Re: Merging 'yacc-work' into master

[Peter Rosin]

Peter Rosin skrev 2012-02-14 00:56:
> Stefano Lattarini skrev 2012-02-12 15:59:
>> I have cloned 'master' into a temporary branch 'yl-work-for-master',
>> merged master in there, and pushed.  You can find the resulting branch
>> here:
>>
>>   http://git.savannah.gnu.org/cgit/automake.git/log/?h=yl-work-for-master
>>
>> Below are the commit message and the stats of the merge, and attached
>> is the (compressed) result of:
>>
>>   git diff --stat master yl-work-for-master
>>
>> I will merge the 'yl-work-for-master' branch back into master in 72
>> hours or so if there is no objection by then.
> 
> I think you should address the following regression caused by
> v1.11-284-g34ca92b before merging.

That's a mis-characterization, sorry for any confusion.  The
regression happened when yacc-work was merged into yl-work-for-master
in v1.11-1910-ge6c40d4.

It appears that the real culprit is that this in defs-static.in

YACC=${AM_TESTSUITE_YACC-${YACC-'@YACC@'}}
LEX=${AM_TESTSUITE_LEX-${LEX-'@LEX@'}}

gets expanded into (in defs-static)

YACC=${AM_TESTSUITE_YACC-${YACC-'yacc'}}
LEX=${AM_TESTSUITE_LEX-${LEX-'lex'}}

when it really should have been

YACC=${AM_TESTSUITE_YACC-${YACC-'false'}}
LEX=${AM_TESTSUITE_LEX-${LEX-'false'}}

I don't get that, because the config.log from the top configure has

LEX='false'
YACC='false'

and config.status has

S["LEX"]="false"
S["YACC"]="false"

Question is, when does those false values turn into the lex/yacc defaults?

Cheers,
Peter