Menu
▾
▴
sbcl-bugs — New bug reports. Moderated, subscription not required for posting.
You can subscribe to this list here.
| 2009 |
Jan
|
Feb
(3) |
Mar
(1) |
Apr
|
May
|
Jun
(5) |
Jul
(3) |
Aug
(5) |
Sep
(2) |
Oct
(6) |
Nov
(2) |
Dec
(1) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2010 |
Jan
(16) |
Feb
(1) |
Mar
(10) |
Apr
(4) |
May
(3) |
Jun
|
Jul
(1) |
Aug
(1) |
Sep
(1) |
Oct
(3) |
Nov
|
Dec
(1) |
| 2011 |
Jan
(3) |
Feb
(6) |
Mar
|
Apr
(2) |
May
(4) |
Jun
(5) |
Jul
(1) |
Aug
(8) |
Sep
(1) |
Oct
(1) |
Nov
(2) |
Dec
(1) |
| 2012 |
Jan
(2) |
Feb
(2) |
Mar
(8) |
Apr
(6) |
May
(13) |
Jun
(6) |
Jul
(3) |
Aug
(2) |
Sep
(7) |
Oct
(1) |
Nov
|
Dec
|
| 2013 |
Jan
(5) |
Feb
(1) |
Mar
(6) |
Apr
(1) |
May
(1) |
Jun
(11) |
Jul
(9) |
Aug
|
Sep
(4) |
Oct
|
Nov
(12) |
Dec
(6) |
| 2014 |
Jan
(6) |
Feb
(17) |
Mar
(3) |
Apr
(3) |
May
|
Jun
(4) |
Jul
|
Aug
(7) |
Sep
(2) |
Oct
(4) |
Nov
(1) |
Dec
(1) |
| 2015 |
Jan
(2) |
Feb
(1) |
Mar
(6) |
Apr
(2) |
May
|
Jun
(5) |
Jul
(7) |
Aug
(2) |
Sep
(5) |
Oct
(3) |
Nov
(16) |
Dec
(15) |
| 2016 |
Jan
(11) |
Feb
(11) |
Mar
(5) |
Apr
|
May
(3) |
Jun
|
Jul
|
Aug
|
Sep
(2) |
Oct
|
Nov
|
Dec
(1) |
| 2017 |
Jan
(2) |
Feb
|
Mar
(3) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2018 |
Jan
|
Feb
|
Mar
(2) |
Apr
(12) |
May
|
Jun
(1) |
Jul
|
Aug
(2) |
Sep
|
Oct
(4) |
Nov
|
Dec
|
| 2019 |
Jan
|
Feb
(6) |
Mar
|
Apr
(4) |
May
|
Jun
(2) |
Jul
(9) |
Aug
|
Sep
(12) |
Oct
(2) |
Nov
|
Dec
(8) |
| 2020 |
Jan
(12) |
Feb
(3) |
Mar
(4) |
Apr
(4) |
May
(27) |
Jun
(14) |
Jul
(3) |
Aug
(7) |
Sep
(1) |
Oct
(1) |
Nov
(2) |
Dec
(3) |
| 2021 |
Jan
(2) |
Feb
(6) |
Mar
(8) |
Apr
(10) |
May
(1) |
Jun
(8) |
Jul
(4) |
Aug
(9) |
Sep
(1) |
Oct
(7) |
Nov
(6) |
Dec
(8) |
| 2022 |
Jan
(7) |
Feb
(4) |
Mar
(3) |
Apr
(2) |
May
(2) |
Jun
(3) |
Jul
(14) |
Aug
(15) |
Sep
(13) |
Oct
(16) |
Nov
(5) |
Dec
(6) |
| 2023 |
Jan
(18) |
Feb
(2) |
Mar
(28) |
Apr
(8) |
May
(3) |
Jun
(24) |
Jul
(11) |
Aug
(22) |
Sep
(20) |
Oct
(27) |
Nov
(12) |
Dec
(2) |
| 2024 |
Jan
(5) |
Feb
(45) |
Mar
(45) |
Apr
(25) |
May
(10) |
Jun
(23) |
Jul
(20) |
Aug
(5) |
Sep
(2) |
Oct
|
Nov
|
Dec
|
|
From: Sean W. <spw...@sp...> - 2024-05-31 15:21:51
|
Hello, On Fri 31 May 2024 at 01:53pm +01, Christophe Rhodes wrote: > Is this the same root cause as > <https://sourceforge.net/p/sbcl/mailman/message/58776268/> (it certainly > looks similar)? You could try applying the patch from > dc890089a5548bc3dcfe11c358bc5a88957ba53b to see if it resolves the > problem. > > If it doesn't, please provide more details on the configuration that the > debian package is building with (what features are being enabled, etc.) That fixes it, thank you. -- Sean Whitton |
|
From: Christophe R. <cs...@ca...> - 2024-05-31 12:54:01
|
Is this the same root cause as <https://sourceforge.net/p/sbcl/mailman/message/58776268/> (it certainly looks similar)? You could try applying the patch from dc890089a5548bc3dcfe11c358bc5a88957ba53b to see if it resolves the problem. If it doesn't, please provide more details on the configuration that the debian package is building with (what features are being enabled, etc.) Christophe Sean Whitton via Sbcl-bugs <sbc...@li...> writes: > Hello, > > I am working to update SBCL in Debian to the latest release. > Unfortunately, hide-packages.test.sh is failing. > I would be grateful for any input on this backtrace: > > // Running hide-packages.test.sh in COMPILE evaluator mode > Path to "SB-FASL": > from "main thread" PC=547A6C9B in #<code id=3FCB [5] SIMPLE-EVAL-IN-LEXENV {547A650F}> > 6 1000036663 [ 2] a HIDDEN-SB-KERNEL::CLASSOID-CELL > 6 100010D093 [ 12] a HIDDEN-SB-KERNEL::CONDITION-CLASSOID > 6 54CD835F [ 3] #<code id=1BE7 [1] (HIDDEN-SB-KERNEL::CONDITION-REPORT HIDDEN-SB-C:COMPILER-ERROR) {54CD835F}> > 6 1000CDEA83 [ 3] a HIDDEN-SB-C::COMPILED-DEBUG-INFO > 6 10000E8083 [ 7] #<PACKAGE "HIDDEN-SB-C"> > 6 100020FEF3 [ 1] a HIDDEN-SB-IMPL::SYMBOL-TABLE > 6 1000350EC7 [ 1] a cons = (# . #) > 6 10003F800F [ 733] a (COMMON-LISP:SIMPLE-VECTOR 6367) > 6 509C936F [ 4] HIDDEN-SB-C::DUMP-FOP > 6 10006A4983 [ 2] a HIDDEN-SB-INT:PACKED-INFO > 6 54D41CEF [ 12] #<code id=2590 [1] (MACRO-FUNCTION HIDDEN-SB-C::DUMP-FOP) {54D41CEF}> > Unhandled SIMPLE-ERROR in thread #<HIDDEN-SB-THREAD:THREAD tid=618776 "main thread" RUNNING > {1000DC8003}>: > The assertion (NULL *WEAK-PTRS*) failed with *WEAK-PTRS* = > (#<weak pointer: "SB-FASL">). > > Backtrace for: #<HIDDEN-SB-THREAD:THREAD tid=618776 "main thread" RUNNING {1000DC8003}> > 0: (HIDDEN-SB-DEBUG::DEBUGGER-DISABLED-HOOK #<SIMPLE-ERROR "~@<The assertion ~S failed~:[.~:; ~ > with ~:*~{~S = ~S~^, ~}.~]~:@>" {100173A533}> #<unused argument> :QUIT T) > 1: (HIDDEN-SB-DEBUG::RUN-HOOK *INVOKE-DEBUGGER-HOOK* #<SIMPLE-ERROR "~@<The assertion ~S failed~:[.~:; ~ > with ~:*~{~S = ~S~^, ~}.~]~:@>" {100173A533}>) > 2: (INVOKE-DEBUGGER #<SIMPLE-ERROR "~@<The assertion ~S failed~:[.~:; ~ > with ~:*~{~S = ~S~^, ~}.~]~:@>" {100173A533}>) > unhandled condition in --disable-debugger mode, quitting |
|
From: Claude RC <cla...@gm...> - 2024-05-31 02:08:48
|
Thanks for the info. You built from source with a previous version? The most recent official built is 1.5.4, right? My Travis ppc64le workflow keeps pending and time out, so I was trying to cross-compile. It fails at target-2 so could very well be qemu The aarch64 build worked like a charm I'll dig a bit more, the fail occurs in package-cold-info apparently, when I try to print spec, I get a mprotect() missed error Thanks On Thu, May 30, 2024, 8:51 PM Douglas Katzman via Sbcl-bugs < sbc...@li...> wrote: > release 2.4.5 on linux ppc64le works fine for me. qemu in user mode > emulation probably has trouble with SBCL. > > This is SBCL 2.4.5.13-649b683a8, an implementation of ANSI Common Lisp. > More information about SBCL is available at <http://www.sbcl.org/>. > > SBCL is free software, provided as is, with absolutely no warranty. > It is mostly in the public domain; some portions are provided under > BSD-style licenses. See the CREDITS and COPYING files in the > distribution for more information. > * *features* > (:PPC64 :GENCGC :64-BIT :ANSI-CL :COMMON-LISP :ELF :IEEE-FLOATING-POINT > :LINUX > :LITTLE-ENDIAN :PACKAGE-LOCAL-NICKNAMES :SB-LDB :SB-PACKAGE-LOCKS > :SB-THREAD > :SB-UNICODE :SBCL :UNIX) > _______________________________________________ > Sbcl-bugs mailing list > Sbc...@li... > https://lists.sourceforge.net/lists/listinfo/sbcl-bugs > |
|
From: Douglas K. <do...@go...> - 2024-05-31 01:50:41
|
release 2.4.5 on linux ppc64le works fine for me. qemu in user mode emulation probably has trouble with SBCL. This is SBCL 2.4.5.13-649b683a8, an implementation of ANSI Common Lisp. More information about SBCL is available at <http://www.sbcl.org/>. SBCL is free software, provided as is, with absolutely no warranty. It is mostly in the public domain; some portions are provided under BSD-style licenses. See the CREDITS and COPYING files in the distribution for more information. * *features* (:PPC64 :GENCGC :64-BIT :ANSI-CL :COMMON-LISP :ELF :IEEE-FLOATING-POINT :LINUX :LITTLE-ENDIAN :PACKAGE-LOCAL-NICKNAMES :SB-LDB :SB-PACKAGE-LOCKS :SB-THREAD :SB-UNICODE :SBCL :UNIX) |
|
From: Claude RC <cla...@gm...> - 2024-05-30 23:10:46
|
Hello, I am having difficulties cross-compiling with x86_64-linux v2.2.11 for ppc64-linux arch. I use qemu-ppc64le-static for target execution. This is to complement the current packaging of sbcl on conda-forge I get the following error report for v2.4.5. It seems to be happening when the package-cold-info is called, when I put debugging prints in that function I get nothing but a mprotect() error //entering make-target-2.sh //doing warm init - compilation phase This is SBCL 2.4.5, an implementation of ANSI Common Lisp. More information about SBCL is available at <http://www.sbcl.org/>. SBCL is free software, provided as is, with absolutely no warranty. It is mostly in the public domain; some portions are provided under BSD-style licenses. See the CREDITS and COPYING files in the distribution for more information. Initial page table: Immobile Object Counts Gen layout fdefn symbol code Boxed Cons Raw Code SmMix Mixed LgRaw LgCode LgMix Waste% Alloc Trig Dirty GCs Mem-age 6 0 0 0 0 0 45 0 264 0 166 0 0 0 0.2 31054464 2000000 0 0 0.0000 Tot 0 0 0 0 0 45 0 264 0 166 0 0 0 0.2 31054464 [0.7% of 4294967296 max] COLD-INIT... |
|
From: Sean W. <spw...@sp...> - 2024-05-30 21:48:42
|
Hello,
I am working to update SBCL in Debian to the latest release.
Unfortunately, hide-packages.test.sh is failing.
I would be grateful for any input on this backtrace:
// Running hide-packages.test.sh in COMPILE evaluator mode
Path to "SB-FASL":
from "main thread" PC=547A6C9B in #<code id=3FCB [5] SIMPLE-EVAL-IN-LEXENV {547A650F}>
6 1000036663 [ 2] a HIDDEN-SB-KERNEL::CLASSOID-CELL
6 100010D093 [ 12] a HIDDEN-SB-KERNEL::CONDITION-CLASSOID
6 54CD835F [ 3] #<code id=1BE7 [1] (HIDDEN-SB-KERNEL::CONDITION-REPORT HIDDEN-SB-C:COMPILER-ERROR) {54CD835F}>
6 1000CDEA83 [ 3] a HIDDEN-SB-C::COMPILED-DEBUG-INFO
6 10000E8083 [ 7] #<PACKAGE "HIDDEN-SB-C">
6 100020FEF3 [ 1] a HIDDEN-SB-IMPL::SYMBOL-TABLE
6 1000350EC7 [ 1] a cons = (# . #)
6 10003F800F [ 733] a (COMMON-LISP:SIMPLE-VECTOR 6367)
6 509C936F [ 4] HIDDEN-SB-C::DUMP-FOP
6 10006A4983 [ 2] a HIDDEN-SB-INT:PACKED-INFO
6 54D41CEF [ 12] #<code id=2590 [1] (MACRO-FUNCTION HIDDEN-SB-C::DUMP-FOP) {54D41CEF}>
Unhandled SIMPLE-ERROR in thread #<HIDDEN-SB-THREAD:THREAD tid=618776 "main thread" RUNNING
{1000DC8003}>:
The assertion (NULL *WEAK-PTRS*) failed with *WEAK-PTRS* =
(#<weak pointer: "SB-FASL">).
Backtrace for: #<HIDDEN-SB-THREAD:THREAD tid=618776 "main thread" RUNNING {1000DC8003}>
0: (HIDDEN-SB-DEBUG::DEBUGGER-DISABLED-HOOK #<SIMPLE-ERROR "~@<The assertion ~S failed~:[.~:; ~
with ~:*~{~S = ~S~^, ~}.~]~:@>" {100173A533}> #<unused argument> :QUIT T)
1: (HIDDEN-SB-DEBUG::RUN-HOOK *INVOKE-DEBUGGER-HOOK* #<SIMPLE-ERROR "~@<The assertion ~S failed~:[.~:; ~
with ~:*~{~S = ~S~^, ~}.~]~:@>" {100173A533}>)
2: (INVOKE-DEBUGGER #<SIMPLE-ERROR "~@<The assertion ~S failed~:[.~:; ~
with ~:*~{~S = ~S~^, ~}.~]~:@>" {100173A533}>)
unhandled condition in --disable-debugger mode, quitting
--
Sean Whitton
|
|
From: Hraban <hr...@0b...> - 2024-05-20 05:57:14
|
When building SBCL from master on a x86_64-linux VM I get the following
error during testing:
/ Running /build/source/tests/debug.pure.lisp in COMPILE evaluator mode
::: Running :FUN-CODE-HEADER-BOGUS
::: Success :FUN-CODE-HEADER-BOGUS
::: Running :C-UNPACK-COMPILED-DEBUG-INFO-FUN-MAP
*** stack smashing detected ***: terminated
fatal error encountered in SBCL pid 1650 tid 1650:
SIGABRT received.
0: fp=0x7ffff7c0f740 pc=0x7ffff7ca3efc Foreign function (null)
test failed, expected 104 return code, got 1
This is for SBCL from https://git.code.sf.net/p/sbcl/sbcl revision
4da0692ae1a326527878425d2e9442ad80bfa880
The same happens on a local x86_64 VM. Not on 2.4.4.
I have used git bisect to pinpoint the bug to
7a813594ce1edaee0eb8330814430c5b3d2afb11 which seems a likely culprit.
Full logs:
https://github.com/hraban/cl-nix-lite/actions/runs/9153245536/job/25161882302
Build config:
linkableRuntime = false;
coreCompression = true;
markRegionGC = false;
threadSupport = false;
(forgive the custom variable names hopefully you get the idea)
|
|
From: Stas B. <sta...@gm...> - 2024-05-06 11:35:06
|
Fixed. Thanks. On Sun, May 5, 2024 at 10:46 AM lisexp--- via Sbcl-bugs < sbc...@li...> wrote: > Hello SBCL developers, > > I have encountered an issue where SBCL fails to compile the function > ALIST-VALUE in the following scenario: > > > (ql:quickload :alexandria) > > (defstruct key) > (defparameter *key* (make-key)) > (alexandria:define-constant +alist+ (list (cons *key* 0)) :test #'equal) > (defun alist-value (key) > (assoc key +alist+)) > > > The compiler emits the following error: > > error: don't know how to dump #S(KEY) (default MAKE-LOAD-FORM method > called). > > However, evaluating the DEFUN form to define the function works fine and > yields the correct result. > > Thank you for taking the time to look into this issue. > > Best regards, > > bohonghuang > > > > _______________________________________________ > Sbcl-bugs mailing list > Sbc...@li... > https://lists.sourceforge.net/lists/listinfo/sbcl-bugs > |
|
From: <li...@qq...> - 2024-05-05 07:46:34
|
Hello SBCL developers, I have encountered an issue where SBCL fails to compile the function ALIST-VALUE in the following scenario: (ql:quickload :alexandria) (defstruct key) (defparameter *key* (make-key)) (alexandria:define-constant +alist+ (list (cons *key* 0)) :test #'equal) (defun alist-value (key) (assoc key +alist+)) The compiler emits the following error: error: don't know how to dump #S(KEY) (default MAKE-LOAD-FORM method called). However, evaluating the DEFUN form to define the function works fine and yields the correct result. Thank you for taking the time to look into this issue. Best regards, bohonghuang |
|
From: Robert B. <rob...@gm...> - 2024-05-03 15:57:13
|
The stupid declaim below took me about a day to put together trying to
start resurrecting
some 40 year old Lisp code that worked great in KCL and Zetalisp.
My 'problem' is that, for good or not, SBCL's maximally wonderful loader
complains, in my
view, somewhat aggressively and prematurely.
If you wanted to make it more kind and polite, I feel sure that you could
get
the loader not to complain until 'later', if necessary, about a function
being possibly undefined. It seems,
little I know, that unless I define or proclaim the 'arity' of a function
before its first use, I will get a complaint
which is certainly justified, but might just possibly be postponed in the
hope that the loader
can see that finally all is well enough. Perhaps any such fix would slow
down the loader, which
is so splendidly fast.
Anyway, the SBCL loader now finds more bugs in my code than I do. Real AI,
if you don't mind my saying so.
Thanks so very, very much. I wonder if it is possible to nominate SBCL for
the Turing Award?
Bob
(declaim (ftype (function () t) next-token get-d
get-definition get-e get-d-top get-exp get-opexp
get-e0 get-e1 get-e2 get-e3 get-e4 get-e5 get-e6 get-e7
get-e8
get-e9 get-combination get-simple
get-formal
get-listexp get-formal-lst get-structure
get-fstructure get-structure get-element)
(ftype (function (t) t)
*lconsp *ltl print-obj *lhd *ltl our-ascii
declare-defs gcf-top parse-file letter-char
code-char1 elim-zf elim-namelists
get-zf gcf gcf-defs
all-identifiers
make--list
new-name
gcf-def group-defs gcf-def-function strip
and-join simple-arglist
lower-case-hd-tl
names-defined
sasl-conjoin)
(ftype (function (t t) t)
*lith gcf-def-body occurs-free add-args-defs close-calls
compute-tests)
(ftype (function (t t t) t)
cram sasl-subst-var-var make-zf-namelist-defs)
(ftype (function (t t t t) t) make-zf-defs)
(ftype (function (&optional t t) t) tty-readline tty-tyi)
(ftype (function (&optional t t t t) t) tty-peek tty-tyipeek)
(ftype (function (t &optional t) t) tty-untyi)
(ftype (function (&rest t) t)
lazy-tyi lazy-tyipeek lazy-untyi lazy-readline)
)
Here is my working directory:
https://drive.google.com/drive/folders/1UoCT8YaisJBE-deMYJ8n1VE1rNg6K_MW?usp=drive_link
The file 0README.txt has more about me than you probably care to know.
|
|
From: Stas B. <sta...@gm...> - 2024-04-30 12:00:38
|
Fixed. Thanks. On Tue, Apr 30, 2024 at 2:57 PM <er...@ig...> wrote: > Hi, > > Running > > (cd tests/ && sh ./run-test.sh) > > results in > > ======================================================================== > ... > // Running /home/tom/software/lisp/sbcl-2.4.4/tests/arith.pure.lisp in > COMPILE evaluator mode > ... > ::: Running :LOGNOT-TYPE-DERIVE > ::: UNEXPECTED-FAILURE :LOGNOT-TYPE-DERIVE due to SIMPLE-ERROR: > "The assertion > (EQUAL > (CADDR > (SB-KERNEL:%SIMPLE-FUN-TYPE > (CHECKED-COMPILE > `(LAMBDA (B) > (LOGNOT > (IF B > -1 > 2)))))) > '(VALUES (OR (INTEGER -3 -3) (INTEGER 0 0)) &OPTIONAL)) > failed with > (CADDR > (SB-KERNEL:%SIMPLE-FUN-TYPE > (CHECKED-COMPILE > `(LAMBDA (B) > (LOGNOT > (IF B > -1 > 2)))))) > = (VALUES (OR (INTEGER 0 0) (INTEGER -3 -3)) &OPTIONAL)." > ... > ======================================================================== > > Note that the type alternatives came in reverse of the expected order, > and that might actually be quite OK. > > There seems to be some nondeterminism either in > test-util::checked-compile or sb-kernel:%simple-fun-type > or in the sbcl compiler itself. I'm guessing that the problem > is in the test, not in the behavior of sbcl. > > When I run just the individual test > > (cd ./tests/ && sh ./run-tests.sh arith.pure.lisp) > > it passes every time, many times. But it has failed 3 out of 3 times > running all tests. > > > This is happening on fedora linux > > 6.8.7-300.fc40.x86_64 #1 SMP PREEMPT_DYNAMIC Wed Apr 17 19:21:08 UTC > 2024 x86_64 GNU/Linux > > running on an AMD Ryzen 5 2600. > Sbcl-2.4.4 built from source tarball on the same machine as the tests > are run. > > Thanks for all your work, > Tom > > > _______________________________________________ > Sbcl-bugs mailing list > Sbc...@li... > https://lists.sourceforge.net/lists/listinfo/sbcl-bugs > |
|
From: <er...@ig...> - 2024-04-30 01:34:21
|
Hi,
Running
(cd tests/ && sh ./run-test.sh)
results in
========================================================================
...
// Running /home/tom/software/lisp/sbcl-2.4.4/tests/arith.pure.lisp in COMPILE evaluator mode
...
::: Running :LOGNOT-TYPE-DERIVE
::: UNEXPECTED-FAILURE :LOGNOT-TYPE-DERIVE due to SIMPLE-ERROR:
"The assertion
(EQUAL
(CADDR
(SB-KERNEL:%SIMPLE-FUN-TYPE
(CHECKED-COMPILE
`(LAMBDA (B)
(LOGNOT
(IF B
-1
2))))))
'(VALUES (OR (INTEGER -3 -3) (INTEGER 0 0)) &OPTIONAL))
failed with
(CADDR
(SB-KERNEL:%SIMPLE-FUN-TYPE
(CHECKED-COMPILE
`(LAMBDA (B)
(LOGNOT
(IF B
-1
2))))))
= (VALUES (OR (INTEGER 0 0) (INTEGER -3 -3)) &OPTIONAL)."
...
========================================================================
Note that the type alternatives came in reverse of the expected order,
and that might actually be quite OK.
There seems to be some nondeterminism either in
test-util::checked-compile or sb-kernel:%simple-fun-type
or in the sbcl compiler itself. I'm guessing that the problem
is in the test, not in the behavior of sbcl.
When I run just the individual test
(cd ./tests/ && sh ./run-tests.sh arith.pure.lisp)
it passes every time, many times. But it has failed 3 out of 3 times
running all tests.
This is happening on fedora linux
6.8.7-300.fc40.x86_64 #1 SMP PREEMPT_DYNAMIC Wed Apr 17 19:21:08 UTC 2024 x86_64 GNU/Linux
running on an AMD Ryzen 5 2600.
Sbcl-2.4.4 built from source tarball on the same machine as the tests
are run.
Thanks for all your work,
Tom
|
|
From: John M. <jo...@ma...> - 2024-04-29 13:01:15
|
Thanks Christophe. In case anybody is wondering what is special about 1.2.7, it's because that release provided a bunch of binaries. see: https://sourceforge.net/projects/sbcl/files/sbcl/1.2.7/ It might be worth generating binary packages for all supported platforms for say 2.44 and give a new starting point for bootstraps. John P.S. set to "bugs" because I'm not registered at "devel" On Mon, Apr 29, 2024 at 4:16 AM Christophe Rhodes <cs...@ca...> wrote: > John Marino <jo...@ma...> writes: > > > Sorry, I thought I grabbed enough. > > The entire log is here: > > https://loki.dragonflybsd.org/raven/logs/sbcl-243-build.log > > Thanks. I have pushed a workaround for the sbcl-1.2.7 misbehaviour that > was triggered by the current sources, so you should be able to build > 2.4.4 with 1.2.7 again. > > (It's not completely clear to me that this is sustainable, but we could > try: since there is demonstrably a community that is using old sbcls in > this way, we could have such an old host as part of our continuous > integration, so that at least we would know if we broke it.) > > Christophe > |
|
From: Christophe R. <cs...@ca...> - 2024-04-28 20:12:57
|
Douglas Katzman via Sbcl-bugs <sbc...@li...> writes: > On Sat, Apr 27, 2024 at 9:40 PM John Marino <jo...@ma...> wrote: > > The reason for this is not known? > > As known as could be under the circumstance: Your bootstrap SBCL was released Dec 31, 2014, and some time last month an SBCL developer > made a change which unwittingly or wittingly caused current sources to become sensitive to a bug fix made any time over the last decade of > development. If you are asking literally which change it is, either target or host code, no developer has time to spend on such a wild goose > chase. You could run a git bisect on both axes. (And then what- figure out the oldest possible released SBCL to bootstrap because it has > exactly the needed patch?) Last month, with-lock and with-recursive-lock were adjusted to have (and use, sometimes) an environment parameter named ENV. Under #-sb-thread, the code that uses ENV is dead in both macros. I would guess that adding (declare (ignorable env)) to the macro definitions would restore buildability. (If that's right, the bug is that the 10-year-old version of SBCL would style-warn for an unused variable even if a function used it, if that use was in dead code.) Christophe |
|
From: John M. <jo...@ma...> - 2024-04-28 20:02:20
|
Thanks. With that patch, sbcl 2.4.3 builds with clisp as the host compiler. It does seem to take twice as long to build as before with clisp. John On Sun, Apr 28, 2024 at 1:01 PM Douglas Katzman <do...@go...> wrote: > I can't commit the same patch to master and the 2.4.3 tag. > The attached diff should cleanly apply against 2.4.3. > > On Sun, Apr 28, 2024 at 12:17 PM John Marino <jo...@ma...> wrote: > >> Understood. I can create a patch for 2.4.3 once I see the commit. >> > |
|
From: Douglas K. <do...@go...> - 2024-04-28 18:08:06
|
I can't commit the same patch to master and the 2.4.3 tag. The attached diff should cleanly apply against 2.4.3. On Sun, Apr 28, 2024 at 12:17 PM John Marino <jo...@ma...> wrote: > Understood. I can create a patch for 2.4.3 once I see the commit. > |
|
From: John M. <jo...@ma...> - 2024-04-28 16:17:41
|
Understood. I can create a patch for 2.4.3 once I see the commit. On Sun, Apr 28, 2024 at 10:52 AM Douglas Katzman <do...@go...> wrote: > The style-warnings from clisp are ignorable, but the "missing perfect > hash" isn't, and can only occur for non-self-native-build. > I thought the fix for that was committed for this month, but it isn't yet, > so I'll do that shortly. It won't help the past release though > > On Sun, Apr 28, 2024 at 11:20 AM John Marino <jo...@ma...> wrote: > >> Unfortunately, 2.4.3 doesn't build with clisp (at least on DragonFly) >> either. >> https://loki.dragonflybsd.org/raven/logs/sbcl-243-build-clisp.log >> > |
|
From: Douglas K. <do...@go...> - 2024-04-28 15:52:45
|
The style-warnings from clisp are ignorable, but the "missing perfect hash" isn't, and can only occur for non-self-native-build. I thought the fix for that was committed for this month, but it isn't yet, so I'll do that shortly. It won't help the past release though On Sun, Apr 28, 2024 at 11:20 AM John Marino <jo...@ma...> wrote: > Unfortunately, 2.4.3 doesn't build with clisp (at least on DragonFly) > either. > https://loki.dragonflybsd.org/raven/logs/sbcl-243-build-clisp.log > |
|
From: John M. <jo...@ma...> - 2024-04-28 15:46:53
|
Unfortunately, 2.4.3 doesn't build with clisp (at least on DragonFly) either. https://loki.dragonflybsd.org/raven/logs/sbcl-243-build-clisp.log On Sun, Apr 28, 2024 at 9:56 AM John Marino <jo...@ma...> wrote: > Well, I assumed the stack trace would prevent that from being a needle in > the haystack search. So I posted it in case there was a bug introduced > between releases. If you are confident the 2.4.3 code is perfect, ok. > > This recipe supports FreeBSD, DragonFly, NetBSD, Linux, MidnightBSD, > Solaris and formerly Darwin. NetBSD and MidnightBSD didn't have suitable > bootstrap compilers so they were built with clisp, so I guess I have to > switch all the operating systems to clisp for the xc-host. > > As for the options that aren't choices - well, like I said, the genesis of > the recipe goes back more than a decade so they were probably choices in > the past and just became obsolete. There's no dedicated maintainer for > this package so there's nobody studying every release's change. > > > On Sun, Apr 28, 2024 at 8:41 AM Douglas Katzman via Sbcl-bugs < > sbc...@li...> wrote: > >> >> >> On Sat, Apr 27, 2024 at 9:40 PM John Marino <jo...@ma...> wrote: >> >>> >>> The reason for this is not known? >>> >> As known as could be under the circumstance: Your bootstrap SBCL was >> released Dec 31, 2014, and some time last month an SBCL developer made a >> change which unwittingly or wittingly caused current sources to become >> sensitive to a bug fix made any time over the last decade of development. >> If you are asking literally which change it is, either target or host code, >> no developer has time to spend on such a wild goose chase. You could run a >> git bisect on both axes. (And then what- figure out the oldest possible >> released SBCL to bootstrap because it has exactly the needed patch?) >> >> And what do you consider "random"? >>> >> You've got options in this list which are effectless: >> --without-sb-thread --without-restore-fs-segment-register-from-tls >> --with-sb-unicode --with-sb-core-compression --without-sb-qshow >> --without-sb-xref-for-internals --without-sb-safepoint >> --without-sb-thruption --without-sb-wtimer --without-sb-simd >> --without-sb-thread --without-restore-fs-segment-register-from-tls >> >> restore-fs-segment isn't a choice, qshow isn't a choice, wtimer isn't a >> choice, thruption isn't a choice, safepoint isn't the default, unicode is >> the default, and sb-thread is already disabled by default on DragonFly. You >> should remove anything that isn't present for a good reason. >> _______________________________________________ >> Sbcl-bugs mailing list >> Sbc...@li... >> https://lists.sourceforge.net/lists/listinfo/sbcl-bugs >> > |
|
From: Hraban <hr...@0b...> - 2024-04-28 15:39:38
|
At least for Linux they release pre-built binaries with every release, and Darwin not as often but still pretty regular. Btw if you run into trouble with clisp I recommend trying out ecl; I recently switched after I kept getting segfaults building sbcl from clisp on darwin x86-64. Works fine now with ecl. On 4/28/24 10:56 AM, John Marino wrote: > Well, I assumed the stack trace would prevent that from being a needle > in the haystack search. So I posted it in case there was a bug > introduced between releases. If you are confident the 2.4.3 code is > perfect, ok. > > This recipe supports FreeBSD, DragonFly, NetBSD, Linux, MidnightBSD, > Solaris and formerly Darwin. NetBSD and MidnightBSD didn't have > suitable bootstrap compilers so they were built with clisp, so I guess I > have to switch all the operating systems to clisp for the xc-host. > > As for the options that aren't choices - well, like I said, the genesis > of the recipe goes back more than a decade so they were probably choices > in the past and just became obsolete. There's no dedicated maintainer > for this package so there's nobody studying every release's change. > > > On Sun, Apr 28, 2024 at 8:41 AM Douglas Katzman via Sbcl-bugs > <sbc...@li... > <mailto:sbc...@li...>> wrote: > > > > On Sat, Apr 27, 2024 at 9:40 PM John Marino <jo...@ma... > <mailto:jo...@ma...>> wrote: > > > The reason for this is not known? > > As known as could be under the circumstance: Your bootstrap SBCL was > released Dec 31, 2014, and some time last month an SBCL developer > made a change which unwittingly or wittingly caused current sources > to become sensitive to a bug fix made any time over the last decade > of development. If you are asking literally which change it is, > either target or host code, no developer has time to spend on such a > wild goose chase. You could run a git bisect on both axes. (And > then what- figure out the oldest possible released SBCL to bootstrap > because it has exactly the needed patch?) > > And what do you consider "random"? > > You've got options in this list which are effectless: > --without-sb-thread --without-restore-fs-segment-register-from-tls > --with-sb-unicode --with-sb-core-compression --without-sb-qshow > --without-sb-xref-for-internals --without-sb-safepoint > --without-sb-thruption --without-sb-wtimer --without-sb-simd > --without-sb-thread --without-restore-fs-segment-register-from-tls > > restore-fs-segment isn't a choice, qshow isn't a choice, wtimer > isn't a choice, thruption isn't a choice, safepoint isn't the > default, unicode is the default, and sb-thread is already disabled > by default on DragonFly. You should remove anything that isn't > present for a good reason. > _______________________________________________ > Sbcl-bugs mailing list > Sbc...@li... <mailto:Sbc...@li...> > https://lists.sourceforge.net/lists/listinfo/sbcl-bugs > <https://lists.sourceforge.net/lists/listinfo/sbcl-bugs> > |
|
From: John M. <jo...@ma...> - 2024-04-28 14:56:34
|
Well, I assumed the stack trace would prevent that from being a needle in the haystack search. So I posted it in case there was a bug introduced between releases. If you are confident the 2.4.3 code is perfect, ok. This recipe supports FreeBSD, DragonFly, NetBSD, Linux, MidnightBSD, Solaris and formerly Darwin. NetBSD and MidnightBSD didn't have suitable bootstrap compilers so they were built with clisp, so I guess I have to switch all the operating systems to clisp for the xc-host. As for the options that aren't choices - well, like I said, the genesis of the recipe goes back more than a decade so they were probably choices in the past and just became obsolete. There's no dedicated maintainer for this package so there's nobody studying every release's change. On Sun, Apr 28, 2024 at 8:41 AM Douglas Katzman via Sbcl-bugs < sbc...@li...> wrote: > > > On Sat, Apr 27, 2024 at 9:40 PM John Marino <jo...@ma...> wrote: > >> >> The reason for this is not known? >> > As known as could be under the circumstance: Your bootstrap SBCL was > released Dec 31, 2014, and some time last month an SBCL developer made a > change which unwittingly or wittingly caused current sources to become > sensitive to a bug fix made any time over the last decade of development. > If you are asking literally which change it is, either target or host code, > no developer has time to spend on such a wild goose chase. You could run a > git bisect on both axes. (And then what- figure out the oldest possible > released SBCL to bootstrap because it has exactly the needed patch?) > > And what do you consider "random"? >> > You've got options in this list which are effectless: > --without-sb-thread --without-restore-fs-segment-register-from-tls > --with-sb-unicode --with-sb-core-compression --without-sb-qshow > --without-sb-xref-for-internals --without-sb-safepoint > --without-sb-thruption --without-sb-wtimer --without-sb-simd > --without-sb-thread --without-restore-fs-segment-register-from-tls > > restore-fs-segment isn't a choice, qshow isn't a choice, wtimer isn't a > choice, thruption isn't a choice, safepoint isn't the default, unicode is > the default, and sb-thread is already disabled by default on DragonFly. You > should remove anything that isn't present for a good reason. > _______________________________________________ > Sbcl-bugs mailing list > Sbc...@li... > https://lists.sourceforge.net/lists/listinfo/sbcl-bugs > |
|
From: Douglas K. <do...@go...> - 2024-04-28 13:41:16
|
On Sat, Apr 27, 2024 at 9:40 PM John Marino <jo...@ma...> wrote: > > The reason for this is not known? > As known as could be under the circumstance: Your bootstrap SBCL was released Dec 31, 2014, and some time last month an SBCL developer made a change which unwittingly or wittingly caused current sources to become sensitive to a bug fix made any time over the last decade of development. If you are asking literally which change it is, either target or host code, no developer has time to spend on such a wild goose chase. You could run a git bisect on both axes. (And then what- figure out the oldest possible released SBCL to bootstrap because it has exactly the needed patch?) And what do you consider "random"? > You've got options in this list which are effectless: --without-sb-thread --without-restore-fs-segment-register-from-tls --with-sb-unicode --with-sb-core-compression --without-sb-qshow --without-sb-xref-for-internals --without-sb-safepoint --without-sb-thruption --without-sb-wtimer --without-sb-simd --without-sb-thread --without-restore-fs-segment-register-from-tls restore-fs-segment isn't a choice, qshow isn't a choice, wtimer isn't a choice, thruption isn't a choice, safepoint isn't the default, unicode is the default, and sb-thread is already disabled by default on DragonFly. You should remove anything that isn't present for a good reason. |
|
From: <pe...@pe...> - 2024-04-28 10:52:05
|
Hi,
I downloaded sbcl-2.4.0-arm64-darwin-binary.tar.bz2 (which seems to be
the latest build for Mac/arm).
I would like to verify the checksum, but the file sbcl-2.4.0-crhodes.asc
only contains checksums for sbcl-2.4.0-documentation-html.tar,
sbcl-2.4.0-source.tar, and sbcl-2.4.0-x86-64-linux-binary.tar.
Would it be possible to get an updated checksum file?
/pem
|
|
From: John M. <jo...@ma...> - 2024-04-28 01:40:23
|
We've built SBCL for at least 5 years exactly the same way with the ancient SBCL as a bootstrap compiler. SBCL 2.4.2 built without issue. 2.4.3 doesn't. The reason for this is not known? And what do you consider "random"? On Sat, Apr 27, 2024 at 8:33 PM Stas Boukarev <sta...@gm...> wrote: > Don't use an ancient SBCL as a compiler host. > Also probably don't put random things into ./make.sh options. > > On Sun, Apr 28, 2024 at 4:04 AM John Marino <jo...@ma...> wrote: > >> Sorry, I thought I grabbed enough. >> The entire log is here: >> https://loki.dragonflybsd.org/raven/logs/sbcl-243-build.log >> >> >> On Sat, Apr 27, 2024 at 2:32 PM Stas Boukarev <sta...@gm...> wrote: >> >>> The STYLE-WARNING is not included in your transcript. >>> >>> On Sat, Apr 27, 2024 at 10:32 PM John Marino <jo...@ma...> wrote: >>> >>>> With the release of version 2.4.3, SBCL fails to build on DragonFly >>>> BSD. The end of the build log is below. Can anybody suggest a patch? >>>> Thanks, >>>> John >>>> >>>> [305/306] src/code/target-format (0.483 sec) >>>> [306/306] src/code/late-globaldb (0.052 sec) >>>> 67.868 >>>> ; Math journal: wrote #P"obj/from-host/xfloat-math.lisp-expr" (19404 entries) >>>> ctype cache metrics: Count LF Seek Hit maxPSL Mask >>>> VALUES : 44282 67.6% 7 FFFF >>>> List : 41612 63.5% 11 FFFF >>>> Set : 27215 41.5% 6 FFFF >>>> UNION : 21349 65.2% 9 7FFF >>>> EQL : 19269 NIL % >>>> NUMERIC : 10860 66.3% 7 3FFF >>>> FUN : 9519 58.1% 7 3FFF >>>> INTERSECTION : 6184 37.7% 5 3FFF >>>> MEMBER : 3695 45.1% 6 1FFF >>>> NEGATION : 2239 54.7% 8 FFF >>>> Key-Info : 1536 75.0% 7 7FF >>>> ARRAY : 630 61.5% 5 3FF >>>> CONS : 460 44.9% 5 3FF >>>> Key-Info-Set : 452 44.1% 5 3FF >>>> CHARACTER-SET : 140 54.7% 6 FF >>>> ALIEN-TYPE : 104 40.6% 3 FF >>>> UNKNOWN : 67 52.3% 4 7F >>>> CONSTANT : 59 46.1% 4 7F >>>> SIMD-PACK : 42 65.6% 9 3F >>>> SIMD-PACK-256 : 42 65.6% 8 3F >>>> HAIRY : 17 26.6% 3 3F >>>> MEMBER/EQ : 13 20.3% 1 3F >>>> FUN-DESIGNATOR : 5 7.8% 1 3F >>>> Total : 189791 >>>> Alien: >>>> FUN : 73 57.0% 5 7F >>>> POINTER : 17 26.6% 3 3F >>>> INTEGER : 8 12.5% 2 3F >>>> ARRAY : 5 7.8% 1 3F >>>> C-STRING : 3 4.7% 1 3F >>>> BOOLEAN : 1 1.6% 1 3F >>>> VALUES : 1 1.6% 1 3F >>>> ENUM : 0 0.0% 0 3F >>>> UNION : 0 0.0% 0 3F >>>> STRUCT : 0 0.0% 0 3F >>>> Total : 108 >>>> Math journal: replaced from #P"obj/from-host/xfloat-math.lisp-expr" >>>> >>>> ; >>>> ; compilation unit finished >>>> ; caught 2 STYLE-WARNING conditions >>>> ; printed 1 note >>>> Unhandled SIMPLE-ERROR: >>>> make-host-2 stopped due to unexpected STYLE-WARNING raised from the host. >>>> >>>> Backtrace for: #<HOST-SB-THREAD:THREAD "main thread" RUNNING {10030CEA33}> >>>> 0: ((LAMBDA NIL :IN HOST-SB-DEBUG::FUNCALL-WITH-DEBUG-IO-SYNTAX)) >>>> 1: (HOST-SB-IMPL::CALL-WITH-SANE-IO-SYNTAX #<CLOSURE (LAMBDA NIL :IN HOST-SB-DEBUG::FUNCALL-WITH-DEBUG-IO-SYNTAX) {10066244BB}>) >>>> 2: (HOST-SB-IMPL::%WITH-STANDARD-IO-SYNTAX #<CLOSURE (LAMBDA NIL :IN HOST-SB-DEBUG::FUNCALL-WITH-DEBUG-IO-SYNTAX) {100662448B}>) >>>> 3: (HOST-SB-DEBUG:PRINT-BACKTRACE :STREAM #<SYNONYM-STREAM :SYMBOL HOST-SB-SYS:*STDERR* {10001528D3}> :START 0 :FROM :INTERRUPTED-FRAME :COUNT NIL :PRINT-THREAD T :PRINT-FRAME-SOURCE NIL :METHOD-FRAME-STYLE NIL) >>>> 4: (HOST-SB-DEBUG::DEBUGGER-DISABLED-HOOK #<SIMPLE-ERROR "make-host-2 stopped due to unexpected ~A raised from the host." {1006622193}> #<unavailable argument>) >>>> 5: (HOST-SB-DEBUG::RUN-HOOK HOST-SB-EXT:*INVOKE-DEBUGGER-HOOK* #<SIMPLE-ERROR "make-host-2 stopped due to unexpected ~A raised from the host." {1006622193}>) >>>> 6: (INVOKE-DEBUGGER #<SIMPLE-ERROR "make-host-2 stopped due to unexpected ~A raised from the host." {1006622193}>) >>>> 7: (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." STYLE-WARNING) >>>> 8: (HOST-SB-INT:SIMPLE-EVAL-IN-LEXENV (LET (FAIL VARIABLES FUNCTIONS TYPES WARNP STYLE-WARNP) (HANDLER-BIND ((STYLE-WARNING (LAMBDA # # #)) (SIMPLE-WARNING (LAMBDA # # #))) (SB-XC:WITH-COMPILATION-UNIT NIL (LOAD "src/cold/compile-cold-sbcl.lisp") (LET (#) (WHEN # # #)) (WHEN SB-C::*UNDEFINED-WARNINGS* (SETF FAIL T) (DOLIST # #)))) (WHEN FAIL (CERROR "Proceed anyway" "Undefined ~:[~;variables~] ~:[~;types~]~ >>>> ~:[~;functions (incomplete SB-COLD::*UNDEFINED-FUN-ALLOWLIST*?)~]" VARIABLES TYPES FUNCTIONS)) (WHEN (AND (OR WARNP STYLE-WARNP) *FAIL-ON-WARNINGS* (NOT (TARGET-FEATUREP :WIN32))) (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." (OR WARNP STYLE-WARNP)))) #<NULL-LEXENV>) >>>> 9: (HOST-SB-EXT:EVAL-TLF (LET (FAIL VARIABLES FUNCTIONS TYPES WARNP STYLE-WARNP) (HANDLER-BIND ((STYLE-WARNING (LAMBDA # # #)) (SIMPLE-WARNING (LAMBDA # # #))) (SB-XC:WITH-COMPILATION-UNIT NIL (LOAD "src/cold/compile-cold-sbcl.lisp") (LET (#) (WHEN # # #)) (WHEN SB-C::*UNDEFINED-WARNINGS* (SETF FAIL T) (DOLIST # #)))) (WHEN FAIL (CERROR "Proceed anyway" "Undefined ~:[~;variables~] ~:[~;types~]~ >>>> ~:[~;functions (incomplete SB-COLD::*UNDEFINED-FUN-ALLOWLIST*?)~]" VARIABLES TYPES FUNCTIONS)) (WHEN (AND (OR WARNP STYLE-WARNP) *FAIL-ON-WARNINGS* (NOT (TARGET-FEATUREP :WIN32))) (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." (OR WARNP STYLE-WARNP)))) 11 #<NULL-LEXENV>) >>>> 10: ((FLET HOST-SB-FASL::EVAL-FORM :IN HOST-SB-INT:LOAD-AS-SOURCE) (LET (FAIL VARIABLES FUNCTIONS TYPES WARNP STYLE-WARNP) (HANDLER-BIND ((STYLE-WARNING (LAMBDA # # #)) (SIMPLE-WARNING (LAMBDA # # #))) (SB-XC:WITH-COMPILATION-UNIT NIL (LOAD "src/cold/compile-cold-sbcl.lisp") (LET (#) (WHEN # # #)) (WHEN SB-C::*UNDEFINED-WARNINGS* (SETF FAIL T) (DOLIST # #)))) (WHEN FAIL (CERROR "Proceed anyway" "Undefined ~:[~;variables~] ~:[~;types~]~ >>>> ~:[~;functions (incomplete SB-COLD::*UNDEFINED-FUN-ALLOWLIST*?)~]" VARIABLES TYPES FUNCTIONS)) (WHEN (AND (OR WARNP STYLE-WARNP) *FAIL-ON-WARNINGS* (NOT (TARGET-FEATUREP :WIN32))) (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." (OR WARNP STYLE-WARNP)))) 11) >>>> 11: (HOST-SB-INT:LOAD-AS-SOURCE #<HOST-SB-SYS:FD-STREAM for "file /construction/sbcl/sbcl-2.4.3/make-host-2.lisp" {10031A29A3}> :VERBOSE NIL :PRINT NIL :CONTEXT "loading") >>>> 12: ((FLET HOST-SB-FASL::LOAD-STREAM :IN LOAD) #<HOST-SB-SYS:FD-STREAM for "file /construction/sbcl/sbcl-2.4.3/make-host-2.lisp" {10031A29A3}> NIL) >>>> 13: (LOAD "make-host-2.lisp" :VERBOSE NIL :PRINT NIL :IF-DOES-NOT-EXIST T :EXTERNAL-FORMAT :DEFAULT) >>>> 14: (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp" T) >>>> 15: (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp" 17459838004) [tl,external] >>>> 16: (HOST-SB-INT:SIMPLE-EVAL-IN-LEXENV (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp") #<NULL-LEXENV>) >>>> 17: (EVAL (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp")) >>>> 18: (HOST-SB-EXT:INTERACTIVE-EVAL (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp") :EVAL NIL) >>>> 19: (HOST-SB-IMPL::REPL-FUN NIL) >>>> 20: ((LAMBDA NIL :IN HOST-SB-IMPL::TOPLEVEL-REPL)) >>>> 21: (HOST-SB-IMPL::%WITH-REBOUND-IO-SYNTAX #<CLOSURE (LAMBDA NIL :IN HOST-SB-IMPL::TOPLEVEL-REPL) {10030D546B}>) >>>> 22: (HOST-SB-IMPL::TOPLEVEL-REPL NIL) >>>> 23: (HOST-SB-IMPL::TOPLEVEL-INIT) >>>> 24: ((FLET #:WITHOUT-INTERRUPTS-BODY-87 :IN HOST-SB-EXT:SAVE-LISP-AND-DIE)) >>>> 25: ((LABELS HOST-SB-IMPL::RESTART-LISP :IN HOST-SB-EXT:SAVE-LISP-AND-DIE)) >>>> >>>> unhandled condition in --disable-debugger mode, quitting >>>> >>>> _______________________________________________ >>>> Sbcl-bugs mailing list >>>> Sbc...@li... >>>> https://lists.sourceforge.net/lists/listinfo/sbcl-bugs >>>> >>> |
|
From: Stas B. <sta...@gm...> - 2024-04-28 01:33:40
|
Don't use an ancient SBCL as a compiler host. Also probably don't put random things into ./make.sh options. On Sun, Apr 28, 2024 at 4:04 AM John Marino <jo...@ma...> wrote: > Sorry, I thought I grabbed enough. > The entire log is here: > https://loki.dragonflybsd.org/raven/logs/sbcl-243-build.log > > > On Sat, Apr 27, 2024 at 2:32 PM Stas Boukarev <sta...@gm...> wrote: > >> The STYLE-WARNING is not included in your transcript. >> >> On Sat, Apr 27, 2024 at 10:32 PM John Marino <jo...@ma...> wrote: >> >>> With the release of version 2.4.3, SBCL fails to build on DragonFly >>> BSD. The end of the build log is below. Can anybody suggest a patch? >>> Thanks, >>> John >>> >>> [305/306] src/code/target-format (0.483 sec) >>> [306/306] src/code/late-globaldb (0.052 sec) >>> 67.868 >>> ; Math journal: wrote #P"obj/from-host/xfloat-math.lisp-expr" (19404 entries) >>> ctype cache metrics: Count LF Seek Hit maxPSL Mask >>> VALUES : 44282 67.6% 7 FFFF >>> List : 41612 63.5% 11 FFFF >>> Set : 27215 41.5% 6 FFFF >>> UNION : 21349 65.2% 9 7FFF >>> EQL : 19269 NIL % >>> NUMERIC : 10860 66.3% 7 3FFF >>> FUN : 9519 58.1% 7 3FFF >>> INTERSECTION : 6184 37.7% 5 3FFF >>> MEMBER : 3695 45.1% 6 1FFF >>> NEGATION : 2239 54.7% 8 FFF >>> Key-Info : 1536 75.0% 7 7FF >>> ARRAY : 630 61.5% 5 3FF >>> CONS : 460 44.9% 5 3FF >>> Key-Info-Set : 452 44.1% 5 3FF >>> CHARACTER-SET : 140 54.7% 6 FF >>> ALIEN-TYPE : 104 40.6% 3 FF >>> UNKNOWN : 67 52.3% 4 7F >>> CONSTANT : 59 46.1% 4 7F >>> SIMD-PACK : 42 65.6% 9 3F >>> SIMD-PACK-256 : 42 65.6% 8 3F >>> HAIRY : 17 26.6% 3 3F >>> MEMBER/EQ : 13 20.3% 1 3F >>> FUN-DESIGNATOR : 5 7.8% 1 3F >>> Total : 189791 >>> Alien: >>> FUN : 73 57.0% 5 7F >>> POINTER : 17 26.6% 3 3F >>> INTEGER : 8 12.5% 2 3F >>> ARRAY : 5 7.8% 1 3F >>> C-STRING : 3 4.7% 1 3F >>> BOOLEAN : 1 1.6% 1 3F >>> VALUES : 1 1.6% 1 3F >>> ENUM : 0 0.0% 0 3F >>> UNION : 0 0.0% 0 3F >>> STRUCT : 0 0.0% 0 3F >>> Total : 108 >>> Math journal: replaced from #P"obj/from-host/xfloat-math.lisp-expr" >>> >>> ; >>> ; compilation unit finished >>> ; caught 2 STYLE-WARNING conditions >>> ; printed 1 note >>> Unhandled SIMPLE-ERROR: >>> make-host-2 stopped due to unexpected STYLE-WARNING raised from the host. >>> >>> Backtrace for: #<HOST-SB-THREAD:THREAD "main thread" RUNNING {10030CEA33}> >>> 0: ((LAMBDA NIL :IN HOST-SB-DEBUG::FUNCALL-WITH-DEBUG-IO-SYNTAX)) >>> 1: (HOST-SB-IMPL::CALL-WITH-SANE-IO-SYNTAX #<CLOSURE (LAMBDA NIL :IN HOST-SB-DEBUG::FUNCALL-WITH-DEBUG-IO-SYNTAX) {10066244BB}>) >>> 2: (HOST-SB-IMPL::%WITH-STANDARD-IO-SYNTAX #<CLOSURE (LAMBDA NIL :IN HOST-SB-DEBUG::FUNCALL-WITH-DEBUG-IO-SYNTAX) {100662448B}>) >>> 3: (HOST-SB-DEBUG:PRINT-BACKTRACE :STREAM #<SYNONYM-STREAM :SYMBOL HOST-SB-SYS:*STDERR* {10001528D3}> :START 0 :FROM :INTERRUPTED-FRAME :COUNT NIL :PRINT-THREAD T :PRINT-FRAME-SOURCE NIL :METHOD-FRAME-STYLE NIL) >>> 4: (HOST-SB-DEBUG::DEBUGGER-DISABLED-HOOK #<SIMPLE-ERROR "make-host-2 stopped due to unexpected ~A raised from the host." {1006622193}> #<unavailable argument>) >>> 5: (HOST-SB-DEBUG::RUN-HOOK HOST-SB-EXT:*INVOKE-DEBUGGER-HOOK* #<SIMPLE-ERROR "make-host-2 stopped due to unexpected ~A raised from the host." {1006622193}>) >>> 6: (INVOKE-DEBUGGER #<SIMPLE-ERROR "make-host-2 stopped due to unexpected ~A raised from the host." {1006622193}>) >>> 7: (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." STYLE-WARNING) >>> 8: (HOST-SB-INT:SIMPLE-EVAL-IN-LEXENV (LET (FAIL VARIABLES FUNCTIONS TYPES WARNP STYLE-WARNP) (HANDLER-BIND ((STYLE-WARNING (LAMBDA # # #)) (SIMPLE-WARNING (LAMBDA # # #))) (SB-XC:WITH-COMPILATION-UNIT NIL (LOAD "src/cold/compile-cold-sbcl.lisp") (LET (#) (WHEN # # #)) (WHEN SB-C::*UNDEFINED-WARNINGS* (SETF FAIL T) (DOLIST # #)))) (WHEN FAIL (CERROR "Proceed anyway" "Undefined ~:[~;variables~] ~:[~;types~]~ >>> ~:[~;functions (incomplete SB-COLD::*UNDEFINED-FUN-ALLOWLIST*?)~]" VARIABLES TYPES FUNCTIONS)) (WHEN (AND (OR WARNP STYLE-WARNP) *FAIL-ON-WARNINGS* (NOT (TARGET-FEATUREP :WIN32))) (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." (OR WARNP STYLE-WARNP)))) #<NULL-LEXENV>) >>> 9: (HOST-SB-EXT:EVAL-TLF (LET (FAIL VARIABLES FUNCTIONS TYPES WARNP STYLE-WARNP) (HANDLER-BIND ((STYLE-WARNING (LAMBDA # # #)) (SIMPLE-WARNING (LAMBDA # # #))) (SB-XC:WITH-COMPILATION-UNIT NIL (LOAD "src/cold/compile-cold-sbcl.lisp") (LET (#) (WHEN # # #)) (WHEN SB-C::*UNDEFINED-WARNINGS* (SETF FAIL T) (DOLIST # #)))) (WHEN FAIL (CERROR "Proceed anyway" "Undefined ~:[~;variables~] ~:[~;types~]~ >>> ~:[~;functions (incomplete SB-COLD::*UNDEFINED-FUN-ALLOWLIST*?)~]" VARIABLES TYPES FUNCTIONS)) (WHEN (AND (OR WARNP STYLE-WARNP) *FAIL-ON-WARNINGS* (NOT (TARGET-FEATUREP :WIN32))) (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." (OR WARNP STYLE-WARNP)))) 11 #<NULL-LEXENV>) >>> 10: ((FLET HOST-SB-FASL::EVAL-FORM :IN HOST-SB-INT:LOAD-AS-SOURCE) (LET (FAIL VARIABLES FUNCTIONS TYPES WARNP STYLE-WARNP) (HANDLER-BIND ((STYLE-WARNING (LAMBDA # # #)) (SIMPLE-WARNING (LAMBDA # # #))) (SB-XC:WITH-COMPILATION-UNIT NIL (LOAD "src/cold/compile-cold-sbcl.lisp") (LET (#) (WHEN # # #)) (WHEN SB-C::*UNDEFINED-WARNINGS* (SETF FAIL T) (DOLIST # #)))) (WHEN FAIL (CERROR "Proceed anyway" "Undefined ~:[~;variables~] ~:[~;types~]~ >>> ~:[~;functions (incomplete SB-COLD::*UNDEFINED-FUN-ALLOWLIST*?)~]" VARIABLES TYPES FUNCTIONS)) (WHEN (AND (OR WARNP STYLE-WARNP) *FAIL-ON-WARNINGS* (NOT (TARGET-FEATUREP :WIN32))) (CERROR "Proceed anyway" "make-host-2 stopped due to unexpected ~A raised from the host." (OR WARNP STYLE-WARNP)))) 11) >>> 11: (HOST-SB-INT:LOAD-AS-SOURCE #<HOST-SB-SYS:FD-STREAM for "file /construction/sbcl/sbcl-2.4.3/make-host-2.lisp" {10031A29A3}> :VERBOSE NIL :PRINT NIL :CONTEXT "loading") >>> 12: ((FLET HOST-SB-FASL::LOAD-STREAM :IN LOAD) #<HOST-SB-SYS:FD-STREAM for "file /construction/sbcl/sbcl-2.4.3/make-host-2.lisp" {10031A29A3}> NIL) >>> 13: (LOAD "make-host-2.lisp" :VERBOSE NIL :PRINT NIL :IF-DOES-NOT-EXIST T :EXTERNAL-FORMAT :DEFAULT) >>> 14: (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp" T) >>> 15: (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp" 17459838004) [tl,external] >>> 16: (HOST-SB-INT:SIMPLE-EVAL-IN-LEXENV (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp") #<NULL-LEXENV>) >>> 17: (EVAL (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp")) >>> 18: (HOST-SB-EXT:INTERACTIVE-EVAL (COMMON-LISP-USER::LOAD-SBCL-FILE "make-host-2.lisp") :EVAL NIL) >>> 19: (HOST-SB-IMPL::REPL-FUN NIL) >>> 20: ((LAMBDA NIL :IN HOST-SB-IMPL::TOPLEVEL-REPL)) >>> 21: (HOST-SB-IMPL::%WITH-REBOUND-IO-SYNTAX #<CLOSURE (LAMBDA NIL :IN HOST-SB-IMPL::TOPLEVEL-REPL) {10030D546B}>) >>> 22: (HOST-SB-IMPL::TOPLEVEL-REPL NIL) >>> 23: (HOST-SB-IMPL::TOPLEVEL-INIT) >>> 24: ((FLET #:WITHOUT-INTERRUPTS-BODY-87 :IN HOST-SB-EXT:SAVE-LISP-AND-DIE)) >>> 25: ((LABELS HOST-SB-IMPL::RESTART-LISP :IN HOST-SB-EXT:SAVE-LISP-AND-DIE)) >>> >>> unhandled condition in --disable-debugger mode, quitting >>> >>> _______________________________________________ >>> Sbcl-bugs mailing list >>> Sbc...@li... >>> https://lists.sourceforge.net/lists/listinfo/sbcl-bugs >>> >> |
6 messages has been excluded from this view by a project administrator.