For what it's worth, I got these failures with sh validate --fast in HEAD this morning:

Unexpected failures:
   ../../libraries/directory/tests  getPermissions001 [bad stdout] (normal)
   perf/compiler                    T1969 [stat too good] (normal)
   perf/compiler                    T3294 [stat not good enough] (normal)
   perf/compiler                    T4801 [stat not good enough] (normal)
   perf/compiler                    T783 [stat not good enough] (normal)
   perf/haddock                     haddock.Cabal [stat not good enough] (normal)
   perf/haddock                     haddock.base [stat not good enough] (normal)
   perf/haddock                     haddock.compiler [stat too good] (normal)
   perf/should_run                  Conversions [stat not good enough] (normal)
   perf/should_run                  T4830 [stat too good] (normal)
   perf/should_run                  T5536 [stat too good] (normal)
   perf/should_run                  T7436 [stat too good] (normal)
   perf/should_run                  lazy-bs-alloc [stat too good] (normal)
   simplCore/should_compile         T3717 [stderr mismatch] (optasm)
   simplCore/should_compile         T4908 [stderr mismatch] (optasm)
   simplCore/should_compile         T7360 [stderr mismatch] (optasm)
   simplCore/should_compile         spec-inline [stderr mismatch] (optasm)
   typecheck/should_compile         holes [stderr mismatch] (normal)
   typecheck/should_compile         holes2 [stderr mismatch] (normal)
   typecheck/should_compile         holes3 [stderr mismatch] (normal)


But then again, looking at it, I'm not sure what to do about the perf ones. Should
we be more flexible in the expected result, or should I have fewer expectations
about my machine being representative?

And getPermissions001 might or might not have to do with the fact that this is on
a virtual machine...


Cheers,
Pedro

On Thu, Feb 7, 2013 at 10:13 AM, Simon Marlow <marlowsd@gmail.com> wrote:
These are still failing, BTW.

Cheers,
        Simon


On 01/02/13 14:51, Simon Peyton-Jones wrote:
yes, see earlier email on this. almost certainly my fault, will fix Monday

|  -----Original Message-----
|  From: ghc-devs-bounces@haskell.org [mailto:ghc-devs-bounces@haskell.org] On
|  Behalf Of Simon Marlow
|  Sent: 01 February 2013 08:57
|  To: ghc-devs@haskell.org
|  Subject: Validate failures
|
|  A batch of testsuite failures this morning - forgotten push to testsuite
|  perhaps?
|
|  Unexpected failures:
|      perf/should_run           T5113 [stat not good enough] (normal)
|      simplCore/should_compile  T3717 [stderr mismatch] (optasm)
|      simplCore/should_compile  T4908 [stderr mismatch] (optasm)
|      simplCore/should_compile  T7360 [stderr mismatch] (optasm)
|      simplCore/should_compile  spec-inline [stderr mismatch] (optasm)
|      typecheck/should_compile  holes [stderr mismatch] (normal)
|      typecheck/should_compile  holes2 [stderr mismatch] (normal)
|      typecheck/should_compile  holes3 [stderr mismatch] (normal)
|
|  Cheers,
|       Simon
|
|  _______________________________________________
|  ghc-devs mailing list
|  ghc-devs@haskell.org
|  http://www.haskell.org/mailman/listinfo/ghc-devs



_______________________________________________
ghc-devs mailing list
ghc-devs@haskell.org
http://www.haskell.org/mailman/listinfo/ghc-devs