| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Add 'make list_brokens' | Ian Lynagh | 2013-02-11 | 1 | -0/+10 |
| | | | | | | | Gives a list of tickets that the testsuite thinks are broken, and what bug it thinks is the reason. This can then be pasted into trac and 'previewed', which will show any closed tickets with strikeout. | ||||
| * | Handles TEST_HC=c:/... on Windows | Ian Lynagh | 2013-01-26 | 1 | -2/+14 |
| | | |||||
| * | Only run dynamicToo001 if we have both vanilla and dynamic libraries | Ian Lynagh | 2012-12-14 | 1 | -1/+8 |
| | | |||||
| * | Use python2 if it exists. Based on a patch from michalt. Fixes #7333. | Ian Lynagh | 2012-11-30 | 1 | -0/+3 |
| | | |||||
| * | Add if_compiler_debugged to testsuite driver | Ian Lynagh | 2012-11-10 | 2 | -0/+3 |
| | | | | | arrowfail001 currently fails only when DEBUG is defined. | ||||
| * | Merge branch 'master' of darcs.haskell.org:/srv/darcs//testsuite | Ian Lynagh | 2012-10-30 | 1 | -24/+40 |
| |\ | | | | | | | | | Conflicts: mk/test.mk | ||||
| | * | fix HAVE_VANILLA when we have both static profiling and vanilla libs | Simon Marlow | 2012-10-30 | 1 | -1/+1 |
| | | | |||||
| | * | make it optional to have DynamicByDefault, to support older GHCs | Simon Marlow | 2012-10-30 | 1 | -24/+40 |
| | | | | | | | | | | | It's often useful to be able to say 'make TEST_HC=ghc' to check that a test really fails with a different GHC. | ||||
| * | | Fix the have-we-got-vanilla-libraries test | Ian Lynagh | 2012-10-30 | 1 | -3/+3 |
| |/ | | | | It thought that the profiling libraries were vanilla | ||||
| * | Testsuite fixes for when we only have dynamic libraries | Ian Lynagh | 2012-10-25 | 1 | -0/+9 |
| | | |||||
| * | ghcilink001 is expected to fail when ghci is dynamic | Ian Lynagh | 2012-10-12 | 1 | -2/+2 |
| | | | | | dynamic ghci can't load .a files | ||||
| * | Add a -static way, enabled only if dynlibs are used by default | Ian Lynagh | 2012-10-03 | 2 | -0/+7 |
| | | |||||
| * | don't include $(EXTRA_HC_OPTS) twice | Simon Marlow | 2012-08-08 | 1 | -1/+0 |
| | | |||||
| * | Avoid using deprecated package-conf GHC flags. | Paolo Capriotti | 2012-06-08 | 2 | -1/+14 |
| | | |||||
| * | Go back to using -no-user-package-conf for a while | Simon Marlow | 2012-05-22 | 1 | -1/+1 |
| | | | | | | | So that we can still use the testsuite with older compilers. We should really test the GHC version, but I don't want to add another $(shell). Maybe modifying ghcconfig.hs would be better. | ||||
| * | Rename package-conf flags to package-db. | Paolo Capriotti | 2012-05-15 | 1 | -1/+1 |
| | | |||||
| * | Fix the way we find the path to 'ar' | Ian Lynagh | 2012-05-03 | 2 | -10/+12 |
| | | | | | Fixes ghcilink001 on Windows | ||||
| * | substitute for $topdir in $(AR) | Simon Marlow | 2012-05-01 | 1 | -2/+12 |
| | | | | | Fixes ghcilink001, ghcilink004 on Windows | ||||
| * | Don't try to run ghc-config when cleaning | Simon Marlow | 2012-05-01 | 1 | -0/+2 |
| | | |||||
| * | Fix llvm check | David Terei | 2012-03-06 | 1 | -1/+1 |
| | | |||||
| * | Wibbles to testsuite/mk/boilerplate.mk | Simon Peyton Jones | 2012-03-05 | 1 | -3/+4 |
| | | | | | ...after consulting with Simon M on a conflict | ||||
| * | Merge branch 'master' of http://darcs.haskell.org//testsuite | Simon Peyton Jones | 2012-03-02 | 1 | -2/+5 |
| |\ | | | | | | | | | Conflicts: mk/boilerplate.mk | ||||
| | * | Follow change in bindisttest install directory name | Ian Lynagh | 2012-02-21 | 1 | -1/+1 |
| | | | |||||
| | * | Remove some Makefile debugging code | Ian Lynagh | 2012-02-10 | 1 | -2/+0 |
| | | | |||||
| | * | Filter colons out of ghc-config-mk | Ian Lynagh | 2012-02-10 | 1 | -1/+3 |
| | | | | | | | | | Otherwise the colon confuses make on Windows | ||||
| | * | Don't use TOP_ABS in Makefile rules | Ian Lynagh | 2012-02-09 | 1 | -4/+7 |
| | | | | | | | | | The : in c:/foo/bar on Windows confuses make. | ||||
| * | | Don't fail if "ar command" is not found | Simon Peyton Jones | 2012-01-20 | 1 | -2/+8 |
| | | | |||||
| * | | Fix for Windows, and delete the .mk file if ghc-config fails | Simon Peyton Jones | 2012-01-20 | 1 | -2/+2 |
| |/ | |||||
| * | fix (non-fast) validate | Simon Marlow | 2012-01-17 | 1 | -2/+4 |
| | | |||||
| * | Revert "The LLVM backend now only works with 2.9+, don't try it if we have 2.7" | David Terei | 2012-01-17 | 1 | -7/+1 |
| | | | | | This reverts commit 0114212f8588211d3dd6735d55127bce2cc1dad7. | ||||
| * | The LLVM backend now only works with 2.9+, don't try it if we have 2.7 | Simon Marlow | 2012-01-16 | 1 | -1/+7 |
| | | | | | | | | This fix is very wrong, but I'm drowning in optllvm failures and need to fix this ASAP (it's somewhat suboptimal that I won't be able to run the LLVM tests here until I upgrade the OS or install a later LLVM manually). | ||||
| * | Parse the ghc --info values using a Haskell program, and cache the results | Simon Marlow | 2012-01-16 | 3 | -33/+48 |
| | | | | | | Should improve startup time of make in the testsuite, and it is simpler. | ||||
| * | fix $(HAVE_PROFILING) on Windows | Simon Marlow | 2011-12-19 | 1 | -1/+1 |
| | | |||||
| * | Get the value of $(AR) from ghc --info | Simon Marlow | 2011-12-13 | 1 | -0/+5 |
| | | | | | On Windows, 'ar' might not necessarily be on $PATH. | ||||
| * | Define the GHC "compiler_always_flags" in make; fixes trac #5213 | Ian Lynagh | 2011-11-17 | 1 | -1/+3 |
| | | | | | We now pass the "always" flags to Makefile tests as well as python tests. | ||||
| * | Don't clean tests that we aren't running | Ian Lynagh | 2011-09-05 | 1 | -0/+6 |
| | | | | | | | | | This can make it a lot faster to run a single test. You can run make CLEANUP=1 CLEAN_ONLY=YES to clean everything without running any tests. | ||||
| * | Add support for checking whether files are written by more than one test | Ian Lynagh | 2011-08-07 | 1 | -0/+4 |
| | | | | | | | | | | | | | | | | | | | | | Work in progress, but largely works. Known issues: * only supported when using the timeout program * 'test.strace' files aren't cleaned, as they end up in the root directory rather than the test's directory * Doesn't yet track what the current directory is, so finds several files like "A.o" being written by multiple tests (and conversely, may be missing writes to the same file from different directories) * Lots of tests write to $HOME/.ghc/ghci_history. We should probably be passing ghci a flag to stop this from happening. * Some .strace lines aren't understood yet, causing framework failures * One .strace file can cause muiltiple framework failures, if it contains lots of lines that aren't understood Performance: Threads fast testsuite time fast testsuite time with checks 1 16:36.14 25:16.07 5 5:33.95 8:04.05 | ||||
| * | Update some more files for ghc-regress removal | David Terei | 2011-07-20 | 1 | -1/+1 |
| | | |||||
| * | TH doesn't work in a stage 1 compiler | Ian Lynagh | 2011-07-13 | 1 | -3/+6 |
| | | |||||
| * | Formatting wibble | David Terei | 2011-07-12 | 1 | -10/+10 |
| | | |||||
| * | Add TEST_HC_OPTS to every Makefile invocation of TEST_HC | Simon Peyton Jones | 2011-04-20 | 1 | -0/+4 |
| | | | | | | | | | | | | | | | | | | | | | | | | In the testsuite there are numerous Makefiles that invoke $(TEST_HC). But, unlike the tests invoked directly by the harness, they didn't pass -dno-debug-output, or $(EXTRA_HC_OPTS). This patch goes part-way towards the goal of making the two consistent: * Almost all invocations in the Makefiles now go '$(TEST_HC)' $(TEST_HC_OPTS) * In test.mk we initialise TEST_HC_OPTS = -dno-debug-output $(EXTRA_HC_OPTS) Still to come (I don't know how to do these) * The invocations of TEST_HC via cabal are not treated (eg cabal01, cabal04), and they should be * TEST_HC_OPTS should probably be initialised the same way as the Python script computes the flags to pass to the TEST_HC that it calls directly | ||||
| * | Pass OUTPUT_SUMMARY on to the testsuite driver | Ian Lynagh | 2010-09-30 | 1 | -1/+8 |
| | | |||||
| * | Fix "/bin/sh: llvmc: command not found" noise when running testsuite | Ian Lynagh | 2010-08-26 | 1 | -1/+1 |
| | | | | | | This also made the "run_command $MAKE" tests fail, as they were also sending this to stderr. | ||||
| * | Enable the optllvm way if llvmc is found | Simon Marlow | 2010-08-23 | 1 | -0/+6 |
| | | |||||
| * | Fix running dyn tests on OS X | Ian Lynagh | 2010-05-03 | 2 | -0/+11 |
| | | |||||
| * | Fix detection of whether we have profiling libs | Ian Lynagh | 2010-03-03 | 1 | -1/+2 |
| | | | | | It broke when the installation path contained a space | ||||
| * | Improve the handling of TEST_HC | Ian Lynagh | 2010-01-08 | 1 | -2/+13 |
| | | | | | | | We now accept TEST_HC=ghc as well as TEST_HC=/usr/bin/ghc The code is also a little more correct, as it will now actually canonicalise the value of TEST_HC if it is given on the commandline. | ||||
| * | Always define IN_TREE_COMPILER in mk/boilerplate.mk | Ian Lynagh | 2010-01-08 | 1 | -0/+2 |
| | | | | | There was one path in which it was not being set. | ||||
| * | Use "test -x" rather than "test -e"; fixes trac #3778 | Ian Lynagh | 2010-01-03 | 1 | -6/+6 |
| | | | | | Solaris doesn't support -e | ||||
| * | Allow tests to behave differently depending on whether the compiler is in-tree | Ian Lynagh | 2009-12-18 | 2 | -0/+8 |
| | | | | | | And skip testwsdeque if it is not in-tree, as we rely on some headers from the build tree. | ||||
 |
index : delta/haskell.git | |
| gitlab.haskell.org: ghc/ghc.git |
| summaryrefslogtreecommitdiff |