This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
dejagnu 980528 is worse than 971225
- To: manfred at s-direktnet dot de, Manfred dot Hollstein at ks dot sel dot alcatel dot de
- Subject: dejagnu 980528 is worse than 971225
- From: hjl at lucon dot org (H.J. Lu)
- Date: Thu, 11 Jun 1998 20:59:17 -0700 (PDT)
- Cc: egcs at cygnus dot com, law at cygnus dot com
Hi,
I just found out dejagnu 980528 had more bugs than 971225. Fisrt
it still needs my lookfor_dir_with_trigger patch. I guess noone
really understands what is going on here. I have tried to explain it
before. I guess it went nowhere. It is very tricky to reproduce it.
It needs a special build directory structure to see it. dejagnu is too
dumb to figure it out. That is why it needs my lookfor_dir_with_trigger
patch.
With dejagnu 980528, there is a new failure:
gcc.c-torture/special/921210-1.c
It is caused by the change in dejagnu 980528 which uses the TCL
built spawn, which generates an extra '\r' at the end. But
gcc.c-torture/special/special.exp
cannot deal with it. I enclose a trace here. In the meantime, I
recommend to use dejagnu 971225 plus my lookfor_dir_with_trigger patch
for egcs.
H.J.
-----
spawn /usr/src/gcc/egcs/gcc/xgcc -B/usr/src/gcc/egcs/gcc/ /home/work/gnu/src/egcs/gcc/testsuite/gcc.c-torture/special/921210-1.c -E
# 1 "/home/work/gnu/src/egcs/gcc/testsuite/gcc.c-torture/special/921210-1.c"
( a->f +1) +1
output is:
# 1 "/home/work/gnu/src/egcs/gcc/testsuite/gcc.c-torture/special/921210-1.c"
( a->f +1) +1
foobar:# 1 "/home/work/gnu/src/egcs/gcc/testsuite/gcc.c-torture/special/921210-1.c"
( a->f +1) +1
:foobar
foobar:# 1 "/home/work/gnu/src/egcs/gcc/testsuite/gcc.c-torture/special/921210-1.c"
( a->f +1) +1
:foobar
foobar:# 1 "/home/work/gnu/src/egcs/gcc/testsuite/gcc.c-torture/special/921210-1.c"
( a->f +1) +1
:foobar
foobar:{# 1 "/home/work/gnu/src/egcs/gcc/testsuite/gcc.c-torture/special/921210-1.c"} {} {} {} {( a->f +1) +1 } {}:foobar
FAIL: 921210-1.c