September 2004 - Glasgow-haskell-users

RE: ppr styles...
by MR K P SCHUPKE 06 Sep '04

06 Sep '04

I have added some of the required deriving clauses, however I now get lots of warnings about: Bad eta expand __coerce () (CmmParse.zgzg {- v rnrp -} @ () @ () (__coerce CmmParse.ExtCode {- tc rB2 -} happyzuxzu1 {- v a1B9 -}) (__coerce CmmParse.ExtCode {- tc rB2 -} happyzuxzu2 {- v a1B8 -}) (__coerce CmmParse.Env {- tc rB6 -} eta {- v skLC -})) () Is this something I should worry about? Keean.

1 0

RE: ppr styles...
by Simon Marlow 06 Sep '04

06 Sep '04

On 06 September 2004 11:11, MR K P SCHUPKE wrote: > Would it cause me any problems if I added "deriving Show" to all the > types in compiler/hsSyn? (Can't think of a problem here... is there > any reason why this is not done by default? - It is really handy if > trying to pattern match on code syntax to be able to see what > abstract syntax is generated by certain code) You'll probably have to derive Show for a *lot* of types, right down to FastString (and in fact deriving Show won't work there, you'll have to write an instance yourself, but it shouldn't be too hard). I guess we don't use Show because it would (a) bloat the compiler a lot, and (b) slow down compiles, and in any case most of the time we want ppr instead. The Show output is going to be pretty hard to read. Cheers, Simon

1 0

RE: ppr styles...
by MR K P SCHUPKE 06 Sep '04

06 Sep '04

Would it cause me any problems if I added "deriving Show" to all the types in compiler/hsSyn? (Can't think of a problem here... is there any reason why this is not done by default? - It is really handy if trying to pattern match on code syntax to be able to see what abstract syntax is generated by certain code) Keean.

1 0

RE: ppr styles...
by Simon Marlow 06 Sep '04

06 Sep '04

On 06 September 2004 11:03, MR K P SCHUPKE wrote: > If I print out HsSyn types with pprTrace, I get: > > test{v} () = id{v} () > > or something similar (depending on the declaration). Is there > any way I can get it to print out the raw data structure? HsSyn doesn't have Show instances, so sorry no. Cheers, Simon

1 0

RE: status of amd64 port?
by Simon Marlow 06 Sep '04

06 Sep '04

On 05 September 2004 22:36, Gabriel Ebner wrote: > Gour <gour(a)mail.inet.hr> writes: > >> Since I plan to upgrade my computer to amd64, I'm interested what is >> the current status of ghc on amd64 platfrom? > > ghc-6.2.1 works fine, but unregisterised. That means it's pretty > slow, and the code it generates isn't any faster, of course. > > Debian packages are available, a gentoo ebuild is in bugzilla, OpenBSD > ports/packages are available too. And if I ever get an amd64 box here, I'll finish the registerised port. It shouldn't be too hard, now that we have a working version of gcc. A fun task for someone would be to port the new native code generator to amd64... any takers? Cheers, Simon

1 0

ppr styles...
by MR K P SCHUPKE 06 Sep '04

06 Sep '04

If I print out HsSyn types with pprTrace, I get: test{v} () = id{v} () or something similar (depending on the declaration). Is there any way I can get it to print out the raw data structure? Keean.

1 0

RE: pprTrace
by Simon Marlow 06 Sep '04

06 Sep '04

On 02 September 2004 17:31, MR K P SCHUPKE wrote: > When writing code in the compiler, how do you use pprTrace? > > I want to print out the type of a "SigD" from HsDecl? There are plenty of examples to be found in the compiler, but briefly you'd use it something like this: myFunction sigd = pprTrace "myFunction" (ppr sigd) $ ... Cheers, Simon

1 0

status of amd64 port?
by Gour 06 Sep '04

06 Sep '04

Since I plan to upgrade my computer to amd64, I'm interested what is the current status of ghc on amd64 platfrom? Sincerely, Gour -- Gour | gour(a)mail.inet.hr Registered Linux User | #278493 GPG Public Key | 8C44EDCD

2 2

RE: Type reps inside GHC...
by Simon Peyton-Jones 03 Sep '04

03 Sep '04

RdrName, Name, Id are the three things it gets instantiated to. * Parser generates HsExpr RdrName * REnamer generates HsExpr Name * Typechecker generates HsExpr Id | -----Original Message----- | From: glasgow-haskell-users-bounces(a)haskell.org [mailto:glasgow-haskell-users- | bounces(a)haskell.org] On Behalf Of MR K P SCHUPKE | Sent: 03 September 2004 16:10 | To: glasgow-haskell-users(a)haskell.org | Subject: Type reps inside GHC... | | | Playing with HsSyn types (like HsModule)... I have got down to the | definition level like: | | Sig id = Sig (Location name) (LHsDecl name) | | what type is name? Presumably it can be a range of types, | what are valid types for name? | | Keean. | _______________________________________________ | Glasgow-haskell-users mailing list | Glasgow-haskell-users(a)haskell.org | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

1 0

Type reps inside GHC...
by MR K P SCHUPKE 03 Sep '04

03 Sep '04

Playing with HsSyn types (like HsModule)... I have got down to the definition level like: Sig id = Sig (Location name) (LHsDecl name) what type is name? Presumably it can be a range of types, what are valid types for name? Keean.

2 1