annotated tag sbcl-2.6.6: created. sbcl-2.6.6

The annotated tag "sbcl-2.6.6" has been created in SBCL:
        at  94b39466945283902c512f5f53d99973f8e95606 (tag)
   tagging  8fbf3824e4614e2dc38e432ffb495cf7e2844092 (commit)
  replaces  sbcl-2.6.5
 tagged by  Christophe Rhodes
        on  Sun Jun 28 09:33:48 2026 +0100

- Log -----------------------------------------------------------------
changes in sbcl-2.6.6 relative to sbcl-2.6.5:
  * minor incompatible change: FDEFINITION now returns the outermost wrapper
    (added e.g. by TRACE, PROFILE) like SYMBOL-FUNCTION.  (lp#799533)
  * minor incompatible change: in unsafe code, C strings with :EXTERNAL-FORMAT
    :ASCII are copied directly as byte-sized quantities without checking
    whether the top bit of the byte is set.
  * platform support:
    ** fix the build on big-endian 64-bit PowerPC with ELFv2.  (thanks to
       Piotr Kubaj)
    ** move the static space address for macOS 27 on ARM64.  (lp#2156072,
       reported by Gary Palter)
    ** optimizations to SB-THREAD:BARRIER for ARM64.  (thanks to Sahil Kang)
    ** fix a compiler crash in MULTIPLE-VALUE-LIST in argument forms on ARM64.
       (lp#2155788, reported by Gary Palter)
  * bug fix: TRACE no longer fails when trying to print a return value that
    cannot be printed readably and *PRINT-READABLY* is true.
  * optimization: the compiler is more precise in its type derivation of
    COERCE given constraints on its inputs.
  * optimization: the compiler is better able to derive the return types of
    AREF and ELT.
  * optimization: faster encoding and decoding of UTF-8 C strings.
  * optimization: (length (intersection a b)) doesn't cons an intermediate
    list.
  * documentation: the manual now includes a section for SB-INTROSPECT, which
    has also seen improvement in its documentation strings and comments.
  * documentation: fixed many typesetting problems and typos in the user
    manual.
-----BEGIN PGP SIGNATURE-----

iQFFBAABCgAvFiEE1oOcoKZ/dNnftwki69WVqRANY80FAmpA3GwRHGNzcjIxQGNh
bnRhYi5uZXQACgkQ69WVqRANY825VQgAmZf+DfiZ/eq/t2egCPLLatDEH76ZuLaz
juSJf4kHiAaOp780VeyTeFzhYAO5yyMbZdt7xnE0vFGuO7MSk3B2t9UqpoKhmn3L
Wc+MFPrYPCmqUD3QVVnOWexVu25+7B1N6nVYknuQiwayHyrUs/xR9318wLUM7A7s
WXouljszIcKFCXB/x6Jgd+7emm+/TptzQhP+VyfhaFKoNBsxnmjp+F5GPPS7NbrI
JOclxiannEaFRBb3e/mLxrsTme3lBHJjvIoR4ElUpO9erXYDfs0GwYTP6SPXKNce
dus9IkbQxgrO02ThbZTHVb/eBfXlH5WdP0S9kLmPPjBc/QG5WyHplQ==
=mpH3
-----END PGP SIGNATURE-----

Christophe Rhodes (3):
      Draft NEWS for sbcl-2.6.6
      Try to defend against releasing binaries with a hash suffix version
      2.6.6: will be tagged as "sbcl-2.6.6"

Douglas Katzman (8):
      Sketch out a way to trace/intercept foreign calls
      Add a missing opaque-identity
      Mix package-id into fun-name hashes
      Improve receiving some C strings from foreign function on 64-bit Unicode
      Return constant string from NATURALIZE if length is 0
      Speed up DO-CONSET-ELEMENTS using CTZ if available
      Fix CONSET-ADJOIN and CONSET-UNION
      Stop trying to make deterministic standard-class hashes

Gabor Melis (9):
      Make FDEFINITION not strip encapsulation
      pcl: Remove GDEFINITION
      doc: document sb-introspect
      Improve reporting of NOTINLINE full calls in make-host-2.lisp
      Housekeeping: whitespace, typos, comments, credits
      Bind *PRINT-READABLY* to NIL when printing TRACE "returned" message
      doc: tweaks
      doc: improve @code{}ification in the user manual
      doc: define and clean up docstring syntax

Piotr Kubaj (1):
      ppc64: fix big-endian ELFv2 build

Sahil Kang (1):
      Optimize sb-thread:barrier for arm64

Stas Boukarev (55):
      Unbreak generate-version.sh
      Remove unused instructions in arm64/simd-copy-utf8-crlf-to-base-string
      Lower :test for search and mismatch
      Lower :test for remove-duplicates
      (coerce (the (not real) x) y) => (not real)
      More coerce type exclusion derivation
      coerce type exclusion: add (and array (not simple-array))
      coerce type derivation: preserve simple array dimensions
      (coerce 1 'list) should signal an error
      coerce derive-type: exclude rational types
      More thorough type checking for concatenate arguments
      check-concatenate: consider each constant value
      Derive AREF type for multiple uses with constants
      Unify the sequence element-type derivers
      Add a type deriver for ELT
      constant-sequence-element-type: ignore circular lists
      More constant modification pass-throughs
      lvar-constants: combine functions and refs from multiple uses
      Turn (copy-list small-constant) into (list ,@constant)
      arm64: can't call sc-is on unused tns
      Check types for concatenate-subseq
      Change static space address for darwin-arm64
      Skip a test
      Don't cons a new list for (length (intersection a b))
      Don't change :test for subst based on the sequence type
      sb-introspect:function-lambda-list: employ sb-ext:unencapsulated-function
      describe: call unencapsulated-function
      arm64: add neon shl
      arm64: unify some neon and ordinary instructions
      Fix do-conset-elements
      arm64: remove the remaining s- prefixes
      arm64: always require vector sizes for simd instructions
      arm64: better umaxv disassembly
      arm64: add uqsub
      arm64: implement FMOV <Xd>, <Vn>.D[1]
      Derive the type of (coerce n `(simple-array ,e (*)))
      arm64: restore bic, orn disassembly
      Optimize read-from-c-string/utf-8/lf
      Don't cons SAPs for sb-alien::c-string-to-string for :utf-8
      Fix interpreted calls to c-string-to-string
      Speed up output-to-c-string/utf-8/lf
      arm64: implement simd-character-string-utf8-length using NEON
      amr64: encode movi immediates without requiring manual shifting
      arm64: optimize simd-character-string-utf8-length
      arm64: optimize simd-character-string-utf8-length some more
      arm64: disassemble shifted immediates directly
      Add tests for simd-utf8-strlen
      Adjust a test
      Add a test for simd-character-string-utf8-length
      Add a test for output-to-c-string/utf-8/lf
      arm64: fix simd-copy-utf8-crlf-to-character-string
      arm64, simd-copy-character-string-to-utf8: use correct lane width
      arm64-simd: off-by-one in check-ascii
      arm64/simd-utf8-strlen: use a different register
      NEWS

-----------------------------------------------------------------------


hooks/post-receive
-- 
SBCL