common-lisp/sbcl.sbcl
1
0
Fork 0
mirror of git://git.code.sf.net/p/sbcl/sbcl synced 2026-09-10 07:26:40 -04:00
Code Issues Actions 8 Packages Projects Releases Wiki Activity
28116 commits 35 branches 284 tags 116 MiB
Commit graph

28116 commits

This branch
This branch
All branches
Author SHA1 Message Date
Christophe Rhodes 1a0e4afd51 2.6.4: will be tagged as "sbcl-2.6.4" 2026-04-29 20:26:57 +01:00
Stas Boukarev 7cabf02c65 Spell-check NEWS 
[skip ci]
 2026-04-28 23:52:26 +03:00
Christophe Rhodes 3fbd492e48 fixes to NEWS 2026-04-28 20:35:36 +01:00
Stas Boukarev 29216741cf Skip a test 2026-04-28 14:24:44 +03:00
Stas Boukarev 7c5b07f38a Add c_ispeed and c_ospeed fields to sb-posix:termios 
Should fix a problem with OpenBSD.

Based on a patch by Robert Palm.

Fixes lp#2150499
 2026-04-28 14:01:14 +03:00
Stas Boukarev 25215818e2 Disable a test 2026-04-28 04:12:05 +03:00
Douglas Katzman 114538b378 Fix input-manifest 2026-04-27 20:40:10 -04:00
Jesse Bouwman 6a39892471 x86-64, struct-by-value: coordinate arglist transforms 2026-04-27 14:05:06 -07:00
Christophe Rhodes 6fff197f35 First draft of NEWS for sbcl-2.6.4 2026-04-27 18:30:32 +01:00
Stas Boukarev 8ecb4fb731 Revert "Persist the :ucrt feature, if present, into the final image" 
Until it's known if users need that.

This reverts commit 4dde13a7c1.
 2026-04-27 14:46:42 +03:00
Jesse Bouwman 314b572626 x86-64, struct-by-value: correct ABI classification for nested structs 2026-04-26 17:44:30 -07:00
Jesse Bouwman f484412bc4 x86-64, struct-by-value: prevent overwrites when copying during callbacks 2026-04-26 11:11:18 -07:00
Stas Boukarev f735fd577c x86-64, struct callbacks: leave registers if a struct doesn't fit 
Which was copied from arm64.
 2026-04-26 19:16:19 +03:00
Jesse Bouwman a190d9710a struct-by-value: don't overwrite when copying values from registers to memory 2026-04-26 12:18:26 +03:00
Jesse Bouwman c4c736c568 tests: failing overwrite cases 2026-04-26 12:18:26 +03:00
Douglas Katzman 71cca79006 Fix strange comment 
which made sense only if you knew it so well as to not need the comment.
 2026-04-26 01:18:17 -04:00
Douglas Katzman b0cdf28901 Improve RELEASE-MUTEX docstring 
"set it to NIL" may have made some sense over 20 years ago when a mutex
was essentially just a container of one value, namely the owning thread.
 2026-04-26 00:22:39 -04:00
Stas Boukarev 5c866b5159 x86-64, struct-by-value: leave registers if a struct doesn't fit 
Something else might go into that register.
 2026-04-25 21:06:13 +03:00
Jesse Bouwman e3d6e48da2 x86-64, struct-by-value: don't read past input struct 2026-04-25 18:33:51 +03:00
Jesse Bouwman fa6a16b5ec tests: struct-by-value parameter overread 2026-04-25 18:27:16 +03:00
Stas Boukarev 0132c442dd Resolve a loop in constraints 
Process newly changed blocks always in the same order.

Reported by _death
 2026-04-25 02:51:06 +03:00
Christophe Rhodes 4dde13a7c1 Persist the :ucrt feature, if present, into the final image 
It's probably not useful for users, but it is necessary for making the
right choice while building our own contribs.  Acting on a report by
Masatoshi SANO.
 2026-04-24 12:12:45 +01:00
Stas Boukarev 8a98e7b59d x86-64, struct-by-value: fix record-arg-tn for win32 
Correctly write to stack tns.
 2026-04-24 00:20:22 +03:00
Stas Boukarev 627064c1dc x86-64, struct-by-value: correctly pass small structs on the stack 2026-04-23 20:16:38 +03:00
Stas Boukarev 467666387e arm64: don't do (ldrsw (32-bit-reg x) ..) 
It's supposed to sign extend into a 64-bit register.
The assembler doesn't care, but there's no reason to create
unnecessary tns.
 2026-04-23 20:16:38 +03:00
Stas Boukarev f1109de3da arm64, callbacks: return float results in the right register 2026-04-23 13:28:19 +03:00
Stas Boukarev 3d222e98b2 arm64, struct-by-value: don't read past the input struct 2026-04-23 12:24:30 +03:00
Stas Boukarev cd3e8721e3 arm64, struct-by-value: don't consume gprs when running out of fprs 2026-04-23 11:27:20 +03:00
Stas Boukarev 0cb8e863d1 arm64, struct-by-value: don't split stack/registers in callbacks 2026-04-23 11:18:59 +03:00
Stas Boukarev 2c30eafc7e x86-64, struct-by-value: correctly pass small structs on the stack 2026-04-23 10:50:34 +03:00
Stas Boukarev 0e7a0d7981 arm64, struct-by-value: correctly pass small structs on the stack 2026-04-23 10:40:05 +03:00
Stas Boukarev 29a23fdce1 arm64, struct-by-value: look at the stack when running out of GPRs 2026-04-23 08:09:39 +03:00
Stas Boukarev 4b3abf2f8c arm64: fix alien-struct-by-value copying 
Don't stop too early.
 2026-04-23 07:48:01 +03:00
Stas Boukarev da85ccc9b4 Fix logtest on ppc64 
Where n-fixnum-tag-bits is different from ppc32 (and from everything else)
 2026-04-23 00:08:42 +03:00
Douglas Katzman b5150429f7 Delete immobile space sigsegv handler 2026-04-22 12:29:11 -04:00
Stas Boukarev 315d1ee568 x86-64: use xorps for conjugate/complex-single-float 2026-04-22 04:21:32 +03:00
Stas Boukarev 77337ebd4a Typos in arm64/float.lisp 2026-04-21 03:11:43 +03:00
Stas Boukarev 82f008f592 arm64: use the right register in ash-inverted 2026-04-20 23:57:16 +03:00
Stas Boukarev 97ecfac9da arm64: fix a potential issue with fast-ash-c/signed=>signed 
Which can't be triggered due to VOP selection.
 2026-04-20 23:28:47 +03:00
Stas Boukarev d81da0bd3f x86-64: fix mask-signed-field-integer for 64 and any-reg result 2026-04-20 22:57:30 +03:00
Stas Boukarev 79752fb42a Address a FIXME in emit-inline-add-sub 2026-04-20 20:44:18 +03:00
Stas Boukarev 8595450051 Typos 2026-04-20 20:16:27 +03:00
Stas Boukarev 0079393bff let-lvar-dest: ignore NIL lvars 
Fixes lp#2148799
 2026-04-20 16:09:23 +03:00
Stas Boukarev 3eb58efbbe Transform floor/ceiling to truncate when used as a divisibility test 2026-04-20 04:26:35 +03:00
Douglas Katzman 914726b494 Implement choice of using fastrem in IR1 transform 2026-04-19 18:05:03 -04:00
Stas Boukarev cfdbd482aa x86-64/truncate-c: correctly determine if the quotient is unused 
Fixes lp#2148782
 2026-04-19 20:06:46 +03:00
Douglas Katzman 566883ab44 Defensively code in type-width-in-bits 
The receiving logic already potentially expected NIL
 2026-04-19 11:59:56 -04:00
Stas Boukarev c9317d951a check-type: don't include ctypes in the expansion 
Which have a hard time when included in inline functions.

Fixes lp#2148777
 2026-04-19 17:26:24 +03:00
Douglas Katzman ae89d7be84 x86-64: TRUNCATE using the Lemire, Kaser, Kurz transform 
2 MULs outperforms one DIV or IDIV by at least 2x based on my
testing and also Agner Fog's instruction latency tables
 2026-04-18 22:17:50 -04:00
Douglas Katzman c41011cb52 Don't forget NIL 2026-04-18 22:00:21 -04:00