| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
to TIP 27. In particular, the "part2" arguments were CONSTified.
|
|
|
|
|
| |
be exploitable in any meaningful way, but crashing Tcl instead of
triggering an error still isn't good.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* win/configure.in: Add configure time test for SEH
support in the compiler.
* win/tclWin32Dll.c (ESP, EBP, TclpCheckStackSpace,
_except_checkstackspace_handler):
* win/tclWinChan.c (ESP, EBP, Tcl_MakeFileChannel,
_except_makefilechannel_handler):
* win/tclWinFCmd.c (ESP, EBP, DoRenameFile,
_except_dorenamefile_handler,
DoCopyFile, _except_docopyfile_handler):
Implement SEH support under gcc using inline asm.
Tcl and Tk should now compile with Mingw 1.1. [Patch 525746]
|
|
|
|
|
|
|
|
| |
an SEH exception with EXCEPTION_EXECUTE_HANDLER instead
of restarting the faulting instruction with
EXCEPTION_CONTINUE_EXECUTION. Bug 466102 provides an
example of how restarting could send Tcl into an
infinite loop. [Patch 525746]
|
|
|
|
|
|
| |
DoRemoveJustDirectory): Make sure we don't pass NULL or ""
as a path name to Win32 API functions since this was
crashing under Windows 98.
|
| |
|
|
|
|
|
| |
a proper list to foreach to avoid munging a Windows
patch like D:\Foo\Bar into D:FooBar before the glob.
|
|
|
|
|
|
|
|
| |
* generic/tclIO.c (DoReadChars, ReadBytes, ReadChars):
Use NULL value instead of pointer set to NULL to make
things more clear. Reorder arguments so that they
match the function signatures. Cleanup little typos
and add more descriptive comment.
|
| |
|
|
|
|
|
|
| |
required to build Tcl. Add section describing new
msys based build process. Update Cygwin build
instructions so users know where to find Mingw 1.1.
|
| |
|
|
|
|
|
|
|
| |
* win/tclWinFCmd.c (DoRenameFile, DoCopyFile): Replace
hard coded constants with Win32 symbolic names.
Move control flow statements out of __try blocks
since the documentation indicates it is frowned upon.
|
|
|
|
|
| |
set/query the recursion limit of an interpreter. Proposal and
implementation from Stephen Trier. [TIP 87, Patch 522849]
|
|
|
|
| |
easier for now.
|
|
|
|
| |
macro [Bug 524802]
|
| |
|
| |
|
|
|
|
|
|
| |
problem introduced by TIP #35. This may not satisfy true serial
channels, but it restores the correct flushing of std* channels on
exit.
|
| |
|
| |
|
|
|
|
| |
but people might find them convenient.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
* tests/encoding.test: corrected iso2022 encoding results.
added encoding-24.*
* generic/tclEncoding.c (EscapeFromUtfProc): corrected output of
escape codes as per RFC 1468. [Patch #474358] (taguchi)
(TclFinalizeEncodingSubsystem): corrected potential double-free
when encodings were finalized on exit. [Bug #219314, #524674]
|
| |
|
|
|
|
|
|
|
| |
* generic/tclIO.c (FilterInputBytes): reset the TCL_ENCODING_START
flags in the ChannelState when using 'gets'. [Bug #523988]
Also reduced the value of ENCODING_LINESIZE from 30 to 20 as this
seems to improve the performance of 'gets' according to tclbench.
|
|
|
|
|
|
|
| |
* library/encoding/iso2022.enc:
* tools/encoding/iso2022-jp.esc:
* tools/encoding/iso2022.esc: gave <ESC>$B precedence over <ESC>$@,
based on comments (point 1) in [Bug #219283] (rfc 1468)
|
| |
|
| |
|
| |
|
|
|
|
|
| |
structure was also deleted when a command was deleted to prevent a
mem leak.
|
|
|
|
| |
correctly.
|
|
|
|
| |
tls on thread exit.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
* generic/tcl.decls (Tcl_GetIndexFromObjStruct):
* generic/tclIndexObj.c (Tcl_GetIndexFromObjStruct): Revised the
prototype of the Tcl_GetIndexFromObjStruct to take its struct
table as a (CONST VOID *) argument, better describing what it is,
maintaining source compatibility, and adding CONST correctness
according to TIP 27. Thanks to Joe English for an elegant
solution. [Bug 520304]
* generic/tclDecls.h: make genstubs
|
|
|
|
|
| |
count management errors on the interactive commandPtr Tcl_Obj found by
Purify. Thanks to Jeff Hobbs for the report and assistance.
|
| |
|
| |
|
| |
|
|
|
|
| |
in error case.
|
|
|
|
|
| |
bug #523470 provided by Don Porter <dgp@users.sourceforge.net>
to avoid timing problems in that test.
|
|
|
|
|
|
|
|
|
|
| |
"/dev/tty" (by name) and to not handle it as tty / serial
line. This is the controlling terminal and is special. Setting
it into raw mode as is done for other tty's is a bad idea. This
is a hackish fix for expect SGF Bug #520624. The fix has
limitation: Tcl_MakeFileChannel handles tty's specially too, but
is unable to recognize /dev/tty as it only gets a file
descriptor, and no name for it.
|
| |
|
| |
|