Re: [commit: ghc] master: Fix non-termination of SpecConstr (see #5550). ForceSpecConstr will now only specialise recursive types a finite number of times. There is a new option -fspec-constr-recursiv

Hi Simon, No problem, I'll commit the extra notes tomorrow. I put the test case in simplCore/should_compile/T5550.hs. Does that seem like a reasonable place? I couldn't see any directory specifically for sanity-checking optimisations or specialise. Amos On Tue, Apr 2, 2013 at 9:37 PM, Simon Peyt

Fwd: Crash on Windows with large data

Please forgive me if you received this message twice. I believe that my first attempt to send it was swallowed due to my not being a subscriber to the list at that time. We now have a Windows server with 23 GB of memory temporarily available this week to work on this GHC crash. Is there anyone in

Re: What promotes to (* :: BOX) ?

Hello, I've been using an existential quantifier for this. It looks weird (and I think we should fix this as discussed) but it is the current workaround: data MyKind = forall star. T star -- T :: * -> MyKind -Iavor On Tue, Apr 2, 2013 at 7:11 AM, Richard Eisenberg wrote: > For more immed

Re: Why do Names defined in the current module lack a module name?

Hi Simon, On Tue, Apr 2, 2013 at 3:31 AM, Simon Peyton-Jones wrote: > > Does this help > > http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/NameType A bit, but it's still not clear to me exactly when user defined exported entities will have full (i.e. including module) names. > Some

Re: What promotes to (* :: BOX) ?

For more immediate satisfaction (i.e., before we get around to updating GHC in the manner suggested on the wiki page), you have to use a parameterized type and then specialize it to *: > data Prom2 a = Typic a Now, we have (Prom2 * :: BOX) with (Typic Int :: Prom2 *). Unfortunately, we have no

Re: What promotes to (* :: BOX) ?

Hi Gabor, See http://hackage.haskell.org/trac/ghc/wiki/GhcKinds/KindsWithoutData Comments on that discussion are welcome. Cheers, Pedro On Tue, Apr 2, 2013 at 2:51 PM, Gabor Greif wrote: > Hi devs! > > I guess I found a gap in the promotion mechanism: > > > data Prom1 = Symic Symbol > > gets

What promotes to (* :: BOX) ?

Hi devs! I guess I found a gap in the promotion mechanism: > data Prom1 = Symic Symbol gets promoted to to a kind (Prom1 :: BOX) with (Symic "AAA" :: Prom1). So far so good. But how can I define by promotion (Prom2 :: BOX) with (Typic Int :: Prom2) ? I'd like to write > data Prom2 = Typic Typ

RE: [commit: ghc] master: Fix non-termination of SpecConstr (see #5550). ForceSpecConstr will now only specialise recursive types a finite number of times. There is a new option -fspec-constr-recursiv

Amos Thanks for doing this. Could you pls add a Note [Limit recursive specialisation] in SpecConstr, that explains (a) the problem, with an example, and (b) the solution. Plus, mention the relevant Trac tickets. It'll take you a few minutes to write, but it's worth it. That way in fiv

RE: Why do Names defined in the current module lack a module name?

Does this help http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/NameType Something that starts top-level may not finish up as top-level. Nested bindings are never qualified. After TidyPgm, externally-visible names (to the linker) are qualified, ones local to the .o file are not. D

Re: build failure on FreeBSD 9.1

Hi, I have created a ticket about "Dynamic linking and libffi": http://hackage.haskell.org/trac/ghc/ticket/7806 --Kazu >> OK, I've just pushed another attempt; can you let me know whether HEAD >> now works for you, please? > > Thanks. > >> That's expected. All the Haskell .so's should