GitLab

There might be threads in foreign calls that will attempt to return
via resumeThread() and grab this lock, so we can't safely destroy it.

Fixes one cause of

   internal error: ASSERTION FAILED: file Capability.c, line 90

although I haven't repeated that assertion failure in the wild, only
with a specially crafted test case, so I can't be sure I really got
it.

To avoid IO requests completing only to discover that the
completed_table_sema has been CloseHandle()'d.  This all looks a bit
wrong, though: we shouldn't really be waiting for these requests to
complete, they might take forever.

Some of the memory allocation calls were being checked for error, but
the RTS was printing a message and continuing.  These error cases
lead to crashes later, so better to just fail immediately.

HpAlloc was not being set when returning to the scheduler via MAYBE_GC(),
which at the least was just wrong (the scheduler might allocate a large
block more than once), and at worst could lead to crashes if HpAlloc contains
garbage.

Fixes at least one threaded2 test on Windows.

This affects -debug only, avoids crash with test conc012.

Patch from: Roman Leshchinskiy

They aren't built as part of a GHC build anyway, it doesn't make much
sense to include them in checkouts or source tarballs.

It's an internal module used by non-GHC implementations only.

Fixes ffi011(opt) on x86_64.  I don't know why this has only just
appeared today, it's apparently been broken for some time.

This is mainly for the benefit of Solaris. I'll fix this properly later.

We only support Sparc V9 and better as V8 lacks an atomic CAS instruction
which we need for SMP. This means that we have to pass -mcpu=v9 to GCC when
compiling and assembling. Hardcoding the flag is hackish but seems to be
our best bet at the moment. It can still be overridden by the user as GCC
picks the best -mcpu flag regardless of the ordering.

See comments for details.  Fixes assertion failures in stage 3 build
which appeared after recent closeMutex() addidion.  May fix other
shutdown issues.

Contributed by: Thiemo Seufer <ths@networkno.de>

patch from #878

Not much has changed really: just the removal of the overlap
restriction, and the re-instatement of the requirement that
-package-name must be used when compiling a package now.