Weird warnings from recent GHC snapshot

I run haskell on Mac OS X (Snow Leopard). After upgrading my Xcode installation to 4.0 I had a tricky time getting ghc working again; the version bundled with the Haskell Platform no longer works and I had to compile a recent snapshot (ghc-7.1.20110315) from source. This worked fine, but now when I compile my code I keep getting really weird warnings. For instance: ld: warning: could not create compact unwind for .LFB3: non-standard register 5 being saved in prolog and SpecConstr Function `$w$j{v s3hL} [lid]' has one call pattern, but the limit is 0 Use -fspec-constr-count=n to set the bound Use -dppr-debug to see specialisations SpecConstr Function `$w$j{v s3ic} [lid]' has two call patterns, but the limit is 1 Use -fspec-constr-count=n to set the bound Use -dppr-debug to see specialisations The code seems to work, but this is quite irritating. Is there anything that can be done about these warnings? Thanks, Mike

1) if you copy the 10.5sdk from developer-old into the new developer folder
sdk folder, the standard ghc distro will work again
2) theres a ghc ticket for that, its apparently partly because theres no
-nowarnunwind flag or the like
On Thu, Mar 17, 2011 at 6:11 AM, Michael Vanier
I run haskell on Mac OS X (Snow Leopard). After upgrading my Xcode installation to 4.0 I had a tricky time getting ghc working again; the version bundled with the Haskell Platform no longer works and I had to compile a recent snapshot (ghc-7.1.20110315) from source. This worked fine, but now when I compile my code I keep getting really weird warnings. For instance:
ld: warning: could not create compact unwind for .LFB3: non-standard register 5 being saved in prolog
and
SpecConstr Function `$w$j{v s3hL} [lid]' has one call pattern, but the limit is 0 Use -fspec-constr-count=n to set the bound Use -dppr-debug to see specialisations SpecConstr Function `$w$j{v s3ic} [lid]' has two call patterns, but the limit is 1 Use -fspec-constr-count=n to set the bound Use -dppr-debug to see specialisations
The code seems to work, but this is quite irritating. Is there anything that can be done about these warnings?
Thanks,
Mike
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
participants (2)
-
Carter Schonwald
-
Michael Vanier