aboutsummaryrefslogtreecommitdiff
path: root/nptl/perf.c
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>2002-11-26 22:50:54 +0000
committerUlrich Drepper <drepper@redhat.com>2002-11-26 22:50:54 +0000
commit76a50749f7af5935ba3739e815aa6a16ae4440d1 (patch)
treec16eac47f220f03fea74d80ef0a4e774809e94b7 /nptl/perf.c
parent6938e63f714b15c377d8cbf8e97b6f15b0e1b692 (diff)
downloadglibc-76a50749f7af5935ba3739e815aa6a16ae4440d1.zip
glibc-76a50749f7af5935ba3739e815aa6a16ae4440d1.tar.gz
glibc-76a50749f7af5935ba3739e815aa6a16ae4440d1.tar.bz2
Initial revisioncvs/initial
2002-11-26 Ulrich Drepper <drepper@redhat.com> * allocatestack.c (queue_stack): Don't remove stack from list here. Do it in the caller. Correct condition to prematurely terminate loop to free stacks. (__deallocate_stack): Remove stack from list here. 2002-11-26 Ulrich Drepper <drepper@redhat.com> * Makefile (tests): Add tst-stack1. * tst-stack1.c: New file. * allocatestack.c (allocate_stack): Initialize the TCB on a user provided stack. * pthread_attr_getstack.c: Return bottom of the thread area. 2002-11-25 Ulrich Drepper <drepper@redhat.com> * Makefile (libpthread-routines): Add pt-allocrtsig and pthread_kill_other_threads. * pt-allocrtsig.c: New file. * pthread_kill_other_threads.c: New file. * sysdeps/unix/sysv/linux/allocrtsig.c: Add additional aliases for all three functions. * sysdeps/unix/sysv/linux/Makefile (sysdep_routines): Remove allocrtsig. * sysdeps/unix/sysv/linux/Versions (libc:GLIBC_PRIVATE): Export __libc_current_sigrtmin_private, __libc_current_sigrtmax_private, and __libc_allocate_rtsig_private. * Versions (libpthread): Export pthread_kill_other_threads_np, __libc_current_sigrtmin, and __libc_current_sigrtmax. 2002-11-24 Ulrich Drepper <drepper@redhat.com> * allocatestack.c (allocate_stack): stackaddr in attribute points to the end of the stack. Adjust computations. When mprotect call fails dequeue stack and free it. * pthread_attr_setstack.c: Store top of the stack in stackaddr attribute. * pthread_getattr_np.c: Likewise. * descr.h (IS_DETACHED): Add some more parenthesis to prevent surprises. 2002-11-23 Ulrich Drepper <drepper@redhat.com> * sysdeps/pthread/pthread.h (pthread_self): __THROW must come before attribute definitions. Patch by Luca Barbieri <ldb@ldb.ods.org>. 2002-11-22 Ulrich Drepper <drepper@redhat.com> * pthread_getspecific.c: Optimize access to first 2nd-level array. * pthread_setspecific.c: Likewise. 2002-11-21 Ulrich Drepper <drepper@redhat.com> * sysdeps/unix/sysv/linux/i386/createthread.c: Remove CLONE_ flags definitions. Get them from the official place. * sysdeps/unix/sysv/linux/i386/fork.c: Likewise. * sysdeps/unix/sysv/linux/i386/createthread.c: Update CLONE_* flags. Use new CLONE_ flags in clone() calls. * sysdeps/unix/sysv/linux/fork.c: Use ARCH_FORK to actually fork. * sysdeps/unix/sysv/linux/i386/fork.c: New file. * Versions: Add pthread_* functions for libc. * forward.c: New file. * sysdeps/pthread/Makefile (libpthread-sysdeps_routines): Add errno-loc. * herrno.c: New file. * res.c: New file. * Makefile (libpthread-routines): Remove sem_post, sem_wait, sem_trywait, and sem_timedwait. Add herrno and res. * sem_init.c: Don't initialize lock and waiters members. * sem_open.c: Likewise. * sem_post.c: Removed. * sem_wait.c: Removed. * sem_trywait.c: Removed. * sem_timedwait.c: Removed. * sysdeps/unix/sysv/linux/i386/i486/lowlevelsem.S: Complete rewrite. Includes full implementations of sem_post, sem_wait, sem_trywait, and sem_timedwait. * sysdeps/unix/sysv/linux/i386/lowlevelsem.h (lll_sem_post): Adjust for new implementation. * sysdeps/unix/sysv/linux/internaltypes.h (struct sem): Remove lock and waiters fields. * tst-sem3.c: Improve error message. * tst-signal3.c: Likewise. * init.c (__pthread_initialize_minimal): Use set_tid_address syscall to tell the kernel about the termination futex and to initialize tid member. Don't initialize main_thread. * descr.h (struct pthread): Remove main_thread member. * cancelllation.c (__do_cancel): Remove code handling main thread. The main thread is not special anymore. * allocatestack.c (__reclaim_stacks): Mark stacks as unused. Add size of the stacks to stack_cache_actsize. * pt-readv.c: Add missing "defined". * pt-sigwait.c: Likewise. * pt-writev.c: Likewise. 2002-11-09 Ulrich Drepper <drepper@redhat.com> * Versions: Export __connect from libpthread. Patch by Luca Barbieri <ldb@ldb.ods.org>. * Makefile (libpthread-routines): Add pt-raise. * sysdeps/unix/sysv/linux/raise.c: New file. * sysdeps/unix/sysv/linux/pt-raise.c: New file. * sysdeps/generic/pt-raise.c: New file. * pthread_cond_init.c: Initialize all data elements of the condvar structure. Patch by Luca Barbieri <ldb@ldb.ods.org>. * pthread_attr_init.c: Actually implement 2.0 compatibility version. * pthread_create.c: Likewise. * Makefile (tests): Add tst-key1, tst-key2, tst-key3. * tst-key1.c: New file. * tst-key2.c: New file. * tst-key3.c: New file. * Versions: Export pthread_detach for version GLIBC_2.0. Reported by Saurabh Desai <sdesai@austin.ibm.com>. 2002-11-08 Ulrich Drepper <drepper@redhat.com> * pthread_key_create.c: Terminate search after an unused key was found. Patch by Luca Barbieri <ldb@ldb.ods.org>. * sysdeps/unix/sysv/linux/i386/pthread_once.S: Return zero. Patch by Luca Barbieri <ldb@ldb.ods.org>. 2002-10-10 Ulrich Drepper <drepper@redhat.com> * sysdeps/unix/sysv/linux/i386/i486/lowlevelsem.S: Use slow generic dynamic lookup for errno in PIC. * allocatestack.c (get_cached_stack): Rearrange code slightly to release the stack lock as soon as possible. Call _dl_allocate_tls_init for TCB from the cache to re-initialize the static TLS block. (allocate_stack): Call _dl_allocate_tls_init for user-provided stack. * cancellation.c: Renamed from cancelation.c. * Makefile: Adjust accordingly. * pthreadP.h (CANCELLATION_P): Renamed from CANCELATION_P. * cleanup_defer.c: Use CANCELLATION_P. * pthread_testcancel.c: Likewise. * descr.h: Fix spelling in comments. * init.c: Likewise. * pthread_getattr_np.c: Likewise. * pthread_getschedparam.c: Likewise. * pthread_setschedparam.c: Likewise. * Versions: Likewise. * pt-pselect.c: New file. * Makefile (libpthread-routines): Add pt-pselect. * Versions: Add pselect. * tst-cancel4.c: New file. * Makefile (tests): Add tst-cancel4. 2002-10-09 Ulrich Drepper <drepper@redhat.com> * pthread_mutex_lock.c: Always record lock ownership. * pthread_mutex_timedlock.c: Likewise. * pthread_mutex_trylock.c: Likewise. * pt-readv.c: New file. * pt-writev.c: New file. * pt-creat.c: New file. * pt-msgrcv.c: New file. * pt-msgsnd.c: New file. * pt-poll.c: New file. * pt-select.c: New file. * pt-sigpause.c: New file. * pt-sigsuspend.c: New file. * pt-sigwait.c: New file. * pt-sigwaitinfo.c: New file. * pt-waitid.c: New file. * Makefile (libpthread-routines): Add pt-readv, pt-writev, pt-creat, pt-msgrcv, pt-msgsnd, pt-poll, pt-select, pt-sigpause, pt-sigsuspend, pt-sigwait, pt-sigwaitinfo, and pt-waitid. * Versions: Add all the new functions. * tst-exit1.c: New file. * Makefile (tests): Add tst-exit1. * sem_timedwait.c: Minor optimization for more optimal fastpath. 2002-10-08 Ulrich Drepper <drepper@redhat.com> * pt-fcntl.c: Only enable asynchronous cancellation for F_SETLKW. * pthread_join.c: Enable asynchronous cancellation around lll_wait_tid call. pthread_join is an official cancellation point. * pthread_timedjoin.c: Likewise. * pthread_cond_wait.c: Revert order in which internal lock are dropped and the condvar's mutex are retrieved. * pthread_cond_timedwait.c: Likewise. Reported by dice@saros.East.Sun.COM. 2002-10-07 Ulrich Drepper <drepper@redhat.com> * pthreadP.h: Cut out all type definitions and move them... * sysdeps/unix/sysv/linux/internaltypes.h: ...here. New file. * pthreadP.h: Include <internaltypes.h>. * sysdeps/unix/sysv/linux/i386/lowlevelsem.h (lll_sem_post): Little performance tweaks. * sem_trywait.c: Shuffle #includes around to get right order. * sem_timedwait.c: Likewise. * sem_post.c: Likewise. * sem_wait.c: Likewise. * nptl 0.3 released. * Makefile (tests): Add tst-signal3. * tst-signal3.c: New file. 2002-10-05 Ulrich Drepper <drepper@redhat.com> * sysdeps/unix/sysv/linux/i386/lowlevelsem.h: Tell the compiler that the asms modify the sem object. (__lll_sem_timedwait): Now takes struct sem* as first parameter. * sysdeps/unix/sysv/linux/i386/bits/semaphore.h (sem_t): Don't expose the actual members. * pthreadP.h (struct sem): New type. Actual semaphore type. * semaphoreP.h: Include pthreadP.h. * sem_getvalue.c: Adjust to sem_t change. * sem_init.c: Likewise. * sem_open.c: Likewise. * sem_post.c: Likewise. * sem_timedwait.c: Likewise. * sem_trywait.c: Likewise. * sem_wait.c: Likewise. 2002-10-04 Ulrich Drepper <drepper@redhat.com> * Makefile (tests): Add tst-basic2, tst-exec1, tst-exec3, tst-exec3. * tst-basic2.c: New file. * tst-exec1.c: New file. * tst-exec2.c: New file. * tst-exec3.c: New file. * tst-fork1.c: Remove extra */. * nptl 0.2 released. The API for IA-32 is complete.
Diffstat (limited to 'nptl/perf.c')
-rw-r--r--nptl/perf.c749
1 files changed, 749 insertions, 0 deletions
diff --git a/nptl/perf.c b/nptl/perf.c
new file mode 100644
index 0000000..e94ccf8
--- /dev/null
+++ b/nptl/perf.c
@@ -0,0 +1,749 @@
+/* Copyright (C) 2002 Free Software Foundation, Inc.
+ This file is part of the GNU C Library.
+ Contributed by Ulrich Drepper <drepper@redhat.com>, 2002.
+
+ The GNU C Library is free software; you can redistribute it and/or
+ modify it under the terms of the GNU Lesser General Public
+ License as published by the Free Software Foundation; either
+ version 2.1 of the License, or (at your option) any later version.
+
+ The GNU C Library 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
+ Lesser General Public License for more details.
+
+ You should have received a copy of the GNU Lesser General Public
+ License along with the GNU C Library; if not, write to the Free
+ Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+ 02111-1307 USA. */
+
+#define _GNU_SOURCE 1
+#include <argp.h>
+#include <error.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <inttypes.h>
+#include <limits.h>
+#include <pthread.h>
+#include <signal.h>
+#include <stdbool.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+#include <unistd.h>
+#include <sys/param.h>
+#include <sys/types.h>
+
+#ifndef MAX_THREADS
+# define MAX_THREADS 100000
+#endif
+#ifndef DEFAULT_THREADS
+# define DEFAULT_THREADS 50
+#endif
+
+
+#define OPT_TO_THREAD 300
+#define OPT_TO_PROCESS 301
+#define OPT_SYNC_SIGNAL 302
+#define OPT_SYNC_JOIN 303
+#define OPT_TOPLEVEL 304
+
+
+static const struct argp_option options[] =
+ {
+ { NULL, 0, NULL, 0, "\
+This is a test for threads so we allow ther user to selection the number of \
+threads which are used at any one time. Independently the total number of \
+rounds can be selected. This is the total number of threads which will have \
+run when the process terminates:" },
+ { "threads", 't', "NUMBER", 0, "Number of threads used at once" },
+ { "starts", 's', "NUMBER", 0, "Total number of working threads" },
+ { "toplevel", OPT_TOPLEVEL, "NUMBER", 0,
+ "Number of toplevel threads which start the other threads; this \
+implies --sync-join" },
+
+ { NULL, 0, NULL, 0, "\
+Each thread can do one of two things: sleep or do work. The latter is 100% \
+CPU bound. The work load is the probability a thread does work. All values \
+from zero to 100 (inclusive) are valid. How often each thread repeats this \
+can be determined by the number of rounds. The work cost determines how long \
+each work session (not sleeping) takes. If it is zero a thread would \
+effectively nothing. By setting the number of rounds to zero the thread \
+does no work at all and pure thread creation times can be measured." },
+ { "workload", 'w', "PERCENT", 0, "Percentage of time spent working" },
+ { "workcost", 'c', "NUMBER", 0,
+ "Factor in the cost of each round of working" },
+ { "rounds", 'r', "NUMBER", 0, "Number of rounds each thread runs" },
+
+ { NULL, 0, NULL, 0, "\
+There are a number of different methods how thread creation can be \
+synchronized. Synchronization is necessary since the number of concurrently \
+running threads is limited." },
+ { "sync-signal", OPT_SYNC_SIGNAL, NULL, 0,
+ "Synchronize using a signal (default)" },
+ { "sync-join", OPT_SYNC_JOIN, NULL, 0, "Synchronize using pthread_join" },
+
+ { NULL, 0, NULL, 0, "\
+One parameter for each threads execution is the size of the stack. If this \
+parameter is not used the system's default stack size is used. If many \
+threads are used the stack size should be chosen quite small." },
+ { "stacksize", 'S', "BYTES", 0, "Size of threads stack" },
+ { "guardsize", 'g', "BYTES", 0,
+ "Size of stack guard area; must fit into the stack" },
+
+ { NULL, 0, NULL, 0, "Signal options:" },
+ { "to-thread", OPT_TO_THREAD, NULL, 0, "Send signal to main thread" },
+ { "to-process", OPT_TO_PROCESS, NULL, 0,
+ "Send signal to process (default)" },
+
+ { NULL, 0, NULL, 0, "Administrative options:" },
+ { "progress", 'p', NULL, 0, "Show signs of progress" },
+ { "timing", 'T', NULL, 0,
+ "Measure time from startup to the last thread finishing" },
+ { NULL, 0, NULL, 0, NULL }
+ };
+
+/* Prototype for option handler. */
+static error_t parse_opt (int key, char *arg, struct argp_state *state);
+
+/* Data structure to communicate with argp functions. */
+static struct argp argp =
+{
+ options, parse_opt
+};
+
+
+static unsigned long int threads = DEFAULT_THREADS;
+static unsigned long int workload = 75;
+static unsigned long int workcost = 20;
+static unsigned long int rounds = 10;
+static long int starts = 5000;
+static unsigned long int stacksize;
+static long int guardsize = -1;
+static bool progress;
+static bool timing;
+static bool to_thread;
+static unsigned long int toplevel = 1;
+
+
+static long int running;
+static pthread_mutex_t running_mutex = PTHREAD_MUTEX_INITIALIZER;
+
+static pid_t pid;
+static pthread_t tmain;
+
+static clockid_t cl;
+static struct timespec start_time;
+
+
+static pthread_mutex_t sum_mutex = PTHREAD_MUTEX_INITIALIZER;
+unsigned int sum;
+
+static enum
+ {
+ sync_signal,
+ sync_join
+ }
+sync_method;
+
+
+/* We use 64bit values for the times. */
+typedef unsigned long long int hp_timing_t;
+
+
+/* Attributes for all created threads. */
+static pthread_attr_t attr;
+
+
+static void *
+work (void *arg)
+{
+ unsigned long int i;
+ unsigned int state = (unsigned long int) arg;
+
+ for (i = 0; i < rounds; ++i)
+ {
+ /* Determine what to do. */
+ unsigned int rnum;
+
+ /* Equal distribution. */
+ do
+ rnum = rand_r (&state);
+ while (rnum >= UINT_MAX - (UINT_MAX % 100));
+
+ rnum %= 100;
+
+ if (rnum < workload)
+ {
+ int j;
+ int a[4] = { i, rnum, i + rnum, rnum - i };
+
+ if (progress)
+ write (STDERR_FILENO, "c", 1);
+
+ for (j = 0; j < workcost; ++j)
+ {
+ a[0] += a[3] >> 12;
+ a[1] += a[2] >> 20;
+ a[2] += a[1] ^ 0x3423423;
+ a[3] += a[0] - a[1];
+ }
+
+ pthread_mutex_lock (&sum_mutex);
+ sum += a[0] + a[1] + a[2] + a[3];
+ pthread_mutex_unlock (&sum_mutex);
+ }
+ else
+ {
+ /* Just sleep. */
+ struct timespec tv;
+
+ tv.tv_sec = 0;
+ tv.tv_nsec = 10000000;
+
+ if (progress)
+ write (STDERR_FILENO, "w", 1);
+
+ nanosleep (&tv, NULL);
+ }
+ }
+
+ return NULL;
+}
+
+
+static void *
+thread_function (void *arg)
+{
+ work (arg);
+
+ pthread_mutex_lock (&running_mutex);
+ if (--running <= 0 && starts <= 0)
+ {
+ /* We are done. */
+ if (progress)
+ write (STDERR_FILENO, "\n", 1);
+
+ if (timing)
+ {
+ struct timespec end_time;
+
+ if (clock_gettime (cl, &end_time) == 0)
+ {
+ end_time.tv_sec -= start_time.tv_sec;
+ end_time.tv_nsec -= start_time.tv_nsec;
+ if (end_time.tv_nsec < 0)
+ {
+ end_time.tv_nsec += 1000000000;
+ --end_time.tv_sec;
+ }
+
+ printf ("\nRuntime: %lu.%09lu seconds\n",
+ (unsigned long int) end_time.tv_sec,
+ (unsigned long int) end_time.tv_nsec);
+ }
+ }
+
+ printf ("Result: %08x\n", sum);
+
+ exit (0);
+ }
+ pthread_mutex_unlock (&running_mutex);
+
+ if (sync_method == sync_signal)
+ {
+ if (to_thread)
+ /* This code sends a signal to the main thread. */
+ pthread_kill (tmain, SIGUSR1);
+ else
+ /* Use this code to test sending a signal to the process. */
+ kill (pid, SIGUSR1);
+ }
+
+ if (progress)
+ write (STDERR_FILENO, "f", 1);
+
+ return NULL;
+}
+
+
+struct start_info
+{
+ unsigned int starts;
+ unsigned int threads;
+};
+
+
+static void *
+start_threads (void *arg)
+{
+ struct start_info *si = arg;
+ unsigned int starts = si->starts;
+ pthread_t ths[si->threads];
+ unsigned int state = starts;
+ unsigned int n;
+ unsigned int i = 0;
+ int err;
+
+ if (progress)
+ write (STDERR_FILENO, "T", 1);
+
+ memset (ths, '\0', sizeof (pthread_t) * si->threads);
+
+ while (starts-- > 0)
+ {
+ if (ths[i] != 0)
+ {
+ /* Wait for the threads in the order they were created. */
+ err = pthread_join (ths[i], NULL);
+ if (err != 0)
+ error (EXIT_FAILURE, err, "cannot join thread");
+
+ if (progress)
+ write (STDERR_FILENO, "f", 1);
+ }
+
+ err = pthread_create (&ths[i], &attr, work,
+ (void *) (rand_r (&state) + starts + i));
+
+ if (err != 0)
+ error (EXIT_FAILURE, err, "cannot start thread");
+
+ if (progress)
+ write (STDERR_FILENO, "t", 1);
+
+ if (++i == si->threads)
+ i = 0;
+ }
+
+ n = i;
+ do
+ {
+ if (ths[i] != 0)
+ {
+ err = pthread_join (ths[i], NULL);
+ if (err != 0)
+ error (EXIT_FAILURE, err, "cannot join thread");
+
+ if (progress)
+ write (STDERR_FILENO, "f", 1);
+ }
+
+ if (++i == si->threads)
+ i = 0;
+ }
+ while (i != n);
+
+ if (progress)
+ write (STDERR_FILENO, "F", 1);
+
+ return NULL;
+}
+
+
+int
+main (int argc, char *argv[])
+{
+ int remaining;
+ sigset_t ss;
+ pthread_t th;
+ pthread_t *ths = NULL;
+ int empty = 0;
+ int last;
+ bool cont = true;
+
+ /* Parse and process arguments. */
+ argp_parse (&argp, argc, argv, 0, &remaining, NULL);
+
+ if (sync_method == sync_join)
+ {
+ ths = (pthread_t *) calloc (threads, sizeof (pthread_t));
+ if (ths == NULL)
+ error (EXIT_FAILURE, errno,
+ "cannot allocate memory for thread descriptor array");
+
+ last = threads;
+ }
+ else
+ {
+ ths = &th;
+ last = 1;
+ }
+
+ if (toplevel > threads)
+ {
+ printf ("resetting number of toplevel threads to %lu to not surpass number to concurrent threads\n",
+ threads);
+ toplevel = threads;
+ }
+
+ if (timing)
+ {
+ if (clock_getcpuclockid (0, &cl) != 0
+ || clock_gettime (cl, &start_time) != 0)
+ timing = false;
+ }
+
+ /* We need this later. */
+ pid = getpid ();
+ tmain = pthread_self ();
+
+ /* We use signal SIGUSR1 for communication between the threads and
+ the main thread. We only want sychronous notification. */
+ if (sync_method == sync_signal)
+ {
+ sigemptyset (&ss);
+ sigaddset (&ss, SIGUSR1);
+ if (sigprocmask (SIG_BLOCK, &ss, NULL) != 0)
+ error (EXIT_FAILURE, errno, "cannot set signal mask");
+ }
+
+ /* Create the thread attributes. */
+ pthread_attr_init (&attr);
+
+ /* If the user provided a stack size use it. */
+ if (stacksize != 0
+ && pthread_attr_setstacksize (&attr, stacksize) != 0)
+ puts ("could not set stack size; will use default");
+ /* And stack guard size. */
+ if (guardsize != -1
+ && pthread_attr_setguardsize (&attr, guardsize) != 0)
+ puts ("invalid stack guard size; will use default");
+
+ /* All threads are created detached if we are not using pthread_join
+ to synchronize. */
+ if (sync_method != sync_join)
+ pthread_attr_setdetachstate (&attr, PTHREAD_CREATE_DETACHED);
+
+ if (sync_method == sync_signal)
+ {
+ while (1)
+ {
+ int err;
+ bool do_wait = false;
+
+ pthread_mutex_lock (&running_mutex);
+ if (starts-- < 0)
+ cont = false;
+ else
+ do_wait = ++running >= threads && starts > 0;
+
+ pthread_mutex_unlock (&running_mutex);
+
+ if (! cont)
+ break;
+
+ if (progress)
+ write (STDERR_FILENO, "t", 1);
+
+ err = pthread_create (&ths[empty], &attr, thread_function,
+ (void *) starts);
+ if (err != 0)
+ error (EXIT_FAILURE, err, "cannot start thread %lu", starts);
+
+ if (++empty == last)
+ empty = 0;
+
+ if (do_wait)
+ sigwaitinfo (&ss, NULL);
+ }
+
+ /* Do nothing anymore. On of the threads will terminate the program. */
+ sigfillset (&ss);
+ sigdelset (&ss, SIGINT);
+ while (1)
+ sigsuspend (&ss);
+ }
+ else
+ {
+ pthread_t ths[toplevel];
+ struct start_info si[toplevel];
+ unsigned int i;
+
+ for (i = 0; i < toplevel; ++i)
+ {
+ unsigned int child_starts = starts / (toplevel - i);
+ unsigned int child_threads = threads / (toplevel - i);
+ int err;
+
+ si[i].starts = child_starts;
+ si[i].threads = child_threads;
+
+ err = pthread_create (&ths[i], &attr, start_threads, &si[i]);
+ if (err != 0)
+ error (EXIT_FAILURE, err, "cannot start thread");
+
+ starts -= child_starts;
+ threads -= child_threads;
+ }
+
+ for (i = 0; i < toplevel; ++i)
+ {
+ int err = pthread_join (ths[i], NULL);
+
+ if (err != 0)
+ error (EXIT_FAILURE, err, "cannot join thread");
+ }
+
+ /* We are done. */
+ if (progress)
+ write (STDERR_FILENO, "\n", 1);
+
+ if (timing)
+ {
+ struct timespec end_time;
+
+ if (clock_gettime (cl, &end_time) == 0)
+ {
+ end_time.tv_sec -= start_time.tv_sec;
+ end_time.tv_nsec -= start_time.tv_nsec;
+ if (end_time.tv_nsec < 0)
+ {
+ end_time.tv_nsec += 1000000000;
+ --end_time.tv_sec;
+ }
+
+ printf ("\nRuntime: %lu.%09lu seconds\n",
+ (unsigned long int) end_time.tv_sec,
+ (unsigned long int) end_time.tv_nsec);
+ }
+ }
+
+ printf ("Result: %08x\n", sum);
+
+ exit (0);
+ }
+
+ /* NOTREACHED */
+ return 0;
+}
+
+
+/* Handle program arguments. */
+static error_t
+parse_opt (int key, char *arg, struct argp_state *state)
+{
+ unsigned long int num;
+ long int snum;
+
+ switch (key)
+ {
+ case 't':
+ num = strtoul (arg, NULL, 0);
+ if (num < MAX_THREADS)
+ threads = num;
+ else
+ printf ("\
+number of threads limited to %u; recompile with a higher limit if necessary",
+ MAX_THREADS);
+ break;
+
+ case 'w':
+ num = strtoul (arg, NULL, 0);
+ if (num <= 100)
+ workload = num;
+ else
+ puts ("workload must be between 0 and 100 percent");
+ break;
+
+ case 'c':
+ workcost = strtoul (arg, NULL, 0);
+ break;
+
+ case 'r':
+ rounds = strtoul (arg, NULL, 0);
+ break;
+
+ case 's':
+ starts = strtoul (arg, NULL, 0);
+ break;
+
+ case 'S':
+ num = strtoul (arg, NULL, 0);
+ if (num >= PTHREAD_STACK_MIN)
+ stacksize = num;
+ else
+ printf ("minimum stack size is %d\n", PTHREAD_STACK_MIN);
+ break;
+
+ case 'g':
+ snum = strtol (arg, NULL, 0);
+ if (snum < 0)
+ printf ("invalid guard size %s\n", arg);
+ else
+ guardsize = snum;
+ break;
+
+ case 'p':
+ progress = true;
+ break;
+
+ case 'T':
+ timing = true;
+ break;
+
+ case OPT_TO_THREAD:
+ to_thread = true;
+ break;
+
+ case OPT_TO_PROCESS:
+ to_thread = false;
+ break;
+
+ case OPT_SYNC_SIGNAL:
+ sync_method = sync_signal;
+ break;
+
+ case OPT_SYNC_JOIN:
+ sync_method = sync_join;
+ break;
+
+ case OPT_TOPLEVEL:
+ num = strtoul (arg, NULL, 0);
+ if (num < MAX_THREADS)
+ toplevel = num;
+ else
+ printf ("\
+number of threads limited to %u; recompile with a higher limit if necessary",
+ MAX_THREADS);
+ sync_method = sync_join;
+ break;
+
+ default:
+ return ARGP_ERR_UNKNOWN;
+ }
+
+ return 0;
+}
+
+
+static hp_timing_t
+get_clockfreq (void)
+{
+ /* We read the information from the /proc filesystem. It contains at
+ least one line like
+ cpu MHz : 497.840237
+ or also
+ cpu MHz : 497.841
+ We search for this line and convert the number in an integer. */
+ static hp_timing_t result;
+ int fd;
+
+ /* If this function was called before, we know the result. */
+ if (result != 0)
+ return result;
+
+ fd = open ("/proc/cpuinfo", O_RDONLY);
+ if (__builtin_expect (fd != -1, 1))
+ {
+ /* XXX AFAIK the /proc filesystem can generate "files" only up
+ to a size of 4096 bytes. */
+ char buf[4096];
+ ssize_t n;
+
+ n = read (fd, buf, sizeof buf);
+ if (__builtin_expect (n, 1) > 0)
+ {
+ char *mhz = memmem (buf, n, "cpu MHz", 7);
+
+ if (__builtin_expect (mhz != NULL, 1))
+ {
+ char *endp = buf + n;
+ int seen_decpoint = 0;
+ int ndigits = 0;
+
+ /* Search for the beginning of the string. */
+ while (mhz < endp && (*mhz < '0' || *mhz > '9') && *mhz != '\n')
+ ++mhz;
+
+ while (mhz < endp && *mhz != '\n')
+ {
+ if (*mhz >= '0' && *mhz <= '9')
+ {
+ result *= 10;
+ result += *mhz - '0';
+ if (seen_decpoint)
+ ++ndigits;
+ }
+ else if (*mhz == '.')
+ seen_decpoint = 1;
+
+ ++mhz;
+ }
+
+ /* Compensate for missing digits at the end. */
+ while (ndigits++ < 6)
+ result *= 10;
+ }
+ }
+
+ close (fd);
+ }
+
+ return result;
+}
+
+
+int
+clock_getcpuclockid (pid_t pid, clockid_t *clock_id)
+{
+ /* We don't allow any process ID but our own. */
+ if (pid != 0 && pid != getpid ())
+ return EPERM;
+
+#ifdef CLOCK_PROCESS_CPUTIME_ID
+ /* Store the number. */
+ *clock_id = CLOCK_PROCESS_CPUTIME_ID;
+
+ return 0;
+#else
+ /* We don't have a timer for that. */
+ return ENOENT;
+#endif
+}
+
+
+#define HP_TIMING_NOW(Var) __asm__ __volatile__ ("rdtsc" : "=A" (Var))
+
+/* Get current value of CLOCK and store it in TP. */
+int
+clock_gettime (clockid_t clock_id, struct timespec *tp)
+{
+ int retval = -1;
+
+ switch (clock_id)
+ {
+ case CLOCK_PROCESS_CPUTIME_ID:
+ {
+
+ static hp_timing_t freq;
+ hp_timing_t tsc;
+
+ /* Get the current counter. */
+ HP_TIMING_NOW (tsc);
+
+ if (freq == 0)
+ {
+ freq = get_clockfreq ();
+ if (freq == 0)
+ return EINVAL;
+ }
+
+ /* Compute the seconds. */
+ tp->tv_sec = tsc / freq;
+
+ /* And the nanoseconds. This computation should be stable until
+ we get machines with about 16GHz frequency. */
+ tp->tv_nsec = ((tsc % freq) * UINT64_C (1000000000)) / freq;
+
+ retval = 0;
+ }
+ break;
+
+ default:
+ errno = EINVAL;
+ break;
+ }
+
+ return retval;
+}