On Tue, Jun 26, 2001 at 03:34:23PM +0900, Bang Jun-Young wrote:
- configure.in:
Remove a redundant check for flex.
It is not redundant! AC_PROG_LEX does more than just look for the 'lex' binary.
Please do not apply.
Ciao, Marcus
-AC_PROG_LEX AC_CHECK_PROGS(XLEX,$LEX flex lex,none) if test "$XLEX" = "none" then
On Tue, Jun 26, 2001 at 08:44:50AM +0200, Marcus Meissner wrote:
On Tue, Jun 26, 2001 at 03:34:23PM +0900, Bang Jun-Young wrote:
- configure.in:
Remove a redundant check for flex.
It is not redundant! AC_PROG_LEX does more than just look for the 'lex' binary.
Please do not apply.
Why does configure check for flex twice? On NetBSD I get:
checking for bison... no checking for byacc... no checking for flex... flex checking for yywrap in -lfl... yes checking for flex... flex
(as seen above, autoconf doesn't search for `yacc' for AC_PROG_YACC -- should be fixed)
Jun-Young
On Tue, Jun 26, 2001 at 05:43:55PM +0900, Bang Jun-Young wrote:
On Tue, Jun 26, 2001 at 08:44:50AM +0200, Marcus Meissner wrote:
On Tue, Jun 26, 2001 at 03:34:23PM +0900, Bang Jun-Young wrote:
- configure.in:
Remove a redundant check for flex.
It is not redundant! AC_PROG_LEX does more than just look for the 'lex' binary.
Please do not apply.
Why does configure check for flex twice? On NetBSD I get:
checking for bison... no checking for byacc... no checking for flex... flex checking for yywrap in -lfl... yes checking for flex... flex
(as seen above, autoconf doesn't search for `yacc' for AC_PROG_YACC -- should be fixed)
It will use it as default if neither byacc nor flex are found.
The reason we check for 'flex' again is that the standard check sets LEX to 'lex' even if none is found, which lead to confusion during compile.
Our check is similar, but it finds out if 'lex' is present and aborts configure if not.
This has reduced compiling support queries :)
Ciao, Marcus
On Tue, Jun 26, 2001 at 11:42:07AM +0200, Marcus Meissner wrote:
On Tue, Jun 26, 2001 at 05:43:55PM +0900, Bang Jun-Young wrote:
Why does configure check for flex twice? On NetBSD I get:
checking for bison... no checking for byacc... no checking for flex... flex checking for yywrap in -lfl... yes checking for flex... flex
(as seen above, autoconf doesn't search for `yacc' for AC_PROG_YACC -- should be fixed)
It will use it as default if neither byacc nor flex are found.
The reason we check for 'flex' again is that the standard check sets LEX to 'lex' even if none is found, which lead to confusion during compile.
Our check is similar, but it finds out if 'lex' is present and aborts configure if not.
This has reduced compiling support queries :)
That's all fine and dandy :), but I think we should add a comment to this configure.in line, given the pretty confusing doubled configure output. I'm about 150% sure other people will want to submit the same "fixes" otherwise.
On Tue, Jun 26, 2001 at 11:42:07AM +0200, Marcus Meissner wrote:
On Tue, Jun 26, 2001 at 05:43:55PM +0900, Bang Jun-Young wrote:
Why does configure check for flex twice? On NetBSD I get:
checking for bison... no checking for byacc... no checking for flex... flex checking for yywrap in -lfl... yes checking for flex... flex
(as seen above, autoconf doesn't search for `yacc' for AC_PROG_YACC -- should be fixed)
It will use it as default if neither byacc nor flex are found.
The reason we check for 'flex' again is that the standard check sets LEX to 'lex' even if none is found, which lead to confusion during compile.
Our check is similar, but it finds out if 'lex' is present and aborts configure if not.
Another question: why doesn't it doublecheck for yacc? There may be a system that doesn't have any parser installed by default and at least one of yacc, byacc, and bison should be installed to compile.
I have made another patch:
diff -u -r1.209 configure.in --- configure.in 2001/06/19 03:30:14 1.209 +++ configure.in 2001/06/26 11:26:50 @@ -54,6 +54,13 @@ AC_PROG_CPP AC_PATH_XTRA AC_PROG_YACC +AC_CHECK_PROGS(XYACC,$YACC yacc byacc bison,none) +if test "$XYACC" = "none" +then + echo "*** Error: No suitable parser found. ***" + echo " Please install 'yacc', 'byacc', or 'bison' package." + exit 1 +fi AC_PROG_LEX AC_CHECK_PROGS(XLEX,$LEX flex lex,none) if test "$XLEX" = "none"
Jun-Young
On Tue, Jun 26, 2001 at 08:43:22PM +0900, Bang Jun-Young wrote:
On Tue, Jun 26, 2001 at 11:42:07AM +0200, Marcus Meissner wrote:
The reason we check for 'flex' again is that the standard check sets LEX to 'lex' even if none is found, which lead to confusion during compile.
Our check is similar, but it finds out if 'lex' is present and aborts configure if not.
Another question: why doesn't it doublecheck for yacc? There may be a system that doesn't have any parser installed by default and at least one of yacc, byacc, and bison should be installed to compile.
Just for reference - I had this very problem just yesterday with some guy. Nearly nothing installed: checking for bison... no checking for byacc... no checking for flex... (cached) flex
configure didn't abort, and compile "crashed" at wrc: make[2]: Entering directory `/home/daniel/tmp/wine/tools/wrc' yacc -bppy -ppp -d -t ./ppy.y make[2]: yacc: Command not found make[2]: *** [ppy.tab.c] Error 127 make[2]: Leaving directory `/home/daniel/tmp/wine/tools/wrc' make[1]: *** [wrc/__depend__] Error 2 make[1]: Leaving directory `/home/daniel/tmp/wine/tools' make: *** [tools/__depend__] Error 2 Command exited with non-zero status 2
-> not good.
--> good if this patch fixes it.
-
Andreas Mohr -
Bang Jun-Young -
Marcus Meissner