summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Added field to tcl_platform to describe the machine-word size.dkf2001-12-025-32/+26
|
* Fixed (hopefully) mistakes in tclTest.c that stopped compilation on Alpha.dkf2001-11-272-5/+14
|
* Missed a #undef...dkf2001-11-271-1/+2
|
* Made 'obsolete fs hooks' version of stat buffers backward compatible.dkf2001-11-264-14/+92
| | | | Thanks to Vince Darley for suggesting this.
* Fixed fault in new Tcl_Stat implementation, and improved docs.dkf2001-11-254-8/+27
|
* Made Tcl_Stat use old-style 'struct stat' in its declaration followingdkf2001-11-237-19/+72
| | | | | | suggestions from Vince Darley and Arjen Markus. Thanks for prompting me to do this, guys! Also tidied up a few spots.
* Following a suggestion by wdi@ccc.uni-erlangen.de (Wolf-Dietrich Ihlenfeldt),dkf2001-11-075-237/+309
| | | | | This implements support for linking wide-int variables to Tcl variables. It also moves the link-var system towards using 8.*-based interfaces!
* Improved backward-compatability on platforms by adding migration entries todkf2001-11-054-19/+84
| | | | the stubs table for Tcl_Seek and Tcl_Tell
* Fixed tesst selection to choose the right behaviour (use 64-bit code if wedkf2001-10-251-6/+6
| | | | can support 64-bit operation.)
* Some places prefer string lengths to be unsigned.dkf2001-10-232-4/+4
|
* Bother! Was losing error messages...dkf2001-10-231-8/+18
|
* Fixed behaviour of [incr] w.r.t. increment amounts that are notdkf2001-10-232-21/+76
| | | | representable in a 'long'.
* Added #include tclPort.h to get definition of ptrdiff_tdgp2001-10-221-1/+2
|
* * win/tclWinMtherr.c (_matherr): Better handling of selecteddkf2001-10-222-2/+9
| | | | Borland C++ math errors, from hgiese@ratiosoft.com
* Avoid an obscure Borland limit...dkf2001-10-221-2/+2
|
* Conditional-define INVALID_SET_FILE_POINTER for VC5dkf2001-10-221-1/+5
|
* Stupid error in code to determine whether a literal is wide or not. Thanks todkf2001-10-222-4/+16
| | | | | | Benjamin Riefenstahl <Benjamin.Riefenstahl@epost.de> for reporting an ensuing weirdness which let me track this problem down. Also made sure that there are definitions for LLONG_MIN and LLONG_MAX...
* SunPro CC *sucks*...dkf2001-10-222-5/+17
|
* Fixed assumptions in testing code relating to internal representation ofdkf2001-10-223-9/+29
| | | | index objects and added note to remind people to keep the code in sync.
* * corrected minor coding errors to get compileable sourcedgp2001-10-191-4/+4
|
* Added clarification that string table used in Tcl_GetIndexFromObj mustdkf2001-10-193-47/+187
| | | | | | be a table of ASCII strings, and rewrote most of tclIndexObj.c for greater clarity and to make it more likely to work on various Cray machines where char* is not a basic C type!
* Keep on trying to get the combination of generic/tcl.h and unix/tcl.m4 right!dkf2001-10-193-107/+105
|
* Added range-checking to strtoll and strtoulldkf2001-10-192-15/+98
|
* OS-level large-file support for Windows.dkf2001-10-192-10/+16
|
* D'oh! Accidental redefinition removed...dkf2001-10-181-3/+1
|
* Simplified tcl.h and removed support for platforms that can understand butdkf2001-10-185-106/+53
| | | | | cannot printf a 64-bit value. If they are widespread, this will be a bit of a retrograde step but it does make it easier to understand the code!
* Corrected dirent64 API detection.dkf2001-10-182-4/+4
|
* Fixed some more stupid errors (resulting from not being able to build on alldkf2001-10-183-18/+19
| | | | platforms) and made sure that strtoll and strtoull are available for [scan]
* Removed some UNIX-specific stuff that had been left in by accidentdkf2001-10-181-6/+1
|
* Various changes to make life much easier on non-UNIX platforms (mostly Win32.)dkf2001-10-188-22/+47
|
* Changes to make more use of what autoconf can discover. Also adds *lots* moredkf2001-10-176-445/+784
| | | | | cacheing of discovered values, which not only speeds up reconfiguring, but also makes mistakes easier to override.
* Trying to do more with configure...dkf2001-10-153-1364/+736
|
* Hopefully will work on AIX now, though this version might be a little on thedkf2001-10-151-9/+17
| | | | aggressively anti-64bit on that platform instead.
* Oops.dkf2001-10-152-4/+4
|
* Start of work to support more platforms (HPUX,IRIX,Win) plus some minor bug ↵dkf2001-10-158-60/+75
| | | | fixes
* Some platforms don't have strtoll or strtoull.dkf2001-10-152-0/+268
| | | | ***WARNING - THESE VERSIONS DON'T TO RANGE CHECKING ***
* [string is integer] test updatedkf2001-10-121-2/+2
|
* Documented (some) 64-bit-related behaviour changes.dkf2001-10-123-22/+64
|
* [string is] support.dkf2001-10-121-2/+10
|
* Removed some debugging code and added a test that exercises a different routedkf2001-10-121-8/+4
| | | | through the code.
* Tests for all sorts of 64-bit support thingsdkf2001-10-114-24/+175
|
* Mod of Wide integers can produce a wide integer. D'oh!dkf2001-10-111-6/+7
|
* More correct handling of wides as booleans, wide literals and [incr] of adkf2001-10-113-58/+98
| | | | variable containing a string that needs to be converted to wide.
* Use more macros from tclInt.h and upper-case file-local macrosdkf2001-10-111-167/+174
|
* Fixes to make test suite cleandkf2001-10-091-2/+2
|
* Fixes to make whole test suite cleandkf2001-10-093-37/+74
|
* Various test fixes - fewer than I'd thought though...dkf2001-10-093-9/+9
|
* Fewer #ifdefs, less warnings, and fixed a stupid sync error in tclCompile.hdkf2001-10-082-174/+131
|
* 64-bit aware binary arithmetic ops (and fixes to get clean build)dkf2001-10-081-17/+92
|
* D'oh! Stupid error...dkf2001-10-081-2/+2
|
generated by cgit v0.12 at 2024-11-20 06:26:09 (GMT)