ports/130218 Maintainer timeout? (lang/ghc update)

Tue Jul 14 00:29:29 UTC 2009

On Mon, Jul 13, 2009 at 1:15 PM, Wesley Shields<wxs at freebsd.org> wrote:
> On Sun, Jul 12, 2009 at 05:55:22PM -0400, L Campbell wrote:
>> *pokes the haskell@ maintainers*
>>
>> Is there anything blocking the update of lang/ghc? We have an
>> *ancient* version of GHC (the primary Haskell compiler) in ports --
>> old enough to cause problems with Haskell library requirements. It
>> really needs to be updated (or, there needs to be a newer version of
>> GHC in ports).
>>
>> The patch in that PR is for GHC 6.10.1; the current
>> latest-and-greatest is 6.10.3 (which came out in May).
>>
>> Is possible, could the lang/ghc port be brought up-to-date? If not,
>> what can I do to make it possible?
>
> The patches have become stale. If you want to submit an updated patch
> (possibly going to the 6.10.3 if you want) I will do my best to make
> sure it builds properly and is committed. I can't do any runtime testing
> of it though.
>

I've got 6.10.2 built properly right now (generating the pkg-plist
now). I had issues with 6.10.1 and 6.10.3 --

6.10.1 has a broken ghci out of the box on amd64 (mmap issues arising
from the shared library support). See:
http://www.haskell.org/pipermail/glasgow-haskell-users/2008-November/016239.html

6.10.3 doesn't build for me due to issues with haskeline (the Haskell
readline library). I suspect there's a cabal package somewhere which
isn't building correctly then silently failing. Haven't investigated
further because 6.10.2 builds fine and that's good enough for my
needs.

I went ahead and ran the GHC regression testsuite on my built -- got a
whopping 300 failures (of ~12000 test cases; 2.5%). I've only skimmed
through the failure list (it's attached for your browsing pleasure) --
some of the things looks like problems on their end, some of them look
like problems on my end (ie, need FreeBSD-specific patches for things
like "getPermissions001", I imagine).

Some of the failed tests look kind of concerning ("space_leak_001").

I've attached my current diff on the Makefile; I'll post a full diff
when I've got the pkg-plist built; definitely want to have some more
eyes testing it (and I still need to test it on an i386 box) to make
sure nothing is completely foobar.
-------------- next part --------------
OVERALL SUMMARY for test run started at Mon Jul 13 22:06:22 UTC 2009
    2378 total tests, which gave rise to
   12670 test cases, of which
       0 caused framework failures
    2528 were skipped

    9548 expected passes
     274 expected failures
      12 unexpected passes
     308 unexpected failures

Unexpected passes:
   2740(ghci)
   break001(ghci)
   break006(ghci)
   break026(ghci)
   hist001(ghci)
   print003(ghci)
   print005(ghci)
   print006(ghci)
   print010(ghci)
   print012(ghci)
   print014(ghci)
   print020(ghci)

Unexpected failures:
   3171(normal)
   3207(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthread
ed)
   3236(normal,threaded1)
   3279(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthread
ed)
   GADT13(normal,profc,profasm)
   NoMatchErr(normal)
   NonLinearSigErr(normal,hpc,profc,profasm)
   SeqRule(optc,optasm)
   Simple16(normal)
   Simple8(normal,profc,profasm)
   T1074(normal)
   T1074a(normal)
   T1148(normal)
   T1633(normal)
   T1792_imports(normal)
   T1900(normal)
   T1969(normal)
   T2267(normal)
   T246(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthread
ed)
   T2627b(normal)
   T2677(normal)
   T2693(normal)
   T2806(normal)
   T2888(normal,hpc,profc,profasm)
   T3013(normal)
   T3066(normal)
   T3095(normal)
   T3118(optc,hpc,optasm,profc,profasm)
   T3126(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthrea
ded)
   T3153(normal)
   T3155(normal)
   T3163(normal)
   T3177(normal)
   T3177a(normal)
   T3219(normal,optc,hpc,optasm,profc,profasm)
   T3221(normal,optc,hpc,optasm,profc,profasm)
   T3262(normal,optc,hpc,optasm,profc,profasm)
   T3265(normal)
   T3319(normal)
   T3323(normal)
   T3342(normal,optc,hpc,optasm,profc,profasm)
   TH_runIO(normal)
   arith013(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profth
readed)
   break011(ghci)
   break017(ghci)
   break018(ghci)
   break024(ghci)
   break025(ghci)
   cabal01(normal)
   ccfail001(normal)
   ccfail002(normal)
   decodingerror001(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded
2,profthreaded)
   derefnull(profc,profthreaded)
   divbyzero(profc,profthreaded)
   ds057(normal)
   dynbrk001(ghci)
   encoding001(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,pro
fthreaded)
   fdReadBuf001(ghci,threaded1,threaded2)
   getPermissions001(normal,optc,hpc,optasm,profc,profasm,threaded1,threaded2,pr
ofthreaded)
   ghci025(ghci)
   ghcpkg02(normal)
   ghcpkg05(normal)
   hClose002(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,proft
hreaded)
   hClose003(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthreaded)
   hDuplicateTo001(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthreaded)
   hSetBuffering004(ghci)
   hSetEncoding001(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthreaded)
   haddock.Test(normal,hpc,profc)
   hpc_ghc_ghci(normal)
   mod176(normal)
   mod177(normal)
   newline001(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthreaded)
   print022(ghci)
   print028(ghci)
   process007(normal,optc,hpc,optasm,profc,profasm,ghci,threaded1,threaded2,profthreaded)
   queryfdoption01(normal,optc,hpc,optasm,profc,profasm,threaded1,threaded2,profthreaded)
   read036(normal)
   read064(normal,optc,hpc,optasm,profc,profasm)
   read066(normal,optc,hpc,optasm,profc,profasm)
   recomp005(normal)
   recomp006(normal)
   records(normal)
   rn037(normal,optc,hpc,optasm,profc,profasm)
   rn046(normal,optc,hpc,optasm,profc,profasm)
   rnfail052(normal)
   space_leak_001(profc,profasm,profthreaded)
   stack002(normal,optc,hpc,optasm,profc,profasm,threaded1,threaded2,profthreaded) 
   tc168(normal,optc,hpc,optasm,profc,profasm)
   tc231(normal,optc,hpc,optasm,profc,profasm)
   tcfail091(normal)
   tcfail166(normal)
   tcfail183(normal)
   tcfail204(normal)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: ghc-6.10.2.diff
Type: application/octet-stream
Size: 4557 bytes
Desc: not available
Url : http://lists.freebsd.org/pipermail/freebsd-ports/attachments/20090714/139e07dd/ghc-6.10.2.obj