Re: memory corruption bug?

[Keith Hopper]

Hi,
     I MAY have found another weird memory problem  in egcs-2.95.69 with a
LIST{ROUT}::append operation - The calling code places a zero word at
%bsp(8) which should be the address of the closure - and, of course, the
program fails woefully with a signal 11.

     This problem does not occur any longer because the LIST should have
been an FLIST - which works properly!

                        Keith

PS A revised 1.3 distribution will be available within 24 hours from our
web-site (http://www.cs.waikato.ac.nz/sather/index.html) and, when Nobbi
has had a chance to look it over it will most likely appear on th Gnu alpha
site.

PPS I have half got Tcl/Tk interim windows working - but the Tk bind
operation appears not to work - and there is a bug in raster_create_raster
C library call.

PPPS Concurrent programs under linux (Slackware) with an unknown signal
(sic!) occurring in pthread_create.

-- 
City Desk
Waikato University
[PGP key available if desired]