Too many local registers...

Simon Marlow simonmarhaskell at gmail.com
Thu Nov 1 05:16:06 EDT 2007


Simon Peyton-Jones wrote:
> We are working on a substantial re-engineering of GHC's code generation, so this problem will eventually go away.  But think "2008" not "Christmas 2007".  Meanwhile does -fasm work?

There's no NCG for IA64, unfortunately.  Unless someone writes one before 
6.10, IA64 will be relegated to unregisterised compilation.

Cheers,
	Simon

> Simon
> 
> | -----Original Message-----
> | From: glasgow-haskell-users-bounces at haskell.org [mailto:glasgow-haskell-users-bounces at haskell.org] On Behalf Of
> | Duncan Coutts
> | Sent: 31 October 2007 23:37
> | To: Cristian Perfumo
> | Cc: glasgow-haskell-users at haskell.org
> | Subject: Re: Too many local registers...
> |
> | On Wed, 2007-10-31 at 14:40 +0100, Cristian Perfumo wrote:
> | > Hi,
> | > I did some modifications to GHC in order to read hardware counters.
> | > Many programs are compiled successfully, but with one of them, I have
> | > a problem:
> | >
> | > Too many local registers allocated by gcc at .../ghc-
> | > 6.6.1-modified/driver/mangler/ghc-asm line 1602, <INASM> line 304643.
> | >
> | > It's ghc 6.6.1 on an Itanium processor.
> |
> | I never found a satisfactory answer. On IA64 we just hope that ghc
> | generates C code that does not turn into object code that uses too many
> | variables. We tried playing with gcc flags to make it only use the
> | limited number of registers but could not make it work. The only
> | workarounds I know of are to use less aggressive optimisation, both in
> | ghc and gcc, so try with -optc-O rather than -optc-O2 and try -O0 rather
> | than -O.
> |
> | Duncan
> | _______________________________________________
> | Glasgow-haskell-users mailing list
> | Glasgow-haskell-users at haskell.org
> | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
> _______________________________________________
> Glasgow-haskell-users mailing list
> Glasgow-haskell-users at haskell.org
> http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
> 



More information about the Glasgow-haskell-users mailing list