can you do an strace -p of the innermost ksh and nmake processes On Sun, 5 Feb 2012 14:21:41 -0500 Bruce Lilly wrote: > Tried building ast-open beta from source on several 32- and 64-bit > machines all running openSUSE 12.1 with updates, Linux kernel > 3.1.9-1.4-desktop.
> All builds run a bunch of stuff then hang. The last few lines printed > before hanging (modulo architectural differences in path) are: > + cc -O -I. -I/usr/include/asm -I- -I../../../include/ast -D_PACKAGE_ast -c > /opt/ast/src/cmd/ss/ssd.c > cc1: note: obsolete option -I- used, please use -iquote instead > + cc -O -Wl,-z,origin '-Wl,-R,$ORIGIN/../lib' -o ssd ssd.o > ../../../lib/libcs.a -ldl ../../../lib/libast.a > + mv -f /opt/ast/arch/linux.i386/bin/ss /opt/ast/arch/linux.i386/bin/ss.old > + ignore cp ss /opt/ast/arch/linux.i386/bin/ss > + mv -f /opt/ast/arch/linux.i386/lib/ssd /opt/ast/arch/linux.i386/lib/ssd.old > + ignore cp ssd /opt/ast/arch/linux.i386/lib/ssd > + chgrp kmem /opt/ast/arch/linux.i386/lib/ssd > A process tree (from another shell) shows: > PID PPID C PRI NI ADDR SZ WCHAN STIME TTY TIME CMD > 15241 2480 0 80 0 - 1301 wait 12:52 pts/1 00:00:00 package > make > 16173 15241 0 80 0 - 970 pipe_w 12:52 pts/1 00:00:00 > /opt/ast/arch/linux.i386/bin/ok/nmake > LOCALRULESPATH=/opt/ast/arch/linux.i386/bin/ok/lib -k -K > 16176 16173 0 80 0 - 873 pipe_w 12:52 ? 00:00:00 ksh > /dev/fd/3 > 14083 16176 0 80 0 - 873 wait 13:35 ? 00:00:00 > ksh /dev/fd/3 > 14084 14083 0 80 0 - 1085 pipe_w 13:35 ? 00:00:00 > /opt/ast/arch/linux.i386/bin/ok/nmake --ignorelock --keepgoing > --errorid=cmd/ss LOCALRUL > 14086 14084 0 80 0 - 873 pipe_w 13:35 ? 00:00:00 > ksh /dev/fd/4 > 16174 15241 0 80 0 - 807 - 12:52 pts/1 00:00:03 > /opt/ast/arch/linux.i386/bin/ok/tee -a > /opt/ast/arch/linux.i386/lib/package/gen/make.out > Everything but the chgrp command appears to have been executed. > N.B. chgrp didn't fail; the process just hangs. > Interrupting, manually running the chgrp (with appropriate > privileges) and rerunning package make seems to work. > Output from package test isn't clear; some parts indicate errors > but programs seem to work reasonably. > _______________________________________________ > ast-developers mailing list > [email protected] > https://mailman.research.att.com/mailman/listinfo/ast-developers _______________________________________________ ast-developers mailing list [email protected] https://mailman.research.att.com/mailman/listinfo/ast-developers
