FYI, with current rawhide, Adobe Reader (I tried 9.1.2 and 9.1.3) crashes on startup. I sent a crash dump to Adobe and suggested that they might want to try to get it fixed before Fedora 12 is released, but I'm not holding my breath.
jik
On Wed, 2009-08-26 at 02:17 -0400, Jonathan Kamens wrote:
FYI, with current rawhide, Adobe Reader (I tried 9.1.2 and 9.1.3) crashes on startup. I sent a crash dump to Adobe and suggested that they might want to try to get it fixed before Fedora 12 is released, but I'm not holding my breath.
Thanks for the feedback Jonathan. If you have the links and supporting details ... this would make a good candidate for https://fedoraproject.org/wiki/Common_F12_bugs.
Thanks, James
Thanks for the feedback Jonathan. If you have the links and supporting details ... this would make a good candidate for https://fedoraproject.org/wiki/Common_F12_bugs.
I don't have any links because Adobe bug reports are not public.
As for supporting details, I've told you all I know.
I would certainly like somebody else to confirm the issue before it gets documented on the Wiki. After all, it could be some sort of idiosyncrasy on my machine that's causing the issue, rather than something that's going to affect everyone.
Jik
On Wed, 2009-08-26 at 19:40 -0400, Jonathan Kamens wrote:
Thanks for the feedback Jonathan. If you have the links and supporting details ... this would make a good candidate for https://fedoraproject.org/wiki/Common_F12_bugs.
I don't have any links because Adobe bug reports are not public.
As for supporting details, I've told you all I know.
I would certainly like somebody else to confirm the issue before it gets documented on the Wiki. After all, it could be some sort of idiosyncrasy on my machine that's causing the issue, rather than something that's going to affect everyone.
If you showed us some details of the crash, we might be able to tell you...
On 08/26/2009 10:11 PM, Matthias Clasen wrote:
On Wed, 2009-08-26 at 19:40 -0400, Jonathan Kamens wrote:
Thanks for the feedback Jonathan. If you have the links and supporting details ... this would make a good candidate for https://fedoraproject.org/wiki/Common_F12_bugs.
I don't have any links because Adobe bug reports are not public.
As for supporting details, I've told you all I know.
I would certainly like somebody else to confirm the issue before it gets documented on the Wiki. After all, it could be some sort of idiosyncrasy on my machine that's causing the issue, rather than something that's going to affect everyone.
If you showed us some details of the crash, we might be able to tell you..
If you run Adobe Reader, it crashes right after putting up its window.
That's enough detail for anyone to be able to attempt to reproduce the problem.
If you /can/ reproduce it, then you the details of the crash are almost certainly not going to help you fix it, because there's no debuginfo and Adobe Reader is closed-source. Sure, it's possible that a sufficiently knowledgeable developer would be able to tell from the crash details what's going wrong in some open-source Fedora component and fix it, but that seems unlikely, and furthermore, if the crash is really caused by a bug in an open-source Fedora component, it seems likely that it'd be affecting at least one other thing besides Reader, in which case it would be easier to debug and fix the issue through that other thing rather than trying to debug through a closed-source application with no debuginfo.
If you /can't/ reproduce the issue, then again, the details of the crash from my end are not going to be particularly useful.
This is why I did not bother to post the crash details here.
Having said all of that, you seem to think that they might be useful, so although I disagree, I will accede to your request:
/usr/bin/acroread [0x84fc739] [@0x8048000] (__kernel_sigreturn+0x0) [0xed5400] [@0xed5000] /usr/bin/acroread [0x85094c8] [@0x8048000] /usr/bin/acroread [0x850956e] [@0x8048000] /usr/bin/acroread [0x822d0c4] [@0x8048000] /usr/bin/acroread [0x850e36d] [@0x8048000] /usr/bin/acroread [0x850fab1] [@0x8048000] /usr/lib/libgtk-x11-2.0.so.0 [0x25370a8] [@0x23e7000] /lib/libgobject-2.0.so.0(g_closure_invoke+0x1b3) [0x445643] [@0x43c000] /lib/libgobject-2.0.so.0 [0x45c4c0] [@0x43c000] /lib/libgobject-2.0.so.0(g_signal_emit_valist+0x678) [0x45d758] [@0x43c000] /lib/libgobject-2.0.so.0(g_signal_emit+0x27) [0x45dd57] [@0x43c000] /usr/lib/libgtk-x11-2.0.so.0 [0x266ee1e] [@0x23e7000] /usr/lib/libgtk-x11-2.0.so.0(gtk_main_do_event+0x3c6) [0x252f956] [@0x23e7000] /usr/lib/libgdk-x11-2.0.so.0 [0x2903187] [@0x28c3000] /usr/lib/libgdk-x11-2.0.so.0 [0x292c4a5] [@0x28c3000] /usr/lib/libgdk-x11-2.0.so.0 [0x28fa53e] [@0x28c3000] /usr/lib/libgdk-x11-2.0.so.0(gdk_window_process_all_updates+0x137) [0x28fe3c7] [@0x28c3000] /usr/lib/libgtk-x11-2.0.so.0 [0x24907f1] [@0x23e7000] /usr/lib/libgdk-x11-2.0.so.0 [0x28d7498] [@0x28c3000] /lib/libglib-2.0.so.0 [0x7a6292] [@0x771000] /lib/libglib-2.0.so.0(g_main_context_dispatch+0x1f8) [0x7a80a8] [@0x771000] /lib/libglib-2.0.so.0 [0x7ab9d8] [@0x771000] /lib/libglib-2.0.so.0(g_main_loop_run+0x1bf) [0x7abe4f] [@0x771000] /usr/lib/libgtk-x11-2.0.so.0(gtk_main+0xb9) [0x252fdd9] [@0x23e7000] /usr/bin/acroread [0x84ffb71] [@0x8048000] /usr/bin/acroread(main+0x87) [0x8563dcd] [@0x8048000] /lib/libc.so.6(__libc_start_main+0xe6) [0x1739b56] [@0x1723000]
I am getting the same problem, running acroread from the command line shows that the following files are missing :
libatk-bridge.so libpk-gtk-module.so libcanberra-gtk-module.so libgnomebreakpad.so
I know that libatk-bridge.so is provided by the ati-spi package which was updated very recently, the others I have not checked. One could therefore argue that this is not an Adobe problem as it was working perfectly OK until recent updates.
Hope the additional information helps
Matthias Clasen wrote:
On Wed, 2009-08-26 at 19:40 -0400, Jonathan Kamens wrote:
Thanks for the feedback Jonathan. If you have the links and supporting details ... this would make a good candidate for https://fedoraproject.org/wiki/Common_F12_bugs.
I don't have any links because Adobe bug reports are not public.
As for supporting details, I've told you all I know.
I would certainly like somebody else to confirm the issue before it gets documented on the Wiki. After all, it could be some sort of idiosyncrasy on my machine that's causing the issue, rather than something that's going to affect everyone.
If you showed us some details of the crash, we might be able to tell you...
-- fedora-test-list mailing list fedora-test-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list
spoffley (spoffley@gmail.com) said:
libatk-bridge.so libpk-gtk-module.so libcanberra-gtk-module.so libgnomebreakpad.so
I know that libatk-bridge.so is provided by the ati-spi package which was updated very recently, the others I have not checked. One could therefore argue that this is not an Adobe problem as it was working perfectly OK until recent updates.
This just sounds like you're on x86_64 and you don't have the same set of GTK modules for 32-bit x86 installed.
Bill
That is indeed the problem, but Adobe Reader was working fine at the end of last week so something changed somewhere to stop it working now, I have never needed those 32 bit libraries in the past. Well change is good so they say :-)
Bill Nottingham wrote:
spoffley (spoffley@gmail.com) said:
libatk-bridge.so libpk-gtk-module.so libcanberra-gtk-module.so libgnomebreakpad.so
I know that libatk-bridge.so is provided by the ati-spi package which was updated very recently, the others I have not checked. One could therefore argue that this is not an Adobe problem as it was working perfectly OK until recent updates.
This just sounds like you're on x86_64 and you don't have the same set of GTK modules for 32-bit x86 installed.
Bill
-- fedora-test-list mailing list fedora-test-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-test-list
On Thu, 2009-08-27 at 09:26 -0700, spoffley wrote:
That is indeed the problem, but Adobe Reader was working fine at the end of last week so something changed somewhere to stop it working now, I have never needed those 32 bit libraries in the past. Well change is good so they say :-)
I'm not convinced those messages are the problem. I've seen similar messages launching apps from a console recently (the exact circumstances escape me), but the apps launched anyway.
On Wed, Aug 26, 2009 at 10:11:42PM -0400, Matthias Clasen wrote:
If you showed us some details of the crash, we might be able to tell you...
By editing 'acroread' script to run binaries under gdb you can find the following:
Program received signal SIGSEGV, Segmentation fault. 0xf67c55b9 in gdk_window_get_geometry () from /usr/lib/libgdk-x11-2.0.so.0 (gdb) where #0 0xf67c55b9 in gdk_window_get_geometry () from /usr/lib/libgdk-x11-2.0.so.0 #1 0x084f4219 in _start ()
This is with gtk2-2.17.9-1.fc12.i686 and does not happen with gtk2-2.16.5-1.fc11.i586 on Fedora 11. I am afraid that in this moment I cannot look at that deeper. Maybe later if somebody is not going to beat me to it.
Another possible problem is that for x86_64 Fedora 11, and earlier, supply gtk2-engines.i?86 and in rawhide repositories you can find only gtk2-engines.x86_64. Acroread appears to be using that even if that is not made explicit in rpm packages from Adobe. No idea if this is really critical. It surely make for tons of complaints.
Michal
On Thu, 2009-08-27 at 11:05 -0600, Michal Jaegermann wrote:
On Wed, Aug 26, 2009 at 10:11:42PM -0400, Matthias Clasen wrote:
If you showed us some details of the crash, we might be able to tell you...
By editing 'acroread' script to run binaries under gdb you can find the following:
Program received signal SIGSEGV, Segmentation fault. 0xf67c55b9 in gdk_window_get_geometry () from /usr/lib/libgdk-x11-2.0.so.0 (gdb) where #0 0xf67c55b9 in gdk_window_get_geometry () from /usr/lib/libgdk-x11-2.0.so.0 #1 0x084f4219 in _start ()
This is with gtk2-2.17.9-1.fc12.i686 and does not happen with gtk2-2.16.5-1.fc11.i586 on Fedora 11. I am afraid that in this moment I cannot look at that deeper. Maybe later if somebody is not going to beat me to it.
Another possible problem is that for x86_64 Fedora 11, and earlier, supply gtk2-engines.i?86 and in rawhide repositories you can find only gtk2-engines.x86_64. Acroread appears to be using that even if that is not made explicit in rpm packages from Adobe. No idea if this is really critical. It surely make for tons of complaints.
Michal
If you have GTK+ 2.17.9, does running
GDK_NATIVE_WINDOWS=1 acroread
work any better ?
On Thu, Aug 27, 2009 at 01:11:34PM -0400, Matthias Clasen wrote:
On Thu, 2009-08-27 at 11:05 -0600, Michal Jaegermann wrote:
On Wed, Aug 26, 2009 at 10:11:42PM -0400, Matthias Clasen wrote:
If you showed us some details of the crash, we might be able to tell you...
By editing 'acroread' script to run binaries under gdb you can find the following:
Program received signal SIGSEGV, Segmentation fault. 0xf67c55b9 in gdk_window_get_geometry () from /usr/lib/libgdk-x11-2.0.so.0 (gdb) where #0 0xf67c55b9 in gdk_window_get_geometry () from /usr/lib/libgdk-x11-2.0.so.0 #1 0x084f4219 in _start ()
If you have GTK+ 2.17.9, does running
GDK_NATIVE_WINDOWS=1 acroread
work any better ?
Yes, indeed, it does. Also editing 'acroread' script to put GDK_NATIVE_WINDOWS=1 in front of 'exec' in a LaunchBinary function provides, in this moment, a workaround.
Missing on x86_64 gtk2-engines.i686 does not seem to be critical but that surely generates a lot and lot of spillage like that:
(acroread:2509): Gtk-WARNING **: Unable to locate theme engine in module_path: "clearlooks",
which does go away after gtk2-engines.i686 package is installed.
Michal
On Thu, Aug 27, 2009 at 01:11:34PM -0400, Matthias Clasen wrote:
On Thu, 2009-08-27 at 11:05 -0600, Michal Jaegermann wrote:
On Wed, Aug 26, 2009 at 10:11:42PM -0400, Matthias Clasen wrote:
If you showed us some details of the crash, we might be able to tell you...
Program received signal SIGSEGV, Segmentation fault. 0xf67c55b9 in gdk_window_get_geometry () from /usr/lib/libgdk-x11-2.0.so.0
If you have GTK+ 2.17.9, does running
GDK_NATIVE_WINDOWS=1 acroread
work any better ?
AFAICT after an update to gtk2-2.17.10-1.fc12.i686 there is no more need for GDK_NATIVE_WINDOWS=1. Thanks!
Michal
-
Adam Williamson -
Bill Nottingham -
James Laska -
Jonathan Kamens -
Matthias Clasen -
Michal Jaegermann -
spoffley