There's some justification for -fwarn-tabs: tabs can lead to confusing
failures when people use the wrong tab setting in their editor. On the
other hand, I don't think there's any good reason for having the
compiler warn about trailing whitespace.
Cheers,
Simon
On 22/08/13 17:18, Edw
On 8/24/13, Herbert Valerio Riedel wrote:
> Hello Gabor,
>
> On 2013-08-22 at 22:51:31 +0200, g...@git.haskell.org wrote:
>> Link :
>> http://ghc.haskell.org/trac/ghc/changeset/cbe3dba89c085f7556fe32126381dac68d93dd8b/ghc
>
>> commit cbe3dba89c085f7556fe32126381dac68d93dd8b
>> Author: Gabor
I am working on improving new comparison primops and implementing Geoffrey's
proposal [1]. I
renamed GHC.Prim to GHC.Prim.BuiltIn and GHC.PrimWrappers (in ghc-prim library)
to GHC.Prim. I
can build the compiler and testsuite passes succesfully, but ./validate fails
with 'haddock
panic' when b
Hello Gabor,
On 2013-08-22 at 22:51:31 +0200, g...@git.haskell.org wrote:
> Link :
> http://ghc.haskell.org/trac/ghc/changeset/cbe3dba89c085f7556fe32126381dac68d93dd8b/ghc
> commit cbe3dba89c085f7556fe32126381dac68d93dd8b
> Author: Gabor Greif
> Date: Mon Aug 19 11:31:15 2013 +0200
>
>
Hello Simon,
On 2013-08-24 at 08:54:20 +0200, GHC wrote:
[...]
> Comment (by simonmar):
>
> @bos: GHC has a `FastMutInt` type for this purpose, which is a 1-element
> mutable unboxed array. I'd be interested to know whether performance
> differs between this and `ForeignPtr`.
>
> source:ghc/c
