This reverts commit f75f26cc.

On second thoughts, this does make sense, for unregisterised via-C
OSes at least.

This reverts commit 2dea11a4.

On second thoughts, this does make sense, for unregisterised via-C
arches at least.

It doesn't make sense. If platformArch is ArchUnknown then we don't know
the answer to any questions about the arch. So now if we don't recognise
the arch we just fail, and the new arch will need to be added to the
datatype.

It doesn't make sense. If platformOS is OSUnknown then we don't know the
answer to any questions about the OS. So now if we don't recognise the
OS we just fail, and the new OS will need to be added to the datatype.

See Note [atomic CAFs] in rts/sm/Storage.c

Ben Gamari authored Oct 12, 2011 and Simon Marlow committed Oct 17, 2011

Fixes build error:

compiler/cmm/PprC.hs:961:33:
    Couldn't match expected type `Platform'
           against inferred type `CmmExpr'
    In the first argument of `pprExpr1', namely `expr'
    In the second argument of `(<+>)', namely `pprExpr1 expr'
    In the first argument of `parens', namely
        `(cast <+> pprExpr1 expr)'

For now, turn off dynlibs when validating

Cabal will use it regardless, and if people are going to use binary then
it's a lot better if they use the one that we ship, rather than
installing a second version.