patch to fix segfault complaints.

others have noted various segfaults,
ex:

"/dev/hde1"

This patch fixes configure script to build good (no segfault) lvm tools for me
theres probably a more comprehensive way to do this, ymmv.

note also that Im only checking for 2.96, not 2.96-85. this may or may not matter

[root@groucho 1.0.1-rc4]# diff -u configure configure~

@@ -796,33 +796,22 @@
 
 fi
 
-${CC-cc} -v 2>&1 | grep -l 2.96
-if [ $? = 0 ] ; then
-    echo "this gcc version breaks with -g -O2"
-    optimize=1
-else
-    optimize=2
-fi
-
 echo "$ac_t""$ac_cv_prog_cc_g" 1>&6
 if test "$ac_test_CFLAGS" = set; then
   CFLAGS="$ac_save_CFLAGS"
 elif test $ac_cv_prog_cc_g = yes; then
   if test "$GCC" = yes; then
-    CFLAGS="-g -O$optimize"
+    CFLAGS="-g -O2"
   else
     CFLAGS="-g"
   fi
 else
   if test "$GCC" = yes; then
-    CFLAGS="-O$optimize"
+    CFLAGS="-O2"
   else
     CFLAGS=
   fi
 fi
 
 # Find a good install program.  We prefer a C program (faster),
 # so one script is as good as another.  But avoid the broken or

yes, I suppose it is, judging from the --- and +++ associated with ARGS.

but before its used, Id like to see discussion on;

whether its accurate enough wrt 2.96-85 or -O1
whether compiler / vendor / phase-of-moon accomodations are appropriate.

how to do it better.

in other words, it probably wont be applied to CVS unless and until tested by us (the users) ..

so -

did you reverse apply it ? successfully ?
did you formerly have a segfault problem due to -O2 ? does -O1 fix it for you ?