[llvm-commits] CVS: llvm/lib/AsmParser/Lexer.l llvmAsmParser.y

Jeff Cohen jeffc at jolt-lang.org
Sun Dec 31 12:42:51 PST 2006


A semi-colon appears to be in the wrong place, as marked below.  Some 
versions of bison complain, while others silently accept it (including 
the newest ones)--though how they interpret the grammar is not clear.

Reid Spencer wrote:
> Changes in directory llvm/lib/AsmParser:
>
> Lexer.l updated: 1.89 -> 1.90
> llvmAsmParser.y updated: 1.292 -> 1.293
> ---
> Log message:
>
> For PR950: http://llvm.org/PR950 :
> Major reorganization. This patch introduces the signedness changes for
> the new integer types (i8, i16, i32, i64) which replace the old signed
> versions (ubyte, sbyte, ushort, short, etc). This patch also implements
> the function type parameter attributes feature. Together these conspired
> to introduce new reduce/reduce errors into the grammar. Consequently, it
> was necessary to introduce a new keyword into the grammar in order to
> disambiguate. Without this, yacc would make incorrect shift/reduce and
> reduce/reduce decisions and fail to parse the intended assembly.
>
> Changes in assembly:
>
> 1. The "implementation" keyword is superfluous but still supported. You
>    can use it as a sentry which will ensure there are no remaining up
>    reference types. However, this is optional as those checks are also
>    performed elsewhere.
>
> 2. Parameter attributes are now implemented using an at sign to
>    indicate the attribute. The attributes are placed after the type
>    in a function declaration or after the argument value in a function
>    call. For example:
>       i8 @sext %myfunc(i16 @zext)
>       call i8 @sext %myfunc(i16 @zext %someVal)
>    The facility is available for supporting additional attributes and
>    they can be combined using the @(attr1,attr2,attr3) syntax. Right
>    now  the only two supported are @sext and @zext
>
> 3. Functions must now be defined with the "define" keyword which is
>    analagous to the "declare" keyword for function declarations. The
>    introduction of this keyword disambiguates situations where a 
>    named result type is confused with a new type or gvar definition.
>    For example:
>       %MyType = type i16
>       %MyType %func(%MyType) { ... }
>    With the introduction of optional parameter attributes between 
>    the function name and the function result type, yacc will pick
>    the wrong rule to reduce unless it is disambiguated with "define"
>    before the function definition, as in:
>       define %MyType @zext %func(%MyType %someArg) { ... }
>
>
>
> ---
> Diffs of the changes:  (+413 -253)
>
>  Lexer.l         |   12 -
>  llvmAsmParser.y |  654 +++++++++++++++++++++++++++++++++++---------------------
>  2 files changed, 413 insertions(+), 253 deletions(-)
>
>
> Index: llvm/lib/AsmParser/llvmAsmParser.y
> diff -u llvm/lib/AsmParser/llvmAsmParser.y:1.292 llvm/lib/AsmParser/llvmAsmParser.y:1.293
> --- llvm/lib/AsmParser/llvmAsmParser.y:1.292	Fri Dec 29 14:29:48 2006
> +++ llvm/lib/AsmParser/llvmAsmParser.y	Sat Dec 30 23:40:12 2006
>
> @@ -2307,18 +2412,31 @@
>    };
>  
>  
> -ValueRefList : ResolvedVal {    // Used for call statements, and memory insts...
> -    $$ = new std::vector<Value*>();
> -    $$->push_back($1);
> +ValueRefList : Types ValueRef OptParamAttrs {    
> +    if (!UpRefs.empty())
> +      GEN_ERROR("Invalid upreference in type: " + (*$1)->getDescription());
> +    // Used for call and invoke instructions
> +    $$ = new ValueRefList();
> +    ValueRefListEntry E; E.Attrs = $3; E.Val = getVal($1->get(), $2);
> +    $$->push_back(E);
>    }
> -  | ValueRefList ',' ResolvedVal {
> +  | ValueRefList ',' Types ValueRef OptParamAttrs {
> +    if (!UpRefs.empty())
> +      GEN_ERROR("Invalid upreference in type: " + (*$3)->getDescription());
>      $$ = $1;
> -    $1->push_back($3);
> +    ValueRefListEntry E; E.Attrs = $5; E.Val = getVal($3->get(), $4);
> +    $$->push_back(E);
>      CHECK_FOR_ERROR
> -  };
> +  }
> +  | /*empty*/ { $$ = new ValueRefList(); };
>  
> -// ValueRefListE - Just like ValueRefList, except that it may also be empty!
> -ValueRefListE : ValueRefList | /*empty*/ { $$ = 0; };
> +IndexList       // Used for gep instructions and constant expressions
> +  : /*empty*/ { $$ = new std::vector<Value*>(); };  <====== bad semi-colon
> +  | IndexList ',' ResolvedVal {
> +    $$ = $1;
> +    $$->push_back($3);
> +    CHECK_FOR_ERROR
> +  } <====== missing semi-colon
>  
>
>   




More information about the llvm-commits mailing list