Re: Change in demand analysis results between 7.10.2 and RC1 (not fixed in RC2 and HEAD)

2016-02-29 Thread Ömer Sinan Ağacan
> Why do you say > > | In our case, we prefer the result in 7.10.2 of course, because that's a > | more precise demand and it gives us more opportunities for > | optimizations. But I guess this could potentially reveal itself in some > > What optimisations do you have in mind? I just had

RE: CoreLint check for case with no alts

2016-02-29 Thread Simon Peyton Jones
| coreToStgExpr (Case scrut bndr ty []) | = coreToStgExpr (Case scrut bndr ty [(DEFAULT, [], mkImpossibleExpr | ty)]) Yes that should be fine. I'm afraid I have no idea why you are getting | Oops! Entered absent arg w_sc1l FilePath or seg-faults. Simon | | But that may be

Re: CoreLint check for case with no alts

2016-02-29 Thread Ömer Sinan Ağacan
> So feel free to make Lint cleverer; make sure you add a Note. But if it > /needs/ to be cleverer, that suggests that the simplifier should be cleverer > instead, and should simplify the code so that even a dumb Core Lint has no > trouble. Good point. The problem is linter is checking every

RE: CoreLint check for case with no alts

2016-02-29 Thread Simon Peyton Jones
| I wanted to ask: Is there a restriction on what kind empty case | expressions are supported by the code generator or can I just improve | the lint check and assume that the code will be handled by the code | generator correctly? I believe the latter. See CoreToStg line 364. So feel free

RE: Change in demand analysis results between 7.10.2 and RC1 (not fixed in RC2 and HEAD)

2016-02-29 Thread Simon Peyton Jones
See Note [Add demands for strict constructors] in DmdAnal, esp the bit that says If the argument is not used at all in the alternative (i.e. it is Absent), then *don't* add a 'seqDmd'. If we do, it makes it look used and hence it'll be passed to the worker when it doesn't

Re: [ANNOUNCE] GHC 8.0.1 release candidate 2

2016-02-29 Thread Jens Petersen
On 26 February 2016 at 02:09, Ben Gamari wrote: >> Thank you for RC2. > I'm happy to help. I think you're doing a great job. >> I finally built ghc-8.0.0 for Fedora Rawhide (quick build): >> https://copr.fedorainfracloud.org/coprs/petersen/ghc-8.0.1 >> (perf build is still