summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Use the Thread package instead of the [testthread] command to do dgp_thread_leaksdgp2011-08-301-11/+8
| | | thread-enabled testing of the http package.
* Prevent leak of interps when using the [testthread exit] command.dgp2011-08-306-8/+49
|\
| * Prevent segfaults attempting to use thread maps after they've been deleted.bug_3397515dgp2011-08-302-0/+6
| |
| * Work in progress plugging thread finalization memory leaks.dgp2011-08-293-8/+35
| |
* | [Bug 3398794]: Use Tcl errors in scripts, not panics.dkf2011-08-303-46/+84
|\ \
| * | [Bug 3398794]: Use Tcl errors in scripts, not panics.dkf2011-08-303-13/+54
| | |
* | | Put back the check for server sockets (bug #3394732).max2011-08-292-0/+14
| |/ |/|
* | Leak of ReflectedTransformMap.dgp2011-08-292-4/+7
| |
* | Fix eval's faulty objProc, it was actually an nreProc [3399564|Bug 3399564]. ↵mig2011-08-293-2/+13
| | | | | | | | Thanks to Joe Mistachkin for detection and analysis.
* | [3396731] inline string reverse: minor further improvementsjan.nijtmans2011-08-291-32/+22
| |
* | 3396731 Revise the [string reverse] implementation to operate on thedgp2011-08-273-109/+129
|\ \ | | | | | | representation that comes in, avoid conversion to other reps.
| * | Repaired the lost performance in the copy loop hotspots. Now meets or revert_3396731dgp2011-08-271-13/+12
| | | | | | | | | beats the former trunk (and current trunk by magnitudes) in tclbench.
| * | 3396731 Another rewrite of TclStringObjReverse() to make it adopt thedgp2011-08-252-70/+114
| | | | | | | | | | | | nijtmans approach for reversing the objPtr->bytes rep without losing performance.
| * | Revert the [string reverse] rewrite in a branch for more review.dgp2011-08-232-84/+80
| |\ \
* | | | [Enh 3396731] Follow-up: special case for Pure-unicode representationjan.nijtmans2011-08-252-0/+28
|/ / /
* | | typodgp2011-08-231-1/+1
| | |
* | | 3396948 Leak of ReflectedChannelMap.dgp2011-08-232-4/+7
| | |
* | | gcc 4.6 (64-bit) warningsjan.nijtmans2011-08-231-4/+4
|\ \ \ | | |/ | |/|
| * | gcc 4.6 (64-bit) warningsjan.nijtmans2011-08-231-2/+2
| | |
| * | merge-markjan.nijtmans2011-08-220-0/+0
| |\ \
| | * | Fixed use of C99 features breaking the build with MSVC6. The problem was ↵andreask2011-08-222-1/+7
| | | | | | | | | | | | | | | | introduced with checkin [e9a08358f4].
* | | | [FRQ 3396731] inline string reversejan.nijtmans2011-08-232-80/+84
| |_|/ |/| |
* | | merge-markjan.nijtmans2011-08-220-0/+0
|\ \ \ | |/ /
| * | [Bug 1774689] async-4.3 sometimes fails. (backport)ferrieux2011-08-192-1/+2
| | |
| * | [Bug 2981154] async-4.3 segfault. (backport)ferrieux2011-08-192-9/+44
| | |
* | | 3393279, 3393280 ReflectClose(.) is missing Tcl_EventuallyFree() callsdgp2011-08-192-0/+5
| | | | | | | | | at some of its exits.
* | | 3394654, 3393276 Revise FlushChannel() to account for the possibilitydgp2011-08-192-1/+7
| | | | | | | | | that the ChanWrite() call might recycle the buffer out from under us.
* | | Preserve the chanPtr during FlushChannel so that channel driversdgp2011-08-192-4/+16
| | | | | | | | | don't yank it away before we're done with it.
* | | [Bug 1774689] async-4.3 sometimes fails.ferrieux2011-08-192-1/+2
| | |
* | | [Bug 2981154] async-4.3 segfault.ferrieux2011-08-192-11/+49
| | |
* | | [Bug 3096275] Sync fcopy buffers input.ferrieux2011-08-182-5/+14
| | |
* | | remove some debugging stuffjan.nijtmans2011-08-181-12/+1
|\ \ \ | |/ /
| * | remove some debugging stuffjan.nijtmans2011-08-181-12/+1
| |\ \ | | |/
| | * remove some debugging stuffjan.nijtmans2011-08-181-12/+1
| | |
* | | [Bug 3393714] overflow in toupper deltajan.nijtmans2011-08-184-35/+48
|\ \ \ | |/ /
| * | [Bug 3393714] overflow in toupper deltajan.nijtmans2011-08-184-35/+48
| |\ \ | | |/
| | * [Bug 3393714] overflow in toupper deltajan.nijtmans2011-08-184-35/+48
| | |
| * | merge-markjan.nijtmans2011-08-180-0/+0
| |\ \ | | |/
| | * A few more compiler warningsjan.nijtmans2011-08-182-4/+4
| | |
* | | [Bug 2946474] Consistently resume backgrounded flushes+closes when exiting.ferrieux2011-08-173-2/+39
| | |
* | | Document TIP 378's one-way-ness.ferrieux2011-08-172-0/+8
| | |
* | | 3393150 Overlooked free of intreps. (It matters for bignums!)dgp2011-08-172-0/+7
|\ \ \ | |/ /
| * | 3393150 Overlooked free of intreps. (It matters for bignums!)dgp2011-08-172-0/+8
| | |
* | | 3392070 More complete prevention of Tcl_Obj reference cyclesdgp2011-08-172-1/+14
|\ \ \ | | | | | | | | when producing an intrep of ByteCode.
| * | | 3392070 More complete prevention of Tcl_Obj reference cycles bug_3392070dgp2011-08-162-1/+14
| | | | | | | | | | | | when producing an intrep of ByteCode.
* | | | separate test for overflowed and negative Tcl_UniCharjan.nijtmans2011-08-171-3/+6
| | | |
* | | | Fixed the C99/C++ comments introduced by [8d3f0fb215] which break strict C89 ↵andreask2011-08-161-2/+2
|/ / / | | | | | | | | | compilers (AIX, cough, cough)
* | | Small changes to quell gcc warnings and make message generation less ugly.dkf2011-08-165-72/+108
| | |
* | | merge-markjan.nijtmans2011-08-160-0/+0
|\ \ \ | |/ /
| * | [Bug 3388350] mingw64 compiler warningsjan.nijtmans2011-08-1612-16/+35
| |\ \ | | |/