On Fri, May 12, 2006 at 09:30:08AM -0400, Igor Peshansky wrote: > After this point, conftest.exe cannot be removed. > Looking around line 4167 of screen's configure should show what exact code > has this behavior. You can then compile it separately and see if you can > reproduce it. After that, it's a matter of debugging.
I copied the problem conftest code out of configure and created a standalone program that reproduces the problem behavior each time it's run. After running, I attached gdb to the process and got this backtrace: (gdb) bt #0 0x77f9193d in ntdll!DbgUiConnectToDbg () from /cygdrive/c/WINNT/system32/NTDLL.DLL #1 0x7c57fe8d in KERNEL32!DebugActiveProcess () from /cygdrive/c/WINNT/system32/KERNEL32.DLL #2 0x61004409 in _cygtls::call2 () from /usr/bin/cygwin1.dll #3 0x00000000 in ?? () from I'm not really sure what this means. If anyone's interested: http://jca.sdf1.org/cygwin/conftest.c http://jca.sdf1.org/cygwin/conftest.exe Meanwhile, I found this thread related to screen 4.0.2 and the same behavior: http://sourceware.org/ml/cygwin/2005-06/msg00843.html If I take the fifo test out of configure, the configure script completes successfully. I installed Andrew's test package which is working fine so far. jca -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/