[sane-devel] [dev] net backend and timing problems? (pretty weird)

Jochen Eisinger jochen.eisinger@gmx.de
Sat, 08 Sep 2001 20:29:43 +0200

Hash: SHA1


Hum... if you have the time, could you send me the debug output from my
new net.c when crashing due the option changes, and the debug output
from my new sanei_wire with the old net backend (while probably crashing
after a preview)

- -- jochen

mh wrote:
| Jochen Eisinger, Sam, 08 Sep 2001:
|>~From the debug output Michael send me, I assume that the bug isn't in
|>the net backend. The data pipe is closed correctly and the sigpipe
|>occurs on the control pipe. Why this doesn't happen with a delay of some
|>seconds isn't clear to me.
|>Anyways, I'd like to use the new net backend version, because it has
|>much more debug output and fixes some minor bugs. I've tested it,
|>Michael tested it. Could some more people test it before it's committed
|>to cvs?
|>I'm adding debug output to the sanei_wire interface at the moment and
|>already stumbled over a serious bug... in sanei_w_space memcpy is used
|>although the areas copied may overlap (which may result in data
| This should really be tested by others; I just discovered, that your
| version of net.c introduces some new problems (at least for me). When
I try to
| change the scan mode (plustek backend) from e.g. "Color" to "Gray",
xsane and
| xscanimage segfault :-(
| The same happens, if I set the options "Grayify", "Three-Pass
Simulation" or
| "Hand-Scanner Simulation" with the pnm backend.
| For some unknown reason, quiteinsane doesn't crash.
| Michael
| Some reading for this long and rainy weekend follows (xscanimages
reaction after
| trying to change the scanmode :'-( ):
| Program received signal SIGSEGV, Segmentation fault.
| panel_destroy (dialog=0x80b2c30) at gtkglue.c:833
| 833     gtkglue.c: Datei oder Verzeichnis nicht gefunden.
| (gdb) bt
| #0  panel_destroy (dialog=0x80b2c30) at gtkglue.c:833
| #1  0x8058b43 in panel_rebuild (dialog=0x80b2c30) at gtkglue.c:1181
| #2  0x80565fa in set_option (dialog=0x80b2c30, opt_num=2, val=0x80b2690,
|     action=SANE_ACTION_SET_VALUE) at gtkglue.c:201
| #3  0x80573fb in option_menu_callback (widget=0x80b04cc, data=0x80b3228)
|     at gtkglue.c:568
| #4  0x400b3e93 in gtk_marshal_NONE__NONE () at gtkmarshal.c:124
| #5  0x400e4ce8 in gtk_handlers_run (handlers=0x80a29a0, signal=0xbfffefc8,
|     object=0x80b04cc, params=0xbffff01c, after=0) at gtksignal.c:1917
| #6  0x400e40bf in gtk_signal_real_emit (object=0x80b04cc, signal_id=75,
|     params=0xbffff01c) at gtksignal.c:1477
| #7  0x400e2017 in gtk_signal_emit () at gtksignal.c:234
| #8  0x4011aacf in gtk_widget_activate () at gtkwidget.c:2707
| #9  0x400bbe7c in gtk_menu_shell_activate_item () at gtkmenushell.c:771
| _______________________________________________
| Sane-devel mailing list
| Sane-devel@www.mostang.com
| http://www.mostang.com/mailman/listinfo/sane-devel

- --
~  [This is a signature virus, please copy me into your signature file!]

~  GnuGP public key for jochen.eisinger@gmx.de:
~      http://home.nexgo.de/jochen.eisinger/pubkey.asc
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org