[GHC] #13091: Build broken on amd64 solaris 11
GHC
ghc-devs at haskell.org
Wed May 2 22:07:07 UTC 2018
#13091: Build broken on amd64 solaris 11
-------------------------------------+-------------------------------------
Reporter: kgardas | Owner: (none)
Type: bug | Status: new
Priority: normal | Milestone: 8.6.1
Component: Build System | Version: 8.0.2
Resolution: | Keywords:
Operating System: Solaris | Architecture: x86_64
Type of failure: Building GHC | (amd64)
failed | Test Case:
Blocked By: | Blocking:
Related Tickets: #15075 | Differential Rev(s):
Wiki Page: |
-------------------------------------+-------------------------------------
Comment (by ru0mad):
Hello,
I managed somehow to build ghc-8.4.2 on solaris 11.4 (x86_64), along with
cabal-install and finally stack.
I had to apply some patches :
- to workaround the file locking problem (flock() missing on solaris) :
* in ghc to catch the FileLockingNotSupported exception raised on
solaris by GHC.IO.Handle.Lock
* similar in cabal-install (which clones more or less the GHC code)
* patch in filelock (used by stack) to refer to the file locking from
GHC base
- to have a working building process (the -optl-optl problem + )
- to include <termios.h> in stack where needed (in unix/Terminal.hsc)
As file locking is made a no-op, concurrent acces to package db should be
broken, but so far I didn't have any problem on my standalone solaris box.
Some tests are failing, but most of it seems quite innocuous (mostly
differences in output) but some are more preoccupying, yet it sort of
works (test summary attached).
The patches and building scripts are attached (for ghc, cabal-install,
stack).
As I'm not very experimented in this, I would be very interested to know
if the tweaks I applied still make a workable version or if it is totally
borked.
Bruno
--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13091#comment:6>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
More information about the ghc-tickets
mailing list