| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
| |
not surprisingly, doesn't have the MSVCRT specific _beginthreadex /
_endthreadex pair. This might have to be revisited for proper
Borland, lcc32, Watcom and other support as well.
|
| |
|
|
|
|
|
|
|
|
| |
* generic/tclBasic.c: added TclCompileListCmd compile proc
* generic/tclCompCmds.c (TclCompileListCmd): function to compile
the 'list' command at parse time.
* generic/tclExecute.c (TclExecuteByteCode): definition of
INST_LIST bytecode.
|
|
|
|
| |
with the Tcl_Append* functions.
|
|
|
|
|
|
|
|
|
| |
* generic/tclUtil.c (TclGetIntForIndex): added Tcl_ResetResult
before Tcl_AppendStringsToObj to prevent shared object crash when
called from bcc instruction. The Tcl_Append* calls that append to
the result object that are invoked by bcc insts must remember to
call Tcl_ResetResult because the bcc doesn't do this for us.
[Bug #456892]
|
|
|
|
| |
Crays. [Bug #419528] (andreasen)
|
| |
|
|
|
|
| |
namespaces within the same interpreter without forcing a new lookup. This speeds up scripts that pass command names in variables ("this" in some OO packages). [Patch 456668].
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclEvent.c:
* generic/tclInt.h: Improper cleanup of asyncMutex in tclAsync.c
repaired. TclFinalizeSynchronization() was trying to remove a
registered mutex that was dumped earlier when the TSD it was stored
in was cleared. This was only surfacing on *nix. Windows was being
masked by mutexes not actually being returned to the system! That
was repaired in a previous patch. Needed to add a private
TclFinalizeAsync() to tclAsync.c and called from Tcl_FinalizeThread().
Pheww.. Is this done yet?
[Bug: 414419] requested by Rob Ratcliff <rrr6399@futuretek.com>
|
|
|
|
| |
[Bug #453872]
|
|
|
|
|
|
| |
inconsistency between the search path for script libraries and
the directory name $DISTNAME into which distributions built
by 'make test' unpack. [Bug 455642]
|
| |
|
| |
|
|
|
|
|
|
|
| |
* generic/tclCompCmds.c (TclCompileStringCmd): changed to return
TCL_OUT_LINE_COMPILE instead of TCL_ERROR when compiling and an
unknown string method is called. This is necessary as the string
command may be never called, or not until 'string' is redefined.
|
| |
|
| |
|
| |
|
|
|
|
| |
to match Tcl's quirky construction of its init library path.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
* win/tclWinPipe.c (BuildCommandLine): Fixed tcl Bug
[432499]. Part of the code used the non-absolute path to the
executable to determine quoting. This failed if the absolute
path cotained spaces, but the application name itselfl not. This
bug caused no trouble on Win NT 5, but does for other variants
in the Win* family. Report and fix due to Ken Poole
<kenpoole@users.sourceforge.net>.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
* tests/winDde.test:
* win/tclWinDde.c (Tcl_DdeObjCmd): added -binary option to dde
request command to allow for returning binary data. [Bug #227482]
Updated dde to 1.2
|
|
|
|
|
| |
[Patch #454050] (stanton)
Simplified unixExecs constriant test.
|
|
|
|
| |
used 'grep' in the test. [Bug #453143]
|
| |
|
|
|
|
| |
revealed by fix of overagressive compiler. [Bug 451200]
|
| |
|
|
|
|
| |
when env vars weren't Tcl list friendly [Patch #454046] (stanton)
|
|
|
|
| |
header to comply with HTTP/1.1 spec (RFC 2068). [Bug #452217]
|
|
|
|
| |
days ago.
|
| |
|
|
|
|
|
| |
.rc.res inference rule. winver.h wasn't getting included.
[Bug: 445630]
|
| |
|
|
|
|
|
|
|
|
|
| |
* tools/tcl.hpj.in:
* win/tcl.hpj.in: Removed -kb storage in CVS to ensure these text
files are checked-out in the translation mode CVS is in. Setting
these as binary as part of an effort to make sure they are always
in CRLF, no matter what the CVS translation, is bypassing how CVS
works and is confusing.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
* win/tcl.hpj.in: Removed -kb storage in CVS to ensure these text
files to make sure they are checked-out in the translation mode CVS
is in. Setting these as binary as part of an effort to make sure they
are always in CRLF no matter what the CVS translation is bypassing
how CVS works.
* tools/genStubs.tcl: Removed LF-only output. Having to reconvert
back to CRLF for committing to CVS was giving me a headache.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
* unix/configure.in:
* unix/tcl.m4: Use GCC variable set by AC_PROG_CC instead
of defining our own using_gcc variable.
|
| |
|