GitLab

This is a bit unpleasant, as "make binary-dist" really shouldn't actually
build anything, but it works.

It wasn't doing the right thing for bindists. Let's rethink...

See comment for details

Ian Lynagh <igloo@earth.li>**20071124231857
 It used to be a synonym for ":r" in 6.6.1, but this wasn't documented or
 known about by the developers. In 6.8.1 it was accidentally broken.
 This patch brings it back, but as "repeat the last command", similar to
 pressing enter in gdb. This is almost as good for people who want it to
 reload, and means that it can also be used to repeat commands like :step.

Ian Lynagh <igloo@earth.li>**20071124171220

It's far too large now, and no-one complained when 6.8.1 didn't have one.

See trac #1910.

configure will set INSTALL to ./install-sh if it can't find it in the path,
so we need to replace the . with the path to our root.

Fri Nov 23 13:15:59 PST 2007  Ian Lynagh <igloo@earth.li>

Should fix Trac #1833

We were failing to trim the type envt in mkBootModDetails, so several
functions all called (*), for example, were getting into the interface.
Result chaos.  It only actually bites when we do the retyping-loop thing,
which is why it's gone so long without a fix.

Adopt Trac #1537.  The patch ended up a bit bigger than I expected,
so I suggest we do not merge this into the 6.8 branch.  But there
is no funadamental reason why not.

With this patch, rebindable do-notation really does type as if you
had written the original (>>) and (>>=) operations in desguared form.

I ended up refactoring some of the (rather complicated) error-context
stuff in TcUnify, by pushing an InstOrigin into tcSubExp and its
various calls. That means we could get rid of tcFunResTy, and the
SubCtxt type.  This should improve error messages slightly
in complicated situations, because we have an Origin to hand
to instCall (in the (isSigmaTy actual_ty) case of tc_sub1).

Thanks to Pepe for the first draft of the patch.

This is intended only for debugging use: it makes it easier to
compare two variants without the variations between uniques mattering.

(Of course, you can't actually feed the output to the C compiler
or assembler and expect anything sensible to happen!)

When deriving an instance, the data constructors should all be in scope.
This patch checks the condition.

File locking (of the Haskell 98 variety) was previously done using a
static table with linear search, which had two problems: the array had
a fixed size and was sometimes too small (#1109), and performance of
lockFile/unlockFile was suboptimal due to the linear search.
Also the algorithm failed to count readers as required by Haskell 98
(#629).

Now it's done using a hash table (provided by the RTS).  Furthermore I
avoided the extra fstat() for every open file by passing the dev_t and
ino_t into lockFile.  This and the improvements to the locking
algorithm result in a healthy 20% or so performance increase for
opening/closing files (see openFile008 test).