[Ohrrpgce] FreeBASIC Mac OS X build

[Ralph Versteegen]

On 9 October 2010 14:00, James Paige <Bob at hamsterrepublic.com> wrote:
>> Odd... why is the linker complaining? Also I'm surprised that it's
>> symlinked to libSystem. Try
>> file /usr/lib/libgcc_s.1.dylib
>> to see the architectures included, and edit Makefile to add -v to
>> FBFLAGS, and see what it passes to as.
>> A desperate possibility for fix would be adding '-Wl -lgcc_s.10.4' to FBFLAGS
>
> So I got this working with -Wl -lgcc_s.10.4
>
> I created a symlink from ~/fw pointing to ~/Library/Frameworks and I
> used that in my FRAMEWORKS_PATH and that made the command line short
> enough that (freebasic?) didn't truncate it, and the build worked!
>
> That gives me ohrrpgce-game and ohrrpgce-custom, but when I tested
> ohrrpgce-custom by opening Vikings of Midgard, I crashed like this:
>
> broaste:~/src/ohr/wip james$ ./ohrrpgce-custom
> ohrrpgce-custom(49712,0xa0523500) malloc: *** error for object
> 0xbfffcdd0: pointer being freed was not allocated
> *** set a breakpoint in malloc_error_break to debug
> ohrrpgce-custom(49712,0xa0523500) malloc: *** error for object
> 0xbfffcd74: pointer being freed was not allocated
> *** set a breakpoint in malloc_error_break to debug
> ohrrpgce-custom(49712,0xa0523500) malloc: *** error for object
> 0xbfffcdec: pointer being freed was not allocated
> *** set a breakpoint in malloc_error_break to debug
> Segmentation fault
>
> ohrrpgce-game crashes like this:
>
> broaste:~/src/ohr/wip james$ ./ohrrpgce-game
> ohrrpgce-game(49758,0xa0523500) malloc: *** error for object 0x8b04c483:
> pointer being freed was not allocated
> *** set a breakpoint in malloc_error_break to debug
> Bus error
>
> Here is what gdbgame.sh tells me:
>
> broaste:~/src/ohr/wip james$ ./gdbgame.sh
> GNU gdb 6.3.50-20050815 (Apple version gdb-1469) (Wed May  5 04:36:56
> UTC 2010)
> Copyright 2004 Free Software Foundation, Inc.
> GDB is free software, covered by the GNU General Public License, and you
> are
> welcome to change it and/or distribute copies of it under certain
> conditions.
> Type "show copying" to see the conditions.
> There is absolutely no warranty for GDB.  Type "show warranty" for
> details.
> This GDB was configured as "x86_64-apple-darwin"...Reading symbols for
> shared libraries ............ done
>
> warning: Trying to remove a section from the ordered section list that
> did not exist at 0x1d8000.
>
> warning: Trying to remove a section from the ordered section list that
> did not exist at 0x1d9000.
>
> Breakpoint 1 at 0x175441: file ../../libfb_exit.c, line 47.
> Breakpoint 2 at 0x8b60a5
> Breakpoint 3 at 0x9a0674
> Running...
> Reading symbols for shared libraries
> .+++++++++++............................................................................................
> done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> Reading symbols for shared libraries . done
> ohrrpgce-game(49808,0xa0523500) malloc: *** error for object 0x8b04c483:
> pointer being freed was not allocated
> *** set a breakpoint in malloc_error_break to debug
>
> Program received signal EXC_BAD_ACCESS, Could not access memory.
> Reason: KERN_PROTECTION_FAILURE at address: 0x001073dd
> fb_StrDelete (str=0x1073dd) at ../../libfb_str_del.c:51
> 51      ../../libfb_str_del.c: No such file or directory.
>        in ../../libfb_str_del.c
> misc/gdbcmds1.txt:13: Error in sourced command file:
> Invalid type combination in equality test.
> Dumping state
> $1 = 0x0
> $2 = <unknown type>
> $3 = <unknown type>
>  4                                 0x9504e13a in
> semaphore_timedwait_signal_trap ()
>  3                                 0x950739b2 in __workq_kernreturn ()
>  2 "com.apple.libdispatch-manager" 0x95074922 in kevent ()
> * 1 "com.apple.main-thread"         fb_StrDelete (str=0x1073dd) at
> ../../libfb_str_del.c:51
> #0  fb_StrDelete (str=0x1073dd) at ../../libfb_str_del.c:51
> No locals.
> #1  0x00111c75 in Gfx_backend_init ()
> No symbol table info available.
> #2  0x0010111a in FONT_UNLOAD ()
> No symbol table info available.
> #3  0x0010124f in FONT_CREATE_EDGED ()
> No symbol table info available.
> #4  0x00101ebf in SETFONT ()
> No symbol table info available.
> #5  0x0001698c in SDL_main ()
> No symbol table info available.
> #6  0x001aa59c in -[SDLMain applicationDidFinishLaunching:] ()
> No symbol table info available.
> #7  0x9496f1c3 in _nsnote_callback ()
> No symbol table info available.
> #8  0x93f433c3 in __CFXNotificationPost ()
> No symbol table info available.
> #9  0x93f42dca in _CFXNotificationPostNotification ()
> No symbol table info available.
> #10 0x94964090 in -[NSNotificationCenter
> postNotificationName:object:userInfo:] ()
> No symbol table info available.
> #11 0x9497146d in -[NSNotificationCenter postNotificationName:object:]
> ()
> No symbol table info available.
> #12 0x9131670a in -[NSApplication _postDidFinishNotification] ()
> No symbol table info available.
> #13 0x9131661a in -[NSApplication _sendFinishLaunchingNotification] ()
> No symbol table info available.
> #14 0x9146d6a9 in -[NSApplication(NSAppleEventHandling) _handleAEOpen:]
> ()
> No symbol table info available.
> #15 0x9146d2c9 in -[NSApplication(NSAppleEventHandling)
> _handleCoreEvent:withReplyEvent:] ()
> No symbol table info available.
> #16 0x949a4400 in -[NSAppleEventManager
> dispatchRawAppleEvent:withRawReply:handlerRefCon:] ()
> No symbol table info available.
> #17 0x949a41c4 in _NSAppleEventManagerGenericHandler ()
> No symbol table info available.
> #18 0x98a1af58 in aeDispatchAppleEvent ()
> No symbol table info available.
> #19 0x98a1ae57 in dispatchEventAndSendReply ()
> No symbol table info available.
> #20 0x98a1ad61 in aeProcessAppleEvent ()
> No symbol table info available.
> #21 0x97323323 in AEProcessAppleEvent ()
> No symbol table info available.
> #22 0x912e6cc6 in _DPSNextEvent ()
> No symbol table info available.
> #23 0x912e62ca in -[NSApplication
> nextEventMatchingMask:untilDate:inMode:dequeue:] ()
> No symbol table info available.
> #24 0x912a855b in -[NSApplication run] ()
> No symbol table info available.
> #25 0x001aad46 in main ()
> No symbol table info available.
> (gdb)
>
>
> ---
> James

fb_StrDelete's argument is equal to its return address. This is
ambiguous; it could be either stack corruption inside fb_StrDelete
(which I doubt) or bad emitted assembly causing a random jump to
fb_StrDelete without setting up the stack. But I only ever saw this
when cross-compiling with the broken version of gas on a Ubuntu
install, and I provided the known-good gas build you're using. I'm not
going to solve this anytime soon.

Could you please email me ohrrpgce-game, ohrrpgce-custom, and the .o
files? Hopefully I can reproduce it on Monday.

OK, anyone else out there with a Mac willing to try: can you compile and run?