Hi Reinhard,
I checked on an ia64-box and can confirm: the code in the bazaar tree
has eliminated the offending warnings and the patch I submitted is no
longer needed.
Thanks,
--david
--
Mosberger Consulting LLC, http://www.mosberger-consulting.com/
--
To UNSUBSCRIBE, email to [EMAIL
onPtr, and NeAACDecFrameInfo. With
these changes, decoder.h can be included in xine_decoder.c and the
problematic implicit declarations are taken care of. It's not the
most elegant patch, but it's the best I was able to come up with that
doesn't require major surgery.
--david
--
ch for bug 360003 applied, I don't see any 64-bit issues
with building xine_decoder.c anymore. I'm not sure how to test the
library on its own though ("make check" doesn't seem to do anything
useful).
--david
--
Mosberger Consulting LLC, http://www.mosberger-consulting.com/
licit-pointer-functions
In addition, src/read.c:ref_val() implicitly declares its "ref"
argument as an "int", when it's really a pointer. That causes a crash
in the io2.scm test case. With the patch below applied, "make check"
succeeds on ia64 linux.
Thanks,
-
lient' implicitly converted to pointer at
gsm-at-startup.c:30
[1] http://people.debian.org/~dannf/check-implicit-pointer-functions
Patch to add the missing declaration is attached below.
Thanks,
--david
--
Mosberger Consulting LLC, http://www.mosberger-consulting.com/
--- gnome-session-2.
Hi Michael,
Thanks for taking care of this problem. I tested the new version and
it's working fine again.
--david
--
Mosberger Consulting LLC, http://www.mosberger-consulting.com/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble?
Package: apt
Version: 0.6.44
Tags: patch
The current version of `apt' causes unaligned accesses whenever
running "apt-get update". On ia64 Linux, this typically shows up as
messages of the form:
http(22944): unaligned access to 0x600062dd, ip=0x200a6e01
The problem comes from t
o answer the inevitable queries.
>
> http://bugs.slimdevices.com/
Can you forward the report and then sets me as the "Reply-to" contact?
Thanks,
--david
--
Mosberger Consulting LLC, http://www.mosberger-consulting.com/
Package: slimserver
Version: 6.2.1-2
My machine has Debian/stable as the base distribution and then I
installed slimserver/unstable on top of that. At least in that
configuration, slimserver failed to load the GD library and hence was
unable to use TrueType fonts. The library couldn't be loaded
Package: slimserver
Version: 6.2.1-2
Tags: patch
Slimserver assumes that in CUE files several tags such as TRACK or
PERFORMER are indented. This indentation is neither necessary nor is
it generated by tools such as "cueconvert". The patch below fixes the
problem by making the leading whitespace
Package: slimserver
Version: 6.2.1-2
Tags: patch
There appear to be two problems with getting SlimServer to recognize
FLACs with multiple tracks (I like to rip CDs into a single FLAC with
a separate CUE sheet).
First, SlimServer ignores songs which don't have an associated
byte-size. This is the
SOURCE need to
be defined because with _XOPEN_SOURCE alone, type "ulong" does NOT get
declared, which then causes problems in zend_hash.h. Fun, fun.
Regards,
--david
--
Mosberger Consulting LLC, http://www.mosberger-consulting.com/
--- ext/standard/datetime.c-orig 2006-03-20 08:20:
Package: mono
Version: 1.1.9.1
Tags: patch
Our automated buildd log filter [1] detected a problem that will cause
your package to segfault on architectures where the size of a pointer
is greater than the size of an integer, such as ia64:
Function `dis_stringify_marshal_spec' implicitly converted
Package: gnome-pilot
Version: 2.0.12-1.3
Tags: patch
Our automated buildd log filter [1] detected a problem that will cause
your package to segfault on architectures where the size of a pointer
is greater than the size of an integer, such as ia64:
Function `crypt' implicitly converted to pointer
Package: gtkdialog
Version: 0.59.4-1
Tags: patch
Our automated buildd log filter [1] detected a problem that will cause
your package to segfault on architectures where the size of a pointer
is greater than the size of an integer, such as ia64:
Function `str_default_name' implicitly converted to p
Package: gtkmathview
Version: 0.7.0-1
Our automated buildd log filter[1] detected a problem that could cause
your package to segfault on architectures where the size of a pointer
is greater than the size of an integer, such as ia64. The specific
issue is:
Function `find_action_element' implici
Package: mrxvt
Version: 0.4.1-2
Tags: patch
Our automated buildd log filter[1] detected a problem that causes your
package to segfault on architectures where the size of a pointer is
greater than the size of an integer, such as ia64 (see warnings
below). The fix for this problem is to add -D_GNU_
>> Third, part-property.c passed a pointer to an unitialized size_t-typed
>> variable to get_macro_name(), which was expecting only an int pointer.
>> Thus, the top 32-bits remained unitialized and on ia64 this caused a
>> crash when attempting to insert any part.
> this was fixed in 0.40.3
Package: oregano
Version: 0.40.0-3
Severity: serious
Tags: patch
Oregano defines GNOME_DISABLE_DEPRECATED which has the effect that
gnome_menu_item_new() does NOT get declared. As a result, the
function gets implicitly defined to return an "int", which is a
problem on 64-bit architectures since
Package: gnuift
Version: 0.1.14-1
> On Tue, 10 May 2005 23:59:36 +0100, Debian/IA64 non-US Build Daemon
> <[EMAIL PROTECTED]> said:
> Function `malloc' implicitly converted to pointer at rgb2hsv_utils.c:186
> Function `malloc' implicitly converted to pointer at quantize.c:29
These ar
Are you sure the bug was fixed? I just got this from the buildd
monitor:
--
From: Debian/IA64 non-US Build Daemon <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Subject: Log for successful build of evolution-data-server1.2_1.2.2-3
(dist=unstable)
Date: Thu, 05 May 2005 14:02:00 +0100
Function `camel
> On Wed, 13 Apr 2005 15:22:22 +0200, Daniel Kobras <[EMAIL PROTECTED]>
> said:
Daniel> Great, thanks for testing! If you need some more material to
Daniel> play around with, the download section of the libdv project
Daniel> on sourceforge carries a few canonical examples in various
Here is the trivial patch to fix the evolution 2.2 crash when clicking
on "Tasks". Please apply.
--david
diff -urN -x Makefile -x '*log' -x config.status
evolution-2.2.2/calendar/gui/e-tasks.c
evolution-2.2.2-davidm/calendar/gui/e-tasks.c
--- evolution-2.2.2/calendar/gui/e-tasks.c
Package: evolution-data-server1.2
Version: 1.2.2-2
Severity: important
Tags: patch
The patch below fixes bug 304708 but since the real culprit is
evolution-data-server1.2, I thought it's better to file it here.
Briefly, clicking on "reply" in the evolution mailer causes crashes on
ia64. This is
>>>>> On Mon, 18 Apr 2005 13:38:08 -0600, dann frazier <[EMAIL PROTECTED]> said:
Dann> On Mon, 2005-04-18 at 11:56 -0700, David Mosberger wrote:
>> >>>>> On Wed, 13 Apr 2005 09:57:53 -0600, dann frazier
>> <[EMAIL PROTECTED]> s
On 4/17/05, Eric Dorland <[EMAIL PROTECTED]> wrote:
> Hmmm, I looked at the mozilla trunk and I didn't see any real
> substantial changes to those in the past year. Are you sure upstream
> fixed this problem themselves? I'd prefer to use their solution rather
> than yours just for consistency's sak
>>>>> On Tue, 12 Apr 2005 10:37:13 -0700, David Mosberger <[EMAIL PROTECTED]>
>>>>> said:
David> Great, thanks for checking this out. Is there a test video
David> somewhere on the net that I could try?
Actually, after googling around a bit wit
> On Tue, 12 Apr 2005 12:02:28 +0200, Daniel Kobras <[EMAIL PROTECTED]>
> said:
Daniel> Moi David! On Mon, Apr 11, 2005 at 03:07:22PM -0700, David
Daniel> Mosberger wrote:
>> Kino currently crashes right on startup on ia64. The crash is
>> due to one of the infamous "guint inste
Package: kino
Version: 0.75-6
Severity: normal
Tags: patch
Kino currently crashes right on startup on ia64. The crash is due to
one of the infamous "guint instead of GType" bugs and is thus likely
to affect all 64-bit platforms. Attached patch fixes the problem.
I'm unsure if this problem still
> On Fri, 08 Apr 2005 18:38:05 +0900, GOTO Masanori <[EMAIL PROTECTED]>
> said:
GOTO> OK, I pulled the Jakub's patch and built it. I put it at:
GOTO>
GOTO> http://www.gotom.jp/~gotom/debian/glibc/2.3.2.ds1-21_ia64.linuxthreads
GOTO> David, could you test this glibc on your ia64
> On Fri, 08 Apr 2005 15:32:29 +0900, GOTO Masanori <[EMAIL PROTECTED]>
> said:
GOTO> I fear to change this interface until sarge release because there
GOTO> might be another packages that uses sched_setaffinity.
Well, yes, schedutils probably would need updating. I don't know of
an
Package: mozilla-firefox
Version: 1.0.2-1
Severity: normal
Tags: patch
See this upstream bugreport for details:
https://bugzilla.mozilla.org/show_bug.cgi?id=289394
For convenience, I'm attaching the patch to fix this problem below.
Thanks.
-- System Information:
Debian Release: 3.1
APT pre
Package: mozilla-firefox
Version: 1.0.2-1
Severity: normal
Tags: patch
Firefox often crashes when running update-mozilla-firefox-chrome (or
mozilla-firefox -register). In particular, this currently happens
reliably after installing any locale package (such as
mozilla-firefox-locale-de-de). The
> On Wed, 6 Apr 2005 16:40:38 -0400, Eric Dorland <[EMAIL PROTECTED]> said:
Eric> Err, I haven't done it yet, 1.0.2-3 doesn't contain the fix, I
Eric> uploaded him last night, but 1.0.2-4 or 1.0.3-1 will.
I just noticed my patch was slightly off (misplaced closing paren). The
correct pat
>>>>> On Wed, 6 Apr 2005 15:49:38 -0400, Eric Dorland <[EMAIL PROTECTED]> said:
Eric> forwarded 303438
Eric> https://bugzilla.mozilla.org/show_bug.cgi?id=289326 tags
Eric> 303438 + patch thanks
Eric> * David Mosberger ([EMAIL PROTECTED]) wrote:
>>
FYI, I also filed a bugreport with the mozilla project:
https://bugzilla.mozilla.org/show_bug.cgi?id=289326
--david
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Package: mozilla-firefox
Version: 1.0.2-1
Severity: important
Tags: patch
I noticed that both plain and firebird mozilla hang when visiting:
http://maps.google.com
and clicking on the "Satellite" link (top right corner). The problem
appears to be a stupid bug in the fdlibm used by JavaS
> On Tue, 5 Apr 2005 06:33:35 +0200, Christian Perrier <[EMAIL PROTECTED]>
> said:
Christian> tags 113508 unreproducible retitle 113508 [TO CLOSE
Christian> 20050405] login: SEGV at login.c:585 when using rlogin
Christian> thanks
>> I can't reproduce it. As root:
>> # dpkg -s
> On Mon, 4 Apr 2005 23:42:00 +0200, Christian Perrier <[EMAIL PROTECTED]>
> said:
Christian> tags 113508 help thanks
Christian> (please keep [EMAIL PROTECTED] CC'ed to answers)
Christian> Dear ia64 users and maintainers,
Christian> Could you please help in trying to validate t
Package: xpdf-reader
Version: 3.00-13
Severity: normal
File: /usr/bin/xpdf
Tags: patch
xpdf crashes on this (large) PDF document:
http://www.getboulder.com/sports/Master_Bike_map.pdf
The problem appears to be due to JPXStream.cc:2213 which looks like
this:
if (y0+y1 > cb->y0)
Package: gnubiff
Version: 2.1.1-3
Severity: normal
Tags: patch
Gnubiff crashes on startup with a SIGSEGV. The problem is due to
truncation of a 64-bit result to a 32-bit value. This patch
fixes the immediate problem:
--- gnubiff-2.1.1/src/support.cc2005-01-11 00:47:40.0 -0800
+
Package: glibc
Severity: grave
Justification: renders package unusable
The current Sarge glibc still provides the obsolete 2-argument
interface for sched_setaffinity(). As more software is starting to
use this system call, this is becoming a real issue because developers
will have to create a sp
> On Mon, 28 Feb 2005 09:32:20 +0900, GOTO Masanori <[EMAIL PROTECTED]>
> said:
GOTO> David, does this problem only occurs on ia64? Or i686? To be honest,
GOTO> I have concerned this kind of problem - the incompatibility between
GOTO> nptl and linuxthreads.
As indicated in my bug
While there hasn't been any discussion for glibc bugzilla report #685
[1], private communication with one of the glibc maintainers indicates
that this issue is not considered to be a glibc bug because,
officially, glibc supports only one thread library at a time:
LinuxThreads _or_ NPTL, but not bot
> On Thu, 13 Jan 2005 17:14:04 -0700, Bjorn Helgaas <[EMAIL PROTECTED]>
> said:
Bjorn> The domain changes to /proc/bus/pci are implemented for
Bjorn> sparc64, ppc64, ia64, alpha, and mips (see pci_name_bus()).
Bjorn> They aren't all identical (sparc64 uses %04x:%02x always,
Bjorn>
Dann,
Try with this patch applied.
--david
--- mutt_ssl_gnutls.c~ 2005-01-13 23:11:09.254391181 -0800
+++ mutt_ssl_gnutls.c 2005-01-13 23:14:42.486848350 -0800
@@ -456,7 +456,7 @@
char helpstr[SHORT_STRING];
char buf[SHORT_STRING];
char fpbuf[SHORT_STRING];
- int buflen;
+
46 matches
Mail list logo
