check-g++ fails

[Michael Ritzert]

Hello all,

I can't run make check-g++ due to this:
Native configuration is i686-pc-linux-gnu

=== g++ tests ===

Schedule of variations:
unix

Running target unix
Using /usr/share/dejagnu/baseboards/unix.exp as board description file for target.
Using /usr/share/dejagnu/config/unix.exp as generic interface file for target.
Using /home/ritzert/compile/gcc/gcc-3_0-branch/gcc/gcc/testsuite/config/default.exp as tool-and-target-specific interface file.
Running /home/ritzert/compile/gcc/gcc-3_0-branch/gcc/gcc/testsuite/g++.dg/dg.exp ...
/tests_flags: /tests_flags: Datei oder Verzeichnis nicht gefunden
while executing
"exec sh ${odir_v3}/tests_flags --compiler ${odir_v3} ${sdir_v3}"
(procedure "g++_include_flags" line 21)
invoked from within
"g++_include_flags"
(procedure "g++_init" line 58)
invoked from within
"${tool}_init $test_file_name"
(procedure "runtest" line 19)
invoked from within
"runtest $test_name"
("foreach" body line 42)
invoked from within
"foreach test_name [lsort [find ${dir} *.exp]] {
if { ${test_name} == "" } {
continue
}
# Ignore this one if asked to.
if { ${ignore..."
("foreach" body line 54)
invoked from within
"foreach dir "${test_top_dirs}" {
if { ${dir} != ${srcdir} } {
# Ignore this directory if is a directory to be
# ignored.
if {[info..."
("foreach" body line 121)
invoked from within
"foreach pass $multipass {

# multipass_name is set for `record_test' to use (see framework.exp).
if { [lindex $pass 0] != "" } {
set multipass_..."
("foreach" body line 51)
invoked from within
"foreach current_target $target_list {
verbose "target is $current_target";
set current_target_name $current_target;
set tlist [split $curr..."
(file "/usr/share/dejagnu/runtest.exp" line 1586)
make: [check-g++] Fehler 1 (ignoriert)

This is on both mainline and branch. However I've seen on gcc-testresults
that other systems do not experience this.
What's going on here? These are two distinct systems and I delete the
complete object tree before each build.
I can't find a file named tests_flags anywhere in my source and object
trees. Should it be there? I always use cvs update -d -P.

This happened for the first time 2001-04-26 midnight CEST. 24h ago it was
Ok.

Regards,
Michael

[Michael Ritzert]

On Fri, Apr 27, 2001 at 01:24:45PM +0200, Michael Ritzert wrote:
> The second bootstrap will take longer because this is a slower machine and
> the error happens later in the bootstrap.

It just died with:
stage2/xgcc -Bstage2/ -B/opt/gcc/i686-pc-linux-gnu/bin/ -c -DIN_GCC -g -O2 -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wtraditional -pedantic -Wno-long-long -DHAVE_CONFIG_H -I. -I. -I../../gcc/gcc -I../../gcc/gcc/. -I../../gcc/gcc/config -I../../gcc/gcc/../include ../../gcc/gcc/combine.c -o combine.o
../../gcc/gcc/combine.c: In function `try_combine':
../../gcc/gcc/combine.c:2786: Internal error: Speicherzugriffsfehler
Please submit a full bug report, with preprocessed source if appropriate.
See <URL:http://www.gnu.org/software/gcc/bugs.html> for instructions.
make[2]: *** [combine.o] Fehler 1

So no luck here either...

Regards,
Michael

[Michael Ritzert]

On Fri, Apr 27, 2001 at 12:12:32PM -0400, Phil Edwards wrote:

> On Fri, Apr 27, 2001 at 12:09:11PM +0200, Michael Ritzert wrote:
> > /tests_flags: /tests_flags: Datei oder Verzeichnis nicht gefunden
>

> This looks like the libstdc++-v3 'configure' did not completely run (or
> maybe it never got started). Is there a libstdc++-v3 directory under your
> i?86-* objdir directory? If so, the last few lines of config.log should
> help point to the error.

You are perfectly right, cf. my other mails. I just didn't notice early
enough because check-gcc actually worked and blurred my sight.
At the moment the situation is that I ran a complete g++ on the 3.0 branch
but with my old "stage1/xgcc not found" bug on a second machine.
Maybe I can figure out what's causing it this time, as I now do daily
builds and have a more limited timeframe to search for changes.

Regards,
Michael