On 7 Jul 1999, Jan Vroonhof wrote:
Oscar Figueiredo <oscar(a)xemacs.org> writes:
> #0 0x403033e5 in memmove (dest=0x836b949, src=0x836b948, len=4294967169)
> at ../sysdeps/generic/memmove.c:98
Note that len=(unsigned int)-127
I know, but couldn't be bothered to investigate, as I had to do some other
work :-)
Note that this should be this call
memmove (addroff + delta, addroff,
/* +1 due to zero-termination. */
string_length (s) + 1 - pos);
> #1 0x807ed2d in resize_string (s=0x839906c, pos=128, delta=1)
> at /home/tigger3/ge204/src/xemacs-21.1.3/src/alloc.c:2261
-127 = 0 + 1 -128
so I'm not sure this an optimization bug. Question: Why is this case
triggering anyway? 128*4 = 512 << 8192 - \epsilon
dunno.
well,
(gdb) p oldfullsize
$7 = 8
(gdb) p newfullsize
$8 = -129
looks strange already -- I haven't really looked at the code to understand
what's going on. I hate all those damned macros! I can have a look
tomorrow afternoon...
Gunnar