[csw-users] evince-2.24.2 in experimental

[Maciej (Matchek) Blizinski]

I'm taking the discussion back on the users list.

I confirmed that it's enough to load a PDF file and hit refresh
(CTRL+R) a couple times to make evince fail.

On Wed, Mar 10, 2010 at 8:45 AM, Maciej (Matchek) Blizinski
<maciej at opencsw.org> wrote:
> I did now, and I can confirm that it crashes after a few refreshes (in
> my case -- 2).
>
> I've built evince with debugging symbols.  It won't help much if the
> error occurs in a library though.  We might need a version of cairo
> with debugging symbols.

I've installed it, opened a PDF and ran it.  Here's the dbx output:

(dbx) run Portugues\ -\ O\ Alquimista\ -\ Paulo\ Coelho.pdf
Running: evince "Portugues - O Alquimista - Paulo Coelho.pdf"
(process id 7801)

** (evince:7801): WARNING **: Failed to create dbus proxy for
org.gnome.SettingsDaemon: Could not get owner of name
'org.gnome.SettingsDaemon':
no such name

(evince:7801): Gtk-WARNING **: Unable to locate theme engine in
module_path: "xfce",

(evince:7801): Gtk-WARNING **: Unable to locate theme engine in
module_path: "xfce",

(evince:7801): Gtk-WARNING **: Unable to locate theme engine in
module_path: "xfce",
t at 4 (l at 4) signal SEGV (no mapping at the fault address) in
FT_Done_Face at 0xd1eaa64c
0xd1eaa64c: FT_Done_Face+0x0024:        movl     0x00000060(%esi),%ecx
Current function is ev_document_render
  241           retval = iface->render (document, rc);
(dbx) where
current thread: t at 4
  [1] FT_Done_Face(0x10), at 0xd1eaa64c
  [2] 0xd1d19ae3(0x8333b70), at 0xd1d19ae3
  [3] _cairo_user_data_array_set_data(0x833229c, 0xd1d37ba0,
0x8502fb8, 0xd1d19ac0), at 0xd1b3b2de
  [4] cairo_font_face_set_user_data(0x8332290, 0xd1d37ba0, 0x8502fb8,
0xd1d19ac0), at 0xd1b418a7
  [5] CairoFont::create(0x82f5438, 0x824b0e8, 0x8232f00, 0x1), at
0xd1d1a02c
  [6] CairoFontEngine::getFont(0x8426890, 0x82f5438, 0x824b0e8), at
0xd1d1a6b4
  [7] CairoOutputDev::updateFont(0x843d248, 0x85b4548), at 0xd1d1b41a
  [8] Gfx::opShowText(0x8315960, 0xd0b4eac0, 0x1), at 0xd13977ea
  [9] Gfx::go(0x8315960, 0x1), at 0xd1383710
  [10] Gfx::display(0x8315960, 0xd0b4ed44, 0x1), at 0xd1383279
  [11] Page::displaySlice(0x87031a8, 0x843d248, 0xe4000000,
0x4048e1e1, 0xe4000000, 0x4048e1e1, 0x0, 0x0, 0x1, 0x0, 0x0, 0x1a7,
0x223, 0x0, 0x82
67cc0, 0x0, 0x0, 0x0, 0x0, 0x86e97c8), at 0xd13e1c18
  [12] 0xd1d12ed8(0x829d368, 0x0, 0x0, 0x1a7, 0x223, 0x20000000,
0x3fe61e1e, 0x0, 0x0, 0x82aa610, 0x83390d0), at 0xd1d12ed8
  [13] poppler_page_render_to_pixbuf(0x829d368, 0x0, 0x0, 0x1a7,
0x223, 0x20000000, 0x3fe61e1e, 0x0, 0x82aa610), at 0xd1d12fbd
  [14] pdf_page_render(0x829d368, 0x1a7, 0x223, 0x84e8600), at 0xd0a14ff5
  [15] pdf_document_render(0x829d340, 0x84e8600), at 0xd0a15137
=>[16] ev_document_render(document = 0x829d340, rc = 0x84e8600), line
241 in "ev-document.c"
  [17] ev_job_render_run(job = 0x81ac3f0), line 512 in "ev-jobs.c"
  [18] ev_job_run(job = 0x81ac3f0), line 203 in "ev-jobs.c"
  [19] ev_job_thread(job = 0x81ac3f0), line 183 in "ev-job-scheduler.c"
  [20] ev_job_thread_proxy(data = (nil)), line 213 in "ev-job-scheduler.c"
  [21] g_thread_create_proxy(0x81b67e8), at 0xd1a10590
  [22] _thr_setup(0xd0d70a00), at 0xd1953885
  [23] _lwp_start(0x8333b70, 0xd1d35e30, 0xd194f594, 0xd197e590, 0x0,
0xd197e590), at 0xd1953b70
(dbx)


I see poppler functions just above the the evince functions on the
stack.  Our poppler package is two years old.  I think that it's a
good idea to update it, link evince against it, and try again.  If it
still crashes, we can compile poppler and libcairo with debugging
symbols and inspect more.

Maciej