bugGNUstep - Bugs: bug #25064, seg.fault on FT_New_Size when...

Group
 
 

bug #25064: seg.fault on FT_New_Size when using art back

Submitter:  Mauro Panigada <shintakezou>
Submitted:  Wed 10 Dec 2008 05:11:18 PM UTC
   
 
Category:  Backend Severity:  3 - Normal
Item Group:  Bug Status:  None
Privacy:  Public Assigned to:  None
Open/Closed:  Open
* Mandatory Fields

Add a New Comment Rich Markup
   

Jump to the original submission

Mon 15 Dec 2008 10:52:22 PM UTC, comment #7: 

ok thanks anyway. hopely it is just a problem with my machine:)

Mauro Panigada <shintakezou>
Mon 15 Dec 2008 07:38:15 PM UTC, comment #6: 

I looked through your configuration and make files and cannot see any problem there. Sorry, but now I am totally clueless.
No idea, where this problem might come from.

Fred Kiefer <FredKiefer>
Group Member
Sat 13 Dec 2008 06:05:52 PM UTC, comment #5: 

sorry i don't know what i've sent instead of makerun2 :)


(file #17051)

Mauro Panigada <shintakezou>
Sat 13 Dec 2008 06:01:06 PM UTC, comment #4: 

I add the output of the ./configure --enable-graphics=art and the stdout and stderr of the make run.


(file #17047, file #17048, file #17049)

Mauro Panigada <shintakezou>
Sat 13 Dec 2008 03:19:34 PM UTC, comment #3: 

As for the cairo problem you see (black bar), this is caused by an old cairo bug that GNUstep had to work around. After cairo fixed that problem in 1.6.0 I adopted GNUstep back to that but this correction was itself corrected immediately after the 0.14 release. You will have to wait for the next GNUstep release to see that fix or switch over to an SVN version of GNUstep.

The output you get when running the art application is really not very helpful. Did you get any useful output when configuring or compiling the art backend? Although things are working for you now, I would still prefer to find the root cause of the issue to prevent others from getting it.

Fred Kiefer <FredKiefer>
Group Member
Thu 11 Dec 2008 12:12:45 AM UTC, comment #2: 

the freetype lib has 6.3.16 (system is GNU/Linux mandriva 2008.0, but after this bug, I've updated freetype, cairo, xft, glibc and some more thing I can't remember ---it did it automatically---, from mandriva 2008.1); but .la file says:

# Version information for libfreetype.
current=9
age=3
revision=16

is this an issue? (current!=6 ?i suppose i should study better what that numbers exactly mean for libtool)

all the output is almost nothing (see attached file); I suppose XIM is not an issue.

Since the previous post, I've did some more test and noticed that even though cairo backend works, it seems not to compute rightly something and the lower part of the window is a dark stripe (to me, it seems it does not consider the title bar, but shouldn't this be a window manager task?)

As last resort I've tried the xlib backend, and it both works and (until now:D) has no graphical issue (but I had a seg. fault testing one of the app in the gnustep-example package)

One thing per step. The application is rather small, as said before I've taken it from gnustep.it, but I attach it too and another app source I've tried and that rise the same problem (access the alert using NSAlert)

(file #17024, file #17025, file #17026)

Mauro Panigada <shintakezou>
Wed 10 Dec 2008 08:30:56 PM UTC, comment #1: 

Thank you for this bug report.

The art backend is still the default backend for GNUstep on X, therfore it will get used even when cairo is available.

The bug you got with FTFontInfo is rather unusual, we will need more information to track it down. Which operation system are you running and what version of FreeType is used? Also it would help to get any output the application made prior to failing. (I expect something like "FTC_Manager_New failed")

Fred Kiefer <FredKiefer>
Group Member
Wed 10 Dec 2008 05:11:18 PM UTC, original submission:  

I am new to all this stuff, hope It's ok this kind of bug report (I've searche for problems with font, found things that seemed different to this issue)...

Today I've downloaded from http://www.gnustep.org/resources/sources.html all the needed sources and compiled it. Everything ok. So the versions are: back 0.14, base 1.16.5, gui 0.14, make 2.0.6. I've left all configure option to their defaults.

Then I tried to compile the example found here:
http://www.gnustep.it/nicola/Tutorials/WritingMakefiles/index.html

Everything went fine, make did its job...

But when I runned the app, obtained a segmentation fault. Running it with openapp --debug option, I've run it in gdb. The seg.fault was into FT_New_Size of the freetype lib. I thought it was of freetype bug... I updated it ...

To make it shorter, I recompiled several type gnustep-gui and base and back after updating zlib, xft and cairo. Then I decided to read it carefully the configure output untill I noticed that, even though the configure script for gnustep-back recognized the cairo lib, and the end it used art (2.3.19). A look at configure options, and recompiled gnustep-back with --enable-graphics=cairo.

The app runned.

So I suppose there's a bug in the art backend (could it be an art bug?). The backtrace from gdb is attached.

I hope this can help killing a bug.


Mauro Panigada <shintakezou>

 

(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach Files:
   
   
Comment:
   

Attached Files
file #17051:  MAKERUN2 added by shintakezou (236B - application/octet-stream - hopely the real MAKERUN2)
file #17048:  MAKERUN added by shintakezou (3KiB - application/octet-stream - configlog1=stdout of ./configure (stderr was void); makerun is the stdout and makerun2 the stderr of make)
file #17049:  MAKERUN2 added by shintakezou (236B - application/octet-stream - configlog1=stdout of ./configure (stderr was void); makerun is the stdout and makerun2 the stderr of make)
file #17047:  CONFIGLOG1 added by shintakezou (4KiB - application/octet-stream - configlog1=stdout of ./configure (stderr was void); makerun is the stdout and makerun2 the stderr of make)
file #17024:  reportcmd.txt added by shintakezou (215B - text/plain)
file #17025:  gnustepalert.m added by shintakezou (327B - application/octet-stream)
file #17026:  mbox.m added by shintakezou (427B - application/octet-stream)
file #17018:  backtrace.txt added by shintakezou (3KiB - text/plain - backtrace by gdb)

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -email is unavailable- added by FredKiefer (Posted a comment)
  • -email is unavailable- added by shintakezou (Submitted the item)
  •  

    There are 0 votes so far. Votes easily highlight which items people would like to see resolved in priority, independently of the priority of the item set by tracker managers.

    Only logged-in users can vote.

     

    Follow 8 latest changes.

    Date Changed by Updated Field Previous Value => Replaced by
    2008-12-13 shintakezou Attached File- Added MAKERUN2, #17051
    2008-12-13 shintakezou Attached File- Added MAKERUN, #17048
        Attached File- Added MAKERUN2, #17049
    2008-12-13 shintakezou Attached File- Added CONFIGLOG1, #17047
    2008-12-11 shintakezou Attached File- Added reportcmd.txt, #17024
        Attached File- Added gnustepalert.m, #17025
        Attached File- Added mbox.m, #17026
    2008-12-10 shintakezou Attached File- Added backtrace.txt, #17018

    Back to the top

    Powered by Savane 3.13-caa5.
    Corresponding source code