Nicholas Clark <[EMAIL PROTECTED]> wrote:
> +$(LIBNCI_SO): $(SRC)/nci_test$(O)
> $(LD) $(LD_SHARED) $(LDFLAGS) \
> $(LD_OUT)$@ $(SRC)/nci_test$(O)
Win32 additionally needs the libnci.def thingy. (Except when we put in
the proposed dll_export macros).
> (gdb) print *obj
> $1 = {
> obj = {
> u = {
> _b = {
> _bufstart = 0x7c0802a6,
^^^^^^^^^^
> _buflen = 3217162232
^^^^^^^^^^^
> },
> },
> flags = 2415984648,
> _pobj_version = 2485256112
^^^^^^^^^^
That's not a 'real' object. It's the outcome of some memory destruction.
Is that before it called into libnci or after?
$ parrot -t xxx.pasm
will it tell you.
> which is now confusing me mightily, because I can change the value of flags
> inside gdb with a set command. Yet attempting to restart the process will
> fail with the EXC_BAD_ACCESS signal straight away.
If ARENA_DOD_FLAGS is on this flag isn't used for "live". See
include/parrot/pobj.h but ...
> Am I confusing what EXC_BAD_ACCESS means? Error is here:
> (gdb) where
> #0 0x0002e354 in pobject_lives (interpreter=0x1000200, obj=0xd13a68) at
> src/dod.c:194
Oops. No ARENA_DOD_FLAGS. But anyway, the object is corrupted.
Which architecture? 64bit?
> Nicholas Clark
leo
