Re: new branch in gm2 git repro for 10.0.2

[Gaius Mulley]

Dragiša Durić <dragisha@m3w.org> writes:

> Hi Gaius,
>
> Not much time for Modula-2, but still problems... I build from
> floopsie, changes from Aug 23 included.
>
> Thread 1 "a.out" received signal SIGSEGV, Segmentation fault.
> findChildAndParent (t=t@entry=0x70e760, a=a@entry=0x73af60, 
> child=@0x7fffffffdeb8: 0xc400c400c400c4, parent=@0x7fffffffdeb0: 0x73af30)
>     at 
> ../../../../../../gm2-floppsie/libgm2/libm2iso/../../gcc/m2/gm2-libs-iso/RTentity.mod:256
> 256         UNTIL (child = NIL) OR (a = child^.entity)
> Warning: the current language does not match this frame.
> (gdb) bt
> #0  findChildAndParent (t=t@entry=0x70e760, a=a@entry=0x73af60, 
> child=@0x7fffffffdeb8: 0xc400c400c400c4, parent=@0x7fffffffdeb0: 0x73af30)
>     at 
> ../../../../../../gm2-floppsie/libgm2/libm2iso/../../gcc/m2/gm2-libs-iso/RTentity.mod:256
> #1  0x0000000000475aee in RTentity_PutKey (g=0x70e760, a=0x73af60, key=168) 
> at 
> ../../../../../../gm2-floppsie/libgm2/libm2iso/../../gcc/m2/gm2-libs-iso/RTentity.mod:110
> #2  0x000000000044e2e5 in Windows_Define (u=@0x6e38c8: 0x73af60, params=0) at 
> Windows.mod:837
> #3  0x0000000000471751 in AdHoc_Init (getyear=1, top=1) at AdHoc.mod:87
> #4  0x000000000047224d in _M2_test5_init () at test5.mod:41
> #5  0x0000000000406116 in init (argc=1, argv=0x7fffffffe128) at 
> a-test5_m2.cpp:285
> #6  0x0000000000406371 in main (argc=1, argv=0x7fffffffe128) at 
> a-test5_m2.cpp:396
>
>
> Relevant code is a lot of it, I will try to make smaller example with
> reproducible problem, as soon as I can.
>
> Regards,
> Dragisa

Hi Dragisa,

thanks for the latest stack trace - will examine this in detail later
on.  This is very strange as all RTentity exported procedures are
protected by a mutex semaphore - unless there is a typo occurring during
initialisation (or an incorrect initialisation order possibly),