On Aug 20, 2007, at 7:22 PM, chromatic wrote:
On Monday 20 August 2007 16:08:42 Will Coleda wrote:
I swear tcl was working this weekend...
with a realclean on r20556
../../parrot tcl.pbc t/cmd_after.t
dies with:
Program received signal EXC_BAD_ACCESS, Could not access memory.
Reason: KERN_INVALID_ADDRESS at address: 0x00ec1894
0x000951b0 in ascii_compare (interp=0x31003b0, lhs=0x356ac94,
rhs=0x1897ad8) at src/charset/ascii.c:328
328 const int ret_val = memcmp(lhs->strstart, rhs-
strstart, min_len);
That looks like the constant string versus COW problem again, but I
can't
reproduce it. Can you pinpoint the revision at which it failed?
(I know a
good CPAN module....)
-- c
More like a mediocre CPAN module. It has an annoying habit of
assuming the test you craft will pass in one chunk of revisions, and
fail in the other; turns out this was assumption was borked by the
float comparison patch which horked the build on OSX x86 in the
middle of a set of passing tcl revisions, throwing me off the scent.
The tcl tests work in 20720 and segfault in 20721
http://perlsix.org/svn/parrot/revision/?rev=20721
If I revert this change against head (r20749), tcl is happy again.
That patch was for RT#44775, which was to fix perl6.
--
Will "Coke" Coleda
[EMAIL PROTECTED]
