I've built a fully optimised HEAD (20110901) and ran a fulltest after 
installing a few libraries required for some tests.

OVERALL SUMMARY for test run started at Fr 2. Sep 00:12:23 CEST 2011
    2935 total tests, which gave rise to
   12690 test cases, of which
       0 caused framework failures
    2380 were skipped

   10031 expected passes
     229 expected failures
       0 unexpected passes
      50 unexpected failures

Breaking down the failures by causes:


   profiling/should_run                5314 [bad exit code] 
(prof_hc_hb,prof_hb)
   profiling/should_run                T3001 [bad exit code] (prof_hb)
   profiling/should_run                heapprof001 [bad exit code] 
(prof_hc_hb,prof_hb)

These all failed with an internal error, e.g.:

T3001: internal error: Invalid object in processHeapClosureForDead(): 
2131671296

The same error occurs pretty instantaneously trying to get a -hb profile 
for other programmes.
That looks quite serious.


   profiling/should_run                T3001-2 [bad exit code] (prof_hb)

That one was timeout-killed, running it manually and watching, it didn't 
produce any output (test.bin remained empty after eight minutes at 100% 
CPU, didn't react to ^C).
strace ends with hundreds of
--- SIGVTALRM (Virtual timer expired) @ 0 (0) ---
rt_sigreturn(0)                         = 140414542696416
--- SIGVTALRM (Virtual timer expired) @ 0 (0) ---
rt_sigreturn(0)                         = 140414542696416
--- SIGVTALRM (Virtual timer expired) @ 0 (0) ---
rt_sigreturn(0)                         = 140414542696416
--- SIGVTALRM (Virtual timer expired) @ 0 (0) ---
rt_sigreturn(0)                         = 140414542696416
^Cdafis@schwartz:~/GHC/clean/testsuite/tests/profiling/should_run>

Changing the compile options from -auto-all to -auto, it also quickly fails 
with an Invalid object in processHeapClosureForDead().
http://hackage.haskell.org/trac/ghc/ticket/5451



   typecheck/should_run                T4809 [exit code non-0] 
(normal,hpc,optasm,profasm,threaded1,threaded2,dyn,profthreaded)
   typecheck/should_run                T4809 [bad stdout or stderr] (ghci)

Fails due to:

T4809_IdentityT.hs:32:10:
    Qualified name in binding position: HSX.XML

T4809_IdentityT.hs:33:13:
    Qualified name in binding position: HSX.Child

That failure occurs also with normal validate builds.




   driver                              5313 [bad exit code] (ghci)

Stderr:
5313: '/home/dafis/GHC/lib/ghc-7.3.20110901'/settings: openFile: does not 
exist (No such file or directory)

The same message (modulo location) occurs with normal validate builds.



   plugins                             plugins05 [exit code non-0] 
(profasm,dyn,profthreaded)
   plugins                             plugins06 [exit code non-0] 
(profasm,dyn,profthreaded)

Fail with:

Loading package ghc-7.3.20110901 ... linking ... done.
Loading package ffi-1.0 ... linking ... done.
<no location info>:
    Dynamic linking required, but this is a non-standard build (eg. prof).
    You need to build the program twice: once the normal way, and then
    in the desired way using -osuf to set the object file suffix.

dyn also fails thus for normal validate builds.

   quasiquotation/qq005                qq005 [exit code non-0] (dyn)

basically the same, but with location info.


   quasiquotation/qq006                qq006 [stderr mismatch] (normal)

More output than expected and the location of the expected error moved one 
line down.



   deriving/should_run                 T3087 [exit code non-0] 
(normal,hpc,optasm,profasm,threaded1,threaded2,dyn,profthreaded)
   deriving/should_run                 T3087 [bad stdout or stderr] (ghci)

Message:

T3087.hs:28:19:
    Ambiguous occurrence `ext2Q'
    It could refer to either `Main.ext2Q', defined at T3087.hs:20:1
                          or `Data.Generics.ext2Q',
                             imported from `Data.Generics' at 
T3087.hs:5:1-20
                             (and originally defined in 
`Data.Generics.Aliases')

T3087.hs:31:20:
    Ambiguous occurrence `ext2Q'
    It could refer to either `Main.ext2Q', defined at T3087.hs:20:1
                          or `Data.Generics.ext2Q',
                             imported from `Data.Generics' at 
T3087.hs:5:1-20
                             (and originally defined in 
`Data.Generics.Aliases')

Change in syb?


   programs/galois_raytrace            galois_raytrace [exit code non-0] 
(normal,hpc,optasm,profasm,threaded1,threaded2,dyn,profthreaded)
   programs/galois_raytrace            galois_raytrace [bad stdout or 
stderr] (ghci)

Fails due to Haskell98 imports (Array, List, ...), the same applies to the 
raytrace test in hpc. Fixing the imports, galois_raytrace passes, but 
hpc_raytrace produces bad stdout:

   ../../libraries/hpc/tests/raytrace  hpc_raytrace [bad stdout] 
(normal,hpc,optasm,profasm,threaded1,threaded2,dyn,profthreaded)


_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc

Reply via email to