Limit qualifier depth

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Fix bash-completion after "--"

#3272 accidentally disabled completions at the start of flags,
e.g. on "cabal install --".
[CI SKIP]

Respect "--" in bash-completion, fixes #3271

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

It needs to be done afterwards, because we need to compute
the ABI hash as part of registration, which can't be done
unless all dependencies are registered, which could include
internal libraries.

This whole affair is very dodgy.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Convenience libraries can have dependencies declared using
build-depends.  Unfortunately, these dependencies are special:
in particular, the version range is meaningless because they
always refer to the same package they were specified in.

So, this infelicity means that the solver has to explicitly
filter out these internal dependencies.  I also had to fix
validation to ignore these internal dependencies as well.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

This comment is not yet implemented, will be the subject
of later patches.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

This are currently failing because there are bugs which need to be
fixed.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

When converting the component graph to operate in terms of UnitIds
instead of CNames I accidentally introduced a regression where we
stopped respecting build-tools when determining an ordering to
build things.  This commit fixes the regression (though perhaps
not in the most clean/performant way you could manage it.)  It
also fixes a latent bug if internal libraries aren't processed
in the correct order.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

In the process, I refactored the test suite a bit to be more robust
and handle the --with-compiler more correctly in most cases.

One note: I turned off the HPC tests with the GHC and WITH_GHC
parameters don't match.  In principle this should work but there
is some sort of bug.  I don't plan on fixing the bug.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Instead of computing the renamings based on componentPackageRenaming
right before it gets to GHC, compute it *during configure time*.
This will make some refactorings for Backpack easier.

NB: I had to keep componentPackageDeps because removing it would
be a big BC break.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

It previously also tracked dependencies from --instantiated-with
but we axed that code so the variable is not necessary.

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>

Signed-off-by: Edward Z. Yang <ezyang@cs.stanford.edu>