On torsdag, feb 20, 2003, at 10:10 Europe/Stockholm, Johan Vromans wrote:
This is the '-Duseithreads' case:
1..11 ok 1 - ok 2 - Check basic returnvalue ok 3 - ok 4 - Check that a array ref works ok 5 - Check that hash refs work ok 6 - Check that we can return FH GLOB(0x812c4d8) ok 7 - ok 8 - ok 9 - ok 10 - # mainthread: $0 = ../ext/threads/t/join.t # subthread: $0 = ../ext/threads/t/join.t # subthread: $0 = foobar # mainthread: $0 = ../ext/threads/t/join.t # pid = 5653 # PID TTY STAT TIME COMMAND # 24571 pts/3 S 0:00 -tcsh # 5653 pts/3 S 0:00 foobar # 5654 pts/3 S 0:00 \_ foobar # 5664 pts/3 R 0:00 \_ ps -f not ok 11 - altering $0 is effective # Failed test at line 114
I suspect a underlying libc problem here, or is this a SMP machine?
No, an ordinary single-CPU (Athlon 1600+).
-- Johan
Ah, this is the global $0 thingy breaking, but it seems like it isn't really breaking, according to the ps output? But why is the second mainthread not changed?
I have no idea what is supposed to be correct behaviour here, would the Real Porter Please Stand Up, (who added this test :-)
Arthur
