| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
* unix/configure.in (Darwin): add plist for tclsh; link the
* unix/Makefile.in (Darwin): Tcl and tclsh plists into their
* macosx/Tclsh-Info.plist.in (new): binaries in all cases.
* macosx/Tcl-Common.xcconfig:
* unix/tcl.m4 (Darwin): fix CF checks in fat 32&64bit builds.
* unix/configure: autoconf-2.59
|
|
|
|
|
|
|
|
| |
* generic/tclInt.h: control the timing with which the global
* generic/tclNamesp.c: variables ::errorCode and ::errorInfo get
* generic/tclProc.c: updated after an error. This keeps more
* generic/tclResult.c: precise compatibility with Tcl 8.4.
* tests/result.test (result-6.2): [Bug 1649062]
|
|
|
|
| |
* generic/tclExecute.c: Tcl-stack reform, [Patch 1701202]
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclCompile.c: TCL_TOKEN_EXPAND_WORD tokens representing
* generic/tclCompile.h: expanded literal words. These sections were
mostly in place to enable [info frame] to discover line information
in expanded literals. Since the parser now generates a token for
each post-expansion word referring to the right location in the
original script string, [info frame] gets all the data it needs.
* generic/tclInt.h: Revised the parser so that it never produces
* generic/tclParse.c: TCL_TOKEN_EXPAND_WORD tokens when parsing an
* tests/parse.test: expanded literal word; that is, something like
{*}{x y z}. Instead, generate the series of TCL_TOKEN_SIMPLE_WORD
tokens to represent the words that expansion of the literal string
produces. [RFE 1725186]
|
|
|
|
| |
pthread_join exit return code storage. [Bug 1712723]
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* win/configure:
* README: Bump version number to 8.5a7
* generic/tcl.h:
* library/init.tcl:
* tools/tcl.wse.in:
* unix/configure.in:
* unix/tcl.spec:
* win/configure.in:
* generic/tclParse.c: Disable and remove the ALLOW_EXPAND sections
* tests/info.test: that continued to support the deprecated
* tests/mathop.test: {expand} syntax. Updated the few remaining
users of that syntax in the test suite.
|
| |
|
| |
|
| |
|
|
|
|
| |
tclStubLib.c MODULE_SCOPE revert. [Bug 1716117]
|
|
|
|
|
|
|
|
| |
pointers back to public visibility.
These symbols need to be exported so that stub-enabled extensions
may be statically linked into an extended tclsh or Big Wish with
a dynamically-linked libtcl. [Bug#1716117]
|
|
|
|
|
|
|
| |
* library/reg/pkgIndex.tcl: Bump to registry 1.2.1 to account for
* win/configure.in: [1682211] bug fix.
* win/makefile.bc:
* win/tclWinReg.c:
|
|
|
|
| |
from tclInt.h now they are in the internal stubs table.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
so make sure it is also always declared (with MODULE_SCOPE).
|
|
|
|
|
| |
* macosx/Tcl.xcodeproj/default.pbxuser: 'DebugLeaks' targets and env
var settings needed to run the 'leaks' tool.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclLink.c (LinkTraceProc): Update Tcl_VarTraceProcs so that
* generic/tclNamesp.c (Error*Read): they call Tcl_InterpDeleted()
* generic/tclTrace.c (Trace*Proc): for themselves, and do not rely
* generic/tclUtil.c (TclPrecTraceProc): on (frequently buggy) setting
of the TCL_INTERP_DESTROYED flag by the trace core.
* generic/tclVar.c: Update callers of TclCallVarTraces to not
pass in the TCL_INTERP_DESTROYED flag. Also apply filters so that
public routines only pass documented flag values down to lower level
routines.
* generic/tclTrace.c (TclCallVarTraces): The setting of the
TCL_INTERP_DESTROYED flag is now done entirely within the
TclCallVarTraces routine, the only place it can be done right.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
in a private namespace, to avoid polluting the global one. This
problem was discovered when running the test suite '-singleproc 1
-skip exec.test' because the 'path' variable in encoding.test
conflicted with the one in io.test.
* tests/io.test: Made more of the working variables private to the
namespace.
|
| |
|
|
|
|
|
|
| |
Corrected several memory leaks that caused refcount imbalances
resulting in memory leaks on Windows. Thanks to Joe Mistachkin
for the patch.
|
| |
|
|
|
|
|
| |
imbalance that affected the filesystem-[147]* tests in the test
suite. Thanks to Don Porter for the patch. [Bug 1710707]
|
|
|
|
|
| |
clean up after the 'testobj' extension to avoid spurious
reports of memory leaks.
|
|
|
|
|
| |
had refCount 0 and was used for appending (but not
lappending). [Bug 1710710], thanks to mistachkin and kbk.
|
|
|
|
|
|
|
| |
DeleteChannelTable tries to close all open channels, not just the
first. [Bug 1710285]
* generic/tclThread.c (TclFinalizeSynchronization): Make sure
that TSD blocks get freed on non-threaded builds. [Bug 1710825]
|
|
|
|
| |
map, a missing Tcl_DictObjDone() call led to a memleak. [Bug 1710709]
|
|
|
|
| |
on tclsh, fixes parallel 'make install' from empty build dir.
|
|
|
|
|
|
| |
mismanagement of newlevel, newcode. Changed to allocate the
Tcl_Obj's as late as possible, and only when actually needed. [Bug
1705778, leak K29].
|
|
|
|
|
| |
reference count mismanagement on the name of the source file in
the TIP 280 code. [Bug 1705778, leak K02 among other manifestations]
|
| |
|
| |
|
|
|
|
|
|
| |
to that the path returned from Tcl_FSGetNativePath is not
duplicated before being stored as the current directory, to avoid
a memory leak. [Bug 1705778, leak K01 among other manifestations]
|
|
|
|
|
|
| |
an error return doesn't prevent all literals getting placed on the
litList to be returned to the caller for freeing. Corrects some
memleaks. [Bug 1705778, leak K23]
|
|
|
|
| |
autoheader bits.
|