summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* [Bug 3606139]: missing error check allows regexp to crash Tcl. Thanks to Tom ↵jan.nijtmans2013-02-273-0/+13
| | | | Lane for providing the test-case and the patch.
* structs ChannelHandler ChannelHandlerEvent GetsState CopyState used locally.dgp2013-02-262-83/+82
| | | Remove from tclIO.h.
* struct NextChannelHandler used only locally. Remove from tclIO.h.dgp2013-02-262-25/+24
|
* Repair linked list management in Tcl_DeleteCloseHandler(). dgp2013-02-252-12/+18
| | | CloseCallback struct is used only locally. Remove from tclIO.h.
* Use iPtr field instead of source field to mark a CompileEnv as uninitialized.dgp2013-02-221-4/+4
| | | envPtr->source == NULL can actually be valid (at least when merging forward).
* Shift more burden of smart cleanup onto the TclFreeCompileEnv() routine.dgp2013-02-223-49/+50
|\ | | | | Stop crashes when the hookProc raises an error.
| * Protect against multiple uses of a CompileEnv with only one initialization.dgp2013-02-212-49/+44
|/ | | | | Make TclFreeCompileEnv smarter about cleanup so all callers do not have to be. Revise TclSetByteCodeFromAny() so that when hookProc raises an error, bytecode is not generated. This was rumored to cause crashes.
* 3605447 Make sure the -clear option to [namespace export] always clears,dgp2013-02-213-40/+35
|\ | | | | whether or not new export patterns are specified.
| * added testbug_3605447dgp2013-02-211-0/+8
| |
| * The flag TCL_LEAVE_ERR_MSG has no effect on the routinedgp2013-02-211-8/+4
| | | | | | | | TclGetNamespaceForQualName() so for goodness sake stop making any special efforts to add it in when making calls.
| * refinementdgp2013-02-201-2/+1
| |
| * 36054447 Convert [namespace export -clear] interface to something less stupid.dgp2013-02-201-32/+18
|/ | | Test suite does not demand the stupidity continue, thank goodness.
* Add test-case for Bug #2438181 (which passes in Tcl 8.4 but fails in 8.5/8.6).jan.nijtmans2013-02-191-0/+10
| | | Provided by Poor Yorick
* revise test numbering for forward mergingdgp2013-02-151-1/+4
|
* 3604074 Fix regexp optimization to stop hanging on the expressiondgp2013-02-153-31/+60
|\ | | | | ((((((((a)*)*)*)*)*)*)*)* . Thanks to Bjørn Grathwohl for discovery.
| * New branch bug-3604074 with improved patch to correct fixempties() failurebug_3604074dgp2013-02-142-31/+54
|/ | | to converge.
* Add some extra paths on Windows for finding tclConfig.sh, for mSys and ↵jan.nijtmans2013-02-141-0/+8
| | | | Cygwin shell.
* Backport various improvements from Tcl 8.5jan.nijtmans2013-02-124-93/+234
|
* [Bug 3603434] Make sure TclpObjNormalizePath() properly declares "a:/" todgp2013-02-052-0/+18
|\ | | | | be normalized, even when no "A:" drive is present on the system.
| * Fix for Bug 3603434.bug_3603434dgp2013-02-051-0/+12
|/
* (::platform::LibcVersion): See [Bug 3599098]: Fixed the RE extractingandreask2013-01-303-3/+11
| | | | | | the version to avoid issues with recent changes to the glibc banner. Now targeting a less variable part of the string. Bumped package to version 1.0.11.
* [Bug 3601804]: platformCPUID segmentation fault on Darwinjan.nijtmans2013-01-262-0/+13
|
* new version of cpuid, which doesn't use the edi register any more. Hopefully ↵jan.nijtmans2013-01-241-4/+3
| | | | that works better on some Darwin.
* revert [273bbe926d]: it doesn't work on i386jan.nijtmans2013-01-241-1/+1
|
* Silence some compiler warnings.dgp2013-01-243-0/+11
|
* Protect Tcl_GetIndexFromObjStruct from invalid "offset" values, like 0 or ↵jan.nijtmans2013-01-231-1/+5
| | | | -1. Undocumented, because I don't want to promote people start using that.
* Bug [3601804]: platformCPUID segmentation fault on Darwinjan.nijtmans2013-01-221-1/+1
|
* and changelogjan.nijtmans2013-01-161-0/+8
|
* Allow win32 build with -DTCL_NO_DEPRECATED, just as the UNIX build, off by ↵jan.nijtmans2013-01-164-14/+18
| | | | | default. Define Tcl_EvalObj and Tcl_GlobalEvalObj as macros, even when TCL_NO_DEPRECATED is defined, so Tk can benefit from it too (this is not what TCL_NO_DEPRECATED is supposed to do).
* Backported [Bug 2882342]: correct struct _REPARSE_DATA_BUFFER in tcl 8.4jan.nijtmans2013-01-091-15/+16
|
* [Bug 3092089]: [file normalize] can remove path components.jan.nijtmans2013-01-082-5/+12
|\ | | | | [Bug 3587096] win vista/7: "can't find init.tcl" when called via junction.
| * new attempt for better fixbug_3092089jan.nijtmans2013-01-081-11/+6
| |
| * suggested fix for Bug 3092089: [file normalize] can remove path component, jan.nijtmans2013-01-031-0/+6
| | | | | | and for Bug 3587096: startup error message when exe in folder with junction with limited rights
| * test case for bug-3598580: Tcl_ListObjReplace may release deleted elements ↵jan.nijtmans2013-01-032-0/+15
| |\ | | | | | | | | | too early
| * | remove stray calls to Tcl_Alloc and friends: the core should only use ↵mig2013-01-023-6/+13
| | | | | | | | | | | | ckalloc to allow MEM_DEBUG to work properly
| * | test Tcl_GetErrorLine() forwards/backwards compatibility in pkgb.so as well. jan.nijtmans2013-01-024-22/+39
| |\ \ | | | | | | | | | | | | Marked some string subcommands as obsolete, following discussion on tcl-core. Don't free ctrl.script if thread creation fails: it is a constant string "testthread wait" normally.
| * \ \ restore refcounts as they were before the Tcl_ListObjReplace call, in the ↵jan.nijtmans2012-12-291-0/+8
| |\ \ \ | | | | | | | | | | | | | | | | | | | | error situation. In Tcl9, make this a true Tcl_DecrRefCount, which properly cleans objects with refCount 0
| | * | | For Tcl9, do a real Tcl_DecrRefCountbug_3598580jan.nijtmans2012-12-291-0/+4
| | | | |
| | * | | restore old refcounts in TCL_ERROR case.jan.nijtmans2012-12-271-0/+3
| |/ / /
| * | | [Bug 3598580]: Tcl_ListObjReplace may release deleted elements too earlyjan.nijtmans2012-12-272-4/+10
| |\ \ \
| * | | | eliminate dependancy of compat/*.h on tcl.hjan.nijtmans2012-12-265-151/+115
| | | | |
| * | | | Make pkgb.so loadable in Tcl 8.4 as well. jan.nijtmans2012-12-213-30/+27
| |\ \ \ \ | | | | | | | | | | | | Eliminate unnessarcy static HasStubSupport() and isDigit() functions, just do the same inline.
| * \ \ \ \ Restore clarity to macro test.dgp2012-12-131-1/+1
| |\ \ \ \ \
| * | | | | | 3595576 Tests/fix for mem corruption: [catch] fails to store options in a var.dgp2012-12-132-1/+6
| | | | | | |
| * | | | | | Changelog entry, and change macro variable to not conflict with possible ↵jan.nijtmans2012-12-132-3/+8
| |\ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | outside variable names
| * \ \ \ \ \ \ Fix Tcl_DecrRefCount macro, not to refer to its objPtr parameter twice.jan.nijtmans2012-12-131-5/+7
| |\ \ \ \ \ \ \
| * \ \ \ \ \ \ \ Restore the initialization of tclStubsPtr from the "Tcl" package clientDatadgp2012-12-101-2/+3
| |\ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | so that we don't close off a potential avenue of future innovations.
| * \ \ \ \ \ \ \ \ only set tclStubsPtr when all version checks passjan.nijtmans2012-12-071-32/+18
| |\ \ \ \ \ \ \ \ \
| * | | | | | | | | | 3593703 Don't crash on bad input to Tcl_PkgRequire*().dgp2012-12-071-0/+4
| | | | | | | | | | |
| * | | | | | | | | | Turn pkgb.so into a Tcl9 interoperability testjan.nijtmans2012-12-072-6/+11
| |\ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | library: Whatever Tcl9 looks like, loading pkgb.so in Tcl 9 should either result in an error-message, either succeed, but never crash.