Date:	Thu, 11 May 2000 09:33:37 -0400 (EDT)
Message-Id:	<200005111333.JAA13195@indy.delorie.com>
From:	Eli Zaretskii <eliz AT delorie DOT com>
To:	muller AT cerbere DOT u-strasbg DOT fr
CC:	djgpp-workers AT delorie DOT com
In-reply-to:	<200005021646.SAA07021@cerbere.u-strasbg.fr> (message from Pierre
	Muller on Tue, 02 May 2000 18:31:45 +0200)
Subject:	Re: interrupt 0x75 limitation in Win95 ?
References:	<200005020750 DOT JAA29985 AT cerbere DOT u-strasbg DOT fr> <200005021646 DOT SAA07021 AT cerbere DOT u-strasbg DOT fr>
Reply-To:	djgpp-workers AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	djgpp-workers AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

> Date: Tue, 02 May 2000 18:31:45 +0200
> From: Pierre Muller <muller AT cerbere DOT u-strasbg DOT fr>
>
>   For instance if you take GNU make Version  3.77 
> and run from it a program that raises such an exception 
> then its make itself that crashes with a SIGFPE !!

FYI: I've just sent a patch to the GNU Make maintainer that prevents
Make from crashing on Windows when a child program triggers an FP
exception.  The patch is reproduced below, for those who would like to
patch their binaries.

Make 3.79.1 is entering a pretest, so I expect this feature to be in
the official release very soon.


2000-05-10  Eli Zaretskii  <eliz AT is DOT elta DOT co DOT il>

	* main.c (main) [__MSDOS__]: Add SIGFPE to signals we block when
	running child programs, to prevent Make from dying on Windows 9X
	when the child triggers an FP exception.

--- main.c~1	Fri Apr  7 20:17:28 2000
+++ main.c	Wed May 10 19:47:28 2000
@@ -853,6 +853,13 @@ int main (int argc, char ** argv)
   FATAL_SIG (SIGINT);
   FATAL_SIG (SIGTERM);
 
+#ifdef __MSDOS__
+  /* Windows 9X delivers FP exceptions in child programs to their
+     parent!  We don't want Make to die when a child divides by zero,
+     so we work around that lossage by catching SIGFPE.  */
+  FATAL_SIG (SIGFPE);
+#endif
+
 #ifdef	SIGDANGER
   FATAL_SIG (SIGDANGER);
 #endif

- Raw text -