-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
According to Elbert Pol on 4/10/2008 2:49 PM:
| Hi Eric,
|
| I attach the check.log it fails :(
I guessed it might, since Bruno's patch for EMX was preliminary.
At any rate, here's the relevant gnulib portions:
Killed by SIGABRT
pid=0x038a ppid=0x0389 tid=0x0001 slot=0x00f5 pri=0x0200 mc=0x0001
U:\TEST\TESTS\TEST-FREADPTR.EXE
Process dumping was disabled, use DUMPPROC / PROCDUMP to enable it.
FAIL: test-freadptr.sh
Yuck - this doesn't say what line number the failure was at. Bruno, could
we replace all the abort()s in the various tests with exit(134), the same
exit status as SIGABRT but which allows stderr to flush on OS/2 and mingw?
~ True, it means you won't get a core file, so debugging a failure is a bit
harder, but I often use ulimit to avoid core files anyway, and it's not
too hard to restart a program.
Elbert, do you have a working debugger, such as gdb, available on your
machine, to single step through the tests to find which line is failing?
If not, can you edit all of the tests/test-*.c files and replace abort()
with exit(134), then rerun the testsuite? Knowing the line numbers of the
failures will help in improving the code.
More failures without line numbers:
Killed by SIGABRT
FAIL: test-freadptr2.sh
Killed by SIGABRT
FAIL: test-freadseek.sh
Killed by SIGFPE
FAIL: test-frexp.exe
Killed by SIGFPE
FAIL: test-frexpl.exe
Killed by SIGFPE
FAIL: test-printf-frexp.exe
Killed by SIGFPE
FAIL: test-printf-frexpl.exe
Killed by SIGFPE
FAIL: test-signbit.exe
Two categories of failures - we aren't handling double and long double
correctly in the printf replacements, and we aren't handling ungetc in the
read-ahead buffer of the stdio enhancements. Both Bruno's areas of expertise.
Now on to the m4 testsuite failures (replies specific to this half of the
email can drop bug-gnulib)
./003.syntax: stdout mismatch
- --- m4-tmp.1115/m4-xout 2008-04-10 22:41:46 +0200
+++ m4-tmp.1115/m4-out 2008-04-10 22:41:46 +0200
@@ -6,7 +6,7 @@
~ builtin: 3
~ defn: odd name: $1
~ divert: [EMAIL PROTECTED]
- -esyscmd: [EMAIL PROTECTED]
+esyscmd: [\0]
OK, a testsuite bug - I assumed that 'printf "[\0]"' is portable, but on
your platform, it prints literal \0 rather than the correct NUL byte. But
not a bug in m4. I might be able to work around it by avoiding that test
if __unix__ is not defined. Or come up with an alternative way to
generate a NUL byte.
@ ../doc/m4.texinfo:4947: Origin of test
./136.diversions: stdout mismatch
@ ../doc/m4.texinfo:6358: Origin of test
./184.sysval: stdout mismatch
All cases where I already skip the test on platforms that do not obey unix
semantics. In other words, once I submit a patch that makes your platform
trigger the __os2__ macro rather than __unix__, these tests should be
resolved.
@ ../doc/m4.texinfo:6308: Origin of test
./183.sysval: stdout mismatch
- -2
+255
Weird - I called system("exit 2") but discovered an exit status of 255.
Do you know how to make your platform exit with a known exit status? Even
mingw does the right thing for a simple 'exit 2'! I'm not sure how to
work around this bug (maybe just leave the testsuite failure in place).
@ ../doc/m4.texinfo:6483: Origin of test
./186.mkstemp: stdout mismatch
+rm: cannot lstat `foo??????': No such file or directory
Ouch. I assume that I created a temporary file, but now I can't remove
it, particularly since the other mkstemp tests passed? Do file globs work
from your command line? Can you check whether a file checks/foo* with 6
random characters still exists in your test directory?
- --
Don't work too hard, make some time for fun as well!
Eric Blake [EMAIL PROTECTED]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iEYEARECAAYFAkf/XScACgkQ84KuGfSFAYBcPgCgjX1jMHt+xuYbIF+7FX8plntw
8EAAoId8IeOsjwgeQWM8VuXnES+DwidY
=VmEg
-----END PGP SIGNATURE-----
