cnt);
|
|
|
|
// Drat. I would use ::qsort() but I would need a global variable and my
|
|
// intro-pascal professor threatened to flunk anyone who used globals.
|
|
// She damaged me for life :-) (That was 1986. See how much influence
|
|
// she had. I don't remember her name but I always feel guilty about globals)
|
|
|
|
// Shell Sort. 581130733 is the max 31 bit value of h = 3h+1
|
|
int mx, i, h;
|
|
for(mx=usefulSymbols/9, h=581130733; h>0; h/=3) {
|
|
if(h<mx) {
|
|
for(i = h-1; i<usefulSymbols; i++) {
|
|
int j, tmp=rankingTable[i], val = countArray[tmp];
|
|
for(j = i; (j>=h) && (countArray[rankingTable[j-h]]<val); j-=h) {
|
|
rankingTable[j] = rankingTable[j-h];
|
|
}
|
|
rankingTable[j] = tmp;
|
|
}
|
|
}
|
|
}
|
|
|
|
// Ok, We are sorted now. Let's go through the table until we get to
|
|
// functions that were never called. Right now we don't do much inside
|
|
// this loop. Later we can get callers and callees into it like gprof
|
|
// does
|
|
fprintf(fp,
|
|
"<h2><A NAME=hier_%d></A><center><a href=\"http://lxr.mozilla.org/mozilla/source/tools/jprof/README.html#hier\">Hierarchical Profile</a></center></h2><hr>\n",
|
|
thread);
|
|
fprintf(fp, "<pre>\n");
|
|
fprintf(fp, "%6s %6s %4s %s\n",
|
|
"index", "Count", "Hits", "Function Name");
|
|
|
|
for(i=0; i<usefulSymbols && countArray[rankingTable[i]]>0; i++) {
|
|
Symbol *sp=&externalSymbols[rankingTable[i]];
|
|
|
|
sp->cntP.printReport(fp, this, rankingTable[i], totalTimerHits);
|
|
|
|
char *symname = htmlify(sp->name);
|
|
fprintf(fp, "%6d %6d (%3.1f%%)%s <a name=%d>%8d (%3.1f%%)</a>%s <b>%s</b>\n",
|
|
rankingTable[i],
|
|
sp->timerHit, (sp->timerHit*1000/totalTimerHits)/10.0,
|
|
(sp->timerHit*1000/totalTimerHits)/10.0 >= 10.0 ? "" : " ",
|
|
rankingTable[i], countArray[rankingTable[i]],
|
|
(countArray[rankingTable[i]]*1000/totalTimerHits)/10.0,
|
|
(countArray[rankingTable[i]]*1000/totalTimerHits)/10.0 >= 10.0 ? "" : " ",
|
|
symname);
|
|
delete [] symname;
|
|
|
|
sp->cntC.printReport(fp, this, rankingTable[i], totalTimerHits);
|
|
|
|
fprintf(fp, "<hr>\n");
|
|
}
|
|
fprintf(fp,"</pre>\n");
|
|
|
|
// OK, Now we want to print the flat profile. To do this we resort on
|
|
// the hit count.
|
|
|
|
// Cut-N-Paste Shell sort from above. The Ranking Table has already been
|
|
// populated, so we do not have to reinitialize it.
|
|
for(mx=usefulSymbols/9, h=581130733; h>0; h/=3) {
|
|
if(h<mx) {
|
|
for(i = h-1; i<usefulSymbols; i++) {
|
|
int j, tmp=rankingTable[i], val = externalSymbols[tmp].timerHit;
|
|
for(j = i;
|
|
(j>=h) && (externalSymbols[rankingTable[j-h]].timerHit<val); j-=h) {
|
|
rankingTable[j] = rankingTable[j-h];
|
|
}
|
|
rankingTable[j] = tmp;
|
|
}
|
|
}
|
|
}
|
|
|
|
// Pre-count up total counter hits, to get a percentage.
|
|
// I wanted the total before walking the list, if this
|
|
// double-pass over externalSymbols gets slow we can
|
|
// do single-pass and print this out after the loop finishes.
|
|
totalTimerHits = 0;
|
|
for(i=0;
|
|
i<usefulSymbols && externalSymbols[rankingTable[i]].timerHit>0; i++) {
|
|
Symbol *sp=&externalSymbols[rankingTable[i]];
|
|
totalTimerHits += sp->timerHit;
|
|
}
|
|
if (totalTimerHits == 0)
|
|
totalTimerHits = 1;
|
|
|
|
if (totalTimerHits != count)
|
|
fprintf(stderr,"Hit count mismatch: count=%d; totalTimerHits=%d",
|
|
count,totalTimerHits);
|
|
|
|
fprintf(fp,"<h2><A NAME=flat_%d></A><center><a href=\"http://lxr.mozilla.org/mozilla/source/tools/jprof/README.html#flat\">Flat Profile</a></center></h2><br>\n",
|
|
thread);
|
|
fprintf(fp, "<pre>\n");
|
|
|
|
fprintf(fp, "Total hit count: %d\n", totalTimerHits);
|
|
fprintf(fp, "Count %%Total Function Name\n");
|
|
// Now loop for as long as we have timer hits
|
|
for(i=0;
|
|
i<usefulSymbols && externalSymbols[rankingTable[i]].timerHit>0; i++) {
|
|
|
|
Symbol *sp=&externalSymbols[rankingTable[i]];
|
|
|
|
char *symname = htmlify(sp->name);
|
|
fprintf(fp, "<a href=\"#%d\">%3d %-2.1f %s</a>\n",
|
|
rankingTable[i], sp->timerHit,
|
|
((float)sp->timerHit/(float)totalTimerHits)*100.0, symname);
|
|
delete [] symname;
|
|
}
|
|
}
|
|
|
|
void leaky::analyze(int thread)
|
|
{
|
|
int *countArray = new int[usefulSymbols];
|
|
int *flagArray = new int[usefulSymbols];
|
|
|
|
//Zero our function call counter
|
|
memset(countArray, 0, sizeof(countArray[0])*usefulSymbols);
|
|
|
|
// reset hit counts
|
|
for(int i=0; i<usefulSymbols; i++) {
|
|
externalSymbols[i].timerHit = 0;
|
|
externalSymbols[i].regClear();
|
|
}
|
|
|
|
// The flag array is used to prevent counting symbols multiple times
|
|
// if functions are called recursively. In order to keep from having
|
|
// to zero it on each pass through the loop, we mark it with the value
|
|
// of stacks on each trip through the loop. This means we can determine
|
|
// if we have seen this symbol for this stack trace w/o having to reset
|
|
// from the prior stacktrace.
|
|
memset(flagArray, -1, sizeof(flagArray[0])*usefulSymbols);
|
|
|
|
// This loop walks through all the call stacks we recorded
|
|
// --last, --start and --end can restrict it, as can excludes/includes
|
|
stacks = 0;
|
|
for(malloc_log_entry* lep=firstLogEntry;
|
|
lep < lastLogEntry;
|
|
lep = reinterpret_cast<malloc_log_entry*>(&lep->pcs[lep->numpcs])) {
|
|
|
|
if ((thread != 0 && lep->thread != thread) ||
|
|
excluded(lep) || !included(lep))
|
|
{
|
|
continue;
|
|
}
|
|
|
|
++stacks; // How many stack frames did we collect
|
|
|
|
// This loop walks through every symbol in the call stack. By walking it
|
|
// backwards we know who called the function when we get there.
|
|
u_int n = (lep->numpcs < stackDepth) ? lep->numpcs : stackDepth;
|
|
char** pcp = &lep->pcs[n-1];
|
|
int idx=-1, parrentIdx=-1; // Init idx incase n==0
|
|
for (int i=n-1; i>=0; --i, --pcp) {
|
|
idx = findSymbolIndex(reinterpret_cast<u_long>(*pcp));
|
|
|
|
if(idx>=0) {
|
|
// Skip over bogus __restore_rt frames that realtime profiling
|
|
// can introduce.
|
|
if (i > 0 && !strcmp(externalSymbols[idx].name, "__restore_rt")) {
|
|
--pcp;
|
|
--i;
|
|
idx = findSymbolIndex(reinterpret_cast<u_long>(*pcp));
|
|
if (idx < 0) {
|
|
continue;
|
|
}
|
|
}
|
|
|
|
// If we have not seen this symbol before count it and mark it as seen
|
|
if(flagArray[idx]!=stacks && ((flagArray[idx]=stacks) || true)) {
|
|
++countArray[idx];
|
|
}
|
|
|
|
// We know who we are and we know who our parrent is. Count this
|
|
if(parrentIdx>=0) {
|
|
externalSymbols[parrentIdx].regChild(idx);
|
|
externalSymbols[idx].regParrent(parrentIdx);
|
|
}
|
|
// inside if() so an unknown in the middle of a stack won't break
|
|
// the link!
|
|
parrentIdx=idx;
|
|
}
|
|
}
|
|
|
|
// idx should be the function that we were in when we received the signal.
|
|
if(idx>=0) {
|
|
++externalSymbols[idx].timerHit;
|
|
}
|
|
}
|
|
|
|
generateReportHTML(outputfd, countArray, stacks, thread);
|
|
}
|
|
|
|
void FunctionCount::printReport(FILE *fp, leaky *lk, int parent, int total)
|
|
{
|
|
const char *fmt = " <A href=\"#%d\">%8d (%3.1f%%)%s %s</A>%s\n";
|
|
|
|
int nmax, tmax=((~0U)>>1);
|
|
|
|
do {
|
|
nmax=0;
|
|
for(int j=getSize(); --j>=0;) {
|
|
int cnt = getCount(j);
|
|
if(cnt==tmax) {
|
|
int idx = getIndex(j);
|
|
char *symname = htmlify(lk->indexToName(idx));
|
|
fprintf(fp, fmt, idx, getCount(j),
|
|
getCount(j)*100.0/total,
|
|
getCount(j)*100.0/total >= 10.0 ? "" : " ",
|
|
symname,
|
|
parent == idx ? " (self)" : "");
|
|
delete [] symname;
|
|
} else if(cnt<tmax && cnt>nmax) {
|
|
nmax=cnt;
|
|
}
|
|
}
|
|
} while((tmax=nmax)>0);
|
|
}
|