Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Created branch dgp-refactor-merge-syntheticdgp_refactor_mergedgp_refactor_merge_synthetic | cvs2fossil | 2011-01-25 | 180 | -0/+50490 |
| | |||||
* | typo | dgp | 2011-01-25 | 1 | -1/+1 |
| | |||||
* | Don't miss 64-bit address bits in panic message. | nijtmans | 2011-01-25 | 2 | -3/+7 |
| | |||||
* | Clarify documentation of format/args checking in Tcl_ObjPrintf. Thanks to ↵ | ferrieux | 2011-01-19 | 1 | -2/+5 |
| | | | | Brad Harder (yukonbob). [Bug 3159920] | ||||
* | [Enh #3159920]: Tcl_ObjPrintf() crashes with bad format specifier. | nijtmans | 2011-01-19 | 4 | -12/+24 |
| | |||||
* | Make sure to use CONST/VOID in stead of | nijtmans | 2011-01-19 | 1 | -1/+7 |
| | | | | | | const/void when appropriate. This allows to use const/void in the *.decls file always, genStubs will do the right thing. | ||||
* | Add missing comment | dkf | 2011-01-18 | 1 | -2/+3 |
| | |||||
* | * generic/tclOOMethod.c (PushMethodCallFrame): [Bug 3001438]: Make | dkf | 2011-01-18 | 3 | -10/+62 |
| | | | | | | | sure that the cmdPtr field of the procPtr is correct and relevant at all times so that [info frame] can report sensible information about a frame after a return to it from a recursive call, instead of probably crashing (depending on what else has overwritten the Tcl stack!) | ||||
* | Various mismatches between Tcl_Panic format string and its arguments, | nijtmans | 2011-01-18 | 8 | -25/+35 |
| | | | | discovered thanks to [Bug 3159920] | ||||
* | [Bug 3148192]: Commands "read/puts" incorrectly interpret parameters. | nijtmans | 2011-01-17 | 5 | -36/+43 |
| | | | | Improved error-message regarding legacy form. | ||||
* | Removed code that suppressed zeroes in a context where zeroes cannot appear. | Kevin B Kenny | 2011-01-15 | 1 | -6/+1 |
| | |||||
* | * doc/tclvars.n: | Kevin B Kenny | 2011-01-15 | 5 | -17/+1986 |
| | | | | | | | | | | * generic/tclStrToD.c: * generic/tclUtil.c (Tcl_PrintDouble): * tests/util.test (util-16.*): Restored full Tcl 8.4 compatibility for the formatting of floating point numbers when $::tcl_precision is not zero. Added compatibility tests to make sure that excess trailing zeroes are suppressed for all eight major code paths. [Bug 3157475] | ||||
* | fall back to assembler for VC++ 32-bit | nijtmans | 2011-01-12 | 1 | -2/+4 |
| | |||||
* | handle --enable-64bit=ia64 for gcc | nijtmans | 2011-01-12 | 4 | -47/+63 |
| | | | | more accurate test for correct <intrin.h> | ||||
* | VS 2005 64-bit does not have intrin.h, and does not need it. | nijtmans | 2011-01-12 | 2 | -8/+5 |
| | |||||
* | VS 2005 64-bit does not have intrin.h, and does not need it. | nijtmans | 2011-01-12 | 2 | -2/+4 |
| | |||||
* | Use _vsnprintf in stead of vsnprintf, because MSVC 6 doesn't have it. ↵ | nijtmans | 2011-01-12 | 2 | -2/+7 |
| | | | | Reported by andreask. | ||||
* | * tests/util.test (util-15.*): Added test cases for floating point | Kevin B Kenny | 2011-01-07 | 2 | -1/+97 |
| | | | | | | | conversion of the largest denormal and the smallest normal number, to avoid any possibility of the failure suffered by PHP in the last couple of days. (They didn't fail, so no actual functional change.) | ||||
* | * tests/package.test, tests/pkg.test: Coalesce these tests into one | dkf | 2011-01-06 | 5 | -1455/+1435 |
| | | | | | | | file that is concerned with the package system. Convert to use tcltest2 properly. * tests/autoMkindex.test, tests/pkgMkIndex.test: Convert to use tcltest2 properly. | ||||
* | Clean up of tests and conversion to tcltest 2. Target has been to get init and | dkf | 2011-01-01 | 21 | -699/+713 |
| | | | | cleanup code out of the test body and into the -setup/-cleanup stanzas. | ||||
* | * tests/execute.test (execute-11.1): [Bug 3142026]: Added test that | dkf | 2011-01-01 | 2 | -94/+129 |
| | | | | fails (with a crash) in an unfixed memdebug build on 64-bit systems. | ||||
* | * generic/tclCmdIL.c (SortElement): Use unions properly in the | dkf | 2011-01-01 | 2 | -31/+41 |
| | | | | | | definition of this structure so that there is no need to use nasty int/pointer type punning. Made it clearer what the purposes of the various parts of the structure are. | ||||
* | Fix [Bug 3148192]: [load] broken. Assure that those files are never compiled ↵ | nijtmans | 2010-12-31 | 7 | -6/+17 |
| | | | | with -DSTATIC_BUILD | ||||
* | * generic/tclExecute.c (GrowEvaluationStack): off-by-one error in | Miguel Sofer | 2010-12-30 | 2 | -4/+11 |
| | | | | | | sizing the new allocation - was ok in comment but wrong in the code. Triggered by [Bug 3142026] which happened to require exactly one more than what was in existence. | ||||
* | * generic/tclCmdIL.c (Tcl_LsortObjCmd): Fix crash when multiple -index | dkf | 2010-12-27 | 3 | -63/+79 |
| | | | | options are used. Simplified memory handling logic. | ||||
* | [Patch 3059922]: fixes for mingw64 - gcc4.5.1 tdm64-1: completed for all ↵ | nijtmans | 2010-12-21 | 2 | -3/+23 |
| | | | | environments. | ||||
* | cannot use '['/']' in configure.in, or re-generation will go wrong | nijtmans | 2010-12-21 | 2 | -5/+3 |
| | |||||
* | fix cpuid determination | nijtmans | 2010-12-20 | 2 | -4/+4 |
| | |||||
* | Explicitely test for intrinsics support in compiler, before assuming only ↵ | nijtmans | 2010-12-20 | 4 | -6/+111 |
| | | | | MSVC has it. | ||||
* | [Patch3059922]: fixes for mingw64 - gcc4.5.1 tdm64-1: | nijtmans | 2010-12-19 | 1 | -1/+1 |
| | | | | Fixed for gcc, not yet for MSVC 64-bit. | ||||
* | [Patch 3124554]: fixes for mingw64 - gcc4.5.1 tdm64-1: | nijtmans | 2010-12-19 | 2 | -5/+41 |
| | | | | Fixed for gcc, not yet for MSVC 64-bit. | ||||
* | Remove unwanted/obsolete 'ddd' target. | stwo | 2010-12-17 | 2 | -7/+6 |
| | |||||
* | Clean up '.PHONY:' targets: Arrange those common to Tcl and Tk as in Tk's ↵ | stwo | 2010-12-17 | 2 | -9/+17 |
| | | | | Makefile.in, add any missing ones and remove duplicates. | ||||
* | [Bug 2446711]: Remove 'allpatch' target. | stwo | 2010-12-17 | 2 | -24/+6 |
| | |||||
* | Use 'rpmbuild', not 'rpm' [Bug 2537626]. | stwo | 2010-12-17 | 2 | -3/+7 |
| | |||||
* | [Patch 3124554]: Move WishPanic from Tk to Tcl | nijtmans | 2010-12-16 | 3 | -16/+42 |
| | | | | Better communication with debugger, if present. | ||||
* | Restore abort() in Tcl_Panic() as it was before. | nijtmans | 2010-12-15 | 3 | -16/+16 |
| | | | | [Patch 3124554] use ExitProcess() here, like in wish. | ||||
* | [Bug 3137454]: Tcl CVS HEAD does not build on GCC 3 | nijtmans | 2010-12-14 | 2 | -2/+6 |
| | |||||
* | * win/tclWinSock.c (CreateSocket): Swap the loops over | rmax | 2010-12-14 | 3 | -24/+32 |
| | | | | | | | * unix/tclUnixSock.c (CreateClientSocket): local and remote addresses, so that the system's address preference for the remote side decides which family gets tried first. Cleanup and clarify some of the comments. | ||||
* | [Bug 3135271] Link error due to hidden symbols (CentOS 4.2) | nijtmans | 2010-12-13 | 4 | -2/+11 |
| | |||||
* | Use gcc's __builtin_trap(), when available, to enter the debugger after a panic. | nijtmans | 2010-12-13 | 3 | -8/+66 |
| | | | | Undocumented feature, see [Patch 3124554]: Move WishPanic from Tk to Tcl. | ||||
* | Better building on OpenBSD. | stwo | 2010-12-12 | 3 | -26/+71 |
| | |||||
* | [Bug 3129448]: Possible over-allocation on 64-bit platforms, part 2 | nijtmans | 2010-12-10 | 8 | -26/+36 |
| | |||||
* | Make sure [fcopy -size ... -command ...] always calls the callback ↵ | ferrieux | 2010-12-10 | 3 | -2/+82 |
| | | | | asynchronously, even for size zero. | ||||
* | Change first parameter of TclSockMinimumBuffers to ClientData, and ↵ | nijtmans | 2010-12-10 | 10 | -36/+358 |
| | | | | TclWin(Get|Set)SockOpt to SOCKET, because on Win64 those are 64-bit, which does not fit. | ||||
* | Fix gcc -Wextra warning: missing initializer | nijtmans | 2010-12-10 | 9 | -167/+179 |
| | |||||
* | * tests/fCmd.test: Improve sanity of constraints now that we don't | dkf | 2010-12-09 | 2 | -22/+42 |
| | | | | support anything before Windows 2000. | ||||
* | * generic/tclCmdAH.c (TclInitFileCmd, TclMakeFileCommandSafe, ...): | dkf | 2010-12-09 | 11 | -930/+1799 |
| | | | | | | | | | | | | | | | | | | Break up [file] into an ensemble. Note that the ensemble is safe in itself, but the majority of its subcommands are not. * generic/tclFCmd.c (FileCopyRename,TclFileDeleteCmd,TclFileAttrsCmd) (TclFileMakeDirsCmd): Adjust these subcommand implementations to work inside an ensemble. (TclFileLinkCmd, TclFileReadLinkCmd, TclFileTemporaryCmd): Move these subcommand implementations from tclCmdAH.c, where they didn't really belong. * generic/tclIOCmd.c (TclChannelNamesCmd): Move to more appropriate source file. * generic/tclEnsemble.c (TclMakeEnsemble): Start of code to make partially-safe ensembles. Currently does not function as expected due to various shortcomings in how safe interpreters are constructed. * tests/cmdAH.test, tests/fCmd.test, tests/interp.test: Test updates to take into account systematization of error messages. | ||||
* | * tests/append.test, tests/appendComp.test: Clean up tests so that | dkf | 2010-12-09 | 3 | -108/+131 |
| | | | | | they don't leave things in the global environment (detected when doing -singleproc testing). | ||||
* | * tests/fCmd.test, tests/safe.test, tests/uplevel.test, | dkf | 2010-12-07 | 6 | -348/+363 |
| | | | | | * tests/upvar.test, tests/var.test: Convert more tests to tcltest2 and factor them to be easier to understand. |