On 05/08/2011 04:21, Manuel M T Chakravarty wrote:
I get
=====> TH_recompile(normal) 1 of 1 [0, 0, 0]
cd .&& $MAKE -s TH_recompile</dev/null>TH_recompile.run.stdout
2>TH_recompile.run.stderr
*** unexpected pass for TH_recompile(normal)
Is there a reason that this is still marked as expected to fail? The
corresponding bug, #481, got fixed a while ago.
Ah! I didn't realise we already had this test, I added the same test in
driver/recomp009.
I haven't been seeing this unexpected pass, due to this:
Actual stdout output differs from expected:
--- ./th/TH_recompile/TH_recompile.stdout 2011-07-21
14:37:55.000000000 +0
100
+++ ./th/TH_recompile/TH_recompile.run.stdout 2011-08-04
16:45:42.341365536 +0100
@@ -1,2 +1,6 @@
+make[3]: Entering directory
`/64playpen/simonmar/validate/testsuite/tests/th/TH_recompile'
+make[4]: Entering directory
`/64playpen/simonmar/validate/testsuite/tests/th/TH_recompile'
+make[4]: Leaving directory
`/64playpen/simonmar/validate/testsuite/tests/th/TH_recompile'
1
2
+make[3]: Leaving directory
`/64playpen/simonmar/validate/testsuite/tests/th/TH_recompile'
cd ./th/TH_recompile && $MAKE -s clean
make[3]: Entering directory
`/64playpen/simonmar/validate/testsuite/tests/th/TH_recompile'
make[3]: Leaving directory
`/64playpen/simonmar/validate/testsuite/tests/th/TH_recompile'
which appears to be due to a missing --no-print-directory in the all.T.
This only happens when you invoke the test through enough levels of
Makefile (:-/).
I'll fix.
Cheers,
Simon
_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc
