| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
| |
to use a heap variable for the message buffer instead of placing
it on the stack.
* Updated "mod-8-3-1-branch" to head of "core-8-3-1-branch" to get
the newest fixes for tcl 8.3.4 into this branch.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* tcl.h:
* tclInt.h: Moved the macros to control feature exclusion and
stack manipulation into the private area of the tcl
headers. Renamed the macros for TCL_STRUCT_ON_HEAP to make their
names more sensible.
* tclCmdAH.c:
* tclCompCmds.c:
* tclCompExpr.c:
* tclCompile.c:
* tclExecute.c:
* tclIO.c:
* tclParse.c:
* tclParseExpr.c:
* tclScan.c:
* tclUnixChan.c: Adapted to changed macro names, added some more
places there structures go on the heap instead of the stack.
Fixed a problem with TCL_FMT_STATIC_FLOATBUFFER_SZ which
caused the interp to crash when actually used to reduce the
usage of the stack.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
interdependencies, implications, etc.
* static.sizes.html: New file. Report on the cuts achieved so
far. Regarding the object files only the files which did change
in size are reported. Usage of the MODULAR_TCL macro currently
cuts about 17 % of the code (measured using strip'ped object
files and libraries).
* Changed files so far .. [cut, see changelog for full list]
* Working on modularization of the tcl core.
***************************
**** mod-8-3-4-branch ****
***************************
|
| | |
|
| |
|
|
|
|
| |
<kennykb@users.sourceforge.net>) two more '(ClientData)' casts
on calls to Tcl_Preserve and Tcl_Release -- ones that Vince
apparently missed.
|
| |
|
|
|
|
|
|
| |
and Tcl_Release() to fix segfault introduced by the 2001-09-26
changes. [Bug 465494]
* doc/TCL_MEM_DEBUG.3: Updated out-of-date reference to
#define GUARD_SIZE.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
more than was in the buffers and then blocked in the OS call as
its pty channel driver provides no blockmodeproc through which
the OS could be notified of blocking-behaviour. Because of this
the general I/O core has to take more care than usual to
preserve the semantics of non-blocking channels.
* generic/tclIO.c (Tcl_ReadRaw): Do not read from the driver if
the channel is non-blocking and the fileevent causing the read
was generated by a timer. We do not know if there is data
available from the OS. Instead of going to the OS for more and
potentially blocking we simply signal EWOULDBLOCK to the higher
levels to cause the system to wait for true fileevents.
(GetInput): Same as before.
(ChannelTimerProc): Added set and clear of CHANNEL_TIMER_FEV.
* generic/tclIO.h (CHANNEL_TIMER_FEV): New flag for channels. Is
set if a fileevent was generated by a timer, the channel is not
blocking and the driver did not provide a blockmodeproc. In that
case the I/O core has to be especially careful about going to
the driver for more data.
|
| |
|
|
|
|
|
| |
* generic/tclIO.c: Aftermath to [SF #427196]. Squash empty buffers
if they are smaller than the requested buffersize, to prevent
reusage of old buffers and to honor changes in the requested
buffersize made by the user.
|
| |
|
|
|
|
|
|
|
| |
overwritten because a buffer was used after a change of the
requested buffersize together with that requested buffersize and
not its actual size, which was smaller. Note that the continous
reuse of the smaller buffer negatively impacts performance. The
system never allocates a buffer with the newly requested bigger
buffersize.
|
| | |
|
| |
|
|
|
| |
that the new channel versioning will be binary compatible with
older channel drivers.
|
| |
|
|
| |
The core-8-3-1-io-rewrite branch should now be considered defunct.
|
| |
|
|
|
|
| |
(TCL_READABLE|TCL_WRITABLE) from the flags, as CloseChannel does
this on the next pass through for the top channel, and it appeared
to be causing hangs by not allowing the final flush.
|
| |
|
|
|
|
|
|
| |
to effect certain operations on the next channel that would have
been done in Tcl_Close. Also added CHANNEL_CLOSED and removed
(TCL_READABLE|TCL_WRITABLE) bits from chanPtr->flags. Changed
final reset of the WatchProc to check the chanDownPtr's (next)
interestMask.
|
| |
|
|
|
| |
unstack a channel during the close process. Fixed a refcount bug
in Tcl_UnstackChannel. [Bug: 5623]
|
| |
|
|
|
|
|
|
| |
* generic/tclIO.c (Tcl_UnstackChannel, line 1831): Forcing interest
mask to the correct value after an unstack and re-initialization
of the notifier via the watchProc. Without this the first
fileevent after an unstack will come through and be processed,
but no more. [Bug: ??].
|
| |
|
|
| |
* generic/tclThreadTest.c: changed thread ids to longs [Bug: 3902]
|
| |
|
|
|
|
|
|
|
|
|
|
| |
* unix/Makefile.in: fixed make gendate to swap const with CONST
so it uses the Tcl defined CONST type [Bug: 3521]
* generic/tclIO.c: removed panic that could occur in FlushChannel
when a "blocking" channel would receive EAGAIN, instead treating
it the same as non-blocking. [Bug: 3773]
* generic/tclUtil.c: fixed Tcl_ScanCountedElement to not step
beyond the end of the counted string [Bug: 3336]
|
| |
|
|
| |
correctly handles resetting translation and encoding.
|
| |
|
|
|
| |
* generic/tclIO.c: changed Tcl_GetChannelNames* to use K&R
style headers (style guide, also pleases HP cc)
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
* doc/OpenFileChnl.3:
* doc/file.n:
* tests/cmdAH.test:
* tclIO.c:
* tclCmdAH.c: added "file channels ?pattern?" tcl command, with
associated Tcl_GetChannelNames and Tcl_GetChannelNamesEx public
C APIs (added to tcl.decls as well), with docs and tests.
* generic/tclCompile.c: add TCL_TOKEN_VARIABLE to the part types
that cause differed compilation for exprs, to correct the expr
double-evaluation problem for vars. Added test cases.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tcl.decls:
* generic/tclDecls.h: Applied patch from Jim Ingham to change the
prototype of Tcl_ListObjGetElements to have the last argument have
a CONST so that you can feed it the objv that you get from the
standard TclObj command proc.
* generic/tclAlloc.c:
* generic/tclCmdIL.c:
* generic/tclIO.c:
* generic/tclThread.c:
* win/tclWinThrd.c:
* unix/tclUnixThrd.c: Fixed Brent's changes so that they work on
Windows (and he fixed the bug in the Unix thread implementation).
|
| |
|
|
|
|
|
|
| |
can be linked against alternate thread packages.
2 Added Tcl_GetChannelNames to tclIO.c
3 Added TclVarTraceExists hook so "info exists" triggers read traces
exactly like it did in Tcl 7.6
4 Stubs table changes to reflect new internal and external APIs
|
| |
|
|
| |
previous patch to our specs.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* win/tclWinPort.h: Block out include of sys/*.h in order to
build extensions with MetroWerks compiler for Win32. [Bug: 2385]
* generic/tclCmdMZ.c:
* generic/tclIO.c: Fix ANSI-style prototypes based on patch from
Ulrich Ring. [Bug: 2391]
* unix/Makefile.in: Need to make install-sh executable before
calling (with chmod +x). [Bug: 2413]
* tests/var.test:
* generic/tclVar.c: Fixed bug that caused a seg. fault when using
"array set a(b) {}", which is a bad array name anyway. Now the
"array set" command will return an error in this case. Added test
case and fixed existing test. [Bug: 2427]
|
| |
|
|
| |
Added new code that modifies the channel in place.
|
| |
|
|
| |
functions.
|
| |
|
|
|
| |
a bad option is used. Because the fconfigure command is configurable (so
more option can be added), the error string is hardcoded.
|
| | |
|
| | |
|
| |
|
|
|
|
| |
Simply replaced Tcl_RegisterChannel/UnregisterChannel with
Tcl_Preserve/Tcl_Release was all it took. Chanels are already
"eventually freed"
|
| | |
|
| | |
|
| | |
|
| |
|
