GitLab

I think the change in optimisation levels due to
d0608373 had the side effect of not
inlining atomic_inc(), which together with the broken way the test was
set up made it fail.

This is because dynamic GHCi can't load archive files, and linker_unload
explicitly does this via loadArchive().
Signed-off-by: Austin Seipp <aseipp@pobox.com>

This test is testing primops, so it should be in the primops/
directory and not numeric.

(fixes #457)

Signed-off-by: Austin Seipp <aseipp@pobox.com>

This caused a failure on some 32bit builds due to floating point
(in)accuracy. But Complex is already strict in both its components, and
we really just want to know the code works, so an easier solution is to
simply use `seq` to force evaluation.
Signed-off-by: Austin Seipp <aseipp@pobox.com>

Signed-off-by: Austin Seipp <aseipp@pobox.com>

Signed-off-by: Austin Seipp <aseipp@pobox.com>

This patch adds tests for new primops and fixes the existing ones.
For a deatiled discussion of this changes please visit the wiki page:
http://hackage.haskell.org/trac/ghc/wiki/PrimBool

Signed-off-by: Austin Seipp <aseipp@pobox.com>

Signed-off-by: Austin Seipp <aseipp@pobox.com>

Signed-off-by: Austin Seipp <aseipp@pobox.com>

Signed-off-by: Austin Seipp <aseipp@pobox.com>

Closes Trac #8117.
Signed-off-by: Austin Seipp <aseipp@pobox.com>

Many of the files modified are just wibbles to output, because now
tycons have roles attached to them, which are produced in the debugging
dumps.