[GHC] #8296: Patch: new primops for byte range copies ByteArray# <-> Addr#

[GHC]

#8296: Patch: new primops for byte range copies ByteArray# <-> Addr#
-------------------------------------+------------------------------------
        Reporter:  duncan            |            Owner:
            Type:  feature request   |           Status:  closed
        Priority:  normal            |        Milestone:
       Component:  Compiler          |          Version:  7.6.3
      Resolution:  fixed             |         Keywords:
Operating System:  Unknown/Multiple  |     Architecture:  Unknown/Multiple
 Type of failure:  None/Unknown      |       Difficulty:  Unknown
       Test Case:                    |       Blocked By:
        Blocking:                    |  Related Tickets:
-------------------------------------+------------------------------------

Comment (by duncan):

 Oh sorry about that. I see that I changed the primops to use `State#
 RealWorld` after writing the tests and then didn't update the tests.

 So actually I think using `RealWorld` here is probably right, because
 these ops work with `Addr#`s, both reading and writing from them, and so
 this does not really make sense in a pure ST context. Admititly, the
 pattern in the existing primops is not totally clear about this, things
 like `touch#` use `State# RealWorld`, but then all the primops like
 `readIntOffAddr#` use `State# s`.

 So not totally clear cut either way. I'll leave the choice to you. If you
 want to switch back to using `State# RealWorld` as in my original patch,
 then I attach the corresponding fix to the tests.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8296#comment:7>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler