clisp not reproducible

OpenSubmitted by Ricardo Wurmus.
Details
One participant
  • Ricardo Wurmus
Owner
unassigned
Severity
normal
R
R
Ricardo Wurmus wrote on 31 Dec 2016 23:48
(name . 'bug-guix@gnu.org')(address . bug-guix@gnu.org)
87lguvzpdr.fsf@elephly.net
The build of CLISP is not reproducible. There are three files that
differ:

diff -r /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49 /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49-check/
Binary files /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49/lib/clisp-2.49/base/lisp.a and /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49-check/lib/clisp-2.49/base/lisp.a differ
Binary files /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49/lib/clisp-2.49/base/lispinit.mem and /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49-check/lib/clisp-2.49/base/lispinit.mem differ
Binary files /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49/lib/clisp-2.49/base/lisp.run and /gnu/store/lf2ybanhxpvbgbxcsxmc733v010g15fp-clisp-2.49-check/lib/clisp-2.49/base/lisp.run differ

This also affects packages that use CLISP to build executables.

--
Ricardo

GPG: BCA6 89B6 3655 3801 C3C6 2150 197A 5888 235F ACAC
R
R
Ricardo Wurmus wrote on 1 Jan 2017 21:58
timestamp
(address . 25314@debbugs.gnu.org)
874m1izedg.fsf@elephly.net
The file ./src/constobj.d contains this:

Toggle snippet (7 lines)
#ifdef __DATE__
LISPOBJ_S(lisp_implementation_version_built_string,__DATE__ __TIME__)
#else
LISPOBJ(lisp_implementation_version_built_string,"NIL")
#endif

diffoscope indicates that this may be the cause of the irreproducible
build of lib/clisp-2.49/base/lisp.run.

│ @@ -25813,16 +25813,16 @@
│ 0x005a6de0 5745414b 2d4f522d 4d415050 494e4700 WEAK-OR-MAPPING.
│ 0x005a6df0 5359533a 3a494e54 45524e41 4c2d5745 SYS::INTERNAL-WE
│ 0x005a6e00 414b2d41 4c495354 00455854 3a3a5359 AK-ALIST.EXT::SY
│ 0x005a6e10 5354454d 2d46554e 4354494f 4e004249 STEM-FUNCTION.BI
│ 0x005a6e20 474e554d 00282243 4f4d4d4f 4e2d4c49 GNUM.("COMMON-LI
│ 0x005a6e30 53502229 00224722 0022434c 49535022 SP")."G"."CLISP"
│ 0x005a6e40 0022322e 34392028 32303130 2d30372d ."2.49 (2010-07-
│ - 0x005a6e50 30372922 00224465 63203331 20323031 07)"."Dec 31 201
│ - 0x005a6e60 3632323a 33393a34 31220028 32303038 622:39:41".(2008
│ + 0x005a6e50 30372922 00224e6f 76203135 20323031 07)"."Nov 15 201
│ + 0x005a6e60 3632313a 30363a33 39220028 32303038 621:06:39".(2008
│ 0x005a6e70 30343330 29002234 2e392e34 2200222a 0430)."4.9.4"."*
│ 0x005a6e80 2a2a202d 20220028 494e5445 47455220 ** - ".(INTEGER
│ 0x005a6e90 30203235 35290028 494e5445 47455220 0 255).(INTEGER
│ 0x005a6ea0 2d313238 20313237 29002849 4e544547 -128 127).(INTEG
│ 0x005a6eb0 45522030 20363535 33352900 28494e54 ER 0 65535).(INT
│ 0x005a6ec0 45474552 202d3332 37363820 33323736 EGER -32768 3276
│ 0x005a6ed0 37290028 494e5445 47455220 30203432 7).(INTEGER 0 42

This is now fixed in c26ba570cce88a9c2529df722042648e34181e42, but clisp
is still not reproducible because the initial memory dump (lispinit.mem)
differs. Diffoscope does not show anything obvious.

--
Ricardo

GPG: BCA6 89B6 3655 3801 C3C6 2150 197A 5888 235F ACAC
?