Receiving multiple lines with g_io_channel_read_line
Hi, I've had some sucess using g_io_channel_read_line to receive terminated lines from a fifo formed with a server app i wrote. However i need to receive unknown amounts of bytes with line terminators until a final escape character is received, then act on this packet. Is this possible to do in one callback, or should i be using multiple callbacks and a global gchar to buffer until the escape character - or would it be better to somehow use g_io_channel_read? I've seen a previous post about receiving the characters one at a time, which i could use - but i couldn't get it to work properly and didn't think this would be too efficient. Any ideas? Thanks Andrew ___ gtk-app-devel-list mailing list gtk-app-devel-list@gnome.org http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
Re: Memory allocation in Windows error
edward hage writes: I wrote a GTK-application in Linux and it works fine. That might just be a coincidence, it might *seem* to work fine, with some input data. Have you run it under valgrind on Linux? Or, download an evaluation copy of Purify or BoundsChecker for Windows and run it under that. (Hmm, except that those tools work well only if you build with MSVC, so this probably is not useful.) I get this error mainly when I am loading a file from disk. That could be just coincidental, the actual dynamic memory handling error (overflowing a malloced block, use of freed memory, freeing an already freed pointer, etc) might have happened much earlier, and the crash then happens at some random later point. This is typical for dynamic memory handling errors (also on Unix). because I don't use malloc, correct ? But do you use any of the GLib memory handling functions, like g_malloc(), g_new(), g_free()? --tml ___ gtk-app-devel-list mailing list gtk-app-devel-list@gnome.org http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
Re: Memory allocation in Windows error
Tor Lillqvist wrote: edward hage writes: I wrote a GTK-application in Linux and it works fine. That might just be a coincidence, it might *seem* to work fine, with some input data. Have you run it under valgrind on Linux? I never tried valgrind, I have it on my computer so just used it. I don't understand all the lingo it outputs on the screen, but the moment I pushed the okay button for loading a file indeed more than 30,000 errors, so valgrind just stopped with its output. Okay, I can't blame the program. It's output: ==22808== Conditional jump or move depends on uninitialised value(s) ==22808==at 0x806110D: DetermineSetpointXandV (generate_setpoint.c:295) ==22808==by 0x80507F2: Create_profiles (calculate.c:753) ==22808==by 0x8051068: Calculate_moves_and_forces (calculate.c:850) ==22808==by 0x80546F7: toolbar_start_calculate (menus.c:471) ==22808== ==22808== Invalid write of size 1 ==22808==at 0x8064819: Load_data_from_file (loadcase.c:222) ==22808==by 0x8064C2A: fileopen_callback (loadcase.c:135) ==22808==by 0x406457C0: g_cclosure_marshal_VOID__VOID (gmarshal.c:77) ==22808==by 0x40636DAA: g_closure_invoke (gclosure.c:437) ==22808==Address 0x424A4168 is 0 bytes after a block of size 4 alloc'd ==22808==at 0x40027CB4: malloc (in /usr/lib/valgrind/vgskin_memcheck.so) ==22808==by 0x4069B0B6: g_malloc (gmem.c:136) ==22808==by 0x8064649: Load_data_from_file (loadcase.c:173) ==22808==by 0x8064C2A: fileopen_callback (loadcase.c:135) ==22808== ==22808== Invalid read of size 1 ==22808==at 0x40720AD0: __GI_strtod_l_internal (in /lib/libc.so.6) ==22808==by 0x4071D860: __GI___strtod_internal (in /lib/libc.so.6) ==22808==by 0x80649D1: Load_data_from_file (stdlib.h:311) ==22808==by 0x8064C2A: fileopen_callback (loadcase.c:135) ==22808==Address 0x424A4168 is 0 bytes after a block of size 4 alloc'd ==22808==at 0x40027CB4: malloc (in /usr/lib/valgrind/vgskin_memcheck.so) ==22808==by 0x4069B0B6: g_malloc (gmem.c:136) ==22808==by 0x8064649: Load_data_from_file (loadcase.c:173) ==22808==by 0x8064C2A: fileopen_callback (loadcase.c:135) The Invalid read of size 1 occures 3 times.I try to find out what happens here. I will try to find on the web where I can find a clear tutorial of valgrind (no manpage is supplied). If anybody has suggestions for a tutorial I am interested !! because I don't use malloc, correct ? But do you use any of the GLib memory handling functions, like g_malloc(), g_new(), g_free()? Yes, I use g_malloc for almost every pointer and GArray I define. Also in the function in which I read the file. I know I don't consequently have a g_free for every g_malloc, but I didn't experience any problems with it (or better: I didn't realise the problems I have with it). Thank you for the info. Edward ___ gtk-app-devel-list mailing list gtk-app-devel-list@gnome.org http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
Re: Compile Line for GnomeVFS Programs.
On Mon, 7 Feb 2005, Tony Freeman wrote: Every time I try to compile a program that has a #include gnome.h I run into problems. For example: I have just finished typing in the sample program located at: http://developer.gnome.org/doc/API/2.0/gnome-vfs-2.0/gnome-vfs-first-steps.html Now I want to compile. I finally got the program to compile with this line: gcc `pkg-config --cflags --libs gnome libgnomevfs` gnome-vfs.c -o test Try pkg-config --cflags --libs gnome gnome-vfs-2.0 If this doesn't work, then you don't have the right files installed. Allin Cottrell ___ gtk-app-devel-list mailing list gtk-app-devel-list@gnome.org http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
