Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | [Bug 967195]: Make tcltest work when tclsh is compiled without using the ↵ | jan.nijtmans | 2011-11-30 | 1 | -1/+1 |
|\ | | | | | | | setargv() function on mingw. | ||||
| * | [Bug 967195]: Make tcltest work when tclsh is compiled without using the ↵ | jan.nijtmans | 2011-11-30 | 2 | -2/+2 |
| |\ | | | | | | | | | | setargv() function on mingw. | ||||
* | | | Revise results of the thread-7.28 through thread-7.31 tests to account for ↵ | Joe Mistachkin | 2011-11-21 | 1 | -236/+322 |
| | | | | | | | | | | | | the fact they are canceled via a script sent to the thread asynchronously, which then impacts the error message handling. Attempt to manually drain the event queue for the main thread after joining the test thread to make sure no stray events are processed at the wrong time on the main thread. Revise all the synchronization and comparison semantics related to the thread id and error message. | ||||
* | | | Remove unnecessary [after] calls from the thread tests. Make error message ↵ | Joe Mistachkin | 2011-11-21 | 1 | -8/+14 |
| | | | | | | | | | | | | matching more robust for tests that may have built-in race conditions. Test thread-7.26 must first unset all thread testing related variables. | ||||
* | | | Remove all use of thread::release from the thread 7.x tests, replacing it ↵ | Joe Mistachkin | 2011-11-18 | 1 | -50/+67 |
| | | | | | | | | | | | | with a script that can easily cause 'stuck' threads to self-destruct for those test cases that require it. Also, make the error message handling far more robust by keeping track of every asynchronous error. | ||||
* | | | Refactor all the remaining thread-7.x tests that were using [testthread]. ↵ | Joe Mistachkin | 2011-11-18 | 1 | -88/+67 |
| | | | | | | | | | | | | Note that this test file now requires the very latest version of the Thread package to pass all tests. In addition, the thread-7.18 and thread-7.19 tests have been flagged as knownBug because they cannot pass without modifications to the [expr] command, persuant to TIP #392. | ||||
* | | | Generalize previous fix to all of zlib.test; also, tidy up [chan configure] ↵ | ferrieux | 2011-11-11 | 1 | -23/+23 |
| | | | | | | | | | | | | flags across zlib.test. | ||||
* | | | Use nonblocking writes in single-threaded IO tests to avoid deadlocks when ↵ | ferrieux | 2011-11-11 | 1 | -1/+1 |
| | | | | | | | | | | | | going beyond OS buffers [Bug 3428756]. | ||||
* | | | More work in progress converting tests from [testthread] to Thread package. | dgp | 2011-11-09 | 1 | -262/+235 |
| | | | |||||
* | | | missing constraint and too brittle result | dgp | 2011-11-08 | 2 | -2/+3 |
| | | | |||||
* | | | missing constraints | dgp | 2011-11-08 | 2 | -18/+138 |
|\ \ \ | |/ / | |||||
| * | | Missing constraints | dgp | 2011-11-08 | 2 | -16/+136 |
| | | | |||||
* | | | Work in progress converting tests from [testthread cancel] to [thread::cancel] | dgp | 2011-11-01 | 2 | -348/+270 |
| | | | |||||
* | | | 3414754 Fix the PATHFLAGS != 0 intrep normalizing trailing slashes. | dgp | 2011-10-31 | 1 | -0/+4 |
|\ \ \ | |/ / | |||||
| * | | Added test case to the test suite | dgp | 2011-10-27 | 1 | -0/+4 |
| | | | |||||
| * | | Backport 2902268 fix.bug_2902268 | dgp | 2011-10-26 | 1 | -6/+27 |
| | | | |||||
| * | | Don't cache the system timezone when it was derived from TCL_TZ or TZ. | max | 2011-10-18 | 1 | -0/+33 |
| | | | |||||
* | | | Resolver fix from Stefan Sobernig. | dkf | 2011-10-20 | 1 | -0/+200 |
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * generic/tclLiteral.c (TclInvalidateCmdLiteral): [Bug 3418547]: Additional code for handling the invalidation of literals. * generic/tclBasic.c (Tcl_CreateObjCommand, Tcl_CreateCommand) (TclRenameCommand, Tcl_ExposeCommand): The four additional places that need extra care when dealing with literals. * generic/tclTest.c (TestInterpResolverCmd): Additional test machinery for interpreter resolvers. | ||||
| * | | | And the failing test file too... | dkf | 2011-10-15 | 1 | -0/+200 |
| | | | | |||||
* | | | | Don't cache the system timezone when it was derived from TCL_TZ or TZ. | max | 2011-10-18 | 1 | -0/+33 |
|/ / / | |||||
* | | | revert test-case change for [Bug 2935503] | jan.nijtmans | 2011-10-13 | 1 | -3/+3 |
|\ \ \ | |/ / | |||||
| * | | revert test-case from [Bug 2935503] | jan.nijtmans | 2011-10-13 | 1 | -3/+3 |
| |\ \ | | |/ | |||||
| | * | revert test-case change from [Bug 2935503] | jan.nijtmans | 2011-10-13 | 1 | -3/+3 |
| | | | |||||
* | | | [Bug 2935503] Incorrect mode field returned by file stat command | jan.nijtmans | 2011-10-11 | 1 | -4/+4 |
|\ \ \ | |/ / | |||||
| * | | [Bug 2935503] Incorrect mode field returned by file stat command | jan.nijtmans | 2011-10-11 | 1 | -4/+4 |
| |\ \ | | |/ | |||||
| | * | [Bug 2935503] Incorrect mode field returned by file stat command | jan.nijtmans | 2011-10-11 | 1 | -3/+3 |
| | | | |||||
* | | | * generic/tclCompCmds.c (TclCompileDictWithCmd): Corrected handling of | dkf | 2011-10-09 | 1 | -0/+48 |
| | | | | | | | | | | | | qualified names, and added spacial cases for empty bodies (used when [dict with] is just used for extracting variables). | ||||
* | | | Fix env.test, when running under wine 1.3 | jan.nijtmans | 2011-10-07 | 1 | -0/+3 |
|\ \ \ | |/ / | |||||
| * | | Fix env.test, when running under wine 1.3 (partly backported from Tcl 8.6) | jan.nijtmans | 2011-10-07 | 1 | -28/+38 |
| |\ \ | | |/ | |||||
| | * | Fix env.test, when running under wine 1.3 (partly backported from Tcl 8.6) | jan.nijtmans | 2011-10-07 | 1 | -19/+33 |
| | | | |||||
* | | | Added some tests. | dkf | 2011-10-05 | 1 | -0/+66 |
| | | | |||||
* | | | [Bug 3414769]: Updated list of default-hidden commands for safe interps. | dkf | 2011-09-29 | 1 | -1/+1 |
| | | | |||||
* | | | * generic/tclIndexObj.c (Tcl_ParseArgsObjv): [Bug 3413857]: Corrected | dkf | 2011-09-27 | 1 | -6/+32 |
|\ \ \ | | | | | | | | | | | | | | | | | the memory management for the code parsing arguments when returning "large" numbers of arguments. Also unbroke the TCL_ARGV_AUTO_REST macro in passing. | ||||
| * | | | Test harness for Tcl_ParseArgsObjvbug_3413857 | dkf | 2011-09-27 | 1 | -6/+32 |
| | | | | |||||
* | | | | Make [file] itself be safe, to reduce breakage in existing code. [Bug 3211758] | dkf | 2011-09-26 | 1 | -0/+16 |
|/ / / | |||||
* | | | Stop using the deprecated thread management commands of the tcltest package. | dgp | 2011-09-23 | 1 | -11/+29 |
| | | | | | | | | | | | | The test suite ought to provide these tools for itself. They do not belong in a testing harness. | ||||
* | | | Revise the thread exit handling of the [testthread] command so that it | dgp | 2011-09-21 | 1 | -1/+0 |
| | | | | | | | | | | | | properly maintains the per-process data structures even when the thread exits for reasons other than the [testthread exit] command. | ||||
* | | | Remove constraint on test which apparently passes | dkf | 2011-09-21 | 1 | -1/+1 |
| | | | |||||
* | | | Conversion from [testthread] to Thread package stops most memory leaks. | dgp | 2011-09-19 | 1 | -94/+89 |
| | | | |||||
* | | | Plug most memory leaks in thread.test. | dgp | 2011-09-19 | 1 | -16/+22 |
|\ \ \ | |||||
| * | | | stop segfaultthread_leaks | dgp | 2011-09-12 | 1 | -1/+5 |
| | | | | |||||
| * | | | Attempt to convert test thread-7.26 | dgp | 2011-09-12 | 1 | -9/+9 |
| | | | | |||||
* | | | | Plug all memory leaks in ioCmd.test exposed by `make valgrind`. | dgp | 2011-09-19 | 1 | -152/+157 |
|\ \ \ \ | |||||
| * | | | | Constrain test iocmd.tf-32.1 to be skipped during valgrinding. It contains a iocmd_leaks | dgp | 2011-09-19 | 1 | -4/+15 |
| | | | | | | | | | | | | | | | memory leak that cannot be plugged while testing what the test aims to test. | ||||
| * | | | | Plug a number of MarshallError memleaks. | dgp | 2011-09-19 | 1 | -5/+0 |
| | | | | | |||||
| * | | | | Revise the tests that confront background flush on close across threads. | dgp | 2011-09-18 | 1 | -4/+7 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Need cleanup code to bring an end to the otherwise endless loop of thread finalization that continually tries to flush before closing, and is continually thwarted by a driver raising EAGAIN. If this dance isn't cleanly terminated, it continues and corrupts any subsequent tests that define a [foo] command. | ||||
| * | | | | revise iocmd.tf-24.16 result to deal with Thread conversion | dgp | 2011-09-12 | 1 | -1/+1 |
| | | | | | |||||
| * | | | | Convert uses of the [testthread] command to uses of the Thread package. | dgp | 2011-09-09 | 1 | -148/+144 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This takes as many leaks as possible out of the testing harness, so the leaks remaining are the fault of the tested code. Committed to a branch because the conversion creates new test failures that need review. | ||||
* | | | | | Revise tests. You can't robustly thread::release a thread that's not ↵ | dgp | 2011-09-16 | 1 | -4/+2 |
| | | | | | | | | | | | | | | | | | | | | thread::wait-ing | ||||
* | | | | | Made test socket-14.1 more robust to stop failure on OS X (Snow Leopard) | dgp | 2011-09-16 | 1 | -0/+1 |
| | | | | |