summaryrefslogtreecommitdiff
path: root/testsuite/driver
Commit message (Collapse)AuthorAgeFilesLines
...
* Make it possible to test the generated assemblyJohan Tibell2012-10-091-1/+53
| | | | | | | | | | | | | | | | | | | | | | | | | | | This test framework feature is inspired by a similar feature in LLVM. The programmer writes a bit of Cmm #include "Cmm.h" // Large memcpy's should lower to calls. callMemcpy { W_ dst, src; prim %memcpy(dst "ptr", src "ptr", 1024, 4) []; } and asserts what the generated assembly should look like, modulo register naming. callMemcpy: movq ; Move arguments into place movq movl movl call memcpy Patch edited and updated by Simon Marlow, and I also added a test for unrolling memcpy and a simple constant-propagation test.
* We weren't generating a stats file for tests that only had ↵Simon Peyton Jones2012-10-031-1/+1
| | | | | | stats_range_fields perf tests I fixed this in a plausible looking way, similar to compiler_stats_range/num fields.
* Add support for per-test timeout adjustment (timeout_multiplier), and ↵Edward Z. Yang2012-09-252-8/+21
| | | | | | tighten up #367 test. Signed-off-by: Edward Z. Yang <ezyang@mit.edu>
* Print stdout diff even if stderr diff alrady fails (#7247)Joachim Breitner2012-09-211-2/+4
|
* Add some haddock space usage testsIan Lynagh2012-08-141-0/+5
|
* Fix a testsuite driver failure on cygwinIan Lynagh2012-05-221-0/+11
|
* dead code: guess_compiler_flags()Simon Marlow2012-05-221-9/+0
|
* Rename package-conf flags to package-db.Paolo Capriotti2012-05-151-2/+2
|
* Driver tweakIan Lynagh2012-05-081-4/+4
|
* Change how we detect whether the terminal is cygwin or msysIan Lynagh2012-05-011-9/+9
| | | | sys.platform now always returns 'win32' for me.
* Allow OS-specific test outputs in the driverIan Lynagh2012-05-011-0/+1
|
* Simplify the stdout-* filename generation in the driverIan Lynagh2012-05-011-17/+8
|
* Testsuite driver tweak for WindowsIan Lynagh2012-04-271-3/+9
| | | | | This handles the case where we redirect stdout/stderr to a file, but the command fails quickly and the file is not actually created.
* fix bug in stats checkingSimon Marlow2012-04-261-2/+2
|
* Add combined_output option.Paolo Capriotti2012-04-232-5/+17
| | | | | | | | | Added an option to combine stdout and stderr into a single file. This is useful for ghci scripts that produce interleaved errors and normal output. Also modified check_stderr_ok so that it normalizes stderr in the same way as compile tests.
* Disable threads with python 2.7.2.Paolo Capriotti2012-03-061-0/+6
|
* Remove some out-of-date normalisationIan Lynagh2012-02-161-4/+0
|
* add req_smpSimon Marlow2012-02-162-0/+7
|
* Allow perf bounds to be specified as base + percentage deviationSimon Marlow2012-02-162-3/+56
| | | | | | | | | | | | | | | | | | | | | | | | | This makes it much easier to update the bounds. Instead of coming up with a suitable (min,max) pair, you just give e.g. (base, 10) to allow 10% deviation from the base figure, which can be pasted from the error. e.g. previously: - # expected value: 458700632 (amd64/Linux): - if_wordsize(64, - compiler_stats_num_field('bytes allocated', 440000000, - 480000000)), now: + if_wordsize(64, # sample from amd64/Linux 15/2/2012 + compiler_stats_range_field('bytes allocated', 360243576, 10)), Note: use stats_range_field rather than stats_num_field. I left support for the old way for now so that we can do a gradual migration. (next I suppose we should make it so that 'make accept' works for perf tests, but that's for another day)
* Make the compiler_always_flags modifiable on a per-test basisSimon Marlow2012-02-162-5/+11
| | | | Motivation: we wanted to disable -dcore-lint for the perf tests
* Add normalise_exe for removing ".exe" extensions in output on WindowsSimon Marlow2012-01-171-0/+7
|
* Count the number of tests that failed due to reqlib()Simon Marlow2012-01-062-6/+19
| | | | | | | | | | | | | | OVERALL SUMMARY for test run started at Fri Jan 6 09:46:46 GMT 2012 82 total tests, which gave rise to 820 test cases, of which 0 caused framework failures 818 were skipped 0 expected passes 2 had missing libraries <--- 0 expected failures 0 unexpected passes 0 unexpected failures
* add test for #5747Simon Marlow2012-01-061-0/+8
|
* Ignore IDLE in .prof.sample (fix spurious profthreaded failures)Simon Marlow2011-11-301-1/+1
|
* add req_shared_libs and use itSimon Marlow2011-11-161-0/+4
|
* Set -no-hs-main if the test is not Haskell codeSimon Marlow2011-11-161-0/+6
| | | | Required by the changes to GHC to generate main() when linking.
* Fix meaning of the only_ways field (fixed cgrun057(normal) in validate)Simon Marlow2011-11-162-5/+5
| | | | | | | | | | | | | The default setting for only_ways was [], which meant "all ways". However, it's really useful to have [] mean "no ways". In cgrun057 I used only_ways(prof_ways) to mean "only run this in the profiling ways", but if profiling is disabled then prof_ways = [] and this doesn't do what we want. Profiling is normally disabled in validate, but I had it enabled because I was testing profiling. So now only_ways defaults to None, which means "all ways", and [] now means "no ways".
* cleaning of testsuiteDavid Terei2011-11-151-4/+0
|
* remove accidentally-left-in debugging outputSimon Marlow2011-11-141-2/+0
|
* shouldn't fail the test if we are accepting outputSimon Marlow2011-11-091-4/+8
|
* normalise slashes in error messagesSimon Marlow2011-11-091-0/+2
| | | | Fixes various Windows test failures (T3103, plugins04)
* allow error message normalisation to be customisedSimon Marlow2011-11-092-9/+26
|
* Don't give normalised diff outputIan Lynagh2011-11-051-5/+0
| | | | | | It is in general illegible (e.g. all newlines have been removed), and in the rare cases you really need it then you can diff the normalised outputs manually.
* some more profiling testsSimon Marlow2011-11-021-0/+4
|
* code for comparing profile outputsSimon Marlow2011-11-021-1/+48
| | | | | | | To add a sample profile, create a file <test>.prof.sample. The test driver normalises the file to remove the actual data (timing and allocations), so that the test will compare the shape of the profile and the entry counts only.
* Add more Safe Haskell tests for new designDavid Terei2011-10-251-3/+3
|
* Merge branch 'master' of http://darcs.haskell.org/testsuiteIan Lynagh2011-10-182-17/+23
|\
| * generalise cmd_prefix to a general function to transform the commandSimon Marlow2011-10-182-17/+23
| |
* | Get rid of the multisrc* test typesIan Lynagh2011-10-181-15/+2
| | | | | | | | | | | | | | They added too much complexity for their very few uses. Fixes: framework failure for tc251(runTest) Unhandled exception: too many values to unpack
* | Catch any uncaught exceptions in test_common_workIan Lynagh2011-10-181-89/+92
|/ | | | | | | | Before, the testsuite just printed Unhandled exception in thread started by <function test_common_thread at 0x188bde8> [...] and carried on when using multiple threads, whereas it would just fall over in single-threaded mode.
* Have multi[src,mod] test ways clean up after themselves.David Terei2011-10-111-0/+6
|
* add a test for foreign imports in GHCiSimon Marlow2011-10-061-0/+3
|
* Driver: define config.libdir for tests to useIan Lynagh2011-10-051-1/+1
| | | | | We also use it internally, which saves us running "ghc --print-libdir" on Windows.
* Tell python to let us use the "with X:" syntaxIan Lynagh2011-09-251-8/+4
| | | | This fixes the testsuite driver with python 2.5. Patch from Florian Weimer.
* Don't use "with lock" syntax in the driverIan Lynagh2011-09-241-1/+8
| | | | | Older versions of python, including that on the Windows nightly builder, throw "SyntaxError: invalid syntax".
* Rejig how performance test stats are storedIan Lynagh2011-09-182-8/+12
| | | | | | In particular, this allows more specific results to be specified, e.g. if_wordsize(64, stats_num_field('foo', 50, 54)), if_platform('x86_64-apple-darwin', stats_num_field('foo', 62, 66)),
* Don't clean tests that we aren't runningIan Lynagh2011-09-051-7/+8
| | | | | | | | 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.
* Improve check-files-written modeIan Lynagh2011-09-041-20/+48
| | | | Also look for files that are written but aren't cleaned
* Improve the check-files-written testsuite driver codeIan Lynagh2011-09-041-17/+23
|
* Check explicitly for writes to ghci_historyIan Lynagh2011-09-011-4/+16
|
View Lorry Controller Status