[GHC] #10536: Clear up how to turn off dynamic linking in build.mk

GHC ghc-devs at haskell.org
Tue Jul 7 11:24:44 UTC 2015


#10536: Clear up how to turn off dynamic linking in build.mk
-------------------------------------+-------------------------------------
        Reporter:  thomie            |                   Owner:
            Type:  task              |                  Status:  new
        Priority:  normal            |               Milestone:  7.12.1
       Component:  Build System      |                 Version:  7.10.1
      Resolution:                    |                Keywords:
Operating System:  Unknown/Multiple  |            Architecture:
 Type of failure:  None/Unknown      |  Unknown/Multiple
      Blocked By:                    |               Test Case:
 Related Tickets:                    |                Blocking:
                                     |  Differential Revisions:  Phab:D1021
-------------------------------------+-------------------------------------
Changes (by thomie):

 * differential:   => Phab:D1021


Comment:

 bgamari mentioned in Phab:D1021:
 > The exact semantics of `DYNAMIC_GHC_PROGRAMS` actually perplexed me
 quite a bit at first, in large part due to this overloading of roles.
 Perhaps it would make sense to just tear off the band-aid: We could retain
 `DYNAMIC_GHC_PROGRAMS` and its associated logic for the time being,
 emitting a warning if we find it set. Alongside it we could introduce a
 new `DYNAMIC_GHC_EXECUTABLES` (or some other name) flag strictly intended
 to control whether we link ghc and friends dynamically. Finally, we could
 introduce another flag for specifying whether dynamic libraries should be
 built. Eventually `DYNAMIC_GHC_PROGRAMS` could be dropped.

 thoughtpolice:
 > I'm somewhat against putting in all this work because it does sound
 confusing, but also because I think the state of shared library support is
 in the air (personally I'm coming around to the "Nuke it from orbit"
 position.) So this improvement is at least a cleanup with no substantial
 semantics change for right now.

 Leaving this ticket open till we make up our mind about this.

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10536#comment:2>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler


More information about the ghc-tickets mailing list