GitLab

It turns out that the `fail` method was unused to begin with.

Fixes #11

This commit was performed automatically by a script.
https://github.com/RyanGlScott/travis-maintenance

This commit was performed automatically by a script.
https://github.com/RyanGlScott/travis-maintenance

Reviewers: O25 HSC2HS, hvr

Subscribers: ckoparkar

Differential Revision: https://phabricator.haskell.org/D5238

Feature/add more archs

See also https://ghc.haskell.org/trac/ghc/ticket/15758

The hsc2hs-0.68.4 release was already revised with a lower bound `base >= 4.12` to
mitigate the issue solver-side
(http://hackage.haskell.org/package/hsc2hs-0.68.4/revisions/)

This improves upon #9

This commit was performed automatically by a script.
https://github.com/RyanGlScott/travis-maintenance

All lower bounds except those for `process` (due to showCommandForUser)
match the versions of boot libraries shipped w/ GHC 7.0

See https://ghc.haskell.org/trac/ghc/ticket/13896

While I was in town, I also:

* Added a cabal.project file
* Updated .gitignore to include cabal new-build artifacts

These are tests to illustrate GHC cross-compilation failures:
    https://ghc.haskell.org/trac/ghc/ticket/14889Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org>

and update GHC 8.2.1 to GHC 8.2.2 while at it

The previous script didn't run the testsuites yet

As this would require us to declare `cabal-version: 2.0` in the
cabal package description.

This regression was introduced in 9c13b192

Fixes #6

Adds a crude at&t assembly parser to resolve constants

update travis.yml to not try to build the testsuite on 7.0 and 7.2
which won't work, because the dependencies 'call-stack' depends
on base >= 4.5, which is a dependency of hspec-core.

Also updates the assembly files, to drop the `\n` test, which might
end up as `\12`, as `\12`, `\x0a` and `\n`, testing this properly
would require some form of normalization, which we don't have yet.