Bug#713035: eatmydata bugs
On Fri, Jun 28, 2013 at 11:55:12AM +1000, Stewart Smith wrote:
> Thomas Preud'homme <robotux@debian.org> writes:
> > And here is a proposed patch although I'm sure you don't need it :)
>
> I think we'll actually need the cancellation points before we check if
> eatmydata is hungry, so I'm actually going to do something slightly
> different than this.
>
> > --- libeatmydata-26/debian/changelog 2011-02-19 13:28:02.000000000
> > +0100
>
> libeatmydata-26 is quite old and really should be updated.
>
> Bugs fixed since then include:
> - added sync_file_range support
> - eatmydata script imported from debian, made to be cross-platform
> - MacOS X support
> - improved test suite
> - fixes bugs related to having/not having large file support, fixes
> 32bit problems
> - merging in most of the debian patches.
>
> so it's probably worth just updating the debian package to the newer
> libeatmydata.
>
> New release up at:
> https://launchpad.net/libeatmydata/trunk/libeatmydata-82
>
> https://launchpad.net/libeatmydata/trunk/libeatmydata-82/+download/libeatmydata-82.tar.gz
>
> sig:
> https://launchpad.net/libeatmydata/trunk/libeatmydata-82/+download/libeatmydata-82.tar.gz.asc
>
Thanks for this new release. I have tested it and I am able to run the
tst-cancel4, tst-cancelx4, tst-cancel5 and tst-cancel5 tests without
problem with eatmydata. Note however that sync() is not a cancellation
point, so the call to pthread_testcancel() should be removed there.
That said, there are still two tests failing when run under eatmydata,
tst-key1 and tst-key4, though I haven't been able to understand what is
the problem. I have attached the later one, so that someone can debug
the issue.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
/* Copyright (C) 2003, 2004 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Contributed by Ulrich Drepper <drepper@redhat.com>, 2003.
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, see
<http://www.gnu.org/licenses/>. */
#include <limits.h>
#include <pthread.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#ifdef PTHREAD_KEYS_MAX
const int max = PTHREAD_KEYS_MAX;
#else
const int max = _POSIX_THREAD_KEYS_MAX;
#endif
static pthread_key_t *keys;
static void *
tf1 (void *arg)
{
int i;
for (i = 0; i < max; ++i)
if (pthread_setspecific (keys[i], (void *) (long int) (i + 1)) != 0)
{
puts ("setspecific failed");
exit (1);
}
return NULL;
}
static void *
tf2 (void *arg)
{
int i;
for (i = 0; i < max; ++i)
if (pthread_getspecific (keys[i]) != NULL)
{
printf ("getspecific for key %d not NULL\n", i);
exit (1);
}
return NULL;
}
int main()
{
keys = alloca (max * sizeof (pthread_key_t));
int i;
for (i = 0; i < max; ++i)
if (pthread_key_create (&keys[i], NULL) != 0)
{
puts ("key_create failed");
exit (1);
}
pthread_attr_t a;
if (pthread_attr_init (&a) != 0)
{
puts ("attr_init failed");
exit (1);
}
if (pthread_attr_setstacksize (&a, 1 * 1024 * 1024) != 0)
{
puts ("attr_setstacksize failed");
return 1;
}
for (i = 0; i < 10; ++i)
{
int j;
#define N 2
pthread_t th[N];
for (j = 0; j < N; ++j)
if (pthread_create (&th[j], NULL, tf1, NULL) != 0)
{
puts ("1st create failed");
exit (1);
}
for (j = 0; j < N; ++j)
if (pthread_join (th[j], NULL) != 0)
{
puts ("1st join failed");
exit (1);
}
for (j = 0; j < N; ++j)
if (pthread_create (&th[j], NULL, tf2, NULL) != 0)
{
puts ("2nd create failed");
exit (1);
}
for (j = 0; j < N; ++j)
if (pthread_join (th[j], NULL) != 0)
{
puts ("2nd join failed");
exit (1);
}
}
if (pthread_attr_destroy (&a) != 0)
{
puts ("attr_destroy failed");
exit (1);
}
return 0;
}
Reply to: