dmd commit, revision 707


user: walter

msg:
fix gcc spurious warning

http://www.dsource.org/projects/dmd/changeset/707

On 10/7/2010 10:35 PM, dsource.org wrote:
> dmd commit, revision 707
> 
> 
> user: walter
> 
> msg:
> fix gcc spurious warning
> 
> http://www.dsource.org/projects/dmd/changeset/707

Error, not warning.

I was curious enough to do some digging.  Here's the relevant snippet..  quite possibly not a direct quote, but repeated several times across several different websites.

============
The reason is that such jump is a scoping violation. C++ standard paragraph 6.7.3:

It is possible to transfer into a block, but not in a way that bypasses declarations with initialization. A program that jumps (77) from a point where a local variable with automatic storage duration is not in scope to a point where it is in scope is ill-formed unless the variable has POD type (3.9) and is declared without an initializer (8.5).

77) The transfer from the condition of a switch statement to a case label is
considered a jump in this respect.
============

So, based on that, it's neither spurious nor a bug in gcc.  That dmd accepts it seems to be a bug. :)

Later,
Brad

Brad Roberts wrote:
>
> Error, not warning.
>
> I was curious enough to do some digging.  Here's the relevant snippet..  quite possibly not a direct quote, but repeated several times across several different websites.
>
> ============
> The reason is that such jump is a scoping violation. C++ standard paragraph 6.7.3:
>
> It is possible to transfer into a block, but not in a way that bypasses declarations with initialization. A program that jumps (77) from a point where a local variable with automatic storage duration is not in scope to a point where it is in scope is ill-formed unless the variable has POD type (3.9) and is declared without an initializer (8.5).
>
> 77) The transfer from the condition of a switch statement to a case label is
> considered a jump in this respect.
> ============
>
> So, based on that, it's neither spurious nor a bug in gcc.  That dmd accepts it seems to be a bug. :)
>
> 

Except that there is no possible way any jump or case statement can bypass the initialization of peekt and refer to it.