[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Problems compiling Guile for Solaris 11.4 SPARC
From: |
Maxime Devos |
Subject: |
Re: Problems compiling Guile for Solaris 11.4 SPARC |
Date: |
Tue, 30 May 2023 00:28:39 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.10.1 |
Op 23-05-2023 om 15:28 schreef Olsson John via General Guile related
discussions:
Hi!
I'm trying to compile Guile 3.0.8 for Solaris 11.4 and I get a core dump for
STAGE0. The output I get is this (paths to build folder has been adjusted to
hide my username)
Making all in stage0
make[2]: Entering directory 'guile-3.0.8/stage0'
GUILE_BOOTSTRAP_STAGE=stage0 ../meta/build-env guild compile --target="sparc-sun-solaris2.11" -W0 -O1 -L
"guile-3.0.8/module" -o "ice-9/eval.go" "../module/ice-9/eval.scm"
Pre-boot error; key: (#f "Wrong type to apply: ~S" (#f) (#f) (#f))
make[2]: Leaving directory 'guile-3.0.8/stage0'
make[1]: *** [Makefile:2061: all-recursive] Error 1
make[1]: Leaving directory 'guile-3.0.8'
make: *** [Makefile:1947: all] Error 2
Any ideas on what is going wrong?
>
$ uname -a
SunOS fnord 5.11 11.4.56.138.2 sun4v sparc sun4v non-global-zone
1. Guile 3.0.8 is out-of-date, best update to 3.0.9.
2. Sparc support was removed in:
commit 570f361c6c8308ddc574d6308b50263175cae59e
Author: Andy Wingo <wingo@igalia.com>
Date: Wed May 8 21:41:46 2019 +0200
Remove Sparc support
Sadly, this is a dead architecture, without an official Debian port.
Rest in peace!
As you just demonstrated, ‘Sparc is dead’ is false, so this should
be reverted. (That's only the JIT though (not the (p-code)
interpreter), so that won't solve your problem; I recommend delaying
this until later.)
3. Unfortunately no backtrace is available.
Perhaps you could run
GUILE_BOOTSTRAP_STAGE=stage0 ../meta/build-env guild compile
--target="sparc-sun-solaris2.11" -W0 -O1 -L "guile-3.0.8/module" -o
"ice-9/eval.go" "../module/ice-9/eval.scm"
under the gdb or the like (with a break on scm_throw) to
see if the backtrace in gdb is informative.
That's the C backtrace, not the Scheme backtrace, so it might be
rather unclear. To avoid this, you could modify scm_throw to
run scm_backtrace() after
scm_puts ("Pre-boot error; key: ", port);
scm_write (key, port);
scm_puts (", args: ", port);
scm_write (args, port);
. I don't know if scm_backtrace is functional in pre-boot, though.
Best regards,
Maxime Devos.
OpenPGP_0x49E3EE22191725EE.asc
Description: OpenPGP public key
OpenPGP_signature
Description: OpenPGP digital signature