(nthcdr most-positive-fixnum (list 1 2 3)) very slow

[Karsten Poeck]

(nthcdr most-positive-fixnum (list 1 2 3)) seems to run for ages
(nthcdr (1+ most-positive-fixnum) (list 1 2 3)) works fine

(nth most-positive-fixnum (list 1 2 3)) same problem
(nth (1+ most-positive-fixnum) (list 1 2 3)) also fine

(time (nthcdr 10000000000 (list 1 2 3))) -> 13 seconds on my machine, so the calculation probably returns eventually

The follwing fix for src/code/list.lisp seems to solve this, running the ansi-tests before and after gives the same result.

diff src/code/list.lisp src/code/list-patched.lisp
204a205,206
> (when (null result)
> (return nil))
250c252
< ((not (plusp i)) result))))
---
> ((or (not (plusp i)) (null result)) result))))

* (lisp-implementation-version)
"1.5.8"
* (machine-type)
"X86-64"
* (machine-version)
"Intel(R) Core(TM) i7-3615QM CPU @ 2.30GHz"
* (print *features*)

(:QUICKLISP :SB-BSD-SOCKETS-ADDRINFO :ASDF3.3 :ASDF3.2 :ASDF3.1 :ASDF3 :ASDF2
 :ASDF :OS-MACOSX :OS-UNIX :NON-BASE-CHARS-EXIST-P :ASDF-UNICODE :X86-64
 :64-BIT :ALIEN-CALLBACKS :ANSI-CL :AVX2 :BSD :C-STACK-IS-CONTROL-STACK
 :CALL-SYMBOL :COMMON-LISP :COMPACT-INSTANCE-HEADER :COMPARE-AND-SWAP-VOPS
 :CYCLE-COUNTER :DARWIN :FP-AND-PC-STANDARD-SAVE :GENCGC :IEEE-FLOATING-POINT
 :IMMOBILE-CODE :IMMOBILE-SPACE :INODE64 :INTEGER-EQL-VOP :LINKAGE-TABLE
 :LITTLE-ENDIAN :MACH-EXCEPTION-HANDLER :MACH-O :OS-PROVIDES-BLKSIZE-T
 :OS-PROVIDES-DLADDR :OS-PROVIDES-DLOPEN :OS-PROVIDES-PUTWC
 :OS-PROVIDES-SUSECONDS-T :PACKAGE-LOCAL-NICKNAMES :RELOCATABLE-HEAP :SB-DOC
 :SB-EVAL :SB-LDB :SB-PACKAGE-LOCKS :SB-SIMD-PACK :SB-SIMD-PACK-256
 :SB-SOURCE-LOCATIONS :SB-THREAD :SB-UNICODE :SBCL :STACK-ALLOCATABLE-CLOSURES
 :STACK-ALLOCATABLE-FIXED-OBJECTS :STACK-ALLOCATABLE-LISTS
 :STACK-ALLOCATABLE-VECTORS :STACK-GROWS-DOWNWARD-NOT-UPWARD
 :UNDEFINED-FUN-RESTARTS :UNIX :UNWIND-TO-FRAME-AND-CALL-VOP)

uname -a
Darwin karsten-poecks-macbook-pro.local 18.7.0 Darwin Kernel Version 18.7.0: Sat Oct 12 00:02:19 PDT 2019; root:xnu-4903.278.12~1/RELEASE_X86_64 x86_64

[Stas Boukarev (stassats)]

But that slows down the normal use case.

[Karsten Poeck (kpoeck)]

What about
(if (> n 100000) ;or any other threshold
    (code-path-testing-end-of-list-too)
    (old-code-path))

That would be 1 test more, but not in the loop

[Paul F. Dietz (paul-f-dietz)]

Similar to 1738071