Christophe Rhodes [Mon, 4 Jul 2005 17:07:41 +0000 (17:07 +0000)]
0.9.2.24:
Better docstring lambda-list treatment (FSVO "better", but at
least gets sb-sprof:with-profiling right)
... also add .cvsignore to internals manual directory
Nikodemus Siivola [Mon, 4 Jul 2005 13:09:20 +0000 (13:09 +0000)]
0.9.2.23: ansi fixes
* fill slots in TYPE-ERRORs from COERCE.
* classes are type designators and must be accepted by MAKE-CONDITION.
* MULTIPLE-VALUE-SETQ returns the primary value of the values form.
* COMPILE must not return NIL.
Christophe Rhodes [Mon, 4 Jul 2005 10:29:04 +0000 (10:29 +0000)]
0.9.2.22:
Merge THSpatch "Minor mips code cleanup" (sbcl-devel 2005-06-19)
... make the code a little prettier.
Christophe Rhodes [Mon, 4 Jul 2005 10:16:22 +0000 (10:16 +0000)]
0.9.2.21:
Merge THS patches for MOVE and branch delay scheduling
(3 parts, "Mips branch delay slot audit" sbcl-devel 2005-06-19)
... prefer MOVE to INST MOVE;
... be more careful with branch delay slots;
... preschedule where possible.
Christophe Rhodes [Mon, 4 Jul 2005 10:08:44 +0000 (10:08 +0000)]
0.9.2.20:
Better LOGNOR on fixnums for MIPS (THS sbcl-devel 2005-05-22)
... just subtract the mask
Christophe Rhodes [Mon, 4 Jul 2005 10:01:40 +0000 (10:01 +0000)]
0.9.2.19:
Merge THS patch "Fix headerptr initialization for cheneygc"
... use the dynamic space that's in use.
... (this is in something called from ldb -- not surprising
no-one noticed)
Nikodemus Siivola [Sun, 3 Jul 2005 13:51:06 +0000 (13:51 +0000)]
0.9.2.18: various error &co reporting improvements and build tweaks
* print the cell name for CELL-ERRORs when *PRINT-ESCAPE* is true.
* print just the filename, not the SHARED-OBJECT structure for failed
LOAD-SHARED-OBJECT.
* improve our Darwin dlerror(3)-shim reporting; also clean up warnings from
ppc-darwin-dlsim.c and reindent it for 80 columns.
* move reporting of failed contribs to make-target-contrib.sh, and re-add
--disable-debugger, which had accidentally fallen off.
Kevin Rosenberg [Sun, 3 Jul 2005 03:07:40 +0000 (03:07 +0000)]
0.9.2.17:
* contrib/sb-aclrepl/repl.lisp: update KILL-CMD for new thread
objects.
Nikodemus Siivola [Sun, 3 Jul 2005 01:11:04 +0000 (01:11 +0000)]
0.9.2.16: restore buildability on PPC
* missing VOPs from arith.lisp are here, I hope.
Kevin Rosenberg [Sat, 2 Jul 2005 21:42:11 +0000 (21:42 +0000)]
0.9.2.15:
* contrib/sb-aclrepl/repl.lisp: PROCESSES-CMD now recognizes
that threads are objects, not numbers.
Christophe Rhodes [Sat, 2 Jul 2005 15:33:46 +0000 (15:33 +0000)]
0.9.2.14:
Make it build, at least, on sparc.
Alexey Dejneka [Sat, 2 Jul 2005 06:49:12 +0000 (06:49 +0000)]
0.9.2.13:
* Add constant folding of ARRAY-HEADER-P for complex-array
argument (fixes spurious warning reported by Eduardo
Mu\~noz).
* FOREIGN-SYMBOL-ADDRESS-AS-INTEGER does not loose second
result (fix bug reported by Luis Oliveira).
* Don't run external format checks on SBCL without Unicode
support.
Gabor Melis [Fri, 1 Jul 2005 14:35:00 +0000 (14:35 +0000)]
0.9.2.12:
* bug fix: no more highly sporadic "couldn't check whether ~S is
readable" when reading a stream and an interrupt hits in the middle
of a select system call
* added with-restarted-syscall macro
* added abcl support to build scripts
Christophe Rhodes [Fri, 1 Jul 2005 11:18:54 +0000 (11:18 +0000)]
0.9.2.11:
Document in an internalsy kind of way the various string
types that exist.
Gabor Melis [Fri, 1 Jul 2005 11:00:26 +0000 (11:00 +0000)]
0.9.2.10:
* threads for x86-64
Gabor Melis [Fri, 1 Jul 2005 08:48:08 +0000 (08:48 +0000)]
0.9.2.9: thread objects
* Public interface changes
** proper thread objects instead of thread ids
** (MAKE-THREAD FN &KEY NAME) => THREAD
** (THREAD-NAME THREAD): threads have names (useful for debugging,
logging)
** (THREAD-ALIVE-P THREAD)
** *CURRENT-THREAD* special is bound in each thread
** (LIST-ALL-THREADS) returns a list of all active threads
* Notes
** thread-init moved earlier in cold-init and reinit
** the lisp side does not ever use os_thread_t (it was problematic due
to pthread_t being opaque) but struct thread *
** threads are reaped (i.e. the thread is pthread_joined and struct
thread* is freed) by the thread object's finalizer. This makes
it easy to implement resetting threads. Running threads are kept
in sb-thread::*all-threads*.
** threads block all blockable signals when going down:
interrupt-thread and others cannot catch it at an inappropriate
moment, for instance calling quit outside the catch %end-of-the-world
** target-thread.lisp renamed target-multithread.lisp,
target-thread.lisp now contains the generic thread support
** new file early-thread.lisp: define *current-thread*
** removed thread state STOPPING that was only used for assertions and
complicated matters
** renumbered thread states
** sb-thread::release-spinlock now releases the locks with
non-fixnum value, but is no longer safe to call multiple times
** much simplified locking for threads and gc
** fixed deadlocking bugs introduced by the pthread merge
Rudi Schlatte [Fri, 1 Jul 2005 05:59:39 +0000 (05:59 +0000)]
0.9.2.8
Make interactive-stream-p a generic function so Gray streams can
specialize on it.
Nathan Froyd [Thu, 30 Jun 2005 20:56:33 +0000 (20:56 +0000)]
0.9.2.7:
Enable slightly more efficient operations with 32-bit constants
on PPC. This reduces our instruction count in the inner loop
of SB-MD5 by nearly 10% and brings us closer to the code that
GCC generates.
Nathan Froyd [Thu, 30 Jun 2005 20:15:59 +0000 (20:15 +0000)]
0.9.2.6:
Fix ASH to fold away shifts of zero based on the actual derived
type, not just the constancy of the arg.
Brian Mastenbrook [Wed, 29 Jun 2005 00:11:37 +0000 (00:11 +0000)]
0.9.2.5: fix failure in OPEN when *print-readably* is T (thanks to Xach)
Nathan Froyd [Tue, 28 Jun 2005 14:56:11 +0000 (14:56 +0000)]
0.9.2.4:
Permit specialized arrays to be placed in a function's constant
vector, enabling faster access (important in, e.g.,
table-driven crypto algorithms).
Rudi Schlatte [Tue, 28 Jun 2005 14:22:37 +0000 (14:22 +0000)]
0.9.2.3:
bivalent streams: streams opened with :element-type :default now allow
character and binary (unsigned-byte 8) I/O
Rudi Schlatte [Tue, 28 Jun 2005 12:27:35 +0000 (12:27 +0000)]
0.9.2.2
Minor test cleanups: remove temporary files after tests finish.
Rudi Schlatte [Tue, 28 Jun 2005 09:09:40 +0000 (09:09 +0000)]
0.9.2.1
Support for koi8-r external format (thanks to Ivan Boldyrev)
William Harold Newman [Mon, 27 Jun 2005 21:10:45 +0000 (21:10 +0000)]
0.9.2:
release, tagged as sbcl_0_9_2
Gabor Melis [Thu, 23 Jun 2005 09:30:22 +0000 (09:30 +0000)]
0.9.1.65:
* link against pthreads only if featurep :sb-thread
Alexey Dejneka [Mon, 20 Jun 2005 18:36:12 +0000 (18:36 +0000)]
0.9.1.64:
* A slightly modified patch by Nikodemus Siivola: compiler
consistency checking did not work with delayed optional
entries.
Christophe Rhodes [Mon, 20 Jun 2005 16:52:45 +0000 (16:52 +0000)]
0.9.1.63:
DX for sparc.
Nikodemus Siivola [Mon, 20 Jun 2005 16:17:01 +0000 (16:17 +0000)]
0.9.1.62: align CSP for dx closures on alpha and ppc
* test-case pending.
* untested on alpha.
A cookie for the first to implement Christophe's suggestion to "adapt
pfdietz' random tester to generate correct dynamic-extent
declarations and calls with many arguments" that produces something that
fails on ppc 0.9.1.61, but passes on .62 ...
Nikodemus Siivola [Mon, 20 Jun 2005 12:43:50 +0000 (12:43 +0000)]
0.9.1.61: really allocate dx closures on stack on ppc and alpha
* add a test-case.
* untested on alpha.
Gabor Melis [Mon, 20 Jun 2005 12:39:12 +0000 (12:39 +0000)]
0.9.1.60:
* sb-thread::release-spinlock now releases the locks with
non-fixnum value, but is no longer safe to call multiple times
Gabor Melis [Sun, 19 Jun 2005 19:35:41 +0000 (19:35 +0000)]
0.9.1.59:
merged most of the pthreads stuff from amd64-pthread-branch except
the amd64 part, plus:
* in the runtime thread_kill, thread_sigmask, thread_self
stand for pthread_kill or kill, sigprocmaks or
pthread_sigmask, and pthread_self or getpid respectively
controlled by the sb-thread feature
* fixed recursive get on session-lock that happened when a gc
interrupting get-foreground reaped a thread
* fixed sigint handling: removed broken (by pthread signal
handling semantics) sigint enable/disable machinery in favor
of sigint-%break looking up the foreground thread and
interrupting it, which is itself racy :-(.
* numerous fixes for interrupt-thread
* threads block signals until they are set up properly
* removed suspend-thread, resume-thread
* destroy-thread is now equivalent to terminate-thread.
Christophe Rhodes [Sun, 19 Jun 2005 17:38:41 +0000 (17:38 +0000)]
0.9.1.58:
Really stack-allocate dx closures on x86-64.
Nikodemus Siivola [Sun, 19 Jun 2005 16:16:48 +0000 (16:16 +0000)]
0.9.1.57: just how hard can "cvs add" be?
...the internals manual, really, this time...
Nikodemus Siivola [Sun, 19 Jun 2005 16:01:59 +0000 (16:01 +0000)]
0.9.1.56: start of an internals manual
* documenting initially "Foreign Linkage": linkage table, lazy
aliens, and callbacks.
Nikodemus Siivola [Sun, 19 Jun 2005 15:57:43 +0000 (15:57 +0000)]
0.9.1.55: trivial cleanup, new documentation directory
* delete stray "for Alpha" from ppc backend.
* doc/internals/ for internals manual.
Nikodemus Siivola [Sun, 19 Jun 2005 13:48:16 +0000 (13:48 +0000)]
0.9.1.54: dynamic-extent lists and closures on ppc
* Implement stack-allocating of lists and closures on PPC, cribbing
liberally from the Alpha backend.
Nikodemus Siivola [Sun, 19 Jun 2005 11:22:59 +0000 (11:22 +0000)]
0.9.1.53: dynamic-extent &rest on ppc
* Implement stack-allocation of &REST lists on PPC, cribbing
liberally from the Alpha backend.
Alexey Dejneka [Sun, 19 Jun 2005 06:30:50 +0000 (06:30 +0000)]
0.9.1.52:
* Implement stack allocation of dynamic extent lists,
&REST-lists and closures for Alpha-32;
... remove obsolete "economic" implementation of stack
allocation.
Nathan Froyd [Fri, 17 Jun 2005 19:23:29 +0000 (19:23 +0000)]
0.9.1.51:
* Improve UB*-BASH-COPY by forcing the compiler to use fixnum
arithmetic on word-sized copying--avoids spurious fixnum
arithmetic overflow checking. This should improve the
performance of REPLACE on character strings on 32-bit platforms;
* ...use ~D instead of ~A to form UB*-BASH-* function names,
protecting ourselves against arbitrary *PRINT-BASE*s;
* ...belatedly improve test suite for bashing functions to
print out the function being tested. Doing this at least
partly assures the user that the test run has not hung.
Nikodemus Siivola [Thu, 16 Jun 2005 20:42:42 +0000 (20:42 +0000)]
0.9.1.50: enable callback tests on x86
Nikodemus Siivola [Thu, 16 Jun 2005 20:39:51 +0000 (20:39 +0000)]
0.9.1.49: callbacks on x86
* thanks to David Lichteblau for massaging the code originally
ported to SBCL by Thomas F. Burdick, based on the work for CMUCL
by Helmut Eller, to the current scheme of things.
...now what just the sparc backend remains to be ported from sbcl-callables,
and a new one for mips...
...and getting the interface straight, and rebustifying the code a bit:
sbcl-callables includes eg. some logic to check that the types given
are compatible -- reinstating this sounds like a good idea one things
settle down...
Christophe Rhodes [Thu, 16 Jun 2005 14:48:00 +0000 (14:48 +0000)]
0.9.1.48:
Implement (most of) the dynamic-extent vops on x86-64.
... no support for vectors yet.
Rudi Schlatte [Thu, 16 Jun 2005 14:19:28 +0000 (14:19 +0000)]
0.9.1.47
Merge EUC-JP external format support and tests
- thanks to NIIMI Satoshi for the patch
Nikodemus Siivola [Thu, 16 Jun 2005 12:43:51 +0000 (12:43 +0000)]
0.9.1.46: refactoring callbacks
* turn ALIEN-CALLBACK into a macro so we avoid a runtime call to the compiler.
* additional bits of interface: ALIEN-CALLBACK-P, ALIEN-CALLBACK-FUNCTION,
(SETF ALIEN-CALLBACK-FUNCTION), and INVALIDATE-ALIEN-CALLBACK.
* more tests.
Christophe Rhodes [Thu, 16 Jun 2005 09:37:29 +0000 (09:37 +0000)]
0.9.1.45:
Merge THS "Mips assembly revisited" sbcl-devel 2005-06-15
Christophe Rhodes [Thu, 16 Jun 2005 09:05:03 +0000 (09:05 +0000)]
0.9.1.44:
Fix for CALL-METHOD / RESTART-CASE interaction.
... patch from Gerd Moellmann cmucl-imp 2005-06-04
... (Bruno's test case is quite neat.)
Nikodemus Siivola [Wed, 15 Jun 2005 17:06:29 +0000 (17:06 +0000)]
0.9.1.43: more callback work
* trivial implementations of DEFINE-ALIEN-CALLBACK and ALIEN-LAMBDA.
D-A-C is essentially OK, A-L sucks in any number of ways.
* tests
Nikodemus Siivola [Wed, 15 Jun 2005 06:13:13 +0000 (06:13 +0000)]
0.9.1.42: partial callback implementation
* SB-ALIEN-INTERNALS:ALIEN-CALLBACK (to be exported from SB-ALIEN at
a later date) makes callbacks for functions designators that can
be ALIEN-FUNCALLed or passes as function pointers to C-code. Based
on patch by Thomas F. Burdick based on work for CMUCL by Helmut
Eller.
PPC/Darwin only for now.
Alexey Dejneka [Mon, 13 Jun 2005 16:36:02 +0000 (16:36 +0000)]
0.9.1.41:
* BIT-* functions are not foldable (reported by Paul F.
Dietz).
Alexey Dejneka [Mon, 13 Jun 2005 11:05:44 +0000 (11:05 +0000)]
0.9.1.40:
* COERCE for objects other than numbers and characters is not
foldable (fix a bug reported by Nikodemus Siivola).
* Set LC_ALL to "C" when building SBCL.
* Set locale to "C" in run-tests.sh.
Christophe Rhodes [Sun, 12 Jun 2005 22:08:23 +0000 (22:08 +0000)]
0.9.1.39:
Restore alpha buildability.
... define two-arg-/= as a static function;
... don't grep out A symbols from nm.
Christophe Rhodes [Sun, 12 Jun 2005 14:02:34 +0000 (14:02 +0000)]
0.9.1.38:
Merge DFL raw-slots patch (sbcl-devel "raw slot changes"
2005-05-18)
... with an amalgam of ths' two mips versions;
... note in OPTIMIZATIONS about the negative index idea, and the
disabledness of HPPA
Christophe Rhodes [Sat, 11 Jun 2005 20:46:19 +0000 (20:46 +0000)]
0.9.1.37:
Maybe fix the build on alpha.
... one more WITH-FIXED-ALLOCATION problem.
Christophe Rhodes [Fri, 10 Jun 2005 09:24:50 +0000 (09:24 +0000)]
0.9.1.36:
Merge THS "Branch relaxation..." sbcl-devel 2005-06-09
Christophe Rhodes [Fri, 10 Jun 2005 09:01:32 +0000 (09:01 +0000)]
0.9.1.35:
Fix PFD "Weird SBCL bug..." sbcl-devel 2005-06-09
... MORE PROGRAM-ERRORs.
Gabor Melis [Thu, 9 Jun 2005 10:44:29 +0000 (10:44 +0000)]
0.9.1.34:
* gcc 2.95: compilation fix
* killed a few innocent looking warnings in the runtime (79 to go)
Gabor Melis [Thu, 9 Jun 2005 08:32:29 +0000 (08:32 +0000)]
0.9.1.33:
* gencgc: don't flood the system with gc interrupts if not
absolutely necessary
* fix compilation on freebsd
Christophe Rhodes [Wed, 8 Jun 2005 14:10:23 +0000 (14:10 +0000)]
0.9.1.32:
Implement direct ub32 -> float on PPC.
Christophe Rhodes [Wed, 8 Jun 2005 11:16:48 +0000 (11:16 +0000)]
0.9.1.31:
Remove sigsetmask() calls on alpha, ppc missed out of previous
signal rearrangement
... (sb-sprof on ppc appears to work now :-)
Juho Snellman [Wed, 8 Jun 2005 10:36:17 +0000 (10:36 +0000)]
0.9.1.30:
* Fix compile on x86-64 (popfl -> popfq).
* Declarations in a DOLIST body are also in scope for the
DOLIST return-form.
* Evaluate the ELEMENT-TYPE keyword argument to
WITH-OUTPUT-TO-STRING even in cases where the element-type is
not used (i.e. FILL-POINTER-OUTPUT-STREAM), in case it has
side-effects.
* COMPILE-FILE accepts all pathname designators as INPUT-FILE
and OUTPUT-FILE (streams weren't accepted before).
Gabor Melis [Wed, 8 Jun 2005 08:49:49 +0000 (08:49 +0000)]
0.9.1.29:
A number of signal handling cleanup/fixes:
* fix gencgc/maybe_defer_handler race (thanks to Thiemo)
* interrupt_maybe_gc (on cheneygc platforms): check for already
pending handler before calling maybe_defer_handler in order
not to lose interrupts
* interrupt_handle_pending: check for the pending handler being
null
* run_deferred_handler: set the pending handler to null, before
calling it to guard against the handler enabling interrupts ...
* more defensiveness: enforce invariants: checks for signal masks,
interrupts
* refactoring: undoably_install_low_level_interrupt_handler wraps
blockable handlers in low_level_maybe_now_maybe_later
* don't unblock signals unconditionally in interrupt_maybe_gc_int just
restore the sigmask from the interrupted context (kludge removed)
* removed misguided sigprocmask calls from mips, hppa and sparc
sig{trap,ill} handlers
* removed non-x86 version of handle_breakpoint (interrupts are enabled
in now common handle_breakpoint)
* fixed arrange_return_to_lisp_function/post_signal_tramp to save and
restore eflags (interrupt-threads seems to work)
Juho Snellman [Tue, 7 Jun 2005 10:49:56 +0000 (10:49 +0000)]
0.9.1.28:
MORE PRETTY:
* "Oops". Change the definition of the recenctly introduced
WITH-CIRCULARITY-DETECTION macro to only include BODY once.
Otherwise code with nested PPRINT-LOGICAL-BLOCKs would
expand into ridiculous amounts of code. (The macroexpansion
of PPRINT-LAMBDA-LIST was >18000 lines).
* Return NIL from PPRINT-POP when OBJECT is NIL.
* Check for malformed LABELS/FLET/MACROLET forms in PPRINT-FLET.
Other stuff:
* Declare a more specific type for *CHARACTER-DATABASE* to
avoid going through HAIRY-DATA-VECTOR-REF. (thanks to
Christophe for noticing this)
* Add missing type predicate for VECTOR-NIL-P. (MISC.596)
Nikodemus Siivola [Mon, 6 Jun 2005 06:54:21 +0000 (06:54 +0000)]
0.9.1.27: (truename "symlink-to-dir") === (truename "symlink-to-dir/")
* it appears that libc doesn't like trailing slashes at the end of symlink
names; paper over this sillyness.
Juho Snellman [Sun, 5 Jun 2005 11:37:01 +0000 (11:37 +0000)]
0.9.1.26:
Fix some circularity detection issues in the pretty printer.
* Move the circularity detection infrastructure into early-print.lisp.
* Do circularity checks in PPRINT-LOGICAL-BLOCK.
* Add a couple of new tests, disable an old test which is (IMHO)
invalid.
Nikodemus Siivola [Sat, 4 Jun 2005 09:54:42 +0000 (09:54 +0000)]
0.9.1.25: source-plist & related
* add :SOURCE-PLIST option to WITH-COMPILATION-UNIT for attaching
arbitrary source information to compiled code, accessible as
SB-INTROSPECT:DEFINITION-SOURCE-PLIST.
* COMPILED-DEBUG-INFO-SOURCE was always a list of one element, make
it just the element.
* rename DEBUG-SOURCE-INFO to DEBUG-SOURCE-FUNCTION for clarity.
Caveat updater: this breaks current Slime.
Nathan Froyd [Fri, 3 Jun 2005 13:19:48 +0000 (13:19 +0000)]
0.9.1.24:
s/32/64/ in :NOTEs for x86-64 arithmetic vops where appropriate.
Christophe Rhodes [Fri, 3 Jun 2005 10:17:25 +0000 (10:17 +0000)]
0.9.1.23:
Merge THS "Fix generic-/= for alpha, mips, ppc, sparc" patch
(sbcl-devel 2005-05-29)
... actually it's mostly a mips optimization, because I don't
think generic-/= can ever be emitted with the source
transforms we have at the moment.
Nathan Froyd [Thu, 2 Jun 2005 14:38:54 +0000 (14:38 +0000)]
0.9.1.22:
*sigh* Correct SPECIF*I*ER-TYPE typo.
Nathan Froyd [Thu, 2 Jun 2005 04:02:07 +0000 (04:02 +0000)]
0.9.1.21:
* Add unsigned bounds derivers for LOGXOR, based on the ones
present in CMUCL;
* Convert existing unsigned bounds derivers to a more idiomatic
CL style, eliminating unnecessary work along the way;
* Belatedly add tests for bounds derivation.
Nathan Froyd [Wed, 1 Jun 2005 22:31:34 +0000 (22:31 +0000)]
0.9.1.20:
Fix typo in LOGAND-DERIVE-UNSIGNED-HIGH-BOUND (thanks to jsnell
for testing and rtoy for his CMUCL work on the same).
(Since CMUCL has essentially the same code, a future commit
will add similar derivers for LOGXOR. A more idiomatic CL
style will also be adopted.)
Juho Snellman [Wed, 1 Jun 2005 21:30:39 +0000 (21:30 +0000)]
0.9.1.19:
* Fix build on multiarch x86-64
* Fix small ansi-bug. The (declare (typespec var*)) abbreviation for
(declare (type typespec var*)) now works with all type specifiers.
Previously only symbols were allowed.
Christophe Rhodes [Wed, 1 Jun 2005 14:46:24 +0000 (14:46 +0000)]
0.9.1.18:
Fix for the modular arithmetic improvement from nfroyd
... lognot is unary :-)
... (I don't have _Hacker's Delight_, so I guessed what was
meant).
Nathan Froyd [Wed, 1 Jun 2005 14:06:49 +0000 (14:06 +0000)]
0.9.1.17:
Disable bit-bashing transforms on (VECTOR NIL).
Christophe Rhodes [Wed, 1 Jun 2005 13:59:02 +0000 (13:59 +0000)]
0.9.1.16:
Implement SOCKET-OPEN-P mostly per Tony Martinez sbcl-devel
2004-10-23.
... add a test or two for it, which look a bit weird to me
but what do I know?
Nathan Froyd [Wed, 1 Jun 2005 13:41:40 +0000 (13:41 +0000)]
0.9.1.15:
Improve modular arithmetic by:
* Making bounds of (unsigned) modular function type derivation
more precise;
* Being less pessimistic when deriving bounds for LOGAND in
the case when its arguments are known to be bounded and
unsigned. As a bonus, LOGIOR type derivation in the same
case is now more precise as well.
The upshot of all this is that a function like:
(defun foo (x y)
(declare (type (integer 0 3) x y))
(mod (- (+ x 4) y) 4))
now uses only fixnum arithmetic.
Christophe Rhodes [Wed, 1 Jun 2005 13:19:36 +0000 (13:19 +0000)]
0.9.1.14:
Fix the canonicalized default initarg protocol
... (KEY FORM FUN) rather than (KEY FUN FORM).
I hope this isn't too painful for anyone out there.
Christophe Rhodes [Wed, 1 Jun 2005 12:08:50 +0000 (12:08 +0000)]
0.9.1.13:
Fix for Haiblebug "ensure-generic-function doesn't support
:declare"
... simply allow both :declare and :declarations;
... add some documentation and index entries.
Christophe Rhodes [Wed, 1 Jun 2005 11:42:43 +0000 (11:42 +0000)]
0.9.1.12:
Merge "Add socket-not-connected-error to sb-bsd-sockets" (Tony
Martinez sbcl-devel 2004-11-06)
... NOT-CONNECTED-ERROR seems to match the naming scheme better
than SOCKET-NOT-CONNECTED-ERROR.
Christophe Rhodes [Wed, 1 Jun 2005 11:22:59 +0000 (11:22 +0000)]
0.9.1.11:
Merge patch ("SOCKET-CLOSE closes different socket's fd stream"
Tony Martinez sbcl-devel 2004-10-25 *sigh*)
... there's a suspicious-looking bit of stuff in the cond clause
above it, too: what happens if closing the stream causes
an error? FIXME.
Christophe Rhodes [Wed, 1 Jun 2005 10:07:39 +0000 (10:07 +0000)]
0.9.1.10:
MORE DIFFSTAT
... merge Thiemo Seufer "Makefile improvements for the C
runtime" sbcl-devel 2005-05-29
... better dependency chasing by gcc;
... s/arch/LISP_FEATURE_ARCH/, hooray.
Add stuff to .cvsignores until it mostly stops emitting ?s
Untested on anything not x86-64. Shout if broken.
Juho Snellman [Wed, 1 Jun 2005 06:06:20 +0000 (06:06 +0000)]
0.9.1.9:
Fix a few ansi-test failures. Symbols can't be both the name of
a type and the name of a declaration (ANSI 3.8.21).
* Add a validate-function slot into the globaldb TYPE-INFO
structure. If defined, the function is called from (SETF INFO)
before SET-INFO-VALUE.
* Define validation functions for (INFO :TYPE :KIND) and
(INFO :DECLARATION :RECOGNIZED).
* Remove (DECLAIM (DECLARATION CLASS)) from PCL. As far as I
can see, PCL uses %CLASS for its internal declarations these
days.
Juho Snellman [Mon, 30 May 2005 05:25:44 +0000 (05:25 +0000)]
0.9.1.8:
* Fix typo (current-dynamic-space-spart -> current-dynamic-space-start)
in the gencgc branch of the "clean up DYNAMIC-SPACE-START
and -END ugliness on cheney-platforms" changes in 0.9.1.5.
* Fix compiling with GCC 4 on x86 and x86-64 (sbcl-devel
"Fixes for gcc4", Sascha Wilde).
* Remove a leftover "with-tn@fp-top(x)" from x86-64/float.lisp
(sbcl-devel , James Knight)
* More x86-64 fp cleanups. (sbcl-devel "x86-64 fp exceptions",
"x86-64 move-*-float-arg bug", James Knight 2005-05-27/29).
Nikodemus Siivola [Sun, 29 May 2005 21:08:45 +0000 (21:08 +0000)]
0.9.1.7: "fix" SB-SPROF on non-gencgc platforms
* block some "potentially dangerous, but not really important"
signals for GC on non-gencgc platforms. see kludge_sigset_for_gc
for commentary.
Gabor Melis [Sun, 29 May 2005 15:55:13 +0000 (15:55 +0000)]
0.9.1.6:
* fixed some lockups due to gc/thread interaction
* removed unused functions: unblock_sigcont_and_sleep, block_sigcont
Nikodemus Siivola [Sun, 29 May 2005 15:48:37 +0000 (15:48 +0000)]
0.9.1.5: bigger dynamic space on PPC plus cleanups
* patch by Cyrus Harmon
* clean up DYNAMIC-SPACE-START and -END ugliness on cheney-platforms.
Nikodemus Siivola [Sun, 29 May 2005 10:37:20 +0000 (10:37 +0000)]
0.9.1.4: ENSURE-CLASS-USING-CLASS patch by Gerd Moellman, from cmucl-imp
* accept a class as :METACLASS as per AMOP.
Rudi Schlatte [Sun, 29 May 2005 10:27:38 +0000 (10:27 +0000)]
0.9.1.3
Small stream cleanups:
- Use same names in generalized-peeking-mechanism as cmucl
- set-fd-stream-routines can now in theory be used to change stream
element-type and external-format (but this clashes with
fast-read-char/-byte, since these work off another layer of buffers)
Christophe Rhodes [Sat, 28 May 2005 09:00:34 +0000 (09:00 +0000)]
0.9.1.2:
Merge a couple of cleanup fixes that I've had tagged for months
if not years.
Juho Snellman [Thu, 26 May 2005 22:53:59 +0000 (22:53 +0000)]
0.9.1.1:
* Invalid dotted lists no longer raise a read error when
*READ-SUPPRESS* is T
* Use a more tasteful :EXPECTED-TYPE for type errors related to
function names
William Harold Newman [Thu, 26 May 2005 19:11:40 +0000 (19:11 +0000)]
0.9.1:
release, tagged as sbcl_0_9_1
Juho Snellman [Sun, 22 May 2005 00:07:45 +0000 (00:07 +0000)]
0.9.0.41:
Increment +FASL-FILE-VERSION+.
Rudi Schlatte [Fri, 20 May 2005 16:43:21 +0000 (16:43 +0000)]
0.9.0.40
Fix peek-char bug reported by Fredrik Sandstrom (sbcl-devel
2005-05-17, "Bug in peek-char")
Christophe Rhodes [Fri, 20 May 2005 09:45:14 +0000 (09:45 +0000)]
0.9.0.39:
Merge amended "64bit FFI for mips/mipsel" patch (Thiemo Seufer
sbcl-devel 2005-05-09)
... including 64-bit return value in call_into_lisp. (Not
currently used)
... other minor MIPS cleanups and bugfixes too as described
in the mail message.
Juho Snellman [Thu, 19 May 2005 02:50:38 +0000 (02:50 +0000)]
0.9.0.38:
Fix a few ansi-test bugs:
* The type-errors signalled for invalid function names now have
a correct (if ugly) expected type.
* Functions taking type names as arguments correctly signal
type-errors (instead of package-lock errors, arg-count-errors,
etc) for some pathological non-function names (e.g (SETF),
(SETF . BAR)).
* (SETF (DOCUMENTATION ... 'STRUCTURE)) no longer signals an error
for structures defined with a :TYPE.
* Documentation strings specified in the DEFSTRUCT form for
typed structures are no longer immediately discarded (not
strictly a bug, just a quality of implementation issue...)
* FILE-STRING-LENGTH and STREAM-EXTERNAL-FORMAT now work on
non-fd-streams too.
* FILE-LENGTH now also works on broadcast streams. The spec
has slightly conflicting opinions on this issue; FILE-LENGTH
description says that stream must be associated with stream
or an error is signalled. BROADCAST-STREAM description
explicitly describes how FILE-LENGTH must be implemented.
Juho Snellman [Wed, 18 May 2005 02:22:48 +0000 (02:22 +0000)]
0.9.0.37:
* Merge sbcl-devel "Unneeded REX prefixes on x86-64"
(Lutz Euler, 2005-05-17)
* Merge sbcl-devel " x86-64 fp exceptions" (James Knight,
2005-05-17)
* Add "--userinit /dev/null --sysinit /dev/null" to the
default SBCL_XC_HOST.
* Clean up some stale x86 code (mostly floating point related)
in the x86-64 port.
Rudi Schlatte [Mon, 16 May 2005 19:40:08 +0000 (19:40 +0000)]
Fix bug 242 for fd-streams
... fd-streams have a method that writes n octets; use it.
Rudi Schlatte [Mon, 16 May 2005 12:19:33 +0000 (12:19 +0000)]
0.9.0.35
Slight depessimization of WRITE-SEQUENCE of byte arrays
... Don't go through full stream dispatch machinery for every byte
Rudi Schlatte [Sun, 15 May 2005 20:09:56 +0000 (20:09 +0000)]
0.9.0.34
Eliminate unnecessary seeks on socket streams
... Introduce dual-channel-p flag in fd-streams to discriminate
between files and socket/pipe-streams
Christophe Rhodes [Sun, 15 May 2005 15:03:07 +0000 (15:03 +0000)]
0.9.0.33:
Make SB-MD5 enforce its requirements.
... also keep lambda lists around, for manual autogroveling
purposes