#8256: adding locality levels to prefetch# and friends -------------------------------------+------------------------------------ Reporter: carter | Owner: carter Type: task | Status: patch Priority: normal | Milestone: 7.8.1 Component: Compiler | Version: 7.7 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: #8252 ,#8107 -------------------------------------+------------------------------------ Comment (by carter): I had a funny problem with validate {{{ == Start post-build package check Timestamp 2013-09-19 17:26:28 UTC for /Users/carter/Desktop/repoScratcher/ghc/inplace/lib/package.conf.d/package.cache Timestamp 2013-09-19 17:26:28 UTC for /Users/carter/Desktop/repoScratcher/ghc/inplace/lib/package.conf.d (same as cache) using cache: /Users/carter/Desktop/repoScratcher/ghc/inplace/lib/package.conf.d/package.cache == End post-build package check rm -f bindist-list make -r --no-print-directory -f ghc.mk bindist BINDIST=YES "cp" -p ghc/stage1/build/tmp/ inplace/lib/bin/ghc-stage1 cp: ghc/stage1/build/tmp/ is a directory (not copied). make[1]: *** [inplace/lib/bin/ghc-stage1] Error 1 make: *** [binary-dist-prep] Error 2 }}} but the teste suite seems to have run fine when i subsequently did a make test {{{ Unexpected results from: TEST="T5313 static001 T1969 T4801" OVERALL SUMMARY for test run started at Thu Sep 19 14:00:24 EDT 2013 3778 total tests, which gave rise to 14853 test cases, of which 11416 were skipped 28 had missing libraries 3344 expected passes 61 expected failures 0 caused framework failures 1 unexpected passes 3 unexpected failures Unexpected passes: driver T5313 (normal) Unexpected failures: driver static001 [bad stderr] (normal) perf/compiler T1969 [stat not good enough] (normal) perf/compiler T4801 [stat too good] (normal) }}} I will upload the current ghc patch shortly. (same testsuite patch as before). will just clean up the documentation a bit. Note also that for now i've given all the prefetch primops a can_fail=True attribute, but it could be argued just as well that they should have no such attribute, or the has_side_effects=True attribute. I'm unsure whats the right story there. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8256#comment:22 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler