Changelog:
User-visible changes between 0.7.2 and 0.7.3:
Bug fixes
Backported protection against glibc math functions clobbering TOS
User-visible changes between 0.7.1 and 0.7.2:
Bug fixes
Makefile fixes for installing
User-visible changes between 0.7.0 and 0.7.1:
Bug fixes
amd64 gdb disassembler works with syntax change (autodetected) workaround for gcc 4.6 and 4.7 problems (newline and superinstructions)
Miscellaneous
changed repository from CVS to git
User-visible changes between 0.6.2 and 0.7.0:
Requirements:
At run-time requires libtool and gcc (for the libcc C interface) and
gdb (for the disassembler (SEE)) on some platforms.
License:
Changed to GPLv3
Bug fixes
Now works with address-space randomization.
The single-step debugger works again in some engines.
Many others.
Ports:
AMD64, ARM, IA-64 (Itanium): better performance
PPC, PPC64: disassembler and assembler
Gforth EC: R8C, 4stack, misc, 8086 work
MacOS X: better support
Invocation:
New flags --ignore-async-signals, --vm-commit (default overcommit)
--print-sequences
Forth 200x:
X:extension-query: produce true for all implemented extensions
X:required REQUIRED etc. (not new)
X:defined: [DEFINED] and [UNDEFINED]
X:parse-name: PARSE-NAME (new name)
X:deferred: deferred words (new: DEFER@ DEFER! ACTION-OF)
X:structures: +FIELD FIELD: FFIELD: CFIELD: etc.
X:ekeys: new: EKEY>FKEY K-SHIFT-MASK K-CTRL-MASK K-ALT-MASK K-F1...K-F12
X:fp-stack (not new)
X:number-prefixes (partially new, see below)
Number prefixes:
0x is a hex prefix: 0xff and 0XfF now produces (decimal) 255
# is a decimal prefix: #10 now produces (decimal) 10
Signs after the number prefix are now accepted, e.g, #-50.
' now only handles a single (x)char: 'ab is no longer accepted,
'a' now produces (decimal) 97
Unicode support (currently supports only uniform encoding):
added xchars words for dealing with variable-width multi-byte characters
provide 8bit (ISO Latin 1) and UTF-8 support for xchars
New words:
\C C-FUNCTION C-LIBRARY END-C-LIBRARY C-LIBRARY-NAME (libcc C interface)
LIB-ERROR (complements OPEN-LIB)
OUTFILE-EXECUTE INFILE-EXECUTE BASE-EXECUTE (limited change of global state)
16-bit and 32-bit memory acces: UW@ UL@ SW@ SL@ W! L! W@ L@ /W /L
NEXT-ARG SHIFT-ARGS (OS command-line argument processing)
NOTHROW (for backtrace control)
FTRUNC FMOD (undocumented)
SEE-CODE SEE-CODE-RANGE (show generated dynamic native code)
Improvements/changes of existing words:
S\", .\" now support \l, \m, \z, and limits hex and octal character specs.
OPEN-FILE with W/O no longer creates or truncates files (no compat. file)
OPEN-LIB now understands ~ at the start, like OPEN-FILE.
TRY...ENDTRY changed significantly, compatibility files available (see docs).
The disassembler (DISCODE) can now use gdb to disassemble code
Uninitialized defered words now give a warning when executed
Division is floored (disable with "configure --enable-force-cdiv")
Gforth (not gforth-fast) reports division by zero and overflow on division
on all platforms.
Newly documented words:
S>NUMBER? S>UNUMBER?
EKEY keypress names: K-LEFT K-RIGHT K-UP K-DOWN K-HOME K-END K-PRIOR
K-NEXT K-INSERT K-DELETE
CLEARSTACKS
FORM
Environment variable GFORTHSYSTEMPREFIX (used by word SYSTEM and friends)
C interface:
exported symbols now start with "gforth_" (for referencing them from C code)
libcc C function call interface (requires libtool and gcc at run-time)
alternative: undocumented libffi-based interface
Libraries:
depth-changes.fs: report stack depth changes during interpretation
ans-report.fs now reports CfV extensions
fsl-util.4th: FSL support files (undocumented)
regexp.fs for regular expressions (undocumented)
complex.fs for complex numbers (undocumented)
fft.fs for Fast Fourier Transform (undocumented)
wf.fs, a Wiki implementation (undocumented)
httpd.fs, a web server (undocumented)
status.fs, show interpreter status in separate xterm (undocumented)
profile.fs for profiling (undocumented, incomplete)
endtry-iferror.fs, recover-endtry.fs to ease the TRY change transition
test/tester.fs: Now works with FP numbers (undocumented)
test/ttester.fs: Version of tester.fs with improved interface (T{...}T).
compat library:
compat/execute-parsing.fs
Speed improvements:
automatic performance tuning on building
static stack caching (good speedup on PPC)
mixed-precision division is now faster
support for int128 types on AMD64
workarounds for gcc performance bugs (in particular, PR 15242)
branch target alignment (good speedup on Alpha).
and then adds it to LDFLAGS. Later, the configure script tests for the
presence of certain functions in libc, and expects to get an error at
link time if the functions are missing. On Interix with -export-dynamic,
the link succeeds even if there are missing functions. This causes
configure to detect all tested functions as present, and produce
binaries that fail at run time.
To work around this issue, postpone adding -export-dynamic to LDFLAGS
until the end of the configure script.
dot uji dot es> in PR 23635.
User-visible changes between 0.6.1 and 0.6.2:
Bug fixes (in particular, gforth-0.6.2 compiles with gcc-3.3)
New words: LATEST, LATESTXT (LASTXT deprecated)
Operating environment: Added optional support for a C interface built
on the ffcall libraries (more portable and powerful than the old
one, but still not documented). To use it, the ffcall libraries
have to be installed before building Gforth (see INSTALL).
Miscellaneous: Gforth-fast now uses static superinstructions (some
speedup on some platforms); generally this is transparent (apart
from the speedup), but there are lots of command-line options for
controlling the static superinstruction generation.
User-visible changes between 0.6.0 and 0.6.1:
Bug fixes (installation on big-endian machines sometimes did not work)
User-visible changes between 0.5.0 and 0.6.0:
Changes in behaviour:
S": interpreted use now ALLOCATEs the string (they live until BYE).
Long word names (512MB on 32-bit systems) are now supported (change to
the header format).
New threaded code execution method: primitive-centric (allows the
following), hybrid direct/indirect threaded (easier portability),
with dynamic superinstructions (typical speedup on Athlon: factor
2). New engine gforth-itc for dealing with some potential
backwards-compatibility problems (see "Direct or Indirect Threaded?"
in the manual).
Operating environment:
Default dictionary size is now 4MB.
Large file support on OSs that support them (i.e., files with more
than 2GB on 32-bit machines).
Gforth can now deal well with broken pipes in most situations.
vi tags files can be built with tags.fs (usage like etags.fs).
gforth.el mostly rewritten.
New image file format.
New words:
Keyboard input: EDIT-LINE K-PRIOR K-NEXT K-DELETE
File input: SLURP-FILE SLURP-FID
Programming tools: ID. .ID WORDLIST-WORDS SIMPLE-SEE
Conditional execution: [DEFINED] [UNDEFINED]
Defining Words: CONST-DOES> ]]
Input stream: PARSE-WORD EXECUTE-PARSING EXECUTE-PARSING-FILE
String comparison: STR= STR< STRING-PREFIX?
String literals: S\" .\" \"-PARSE
Floating point output: F.RDP F>STR-RDP F>BUF-RDP
Miscellaneous:
Generalized prims2x.fs into Vmgen (see README.vmgen etc.); used the
new capabilities in prims (e.g., automatic handling of the return
stack and instruction stream).
User-visible changes between 0.4.0 and 0.5.0:
Changes in behaviour:
There are now two engines: the fast engine (gforth-fast) is at least
as fast as gforth in earlier releases; the debugging engine (gforth)
supports precise backtracing for signals (e.g., illegal memory
access), but is slower by a factor of 1-2.
Block files now start at block 0 by default (instead of block 1). If
you have block files around, prepend 1024 bytes to convert them, or
do a "1 OFFSET !" to establish the old behaviour.
Gforth now does not translate newlines to LFs on reading. Instead,
READ-LINE now interprets LF, CR, and CRLF as newlines. Newlines on
output are in the OSs favourite format.
SEE now disassembles primitives (or hex-DUMPs the code if no
disassembler is available).
>HEAD (aka >NAME) now returns 0 (instead of the nt of ???) on failure.
Syntax of prim changed: stack effects are now surrounded by
parentheses, tabs are insignificant.
Operating environment:
Gforth now produces a backtrace when catching an exception.
On platforms supporting the Unix 98 SA_SIGINFO semantics, you get more
precise error reports for SIGSEGV and SIGFPE (e.g., "stack
underflow" instead of "Invalid memory address").
Gforth now produces exit code 1 if there is an error (i.e., an
uncaught THROW) in batch processing.
You can use "gforthmi --application ..." to build an image that
processes the whole command-line when invoked directly (instead of
through gforth -i).
Ports:
AIX.
20% speedup on 604e under powerpc-unknown-linux-gnu,
19%-29% speedup on Celeron with gcc-2.95.
New words:
Missing ANS Forth words: EKEY EKEY? EKEY>CHAR
Timing words: CPUTIME UTIME
Vector arithmetic: V* FAXPY
FP comparison: F~ABS F~REL
Deferred words: <IS> [IS]
Nested number output: <<# #>>
Exception handling: TRY RECOVER ENDTRY
Directory handling: OPEN-DIR READ-DIR CLOSE-DIR FILENAME-MATCH
Other: ]L PUSH-ORDER
Miscellaneous:
Significant extensions to the manual (added an introduction, among
other things), many of them due to a new team member: Neal Crook.
Added assemblers and disassemblers for 386, Alpha, MIPS (thanks to
contributions by Andrew McKewan, Bernd Thallner, and Christian
Pirker). Contributions of assemblers and disassemblers for other
architectures are welcome.
