Bugzilla Bug 246262

libIDL-0.8.11 fails to compile, so I have modified the MakeFile to see the
errors and I've seen this:
bison -y -d -v ./parser.y
./parser.y:610.9-10: $$ for the midrule at $6 of `struct_type' has no declared
type
./parser.y:627.9-10: $$ for the midrule at $10 of `union_type' has no declared
type
make: *** [stamp-parser] Error 1
In the attached file it's my emerge --info

Reproducible: Always

Created an attachment (id=171290) [edit]
emerge --info

(In reply to comment #0)
> libIDL-0.8.11 fails to compile, so I have modified the MakeFile to see the
> errors and I've seen this:
> bison -y -d -v ./parser.y
> ./parser.y:610.9-10: $$ for the midrule at $6 of `struct_type' has no declared
> type
> ./parser.y:627.9-10: $$ for the midrule at $10 of `union_type' has no declared
> type
> make: *** [stamp-parser] Error 1
> In the attached file it's my emerge --info
> 
> Reproducible: Always
> 

Temp fix : downgrade to bison 2.3
emerge -a =bison-2.3 libIDL

(Then updrade to latest bison again).

While it's revealed by bison 2.4,
according to bug-bison list,
bug lies in libIDL, cause it's using a construct,
that, while accepted in 2.3, was invalid even then.

this is a problem still.

So basically bison-2.4.1 is supposed to be made to still accept this case, but
not a bison-2.5 in the future, if I remember right from reading the bug-bison
mailing list for the purpose of figuring out a fix for another project
(coreboot).

As bison-2.4(.0) has other issues, some are even suggesting to not distribute
it and wait for 2.4.1 anyway. Hence if bison-2.4 would be masked and 2.4.1
entered the tree first from the 2.4 series we'd have some time for libIDL fix
being included in an upstream release. This all was under the assumption 2.4.1
will accept what current libIDL does, but as I said I'm not sure I remember
that one right.

Should be fixed with bison-2.4-r1. Please test.

There's a good chance it won't.
There are two outstanding types of bugs with 2.4:
first is missing semicolon - and r1 will probably fix that
the other is this one - I don't think r1 will fix it, cause 
according to bison upstream, bison 2.3 has incorrectly accepted this
Points of reference:
http://lists.gnu.org/archive/html/bug-bison/2008-11/msg00013.html
http://lists.gnu.org/archive/html/bug-bison/2008-11/msg00015.html

Quoting:
It is a bug that Bison 2.3 accepts this line.  The user failed to specify 
the semantic type for the mid-rule value (referenced by $$ here but $6 in 
later actions).  Bison 2.3 accidentally uses the semantic type from the 
LHS (which would be referenced by $$ if this were not a mid-rule).  The 
user is lucky that this just happens to be the right semantic type.

Bison 2.3a and later correctly spot this as a user error and fail because 
they realize there's no logical way to proceed. 
End quote.

I yesterday created a patch that solves the problem as stated in the above
referred mail thread?

It is located at: http://sacu-consulting.de/parser.y.patch

I applied the patch given by Sascha Cunz in #8. libIDL compiles and installs ok
now using bison-2.4. Thank you very much for providing it!

Okay, this is in portage now, without a bump.  Thanks everyone.

*** Bug 246681 has been marked as a duplicate of this bug. ***