Akim Demaille | 4 Nov 21:56
Picon
Picon
Picon
Gravatar

Weird message from configure

I see this when configuring bison:

checking whether g++ builds executables that work... yes
checking for Java virtual machine... rm: conftest.dSYM: is a directory
java
rm: conftest.dSYM: is a directory
checking for Java compiler... javac -source 1.3
checking for Java virtual machine... rm: conftest.dSYM: is a directory
(cached) java
configure: creating ./config.status
  /bin/sh ./config.status

is this known?

Paul Eggert | 6 Nov 09:21
Favicon

Re: Weird message from configure

Akim Demaille <akim <at> lrde.epita.fr> writes:

> I see this when configuring bison:
>
> checking whether g++ builds executables that work... yes
> checking for Java virtual machine... rm: conftest.dSYM: is a directory
> java
> rm: conftest.dSYM: is a directory
> checking for Java compiler... javac -source 1.3
> checking for Java virtual machine... rm: conftest.dSYM: is a directory
> (cached) java
> configure: creating ./config.status
>  /bin/sh ./config.status
>
> is this known?

First I've seen of it.  The Java part of Bison isn't that well tested
yet, I'm afraid.

Wilson Snyder | 6 Nov 17:02
Favicon

Incorrect(?) useless nonterminal

Hello,

I'm using

$ bison --version
bison (GNU Bison) 1.875

Running bison on the attached good file works fine:

$ bison -d -v --debug --verbose -d -k VParseBisonGood.y -pVParseBisonGood
-o VParseBisonGood.c

While the bad file gives:

$ bison -d -v --debug --verbose -d -k VParseBisonBad.y -pVParseBisonBad -o
VParseBisonBad.c
VParseBisonBad.y: warning: 2 useless nonterminals and 3 useless rules
VParseBisonBad.y:445.17-28: warning: useless nonterminal: sequenceDecl
VParseBisonBad.y:882.63-74: warning: useless nonterminal: sequenceExpr
VParseBisonBad.y:445.17-59: warning: useless rule: modItem: sequenceDecl
VParseBisonBad.y:882.17-107: warning: useless rule: sequenceDecl:
"sequence" "IDENTIFIER" sequenceArgsE ';' funcVarListE sequenceExpr
"endsequence" endLabelE
VParseBisonBad.y:893.17-74: warning: useless rule: sequenceExpr:
"first_match" '(' sequenceExpr ')'

The only difference is line 893:

sequenceExpr:	yFIRST_MATCH '(' ')'	{ $<fl>$=$<fl>1; }

(Continue reading)

Akim Demaille | 7 Nov 09:16
Picon
Picon
Picon
Gravatar

Re: Incorrect(?) useless nonterminal


Le 6 nov. 07 à 17:02, Wilson Snyder a écrit :

> The only difference is line 893:
>
> sequenceExpr:	yFIRST_MATCH '(' ')'	{ $<fl>$=$<fl>1; }
>
> versus the bad version:
>
> sequenceExpr:	yFIRST_MATCH '(' sequenceExpr ')'	{ $<fl>$=$<fl>1; }
>
> While the rule is recursive, it doesn't seem significantly different  
> from
> other rules with parenthesis handling.

What differs is that you don't provide another rule to
anchor the recursion, so sequenceExpr is useless (it
does not produce any word).

Joel E. Denny | 10 Nov 00:17

Re: Weird message from configure

On Tue, 6 Nov 2007, Paul Eggert wrote:

> Akim Demaille <akim <at> lrde.epita.fr> writes:
> 
> > I see this when configuring bison:
> >
> > checking whether g++ builds executables that work... yes
> > checking for Java virtual machine... rm: conftest.dSYM: is a directory
> > java
> > rm: conftest.dSYM: is a directory
> > checking for Java compiler... javac -source 1.3
> > checking for Java virtual machine... rm: conftest.dSYM: is a directory
> > (cached) java
> > configure: creating ./config.status
> >  /bin/sh ./config.status
> >
> > is this known?
> 
> First I've seen of it.

I've not seen it either.

Micah Henning | 9 Nov 20:08
Picon

[GNU Bison 2.3] testsuite: 84 85 87 88 89 90 94 108 126 129 133 135 137 139 141 142 143 144 145 146 147 148 150 151 152 153 156 157 failed


## ------------------------- ##
## GNU Bison 2.3 test suite. ##
## ------------------------- ##

testsuite: command line was:
  $ ./testsuite 

## ----------- ##
## ChangeLogs. ##
## ----------- ##

testsuite: ../ChangeLog:
| 2006-06-05  Paul Eggert  <eggert <at> cs.ucla.edu>
| 
| 	* NEWS: Version 2.3.
| 	* configure.ac (AC_INIT): Likewise.
| 
| 2006-05-30  Paul Eggert  <eggert <at> cs.ucla.edu>
| 
| 	* data/glr.c (YYRECOVERING): Define to be a function-like macro
| 	with no arguments, not as an object-like macro.  This is for
| 	compatibility with data/yacc.c.  Problem reported by John P. Hartmann in

## --------- ##
## Platform. ##
## --------- ##

(Continue reading)

Hans Aberg | 10 Nov 21:34
Picon
Picon

Re: Weird message from configure

On 10 Nov 2007, at 00:17, Joel E. Denny wrote:

> On Tue, 6 Nov 2007, Paul Eggert wrote:
>
>> Akim Demaille <akim <at> lrde.epita.fr> writes:
>>
>>> I see this when configuring bison:
>>>
>>> checking whether g++ builds executables that work... yes
>>> checking for Java virtual machine... rm: conftest.dSYM: is a  
>>> directory
>>> java
>>> rm: conftest.dSYM: is a directory
>>> checking for Java compiler... javac -source 1.3
>>> checking for Java virtual machine... rm: conftest.dSYM: is a  
>>> directory
>>> (cached) java
>>> configure: creating ./config.status
>>>  /bin/sh ./config.status
>>>
>>> is this known?
>>
>> First I've seen of it.
>
> I've not seen it either.

Was it a typo in Pauls's message? - I interpreted it as though he had  
seen it.

If you mean something being a directory when it shouldn't, yes I have  
(Continue reading)

Hans Aberg | 12 Nov 12:15
Picon
Picon

Re: rm: conftest.dSYM: is a directory (Was: Weird message from configure)

On 12 Nov 2007, at 11:36, Akim Demaille wrote:

> I did have a problem when installing XCode 3.0,

I experienced the problem (in connection with Fink) when updating the  
Xcode that came with the Mac OS X 10.4.0 installation DVD (don't know  
about 10.5).

So one fix was to deinstall Xcode using the script that came with it  
(at the end of some readme file), and then install the later Xcode  
version from scratch. I decided though to erase the hard disk,  
install Mac OS X without development package, update it, and then  
install latest Xcode. Then the problem went away. Have fun! :-)

   Hans Aberg

Ralf Wildenhues | 12 Nov 12:39
Picon
Picon

Re: rm: conftest.dSYM: is a directory (Was: Weird message from configure)

Hello,

* Hans Aberg wrote on Mon, Nov 12, 2007 at 12:15:14PM CET:
> On 12 Nov 2007, at 11:36, Akim Demaille wrote:
>
>> I did have a problem when installing XCode 3.0,

See:
<http://thread.gmane.org/gmane.comp.sysutils.autoconf.bugs/5811/focus=5063>

Cheers,
Ralf

Hans Aberg | 12 Nov 13:08
Picon
Picon

Re: rm: conftest.dSYM: is a directory (Was: Weird message from configure)

On 12 Nov 2007, at 12:20, Akim Demaille wrote:

>>> I did have a problem when installing XCode 3.0,
>>
>> I experienced the problem (in connection with Fink) when updating  
>> the Xcode that came with the Mac OS X 10.4.0 installation DVD  
>> (don't know about 10.5).
>>
>> So one fix was to deinstall Xcode using the script that came with  
>> it (at the end of some readme file), and then install the later  
>> Xcode version from scratch. I decided though to erase the hard  
>> disk, install Mac OS X without development package, update it, and  
>> then install latest Xcode. Then the problem went away. Have fun! :-)
>
> I checked on another machine with perfect setup: that appears
> to be now the default behavior of Leopard's GCC.
>
> And it also appears that Autoconf folks already know about it :)
>
> http://www.nabble.com/Mac-OS-X-Leopard-and-conftest.dSYM- 
> directories-t4782282.html

I do not know about Mac OS X 10.5, but in the case of 10.4 the Fink  
developers pointed out that what was directory shouldn't, and when I  
did that reinstallation indicated above, the problem went away. So it  
was verified to be a bug in the updating of Xcode, but I never found  
exactly what caused it.

> Sorry for the noise.

(Continue reading)


Gmane