Re: [Patch, Fortran] PR 48588 - (4.6/4.7 regression) Resolve whole TU before generating code

[Tobias Burnus <burnus at net-b dot de>]

On 04/19/2011 05:24 PM, Mikael Morin wrote:
> On Tuesday 19 April 2011 12:36:11 Tobias Burnus wrote:
> > Build and regtested on x86-64-linux.
> > OK for the trunk - and after some grace period - for the 4.6 branch?
> Yes.
> Do you think we could have a case where we have to delay module namespace
> resolving as well (which we can't as we have to generate mod files)?

Committed as Rev. 172718.

Thanks for the review. I don't think delaying the module namespace makes 
sense. In terms of the Fortran language, each module and each 
subroutine/function/PROGRAM/BLOCK have their own scope. Thus, in 
principle, a Fortran compiler does not know about any other item in the 
program. As modules might be needed later, it makes sense to resolve 
them in the initial order - and immediately.

The global view is only needed at resolve stage for argument checking 
(i.e. diagnostics). For the code generation that's different. At that 
stage one needs a global (whole file) view as the middle end regards 
everything as a single translation unit.

Tobias