| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
|
|
|
|
|
| |
TclGetIntFromObj, and TclGetIntForIndexM macros so that they
retrieve the internalRep.longValue field instead of casting the
internalRep.otherValuePtr field to type long.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
* library/init.tcl: CVS development snapshots from the 8.5.4 and
* unix/configure.in: 8.5.5 releases.
* unix/tcl.spec:
* win/configure.in:
* tools/tcl.wse.in:
* README
* unix/configure: autoconf (2.59)
* win/configure:
|
|
|
|
|
| |
*exactPtr value to really fix [Bug 2065115]. Also avoid possible
DString overflow.
|
|
|
|
|
| |
*exactPtr value to really fix [Bug 2065115].
* tests/regexpComp.test: Correct duplicate test names.
|
| |
|
|
|
|
|
| |
* generic/tclUtil.c (TclReToGlob): translation from exact
to anywhere-in-string match. [Bug 2065115]
|
|
|
|
|
| |
TclInvalidateStringRep macro.
[Bug 2057479]
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclUtil.c (Tcl_ConcatObj):
* tests/util.test (util-4.7):
fix [Bug 1447328]; the original "fix" turned Tcl_ConcatObj() into
a hairy monster. This was exposed by [Bug 2055782]. Additionally,
Tcl_ConcatObj could corrupt its input under certain conditions!
*** NASTY BUG FIXED ***
|
| |
|
|
|
|
|
| |
* generic/tclBasic.c: 'proc', 'cmd' and 'inst' probes (does
_not_ require a platform with DTrace).
|
|
|
|
|
|
| |
dereferencing as line info may
not exists when TclInfoFrame()
is called from a DTrace probe.
|
|
|
|
|
| |
* tests/fileName.test: option to [glob]. [Bug 1750300]
Thanks to Matthias Kraft and George Peter Staplin.
|
|
|
|
|
|
| |
* tests/proc.test: by procbody::test::proc. See [Bug 2043636].
Added a test case demonstrating the leak before the fix. Fixed a
few spelling errors in test descriptions as well.
|
|
|
|
|
|
|
|
|
|
|
| |
* library/init.tcl:
* tools/tcl.wse.in:
* unix/configure.in:
* unix/tcl.spec:
* win/configure.in:
* unix/configure: autoconf-2.59
* win/configure:
|
|
|
|
|
|
| |
locals in the localCache from the CallFrame and not from the Proc
which may have been mangled by a (broken?) recompile. Backport
from the HEAD.
|
|
|
|
|
|
| |
* tests/execute.test: stack trace when a compile epoch bump triggers
fallback to direct evaluation of commands in a compiled script.
[Bug 2037338]
|
|
|
|
| |
TCL_ALLOW_EXCEPTIONS is tested.
|
|
|
|
|
| |
TclStackRealloc to panic on rare corner cases. Thx ajpasadyn for
diagnose and patch.
|
|
|
|
|
|
|
| |
empty string as path (TclEvalEx). In 8.4 the missing code caused
panics in the testsuite. It doesn't in 8.5. I am guessing that the
code path with the missing the incr-refcount is not invoked any
longer. Because the bug in itself is certainly the same.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
of location for uplevel scripts.
* generic/tclCompile.c (TclInitCompileEnv): Reorganized the
initialization of the #280 location information to match the flow
in TclEvalObjEx to get more absolute contexts.
* generic/tclBasic.c (TclEvalObjEx): Moved the pure-list
optimization out of the eval-direct code path to be done always,
i.e. even when a compile is requested. This way we do not loose
the association between #280 location information and the list
elements, if any.
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclCmdIL.c: immediately, without search. Reworked setup
* generic/tclCompile.c: of eoFramePtr, doesn't need the line
* tests/info.test: information, more sensible to have everything
on line 1 when eval'ing a pure list. Updated the users of the line
information to special case this based on the frame type (i.e.
TCL_LOCATION_EVAL_LIST). Added a testcase demonstrating the new
behaviour.
|
| |
|
|
|
|
| |
* generic/tclCompile.h: static, and ansified.
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclCompile.c: for #280 to fix the abysmal performance
* generic/tclCompile.h: for deep recursion, replaced the linear
* generic/tclExecute.c: search through the whole stack with
* generic/tclInt.h: another hashtable and simplified the data
structure used by the compiler (array instead of hashtable).
Incidentially this also fixes the memory leak reported via [Bug
2024937].
|
|
|
|
|
| |
This allows extensions to be compiled with -fvisiblity=hidden and still all
symbols decorated with EXPORT will be exported. See: <http://gcc.gnu.org/wiki/Visibility> (backport from trunk tcl.h 1.256)
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tclCmdAH.c: frame), added the ability to track the
* generic/tclCompCmds.c: absolute location of literal procedure
* generic/tclCompile.c: arguments, and making this information
* generic/tclCompile.h: available to uplevel, eval, and
* generic/tclInterp.c: siblings. This allows proper tracking of
* generic/tclInt.h: absolute location through custom (Tcl-coded)
* generic/tclNamesp.c: control structures based on uplevel, etc.
* generic/tclProc.c:
|
| |
|
| |
|
| |
|
|
|
|
| |
the interp result found by Don Porter.
|
|
|
|
| |
sequences. See [Bug 1908443] for details.
|
|
|
|
|
| |
reported in [Bug 1987821]. Thanks to Miguel for the rpeort and
Don Porter for tracking the cause down.
|
|
|
|
|
|
|
|
|
|
|
| |
* library/init.tcl:
* tools/tcl.wse.in:
* unix/configure.in:
* unix/tcl.spec:
* win/configure.in:
* unix/configure: autoconf-2.59
* win/configure:
|
|
|
|
| |
Rolf Ade for detecting.
|
|
|
|
| |
Rolf Ade for detecting and Dan Steffen for the fix [Bug 2004654].
|
|
|
|
|
| |
Tcl_ObjType for the empty string value. Problem led to a crash in
the command [glob -dir {} a]. [Bug 1999176].
|
|
|
|
|
|
|
| |
operating on the "Special path" variant of the "path" Tcl_ObjType
intrep. A full normalization was getting done, in particular, coercing
relative paths to absolute, contrary to what the function of
producing the "translated path" is supposed to do. [Bug 1972879].
|
|
|
|
|
|
|
|
|
| |
* tests/interp.test: [interp bgerror $slave] and [$slave bgerror]
would always act like [interp bgerror {}]. [Bug 1999035].
* tests/chanio.test: Corrected flawed tests revealed by a -debug 1
* tests/event.test: -singleproc 1 test suite run.
* tests/io.test:
|
|
|
|
| |
over debug output.
|
|
|
|
|
|
|
|
|
| |
* generic/tclClock.c (ConvertLocalToUTC):
* tests/clock.test (clock-63.1): Fixed a bug where the
internal ConvertLocalToUTC command segfaulted if passed a
dictionary without the 'localSeconds' key. To the best of
my knowledge, the bug was not observable in the [clock]
command itself.
|
|
|
|
|
|
|
|
|
| |
* tests/info.test: head branch :: Moved the code looking up the
information for key 'proc' out of the TCL_LOCATION_BC branch to
after the switch, this is common to all frame types. Updated the
testsuite to match. This was exposed by the 2008-06-08 commit
(Miguel), switching uplevel from direct eval to compilation. Fixes
[Bug 1987851].
|
|
|
|
|
|
|
|
|
|
|
|
| |
Alexandre Ferrieux <ferrieux@users.sourceforge.net> to fix the
[Bug 1965787]. 'tell' now works for locations > 2 GB as well
instead of going negative.
* generic/tclIO.c (Tcl_SetChannelBufferSize): Accepted a patch by
* tests/io.test: Alexandre Ferrieux <ferrieux@users.sourceforge.net>
* tests/chanio.test: to fix the [Bug 1969953]. Buffersize outside
of the supported range are now clipped to nearest boundary instead
of ignored.
|
|
|
|
|
| |
handle the argument value length = -1. Thanks to Chris Darroch for
discovering the bug and providing the fix. [Bug 1968245].
|
|
|
|
|
|
|
| |
* tests/parse.test (parse-15.60): routine has no mechanism to
return the "incomplete" status of "\\\n" so calling this routine
anywhere that can be reached within a Tcl_ParseCommand() call is a
mistake. In particular, ParseComment() must not use it. [Bug 1968882].
|
| |
|