error building fricas 1.3.5-2 on Linux

Nasser M. Abbasi

unread,

Jul 4, 2019, 8:47:18 PM7/4/19

to FriCAS - computer algebra system

I first tried installing Latest fricas using Linux Manjaro GUI, and that failed.

Next tried to install it using yay command. This also fail. Below is the start of the log file and the end of the log file showing the error

>uname -r

4.19.36-1-MANJARO

>lsb_release -a

LSB Version: n/a

Distributor ID: ManjaroLinux

Description: Manjaro Linux

Release: 18.0.4

Codename: Illyria

Any idea why it failed and how to correct it?

====================================

>yay -S fricas

:: Checking for conflicts...

:: Checking for inner conflicts...

[Aur: 1] fricas-1.3.5-2

:: Downloaded PKGBUILD (1/1): fricas

1 fricas (Installed) (Build Files Exist)

==> Diffs to show?

==> [N]one [A]ll [Ab]ort [I]nstalled [No]tInstalled or (1 2 3, 1-3, ^4)

==> N

:: Parsing SRCINFO (1/1): fricas

==> Making package: fricas 1.3.5-2 (Thu 04 Jul 2019 07:13:03 PM CDT)

==> Retrieving sources...

-> Downloading fricas-1.3.5-full.tar.bz2...

% Total % Received % Xferd Average Speed Time Time Time Current

Dload Upload Total Spent Left Speed

100 365 100 365 0 0 2005 0 --:--:-- --:--:-- --:--:-- 1994

100 9.9M 100 9.9M 0 0 981k 0 0:00:10 0:00:10 --:--:-- 1107k

-> Downloading copyright.txt...

% Total % Received % Xferd Average Speed Time Time Time Current

Dload Upload Total Spent Left Speed

100 2407 100 2407 0 0 2610 0 --:--:-- --:--:-- --:--:-- 2607

==> Validating source files with md5sums...

fricas-1.3.5-full.tar.bz2 ... Passed

copyright.txt ... Passed

==> Cleaning up...

==> Making package: fricas 1.3.5-2 (Thu 04 Jul 2019 07:13:15 PM CDT)

==> Checking runtime dependencies...

==> Checking buildtime dependencies...

==> Retrieving sources...

-> Found fricas-1.3.5-full.tar.bz2

-> Found copyright.txt

==> Validating source files with md5sums...

fricas-1.3.5-full.tar.bz2 ... Passed

copyright.txt ... Passed

==> Removing existing $srcdir/ directory...

==> Extracting sources...

-> Extracting fricas-1.3.5-full.tar.bz2 with bsdtar

==> Sources are ready.

==> Making package: fricas 1.3.5-2 (Thu 04 Jul 2019 07:13:18 PM CDT)

==> Checking runtime dependencies...

==> Checking buildtime dependencies...

==> WARNING: Using existing $srcdir/ tree

==> Starting build()...

checking build system type... x86_64-linux-gnu

checking host system type... x86_64-linux-gnu

checking target system type... x86_64-linux-gnu

checking for make... make

checking for gcc... gcc

checking whether the C compiler works... yes

checking for C compiler default output file name... a.out

checking for suffix of executables...

checking whether we are cross compiling... no

checking for suffix of object files... o

checking whether we are using the GNU C compiler... yes

checking whether gcc accepts -g... yes

checking for gcc option to accept ISO C89... none needed

checking for a BSD-compatible install... /bin/install -c

checking for touch... touch

checking for mktemp... mktemp

checking for gawk... gawk

checking for gtar... no

checking for tar... tar

checking for ranlib... ranlib

checking for ar... ar

checking for latex... /usr/local/texlive/2018/bin/x86_64-linux/latex

checking for makeindex... makeindex

checking PREGENERATED... "/home/me/.cache/yay/fricas/src/fricas-1.3.5/pre-generated"

checking Lisp implementation... This is SBCL 1.5.3, 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.

* sbcl

checking how to run the C preprocessor... gcc -E

checking for grep that handles long lines and -e... /bin/grep

checking for egrep... /bin/grep -E

checking for ANSI C header files... yes

checking for sys/types.h... yes

checking for sys/stat.h... yes

checking for stdlib.h... yes

checking for string.h... yes

checking for memory.h... yes

checking for strings.h... yes

checking for inttypes.h... yes

checking for stdint.h... yes

checking for unistd.h... yes

fasl

checking dirent.h usability... yes

checking dirent.h presence... yes

checking for dirent.h... yes

checking whether closedir is declared... yes

checking whether opendir is declared... yes

checking whether readdir is declared... yes

checking whether dirfd is declared... yes

checking whether fchdir is declared... yes

checking signal.h usability... yes

checking signal.h presence... yes

checking for signal.h... yes

checking whether sigaction is declared... yes

checking for sys/stat.h... (cached) yes

checking for unistd.h... (cached) yes

checking whether getuid is declared... yes

checking whether geteuid is declared... yes

checking whether getgid is declared... yes

checking whether getegid is declared... yes

checking whether kill is declared... yes

checking sys/socket.h usability... yes

checking sys/socket.h presence... yes

checking for sys/socket.h... yes

checking util.h usability... no

checking util.h presence... no

checking for util.h... no

checking pty.h usability... yes

checking pty.h presence... yes

checking for pty.h... yes

checking whether openpty is declared... yes

checking for openpty in -lc... no

checking for openpty in -lutil... yes

checking sys/wait.h usability... yes

checking sys/wait.h presence... yes

checking for sys/wait.h... yes

checking whether wait is declared... yes

checking whether fork is declared... yes

checking for X... libraries , headers

checking for gethostbyname... yes

checking for connect... yes

checking for remove... yes

checking for shmat... yes

checking for IceConnectionNumber in -lICE... yes

checking for XpmReadFileToImage in -lXpm... yes

checking regex.h usability... yes

checking regex.h presence... yes

checking for regex.h... yes

checking for xvfb-run... no

configure: HyperDoc graphics must be built separately.

configure: WARNING: Aldor interface will not be built.

configure: creating ./config.status

config.status: creating src/clef/Makefile

config.status: creating src/sman/Makefile

config.status: creating src/graph/Makefile

config.status: creating src/graph/Gdraws/Makefile

config.status: creating src/graph/view2D/Makefile

config.status: creating src/graph/view3D/Makefile

config.status: creating src/graph/viewAlone/Makefile

config.status: creating src/graph/viewman/Makefile

config.status: creating src/hyper/Makefile

config.status: creating src/doc/Makefile

config.status: creating Makefile

config.status: creating src/Makefile

config.status: creating src/lib/Makefile

config.status: creating src/lisp/Makefile

config.status: creating src/boot/Makefile

config.status: creating src/interp/Makefile

config.status: creating src/algebra/Makefile

config.status: creating src/input/Makefile

config.status: creating src/etc/Makefile

config.status: creating src/aldor/Makefile

config.status: creating src/aldor/Makefile2

config.status: creating src/aldor/Makefile3

config.status: creating contrib/emacs/Makefile

config.status: creating config/fricas_c_macros.h

extracting list of SPAD type definitions

Type 'make' (without quotes) to build FriCAS

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/bin

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/lib

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/bin

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/lib

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/share/spadhelp

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/share/msgs

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/share/texmf/tex

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/algebra/

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/autoload

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/input

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/src/algebra

mkdir -p /home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu

echo timestamp > stamp-rootdirs

echo "PREGENERATED=[/home/me/.cache/yay/fricas/src/fricas-1.3.5/pre-generated]"

PREGENERATED=[/home/me/.cache/yay/fricas/src/fricas-1.3.5/pre-generated]

if test -z "/home/me/.cache/yay/fricas/src/fricas-1.3.5/pre-generated"; then \

echo "Not using pre-generated data."; \

elif test ! -d "/home/me/.cache/yay/fricas/src/fricas-1.3.5/pre-generated"; then \

echo "PREGENERATED directory does not exist."; \

exit 1; \

else \

if test -d /home/me/.cache/yay/fricas/src/fricas-1.3.5/pre-generated/target; then \

cp -r /home/me/.cache/yay/fricas/src/fricas-1.3.5/pre-generated/target/* /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu; \

fi && \

echo timestamp > pre-gen; \

fi

.

wrote /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/algebra/exposed.fasl

; compilation finished in 0:00:00.014

echo '(load "makeint.lisp") #-:ecl(BOOT::reclaim)' \

'#+:ecl(FRICAS-LISP::make-program "/home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/bin/interpsys" nil)' \

'#-:ecl(BOOT::spad-save "/home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/bin/interpsys" t)' | \

DAASE='/home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share' /home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/bin/bootsys

* Using local database /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/compress.daase.. Re-reading compress.daase Using local database /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/interp.daase.. Re-reading interp.daase

FriCAS initialization: interpreter

FriCAS initialization: database

FriCAS initialization: constructors

FriCAS initialization: history

FriCAS Computer Algebra System

Version: FriCAS 1.3.5

Timestamp: Thu 04 Jul 2019 07:13:20 PM CDT

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

Issue )copyright to view copyright notices.

Issue )summary for a summary of useful system commands.

Issue )quit to leave FriCAS and return to shell.

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

Using local database /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/compress.daase.. Re-reading compress.daase Using local database /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/interp.daase.. Re-reading interp.daase

Using local database /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/operation.daase.. Re-reading operation.daase

Using local database /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/browse.daase.. Re-reading browse.daase

Using local database /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/category.daase.. Re-reading category.daase

Initial getdatabase

preloading /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/algebra/NIL.fasl..skipped.

T

* NIL

*

debugger invoked on a SB-IMPL::SAVE-WITH-MULTIPLE-THREADS-ERROR in thread

#<THREAD "main thread" RUNNING {10005285B3}>:

Cannot save core with multiple threads running.

Interactive thread (of current session):

#<THREAD "main thread" RUNNING {10005285B3}>

Other thread:

#<THREAD "finalizer" RUNNING {10016A8053}>

Ralf Hemmecke

unread,

Jul 5, 2019, 1:23:30 AM7/5/19

to 'Nasser M. Abbasi' via FriCAS - computer algebra system

Hi,

archive the whole directory for investigations that other people here on
the list may have for you.

To me it looks like a problem with parallel build that also happens to
me. It only appears sporadically, so that when I compile the next time,
everything is fine, i.e., it's not reproducible. Maybe you are in the
same situation. Just try the build in a new copy.

Ralf

Nasser M. Abbasi

unread,

Jul 5, 2019, 6:25:32 AM7/5/19

to FriCAS - computer algebra system

I tried it again, on a new fresh installation of Linux Manjaro (under VBox) and got same error in same place.

I put everything under /home/me/.cache/yay/fricas, which is where yay does the build, into one tar.gz

file, It is 70 MB. Here is the link

https://www.12000.org/tmp/070519/fricas.tar.gz

I've always had trouble building Fricas on Linux for some reason.

Waldek Hebisch

unread,

Jul 6, 2019, 6:45:06 AM7/6/19

to fricas...@googlegroups.com

Nasser M. Abbasi wrote:
>
> I first tried installing Latest fricas using Linux Manjaro GUI, and that
> failed.
>
> Next tried to install it using yay command. This also fail. Below is the
> start of the log file and the end of the log file showing the error
>
> >uname -r
> 4.19.36-1-MANJARO
>
> >lsb_release -a
> LSB Version: n/a
> Distributor ID: ManjaroLinux
> Description: Manjaro Linux
> Release: 18.0.4
> Codename: Illyria
>
>
> Any idea why it failed and how to correct it?
>
> ====================================
> >yay -S fricas
> :: Checking for conflicts...
> :: Checking for inner conflicts...
> [Aur: 1] fricas-1.3.5-2

<snip>

> checking Lisp implementation... This is SBCL 1.5.3, an implementation of
> ANSI Common Lisp.

<snip>

> echo '(load "makeint.lisp") #-:ecl(BOOT::reclaim)' \
> '#+:ecl(FRICAS-LISP::make-program
> "/home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/bin/interpsys"
> nil)' \
> '#-:ecl(BOOT::spad-save
> "/home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/bin/interpsys"
> t)' | \
> DAASE='/home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share'
> /home/me/.cache/yay/fricas/src/fricas-1.3.5/build/x86_64-linux-gnu/bin/bootsys
> * Using local database
> /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/compress.daase..
> Re-reading compress.daase Using local database
> /home/me/.cache/yay/fricas/src/fricas-1.3.5/src/share/algebra/interp.daase..

<snip>

>
> T
> * NIL
> *
> debugger invoked on a SB-IMPL::SAVE-WITH-MULTIPLE-THREADS-ERROR in thread
> #<THREAD "main thread" RUNNING {10005285B3}>:
> Cannot save core with multiple threads running.
>
> Interactive thread (of current session):
> #<THREAD "main thread" RUNNING {10005285B3}>
>
> Other thread:
> #<THREAD "finalizer" RUNNING {10016A8053}>
> See also:
> The SBCL Manual, Node "Saving a Core Image"
>
> Type HELP for debugger help, or (SB-EXT:EXIT) to exit from SBCL.
>
> restarts (invokable by number or by possibly-abbreviated name):
> 0: [ABORT] Exit debugger, returning to top level.
>
> (SB-IMPL::DEINIT)
> 0]
>
> 0

This error happens when sbcl is creating new executable (image).
This is rather complex operation done entirely by internal sbcl
code. Everthing before looked OK. The error message about
threads is puzzling: FriCAS uses only one thread and any other
thread can appear only for internal needs of sbcl or due to
some extensions. In other words it seems that on Manjaro
sbcl is not working correctly. I have tried on my machine
using the same version (1.5.3) of sbcl and it works.

One useful test is to fetch sbcl sources, build sbcl from
sources and then use this sbcl to build FriCAS. That would
isolate general Manjaro featurs (libraries, "security
enhancements", etc) from Manjaro-specific changes to sbcl.

--
Waldek Hebisch

oldk1331

unread,

Jul 6, 2019, 9:19:54 AM7/6/19

to fricas...@googlegroups.com

Hi, I installed Manjaro and run "yay - S fricas", it builds successfully. Can you try again?

Nasser M. Abbasi

unread,

Jul 6, 2019, 9:40:34 PM7/6/19

to FriCAS - computer algebra system

On Saturday, July 6, 2019 at 8:19:54 AM UTC-5, oldk1331 wrote:

Hi, I installed Manjaro and run "yay - S fricas", it builds successfully. Can you try again?

I tried it again, and it failed same place:

======================================

>yay -S fricas

:: Checking for conflicts...

:: Checking for inner conflicts...

[Aur: 1] fricas-1.3.5-2

1 fricas (Build Files Exist)

==> Packages to cleanBuild?

==> [N]one [A]ll [Ab]ort [I]nstalled [No]tInstalled or (1 2 3, 1-3, ^4)

==>

:: PKGBUILD up to date, Skipping (1/1): fricas

1 fricas (Build Files Exist)

==> Diffs to show?

==> [N]one [A]ll [Ab]ort [I]nstalled [No]tInstalled or (1 2 3, 1-3, ^4)

==>

:: Parsing SRCINFO (1/1): fricas

==> Making package: fricas 1.3.5-2 (Sat 06 Jul 2019 08:35:12 PM CDT)

==> Retrieving sources...

-> Found fricas-1.3.5-full.tar.bz2

-> Found copyright.txt

==> Validating source files with md5sums...

fricas-1.3.5-full.tar.bz2 ... Passed

copyright.txt ... Passed

==> Cleaning up...

==> Making package: fricas 1.3.5-2 (Sat 06 Jul 2019 08:35:13 PM CDT)

....

preloading /home/me/.cache/yay/fricas/src/fricas-1.3.5/target/x86_64-linux-gnu/algebra/NIL.fasl..skipped.

T

* NIL

*

debugger invoked on a SB-IMPL::SAVE-WITH-MULTIPLE-THREADS-ERROR in thread

#<THREAD "main thread" RUNNING {10005285B3}>:

Cannot save core with multiple threads running.

Interactive thread (of current session):

#<THREAD "main thread" RUNNING {10005285B3}>

Other thread:

#<THREAD "finalizer" RUNNING {10016A8053}>

Ralf Hemmecke

unread,

Sep 17, 2019, 5:25:09 AM9/17/19

to FriCAS - computer algebra system

Interesting, now I seem to be in the same situation as Nasser.

I've created a fresh Debian 10 in Virtualbox and try to compile FriCAS from source.

It stops with

debugger invoked on a SB-IMPL::SAVE-WITH-MULTIPLE-

THREADS-ERROR in thread

This happens no matter whether I use the debian package for sbcl (1.4.16) or the latest (binary) 1.5.6 from sbcl.org.

Needless to say that it is not a problem with parallel builds, since my vbox has only one processor assigned.

Can someone successfully build FriCAS inside VirtualBox? How?

Thank you,

Ralf

Ralf Hemmecke

unread,

Sep 17, 2019, 6:01:26 AM9/17/19

to fricas...@googlegroups.com

It seems that this thread is relevant for FriCAS.

https://stackoverflow.com/questions/47888877/sbcl-building-a-standalone-executable

Maybe our build process has to be more careful of how to save the image.

Ralf

oldk1331

unread,

Sep 17, 2019, 6:11:54 AM9/17/19

to fricas...@googlegroups.com

On 9/17/19 6:01 PM, Ralf Hemmecke wrote:
> It seems that this thread is relevant for FriCAS.
>
> https://stackoverflow.com/questions/47888877/sbcl-building-a-standalone-executable
>
> Maybe our build process has to be more careful of how to save the image.
>
> Ralf
>

In your link it shows
Other threads:
#<THREAD "Swank Sentinel" RUNNING {100329E073}>,......
So it is running in SLIME (inside emacs).

From Nasser's original post, it shows

Other thread:
#<THREAD "finalizer" RUNNING {10016A8053}>

Which is strange and should be a different matter than SLIME.

Hi Ralf, what does your error log shows?

Somehow I wonder this bug is related with VirtualBox?

Ralf Hemmecke

unread,

Sep 17, 2019, 7:37:05 AM9/17/19

to fricas...@googlegroups.com

Hi Qian,

it's as unreliable as before. It seems that it depends on the time on my
virtualbox. When I try to compile now, it goes over that problematic step.

The last view lines from a failing build look like this.

======================
preloading
/home/hemmecke/scratch/build/fricas/target/x86_64-linux-gnu/algebra/NIL.fasl..skipped.
preloading
/home/hemmecke/scratch/build/fricas/target/x86_64-linux-gnu/algebra/NIL.fasl..skipped.

T

*
debugger invoked on a SB-IMPL::SAVE-WITH-MULTIPLE-THREADS-ERROR in thread

#<THREAD "main thread" RUNNING {10004F84C3}>:

Cannot save core with multiple threads running.

Interactive thread (of current session):

#<THREAD "main thread" RUNNING {10004F84C3}>

Other thread:
#<THREAD "finalizer" RUNNING {1001640053}>

See also:
The SBCL Manual, Node "Saving a Core Image"

Type HELP for debugger help, or (SB-EXT:EXIT) to exit from SBCL.

restarts (invokable by number or by possibly-abbreviated name):
0: [ABORT] Exit debugger, returning to top level.

(SB-IMPL::DEINIT)

0] NIL
*
(quit)
ls /home/hemmecke/scratch/build/fricas/build/x86_64-linux-gnu/bin/interpsys
ls: cannot access
'/home/hemmecke/scratch/build/fricas/build/x86_64-linux-gnu/bin/interpsys':

No such file or directory
make[2]: *** [Makefile:327:

/home/hemmecke/scratch/build/fricas/build/x86_64-linux-gnu/bin/interpsys] Error
2
make[2]: Leaving directory '/home/hemmecke/scratch/build/fricas/src/interp'

make[1]: *** [Makefile:229: all-interpsys] Error 2

make: *** [Makefile:246: all-src] Error 2

make[1]: Leaving directory '/home/hemmecke/scratch/build/fricas/src'

real 61m31.510s
user 0m0.172s
sys 0m0.019s
deb10:~/v/git/fricas-bisect:master %>
======================

Well, while I was writing that mail, a fresh the compilation was running
and that was indeed stepping over the above problem. However, the same
problem occured then at a later time. :-(

I've temporarily uploaded the log here:

https://www.risc.jku.at/people/hemmecke/tmp/make.log.gz

You have to search in the file for "THREAD", because when I was put into
the sbcl debugger, I've simply typed "(quit)" and strangely enough, the
compilation went on.

I've used the attached script to build fricas (build-fricas.sh).

======================

I also thought that it might be connected to VirtualBox, but I remember
having seen this error also outside a VM.

I also thought it might be connected to "--dynamic-space-size", because
my call is:

$ /home/hemmecke/g/fricas/configure --prefix= --with-lisp=sbcl
--dynamic-space-size 2643 --prefix=/home/hemmecke/software

and I gave the VM 4GB of RAM.

I'd be happy if that THREAD problem can be solve once and for all.

Qian, if you want, I can make the VM image available to you.

All the best
Ralf

build-fricas.sh

oldk1331

unread,

Sep 18, 2019, 1:27:12 AM9/18/19

to fricas...@googlegroups.com

On 9/17/19 7:37 PM, Ralf Hemmecke wrote:
> Hi Qian,
>
> it's as unreliable as before. It seems that it depends on the time on my
> virtualbox. When I try to compile now, it goes over that problematic step.
>

I tried locally on my machine, here are my findings:

1. This bug is not related with Linux distros.
2. This bug is not related with virtualbox, it also happens on QEMU.
3. This bug happens when the virtual machine has only 1 CPU.
(I suspect it also happens on real 1 core machine.)
4. This bug happens since SBCL 1.4.14. Older versions are fine.

So something related with threads in SBCL during 1.4.13~1.4.14
got wrong.

Ralf Hemmecke

unread,

Sep 18, 2019, 2:50:31 AM9/18/19

to fricas...@googlegroups.com

Hi Qian,

you are a hero. THANK YOU!!!

I've now increased the number of processors of the VM to 4 and the
compilation succeeded.

That doesn't, of course, deal with the actual bug, but as you suggested,
the best we can do is to report that problem to the sbcl developers.

===============

Cannot save core with multiple threads running.

Interactive thread (of current session):
#<THREAD "main thread" RUNNING {10004F84C3}>

Other thread:
#<THREAD "finalizer" RUNNING {1001640053}>

===============

I wonder at which point these two threads are created. I cannot imagine
that FriCAS code is doing it, because I don't see "thread" in the code.

Ralf

oldk1331

unread,

Sep 18, 2019, 2:57:19 AM9/18/19

to fricas...@googlegroups.com

On 9/18/19 2:50 PM, Ralf Hemmecke wrote:
> Hi Qian,
>

> you are a hero. THANK YOU!!!
>
> I've now increased the number of processors of the VM to 4 and the
> compilation succeeded.
>
> That doesn't, of course, deal with the actual bug, but as you suggested,
> the best we can do is to report that problem to the sbcl developers.

I did a git bisect, this bug is caused by sbcl commit
c63557933adae1f6a46250056a8ba53bfb7ead02, I think this part
is relevant:

"Also move TUNE-IMAGE-FOR-DUMP into DEINIT so that any
compiling caused by user code as a consequence of
calling *save-hooks* is done first."

See
https://github.com/sbcl/sbcl/commit/c63557933adae1f6a46250056a8ba53bfb7ead02
they move "(tune-image-for-dump)" inside "(deinit)",
which is after "#+sb-thread (finalizer-thread-stop)".

If we move "(tune-image-for-dump)" before
"#+sb-thread (finalizer-thread-stop)", then there is no problem.

This is very strange after all, and I will report to sbcl
developers soon.

oldk1331

unread,

Sep 19, 2019, 7:34:07 AM9/19/19

to fricas...@googlegroups.com

Fixed in sbcl master now.

Reply all

Reply to author

Forward