457850 - Firefox freezes when trying to print a document

Status: RESOLVED INCOMPLETE

(Firefox :: General, defect)

Description

[info]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.0.3) Gecko/2008092417 Firefox/3.0.3
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.0.3) Gecko/2008092417 Firefox/3.0.3

In any page at any time, firefox will freeze when trying to print a document. This occurs when going through the file tab and print option as well as html link on web pages for print. Does not eventually come out of freeze firefox must be restarted and unable to print this fault is reproduced everytime. This fault has boccured since last update.

Reproducible: Always

Steps to Reproduce:
1. print command
2.
3.
Actual Results:  
no special steps just try to print a document

Expected Results:  
firefox will freeze

bring up the print menu for printing and then print document

Comment 1

[Tom Ressel]

I am experiencing identical symptoms - possibly the same issue. It was impossible to print anything in FF3 3.05 & now updated to 3.06.
I thought I'd fixed it by restarting cups. 

It seems counter-productive to expect users to troubleshoot the most elementary things such as getting an important application to print.

Firstly, Thunderbird 2.0.0.19 (20081227) works absolutely fine.

Other versions available to me:
Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.9.0.6) Gecko/2009011913 Firefox/3.0.6 which works fine.
Mozilla/5.0 (X11; U; Linux i686; en-GB; rv:1.9.1b2) Gecko/20081201 Firefox/3.1b2 which also works fine.
Mozilla/5.0 (X11; U; Linux x86_64; en-GB; rv:1.9.0.6) Gecko/2009012700 SUSE/3.0.6-1.4 Firefox/3.0 does NOT work!
Firefox simply freezes when prompted to print. That's all there is. 
I have to kill FF.

Comment 2

[yagood]

I can confirm that with the latest Linux trunk, same symptoms.

There's a similar report in Ubuntu's Launchpad:

https://bugs.launchpad.net/ubuntu/+source/firefox/+bug/291073

Removing printing-related preferences doesn't help, at least in my case, though.

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.0.7pre) Gecko/2009021504 GranParadiso/3.0.7pre

Comment 3

[yagood]

I switched to "real" trunk (mozilla-central), same result.

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.2a1pre) Gecko/20090215 Minefield/3.2a1pre

After attaching gdb, I get:

(gdb) bt
#0  0xb80c0424 in __kernel_vsyscall ()
#1  0xb8095e28 in connect () from /lib/libpthread.so.0
#2  0xb01e033f in httpAddrConnect () from /usr/lib/libcups.so.2
#3  0xb01deaa6 in httpReconnect () from /usr/lib/libcups.so.2
#4  0xb01df188 in httpConnectEncrypt () from /usr/lib/libcups.so.2
#5  0xb38318b6 in cups_printer_request_details () from /usr/lib/gtk-2.0/2.10.0/printbackends/libprintbackend-cups.so
#6  0xb712e64e in gtk_printer_request_details () from /usr/lib/libgtk-x11-2.0.so.0
#7  0xb7135da5 in selected_printer_changed () from /usr/lib/libgtk-x11-2.0.so.0
#8  0xb6cdd264 in g_cclosure_marshal_VOID__VOID () from /usr/lib/libgobject-2.0.so.0
#9  0xb6ccfbe3 in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#10 0xb6ce3e5d in signal_emit_unlocked_R () from /usr/lib/libgobject-2.0.so.0
#11 0xb6ce5374 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#12 0xb6ce57d6 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0
#13 0xb708bd05 in _gtk_tree_selection_internal_select_node () from /usr/lib/libgtk-x11-2.0.so.0
#14 0xb708c6d6 in gtk_tree_selection_select_path () from /usr/lib/libgtk-x11-2.0.so.0
#15 0xb708c7b2 in gtk_tree_selection_select_iter () from /usr/lib/libgtk-x11-2.0.so.0
#16 0xb7138014 in printer_added_cb () from /usr/lib/libgtk-x11-2.0.so.0
#17 0xb6cdc76c in g_cclosure_marshal_VOID__OBJECT () from /usr/lib/libgobject-2.0.so.0
#18 0xb6ccfbe3 in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#19 0xb6ce3e5d in signal_emit_unlocked_R () from /usr/lib/libgobject-2.0.so.0
#20 0xb6ce5374 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#21 0xb6ce567d in g_signal_emit_by_name () from /usr/lib/libgobject-2.0.so.0
#22 0xb3833127 in cups_request_printer_list_cb () from /usr/lib/gtk-2.0/2.10.0/printbackends/libprintbackend-cups.so
#23 0xb3831acd in cups_dispatch_watch_dispatch () from /usr/lib/gtk-2.0/2.10.0/printbackends/libprintbackend-cups.so
#24 0xb6c46311 in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0
#25 0xb6c499a3 in g_main_context_iterate () from /usr/lib/libglib-2.0.so.0
#26 0xb6c49b61 in g_main_context_iteration () from /usr/lib/libglib-2.0.so.0
#27 0xb7c20ae0 in ?? () from /stuff/bin/firefox/libxul.so
#28 0x00000000 in ?? ()
(gdb)

Comment 4

[yagood]

Just for the record - I don't have problems with printing from GTK applications (like evince, gedit etc.) other than Firefox.

Comment 5

[yagood]

OK, after further investigation it turned out that there were some weird printers in cups cache so I just removed /var/cache/cups and Firefox printing works again!

Comment 6

[Tom Ressel]

(In reply to comment #5)
> OK, after further investigation it turned out that there were some weird
> printers in cups cache so I just removed /var/cache/cups and Firefox printing
> works again!

Is that all you did?
I have nothing in /cache/cups and still no printing.
I wish I could remember how this started i.e. cups or firefox update.
Can you confirm if you are using KDE or not?
The latest firefox prevents kde printing - perhaps something to do with the issue?

Comment 7

[yagood]

I'm using GNOME, I removed whole /var/cache/cups directory and restarted cups.

I noticed that there were some printers in cups with device URI "file:/dev/null" - seemed suspicious. They even stayed there after I removed /etc/cups/printers.conf. Also I couldn't remove them from cups WWW interface. It turned out they were in the cache all this time. Removing the cache solved my problem entirely.

Comment 8

[Tom Ressel]

(In reply to comment #7)
Thanks, That's really frustrating & lucky you!
This is CUPS 1.3.7 & Mozilla/5.0 (X11; U; Linux x86_64; en-GB; rv:1.9.0.6) Gecko/2009012700
SUSE/3.0.6-1.4 Firefox/3.0 only
I have nothing in the cache at all - although each print job is recorded in /var/spool/cups/ and none print to null. 
/etc/cups/yes/ contains job.cache & ppds.dat and /var/spool/cups/ holds the description files of those that don't print (I think)
It is xxxxing annoying that everything else can print but not FF.
Could it be yet more security (in cups, FF or both) that is turning Linux into something quite impossible for Joe the Plumbers such as myself to use?

Comment 9

[yagood]

Try attaching a debugger (gdb) to it and printing back trace - this way we will know where does it hang in your case.

Launch Firefox, make it freeze, then find the PID of firefox-bin process and run command "gdb --pid=<firefox-bin pid you found>", then type "bt" when in gdb, paste the output.

Comment 10

[Tom Ressel]

Attachment: two gdb outputs

gdp debugger output.

Comment 11

[Tom Ressel]

(In reply to comment #9)

Thanks. I got 2 PID outputs - one is huge, so added as attachment.
correction - once the print pop-up locks you can wait about 10+ minutes and the
window is finally completed. At this point you may notice that the print option
is greyed out for any printer [mine were all fine i.e. no new, null or
extraneous printers]. Pressing the cancel button [rather than closing the
window] returns you to FF which - apart from the print nightmare - works as
fine as it did prior to the print attempt...

Comment 12

[Tom Ressel]

update:
another one has gone down with the same affliction:
Mozilla/5.0 (X11; U; Linux i686; en-GB; rv:1.9.1b2) Gecko/20081201 Firefox/3.1b2 previously reported as working, now freezes when commanded to print. 
It's not been subjected to any fiddling or twiddling before failing [if I remember] I'll try the suggested solutions...

Comment 13

[yagood]

It seems that the freeze you're experiencing has the same cause as mine had - it hangs on HTTP connection trying to get printer details or something.

My other symptom was that I couldn't delete/stop/start/whatever some printers from cups WWW interface. Go to http://localhost:631 and to the "Printers" tab. See if there are some "strange" printers or printers you can't modify.

Comment 14

[Tom Ressel]

(In reply to comment #13)
I think you are absolutely right... and hope you have Cups-PDF installed??

The problem is that cups is a total pain on IA64 opensuse. Due to nutcase bizarro security I can rarely get into localhost:631 and configuring is mostly done using kde tools. There has been no printer change for some time now.. I simply do not have the will to break back into cups again - I have a non-conformist printer which takes about two days to install and configure. For the love of humanity don't make me do this! [nb printing works for every other app]

However, there is another development.. I had mentioned that FF 3.1 (Mozilla/5.0 (X11; U; Linux i686; en-GB; rv:1.9.1b2) Gecko/20081201 Firefox/3.1b2) began to show the same fault. This is on a netbook which has FF2 & FF3 installed. FF2 does not have the same symptoms. I started FF2 and looked at available printers. You are right, there is an additional CUPS-PDF printer. I have my original Cups-PDF printer and a new presumably Mozilla spawned, CUPS-PDF@cups. Printing to Cups-PDF is fine but output stalls if I print to [Mozilla] CUPS-PDF@cups. Interestingly the [fedora based] printer configuration panel did show this new printer but it’s now disappeared although the spooler is still visible and shows two stalled print jobs – one from OpenOffice as a test. I returned to FF3 on the netbook and printing became possible again – the new printer remains visible in cups admin, FF3 & other apps print dialogues. Weird.

Going back to IA64 opensuse and doing the same suggests that the hang/freeze/breakage is for a similar reason except that it does not visibly spawn this new printer. I seem to remember that installing Cups-PDF on IA64 opensuse was a sheer joy in that it was only intensely irritating and mindlessly time-consuming rather than the brain-exploding tear-inducing nightmare normally associated with cups. I’ll look at removing Cups-PDF.

I’m guessing this might have to do with Cups-PDF and think this should be a new bug report as it’s being buried as a windows fault that’s gone.

Comment 15

[Tom Ressel]

I've now done the following: 
removed CUPS entirely. [with the exception of cups-libs]
replaced FF 3.07 with FF 3.1b
& installed seamonkey
Used & created a variety of new or existing profiles

In every circumstance FF3 freezes - even in the absence of CUPS.
seamonkey works fine.

Comment 16

[Tom Ressel]

I'm sorry I wish I could have posted a solution for this.
I spent loads and loads of time on it before I simply gave up.
No-one has posted a solution.
All I have are pointers and clues

Deleting FF settings did NOT work for me - it was a complete waste of time and effort - I wish people did some fact checking before posting time-wastey 'tps' that are merely mindless repeats of solutions that do not work.

I also removed CUPS and every reference to it that I could find.
Nothing worked.
I just had to put up with this incredibly stupid flaw [I'm not sure if it's actually a bug] until it went away at some unknown point for some equally unknown reason.

I never found a solution and the problem was replicated in other installations.
I simply updated Firefox each time there was an update and used Konqueror as the print front end.
At some point the problem went away.
I cannot tell you when as by that time I'd stopped fixing it.

My guess/theory is that this is a result of another security lockdown somewhere else that affects CUPS badly.

If you log in as root and try your current & broken installation you may find that CUPS prints.

This seems to suggest that something else has buggered up permissions yet again.

I'm a 'normal' loonix user - in that I use it for everyday business purposes - I am not designing nuclear launch sites or administering 1500 mail servers - and I have found that the primary cause of pretty much every stop or problem I have come across ended up as the result of some security enhancement.
You know, sometimes people just want to use loonix as a general purpose tool.
fanatical security dogma is killing desktop loonix.

Comment 17

[Tyler Downer [He/Him]]

This bug was originally reported on Firefox 3.0 or older, which is no longer supported and will not be receiving any more updates. I strongly suggest that you update to Firefox 4.0.1 or later, update your plugins (flash, adobe, etc.), and retest in a new profile. If you still see the issue with the updated Firefox, please post here. Otherwise, please close as RESOLVED > WORKSFORME
http://www.mozilla.com
http://support.mozilla.com/kb/Managing+profiles
http://support.mozilla.com/kb/Safe+mode

Comment 18

[Tyler Downer [He/Him]]

No reply, INCOMPLETE. Please retest with Firefox 4 or later and a new profile (http://support.mozilla.com/kb/Managing+profiles). Also, ensure you have the most up to date graphics drivers, operating system updates, and plugin versions (flash, java, etc). If you continue to see this issue with the newest Firefox and a new profile, then please comment on this bug.

Comment 19

[Tom Ressel]

yes.. this one disappeared at some point.