From: vischne AT ibm DOT net
Subject: Exception: STATUS_ACCESS_VIOLATION
6 Mar 1998 19:06:58 -0800
Message-ID: <199803040347.DAA25008.cygnus.gnu-win32@out5.ibm.net>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
To: John DOT Cook AT kla-tencor DOT com, gnu-win32 AT cygnus DOT com

>   Cook, John (John DOT Cook AT kla-tencor DOT com)
>   Mon, 2 Mar 1998 15:28:31 -0800 
>I succeeded in building ncurses-4.1 once under b19, but am now trying it
>again
>and getting this error frequently during make:
>
>(e:\Cygnus\B19\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32\2.7-B19\cc1.exe
>1097) Exception: STATUS_ACCESS_VIOLATION
>(e:\Cygnus\B19\H-i386-cygwin32\lib\gcc-lib\i386-cygwin32\2.7-B19\cc1.exe
>1097) Dumping stack trace to cc1.exe.core
>gcc: Internal compiler error: program cc1 got fatal signal 0
>make[1]: *** [../obj_g/lib_pad.o] Error 1
>make: *** [all] Error 2
>
>Each time the error occurs, I type 'make' again, and make gets a little
>farther,
>then barfs the same error again.  Any ideas what causes such an error?

Isn't that strange?  I sent a reply to this a day ago, but it never got
on the gnu-win32 list.

This sounds like an old gcc2.6.3 bug, in which the compiler ran out of
memory during deeply nested makes at irregular intervals.  Re-starting
the make after the crash usually led to successful completion.

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request AT cygnus DOT com" with one line of text: "help".