[cairo-commit] 2 commits - perf/README test/cairo-test-trace.c
Chris Wilson
ickle at kemper.freedesktop.org
Sat Jun 13 02:42:43 PDT 2009
perf/README | 16 +
test/cairo-test-trace.c | 400 ++++++++++++++++++++++++++++--------------------
2 files changed, 252 insertions(+), 164 deletions(-)
New commits:
commit 3dde883b779b81b95f420039c02b51b029311f78
Author: Chris Wilson <chris at chris-wilson.co.uk>
Date: Sat Jun 13 10:13:20 2009 +0100
[test] Code review after sleep
Review cairo-test-trace.c and rewrite parts to ease understanding and fix
various bugs - such as failure to notice the slaves crashing and not
releasing our shared memory after an interrupt.
diff --git a/test/cairo-test-trace.c b/test/cairo-test-trace.c
index b156895..ffc7235 100644
--- a/test/cairo-test-trace.c
+++ b/test/cairo-test-trace.c
@@ -24,12 +24,12 @@
*/
/*
- * The basic premise is that we feed the trace to multiple backends in
- * parallel and compare the output at the end of each context (based on the
- * premise that contexts demarcate expose events, or their logical
- * equivalents) with that of the image[1] backend. Each backend is executed in
- * a separate process, for robustness, with the image data residing in shared
- * memory and synchronising over a socket.
+ * The basic idea is that we feed the trace to multiple backends in parallel
+ * and compare the output at the end of each context (based on the premise
+ * that contexts demarcate expose events, or their logical equivalents) with
+ * that of the image[1] backend. Each backend is executed in a separate
+ * process, for robustness and to isolate the global cairo state, with the
+ * image data residing in shared memory and synchronising over a socket.
*
* [1] Should be reference implementation, currently the image backend is
* considered to be the reference for all other backends.
@@ -45,7 +45,14 @@
* So for each backend spawn two processes, a reference and xlib
* (obviously minimising the number of reference processes when possible)
*/
-#define _GNU_SOURCE 1 /* for sched_getaffinity() and getline() */
+
+/*
+ * XXX Handle show-page as well as cairo_destroy()? Though arguably that is
+ * only relevant for paginated backends which is currently outside the
+ * scope of this test.
+ */
+
+#define _GNU_SOURCE 1 /* getline() */
#include "cairo-test.h"
@@ -77,32 +84,38 @@
/* manage the shared memory using Doug Lea's malloc */
#define USE_LOCKS 0
#define USE_DL_PREFIX 1
+#if HAVE_FFS
#define USE_BUILTIN_FFS 1
+#endif
/* We only use the segment created my create_mspace_with_base. */
#define MSPACES 1
#define ONLY_MSPACES 1
+/* Disable manipulation of the memory segment */
#define HAVE_MORECORE 0
#define HAVE_MMAP 1
#define HAVE_MREMAP 0
+#define mmap(a, b, c, d, e, f) MFAIL
+#define munmap(a, b) (-1)
+#define DEFAULT_MMAP_THRESHOLD MAX_SIZE_T
/* We have no use for this, so save some code and data. */
#define NO_MALLINFO 1
-
-/* We need all allocations to be in regular segments. */
-#define DEFAULT_MMAP_THRESHOLD MAX_SIZE_T
+#define ABORT_ON_ASSERT_FAILURE 0
/* Don't allocate more than a page unless needed. */
#define DEFAULT_GRANULARITY ((size_t)malloc_getpagesize)
-#define mmap(a, b, c, d, e, f) MFAIL
-#define munmap(a, b) (-1)
#include "dlmalloc.c"
#undef mmap
#undef munmap
+#undef assert
+#include <assert.h>
+
#define DATA_SIZE (64 << 20)
+#define SHM_PATH_XXX "/shmem-cairo-trace"
typedef struct _test_runner {
/* Options from command-line */
@@ -144,12 +157,19 @@ struct slave {
const cairo_boilerplate_target_t *target;
};
+struct request_image {
+ unsigned long id;
+ cairo_format_t format;
+ int width;
+ int height;
+ int stride;
+};
+
struct surface_tag {
int width, height;
};
static const cairo_user_data_key_t surface_tag;
-
static cairo_bool_t
writen (int fd, const void *ptr, int len)
{
@@ -222,14 +242,6 @@ format_for_content (cairo_content_t content)
}
}
-struct request_image {
- unsigned long id;
- cairo_format_t format;
- int width;
- int height;
- int stride;
-};
-
static void *
request_image (test_runner_thread_t *thread,
unsigned long id,
@@ -241,8 +253,8 @@ request_image (test_runner_thread_t *thread,
writen (thread->sk, &rq, sizeof (rq));
readn (thread->sk, &offset, sizeof (offset));
-
- assert (offset != (size_t) -1);
+ if (offset == (size_t) -1)
+ return NULL;
return thread->base + offset;
}
@@ -253,29 +265,47 @@ push_surface (test_runner_thread_t *thread,
unsigned long id)
{
cairo_surface_t *image;
- cairo_format_t format;
+ cairo_format_t format = (cairo_format_t) -1;
cairo_t *cr;
int width, height, stride;
void *data;
unsigned long serial;
- format = format_for_content (cairo_surface_get_content (source));
if (cairo_surface_get_type (source) == CAIRO_SURFACE_TYPE_IMAGE) {
width = cairo_image_surface_get_width (source);
height = cairo_image_surface_get_height (source);
+ format = cairo_image_surface_get_format (source);
} else {
struct surface_tag *tag;
tag = cairo_surface_get_user_data (source, &surface_tag);
- assert (tag != NULL);
- width = tag->width;
- height = tag->height;
+ if (tag != NULL) {
+ width = tag->width;
+ height = tag->height;
+ } else {
+ double x0, x1, y0, y1;
+
+ /* presumably created using cairo_surface_create_similar() */
+ cr = cairo_create (source);
+ cairo_clip_extents (cr, &x0, &y0, &x1, &y1);
+ cairo_destroy (cr);
+
+ tag = xmalloc (sizeof (*tag));
+ width = tag->width = x1 - x0;
+ height = tag->height = y1 - y0;
+
+ if (cairo_surface_set_user_data (source, &surface_tag, tag, free))
+ exit (-1);
+ }
}
+ if (format == (cairo_format_t) -1)
+ format = format_for_content (cairo_surface_get_content (source));
stride = cairo_format_stride_for_width (format, width);
data = request_image (thread, id, format, width, height, stride);
- assert (data != NULL);
+ if (data == NULL)
+ exit (-1);
image = cairo_image_surface_create_for_data (data,
format,
@@ -314,7 +344,8 @@ _surface_create (void *closure,
tag = xmalloc (sizeof (*tag));
tag->width = width;
tag->height = height;
- cairo_surface_set_user_data (surface, &surface_tag, tag, free);
+ if (cairo_surface_set_user_data (surface, &surface_tag, tag, free))
+ exit (-1);
}
return surface;
@@ -349,7 +380,8 @@ _context_destroy (void *closure, void *ptr)
if (cairo_surface_status (l->surface) == CAIRO_STATUS_SUCCESS) {
push_surface (thread, l->surface, l->id);
} else {
- fprintf (stderr, "error during replay: %s!\n",
+ fprintf (stderr, "%s: error during replay: %s!\n",
+ thread->target->name,
cairo_status_to_string (cairo_surface_status
(l->surface)));
exit (1);
@@ -382,7 +414,8 @@ execute (test_runner_thread_t *thread,
cairo_script_interpreter_run (csi, trace);
cairo_script_interpreter_finish (csi);
- cairo_script_interpreter_destroy (csi);
+ if (cairo_script_interpreter_destroy (csi))
+ exit (1);
}
static int
@@ -408,24 +441,21 @@ spawn_socket (const char *socket_path, pid_t pid)
return sk;
}
-
static void *
spawn_shm (const char *shm_path)
{
- void *base = MAP_FAILED;
+ void *base;
int fd;
fd = shm_open (shm_path, O_RDWR, 0);
if (fd == -1)
- return base;
+ return MAP_FAILED;
base = mmap (NULL, DATA_SIZE,
PROT_READ | PROT_WRITE,
MAP_SHARED | MAP_NORESERVE,
fd, 0);
close (fd);
- if (base == MAP_FAILED)
- return base;
return base;
}
@@ -490,6 +520,7 @@ spawn_target (const char *socket_path,
exit (0);
}
+/* XXX imagediff - is the extra expense worth it? */
static cairo_bool_t
matching_images (cairo_surface_t *a, cairo_surface_t *b)
{
@@ -521,39 +552,17 @@ matching_images (cairo_surface_t *a, cairo_surface_t *b)
}
static cairo_bool_t
-check_images_and_ack (mspace msp, struct slave *slaves, int num_slaves)
+check_images (struct slave *slaves, int num_slaves)
{
int n;
- /* wait for completion */
- if (slaves[0].image_ready == 0)
- return TRUE;
-
for (n = 1; n < num_slaves; n++) {
- if (slaves[n].image_ready != slaves[0].image_ready)
- return TRUE;
- }
+ assert (slaves[n].image_ready == slaves[0].image_ready);
- /* compare */
- for (n = 1; n < num_slaves; n++) {
if (! matching_images (slaves[n].image, slaves[0].image))
return FALSE;
}
- /* ack */
- for (n = 0; n < num_slaves; n++) {
- cairo_surface_finish (slaves[n].image);
- mspace_free (msp, cairo_image_surface_get_data (slaves[n].image));
- cairo_surface_destroy (slaves[n].image);
- slaves[n].image = NULL;
-
- writen (slaves[n].fd,
- &slaves[n].image_serial,
- sizeof (slaves[n].image_serial));
- slaves[n].image_serial = 0;
- slaves[n].image_ready = 0;
- }
-
return TRUE;
}
@@ -596,7 +605,7 @@ allocate_image_for_slave (uint8_t *base, mspace *msp, struct slave *slave)
return data - base;
}
-static void
+static cairo_bool_t
test_run (void *base,
mspace msp,
int sk,
@@ -606,14 +615,16 @@ test_run (void *base,
{
struct pollfd *pfd;
int npfd, cnt, n, i;
+ int completion;
+ cairo_bool_t ret = FALSE;
pfd = xcalloc (num_slaves+2, sizeof (*pfd));
pfd[0].fd = sk;
pfd[0].events = POLLIN;
- pfd[0].revents = 0; /* valgrind */
npfd = 1;
+ completion = 0;
while ((cnt = poll (pfd, npfd, -1)) > 0) {
if (pfd[0].revents) {
int fd;
@@ -638,21 +649,22 @@ test_run (void *base,
if (! pfd[n].revents)
continue;
+ if (pfd[n].revents & POLLHUP)
+ goto done;
+
for (i = 0; i < num_slaves; i++) {
if (slaves[i].fd == pfd[n].fd) {
- if (pfd[n].revents & POLLHUP)
- /* XXX check exitcode? */
- goto done;
-
/* Communication with the slave is done in three phases,
* and we do each pass synchronously.
*
* 1. The slave requests an image buffer, which we
- * allocate and then return the offset into the shared
- * memory segment.
+ * allocate and then return to the slave the offset into
+ * the shared memory segment.
*
* 2. The slave indicates that it has finished writing
- * into the shared image buffer.
+ * into the shared image buffer. The slave now waits
+ * for the server to collate all the image data - thereby
+ * throttling the slaves.
*
* 3. After all slaves have finished writing their images,
* we compare them all against the reference image and,
@@ -664,30 +676,18 @@ test_run (void *base,
offset =
allocate_image_for_slave (base, msp,
&slaves[i]);
- if (! writen (pfd[n].fd, &offset, sizeof (offset))) {
- fprintf (stderr,
- "communication error with slave\n");
+ if (! writen (pfd[n].fd, &offset, sizeof (offset)))
goto out;
- }
} else {
readn (pfd[n].fd,
&slaves[i].image_ready,
sizeof (slaves[i].image_ready));
- if (slaves[i].image_ready != slaves[i].image_serial) {
- fprintf (stderr,
- "communication error with slave: "
- "expected serial %lu, got %lu\n",
- slaves[i].image_serial,
- slaves[i].image_ready);
+ if (slaves[i].image_ready != slaves[i].image_serial)
goto out;
- }
- if (! check_images_and_ack (msp, slaves, num_slaves)) {
- printf ("FAIL: backends differ after context %lu\n",
- slaves[0].image_serial);
- write_images (trace, slaves, num_slaves);
- goto out;
- }
+ /* Can anyone spell 'P·E·D·A·N·T'? */
+ cairo_surface_mark_dirty (slaves[i].image);
+ completion++;
}
break;
@@ -696,9 +696,37 @@ test_run (void *base,
cnt--;
}
+
+ if (completion == num_slaves) {
+ if (! check_images (slaves, num_slaves)) {
+ write_images (trace, slaves, num_slaves);
+ goto out;
+ }
+
+ /* ack */
+ for (i = 0; i < num_slaves; i++) {
+ cairo_surface_finish (slaves[i].image);
+ mspace_free (msp,
+ cairo_image_surface_get_data (slaves[i].image));
+ cairo_surface_destroy (slaves[i].image);
+ slaves[i].image = NULL;
+
+ if (! writen (slaves[i].fd,
+ &slaves[i].image_serial,
+ sizeof (slaves[i].image_serial)))
+ {
+ goto out;
+ }
+
+ slaves[i].image_serial = 0;
+ slaves[i].image_ready = 0;
+ }
+
+ completion = 0;
+ }
}
done:
- printf ("PASS\n");
+ ret = TRUE;
out:
for (n = 0; n < num_slaves; n++) {
@@ -715,9 +743,12 @@ out:
slaves[n].image_serial = 0;
slaves[n].image_ready = 0;
+ ret = FALSE;
}
free (pfd);
+
+ return ret;
}
/* Paginated surfaces require finalization and external converters and so
@@ -763,93 +794,96 @@ target_is_measurable (const cairo_boilerplate_target_t *target)
return TRUE;
}
-/* XXX cleanup on SIGINT. */
-static void
-test_trace (test_runner_t *test, const char *trace)
+static int
+server_socket (const char *socket_path)
{
- const char *shm_path = "/shmem";
- const cairo_boilerplate_target_t *target, *image;
- char *trace_cpy, *name, *dot;
- struct slave *slaves, *s;
- char socket_dir[] = "/tmp/cairo-test-trace.XXXXXX";
- char *socket_path;
- int sk, fd;
long flags;
struct sockaddr_un addr;
- int i, num_slaves;
- void *base;
- mspace msp;
-
- trace_cpy = xstrdup (trace);
- name = basename (trace_cpy);
- dot = strchr (name, '.');
- if (dot)
- *dot = '\0';
-
- if (test->list_only) {
- printf ("%s\n", name);
- free (trace_cpy);
- return;
- }
-
- printf ("%s: ", name);
- fflush (stdout);
-
- /* create a socket to control the test runners */
- if (mkdtemp (socket_dir) == NULL) {
- fprintf (stderr, "unable to create temporary name for socket\n");
- return;
- }
-
- xasprintf (&socket_path, "%s/socket", socket_dir);
+ int sk;
sk = socket (PF_UNIX, SOCK_STREAM, 0);
- if (sk == -1) {
- fprintf (stderr, "unable to create socket\n");
- return;
- }
+ if (sk == -1)
+ return -1;
memset (&addr, 0, sizeof (addr));
addr.sun_family = AF_UNIX;
strcpy (addr.sun_path, socket_path);
if (bind (sk, (struct sockaddr *) &addr, sizeof (addr)) == -1) {
close (sk);
- fprintf (stderr, "unable to bind socket\n");
- return;
+ return -1;
}
flags = fcntl (sk, F_GETFL);
if (flags == -1 || fcntl (sk, F_SETFL, flags | O_NONBLOCK) == -1) {
close (sk);
- fprintf (stderr, "unable to set socket to non-blocking\n");
- return;
+ return -1;
}
if (listen (sk, 5) == -1) {
- fprintf (stderr, "unable to listen on socket\n");
- goto cleanup_sk;
+ close (sk);
+ return -1;
}
- shm_path="/shmem";
+ return sk;
+}
+
+static int
+server_shm (const char *shm_path)
+{
+ int fd;
+
fd = shm_open (shm_path, O_RDWR | O_EXCL | O_CREAT, 0777);
- if (fd == -1) {
- fprintf (stderr,
- "unable to create shared memory '%s': %s\n",
- shm_path, strerror (errno));
- goto cleanup_sk;
- }
+ if (fd == -1)
+ return -1;
if (ftruncate (fd, DATA_SIZE) == -1) {
- fprintf (stderr, "unable to resize shared memory\n");
close (fd);
+ return -1;
+ }
+
+ return fd;
+}
+
+static cairo_bool_t
+_test_trace (test_runner_t *test, const char *trace, const char *name)
+{
+ const char *shm_path = SHM_PATH_XXX;
+ const cairo_boilerplate_target_t *target, *image;
+ struct slave *slaves, *s;
+ char socket_dir[] = "/tmp/cairo-test-trace.XXXXXX";
+ char *socket_path;
+ int sk, fd;
+ int i, num_slaves;
+ void *base;
+ mspace msp;
+ cairo_bool_t ret = FALSE;
+
+ /* create a socket to control the test runners */
+ if (mkdtemp (socket_dir) == NULL) {
+ fprintf (stderr, "Unable to create temporary name for socket\n");
+ return FALSE;
+ }
+
+ xasprintf (&socket_path, "%s/socket", socket_dir);
+ sk = server_socket (socket_path);
+ if (sk == -1) {
+ fprintf (stderr, "Unable to create socket for server\n");
+ goto cleanup_paths;
+ }
+
+ /* allocate some shared memory */
+ fd = server_shm (shm_path);
+ if (fd == -1) {
+ fprintf (stderr, "Unable to create shared memory '%s'\n",
+ shm_path);
goto cleanup_sk;
}
image = cairo_boilerplate_get_image_target (CAIRO_CONTENT_COLOR_ALPHA);
assert (image != NULL);
+ /* spawn slave processes to run the trace */
s = slaves = xcalloc (test->num_targets + 1, sizeof (struct slave));
-
s->pid = spawn_target (socket_path, shm_path, image, trace);
if (s->pid < 0)
goto cleanup;
@@ -879,36 +913,66 @@ test_trace (test_runner_t *test, const char *trace)
goto cleanup;
}
+ /* map our shared memory and manage using a dlmalloc pool */
base = mmap (NULL, DATA_SIZE, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
- close (fd);
-
if (base == MAP_FAILED) {
- fprintf (stderr, "unable to mmap shared memory\n");
+ fprintf (stderr, "Unable to mmap shared memory\n");
goto cleanup;
}
-
msp = create_mspace_with_base (base, DATA_SIZE, 0);
- test_run (base, msp, sk, name, slaves, num_slaves);
+ ret = test_run (base, msp, sk, name, slaves, num_slaves);
destroy_mspace (msp);
munmap (base, DATA_SIZE);
cleanup:
+ close (fd);
while (s-- > slaves) {
int status;
kill (s->pid, SIGKILL);
waitpid (s->pid, &status, 0);
+ ret &= WIFEXITED (status) && WEXITSTATUS (status) == 0;
+ if (WIFSIGNALED (status) && WTERMSIG(status) != SIGKILL) {
+ fprintf (stderr, "%s crashed\n", s->target->name);
+ }
}
free (slaves);
shm_unlink (shm_path);
cleanup_sk:
close (sk);
+cleanup_paths:
remove (socket_path);
remove (socket_dir);
free (socket_path);
+ return ret;
+}
+
+static void
+test_trace (test_runner_t *test, const char *trace)
+{
+ char *trace_cpy, *name, *dot;
+
+ trace_cpy = xstrdup (trace);
+ name = basename (trace_cpy);
+ dot = strchr (name, '.');
+ if (dot)
+ *dot = '\0';
+
+ if (test->list_only) {
+ printf ("%s\n", name);
+ } else {
+ cairo_bool_t ret;
+
+ printf ("%s: ", name);
+ fflush (stdout);
+
+ ret = _test_trace (test, trace, name);
+ printf ("%s\n", ret ? "PASS" : "FAIL");
+ }
+
free (trace_cpy);
}
@@ -1008,7 +1072,7 @@ usage (const char *argv0)
"Usage: %s [-x exclude-file] [test-names ... | traces ...]\n"
" %s -l\n"
"\n"
-"Run the cairo performance test suite over the given tests (all by default)\n"
+"Run the cairo test suite over the given traces (all by default).\n"
"The command-line arguments are interpreted as follows:\n"
"\n"
" -x exclude; specify a file to read a list of traces to exclude\n"
@@ -1063,7 +1127,7 @@ parse_options (test_runner_t *test, int argc, char *argv[])
}
static void
-cairo_test_reset (test_runner_t *test)
+test_reset (test_runner_t *test)
{
cairo_debug_reset_static_data ();
#if HAVE_FCFINI
@@ -1072,9 +1136,9 @@ cairo_test_reset (test_runner_t *test)
}
static void
-cairo_test_fini (test_runner_t *test)
+test_fini (test_runner_t *test)
{
- cairo_test_reset (test);
+ test_reset (test);
cairo_boilerplate_free_targets (test->targets);
if (test->exclude_names)
@@ -1082,7 +1146,7 @@ cairo_test_fini (test_runner_t *test)
}
static cairo_bool_t
-have_trace_filenames (test_runner_t *test)
+test_has_filenames (test_runner_t *test)
{
unsigned int i;
@@ -1097,8 +1161,7 @@ have_trace_filenames (test_runner_t *test)
}
static cairo_bool_t
-cairo_test_can_run (test_runner_t *test,
- const char *name)
+test_can_run (test_runner_t *test, const char *name)
{
unsigned int i;
char *copy, *dot;
@@ -1151,6 +1214,15 @@ warn_no_traces (const char *message, const char *trace_dir)
message, trace_dir);
}
+static void
+interrupt (int sig)
+{
+ shm_unlink (SHM_PATH_XXX);
+
+ signal (sig, SIG_DFL);
+ raise (sig);
+}
+
int
main (int argc, char *argv[])
{
@@ -1159,6 +1231,7 @@ main (int argc, char *argv[])
unsigned int n;
signal (SIGPIPE, SIG_IGN);
+ signal (SIGINT, interrupt);
parse_options (&test, argc, argv);
@@ -1167,14 +1240,13 @@ main (int argc, char *argv[])
test.targets = cairo_boilerplate_get_targets (&test.num_targets, NULL);
- /* do we have a list of filenames? */
- if (have_trace_filenames (&test)) {
+ if (test_has_filenames (&test)) {
for (n = 0; n < test.num_names; n++) {
- if (cairo_test_can_run (&test, test.names[n]) &&
+ if (test_can_run (&test, test.names[n]) &&
access (test.names[n], R_OK) == 0)
{
test_trace (&test, test.names[n]);
- cairo_test_reset (&test);
+ test_reset (&test);
}
}
} else {
@@ -1185,7 +1257,7 @@ main (int argc, char *argv[])
dir = opendir (trace_dir);
if (dir == NULL) {
warn_no_traces ("Failed to open directory", trace_dir);
- cairo_test_fini (&test);
+ test_fini (&test);
return 1;
}
@@ -1200,12 +1272,12 @@ main (int argc, char *argv[])
continue;
num_traces++;
- if (! cairo_test_can_run (&test, de->d_name))
+ if (! test_can_run (&test, de->d_name))
continue;
xasprintf (&trace, "%s/%s", trace_dir, de->d_name);
test_trace (&test, trace);
- cairo_test_reset (&test);
+ test_reset (&test);
free (trace);
@@ -1214,12 +1286,12 @@ main (int argc, char *argv[])
if (num_traces == 0) {
warn_no_traces ("Found no traces in", trace_dir);
- cairo_test_fini (&test);
+ test_fini (&test);
return 1;
}
}
- cairo_test_fini (&test);
+ test_fini (&test);
return 0;
}
commit 1f542965f0310aed493651b8ddc1e98a4192b977
Author: Chris Wilson <chris at chris-wilson.co.uk>
Date: Sat Jun 13 08:24:22 2009 +0100
[perf] Mention cairo-perf-compare-backends in the README
diff --git a/perf/README b/perf/README
index fd6eb30..0fc5281 100644
--- a/perf/README
+++ b/perf/README
@@ -132,6 +132,22 @@ those tests:
./cairo-perf-diff -f HEAD -- text
+Generating comparisons of different backends
+--------------------------------------------
+An alternate question that is often asked is, "how does the speed of one
+backend compare to another?". cairo-perf-compare-backends can read files
+generated by cairo-perf and produces a comparison of the backends for every
+test.
+
+Again, by way of example:
+
+ # Show relative performance of the backends
+ ./cairo-perf-compare-backends cairo.perf
+
+This will work whether the data files were generate in raw mode (with
+cairo-perf -r) or cooked, (cairo-perf without -r).
+
+
Creating a new performance test
-------------------------------
This is where we could use everybody's help. If you have encountered a
More information about the cairo-commit
mailing list