Re: [PATCH] libgst: Add built-in for Behavior>>#new and >>#new:

On Sun, Aug 03, 2014 at 10:56:19AM +0200, Holger Hans Peter Freyther wrote:

Dear Paolo,

> Okay, before things go wrong Behavior_newInitialize is called while the
> processor scheduler is being created.

valgrind is giving me these with the plain JIT (without my patch
as far as I can see)



==13940== Invalid read of size 4
==13940==    at 0x40CE5E6: sync_wait_process (interp.c:1697)
==13940==    by 0x884D7FF: ???
==13940==    by 0x40CF6AE: _gst_nvmsg_send (interp.c:2318)
==13940==    by 0x404B069: _gst_execute_statements (comp.c:588)
==13940==    by 0x403D3C4: execute_doit (gst-parse.c:592)
==13940==    by 0x403DBFA: make_attribute (gst-parse.c:801)
==13940==    by 0x403F2B6: parse_attribute (gst-parse.c:1497)
==13940==    by 0x403F126: parse_attributes (gst-parse.c:1461)
==13940==    by 0x403ED35: parse_method (gst-parse.c:1333)
==13940==    by 0x403E090: parse_class_definition (gst-parse.c:907)
==13940==    by 0x403D885: parse_scoped_definition (gst-parse.c:718)
==13940==    by 0x403D51E: parse_doit (gst-parse.c:624)


Program received signal SIGTRAP, Trace/breakpoint trap.
sync_wait_process (semaphoreOOP=0x884e460, processOOP=0x410bb64) at interp.c:1697
1697	  if (TO_INT (sem->signals) <= 0)
(gdb) bt
#0  sync_wait_process (semaphoreOOP=0x884e460, processOOP=0x410bb64) at interp.c:1697
#1  0x0884d800 in ?? ()
#2  0x040cf5b4 in _gst_nvmsg_send (receiver=0x884d800, sendSelector=0x8852fe8, args=0x0, 
    sendArgs=0) at interp.c:2318
#3  0x0404b052 in _gst_execute_statements (receiverOOP=0x884d800, method=0x480c1b0, 
    undeclared=false, quiet=true) at comp.c:588
#4  0x0403d3c5 in execute_doit (p=0xbecae4cc, temps=0x0, stmts=0x480c174, 
    receiverOOP=0x884d800, undeclared=false, quiet=true) at gst-parse.c:592
#5  0x0403dbfb in make_attribute (p=0xbecae4cc, classOOP=0x884d800, 
    attribute_keywords=0x480c138) at gst-parse.c:801
#6  0x0403f2b7 in parse_attribute (p=0xbecae4cc) at gst-parse.c:1497
#7  0x0403f127 in parse_attributes (p=0xbecae4cc, prev_attrs=0x480c090) at gst-parse.c:1461
#8  0x0403ed36 in parse_method (p=0xbecae4cc, at_end=93) at gst-parse.c:1333
#9  0x0403e091 in parse_class_definition (p=0xbecae4cc, classOOP=0x884e460, extend=true)
    at gst-parse.c:907
#10 0x0403d886 in parse_scoped_definition (p=0xbecae4cc, first_stmt=0x480be00)
    at gst-parse.c:718
#11 0x0403d51f in parse_doit (p=0xbecae4cc, fail_at_eof=false) at gst-parse.c:624
#12 0x0403cd07 in parse_chunks (p=0xbecae4cc) at gst-parse.c:475
#13 0x0403cbfd in _gst_parse_chunks (currentNamespace=0x0) at gst-parse.c:449
#14 0x04042df0 in _gst_parse_stream (currentNamespace=0x0) at lex.c:1209
#15 0x04087987 in _gst_process_file (fileName=0x40e7000 <standard_files> "Builtins.st", 
    dir=GST_DIR_KERNEL) at input.c:855
#16 0x0403bb7d in load_standard_files () at files.c:582
#17 0x0403b9c2 in _gst_initialize (
    kernel_dir=0x448dc10 "/home/ich/source/smalltalk/smalltalk-jit/kernel", 
    image_file=0x40e7696 "gst.im", flags=11) at files.c:524
#18 0x0403acfb in gst_initialize (kernel_dir=0x0, image_file=0x0, flags=3) at gstpub.c:156
#19 0x0804929b in main (argc=2, argv=0xbecae844) at main.c:386
(gdb) c
Continuing.

so this jump just doesn't make any sense. Would you have some time to
go through this with me? :}