Re: GUILE 2.2 progress

[David Kastrup]

Han-Wen Nienhuys <address@hidden> writes:

> On Sat, Jan 25, 2020 at 3:50 PM David Kastrup <address@hidden> wrote:
>> > That patch is papering over a problem rather than fixing it.  It might
>> > be a reoccurence of something like
>> >
>> > commit c01a2a757e3c59727bdfa8d77568bf42525fbe05
>> > Author: Andy Wingo <address@hidden>
>> > Date:   Thu Jun 23 11:47:42 2016 +0200
>> >
>> >     Fix race between SMOB marking and finalization
>> >
>> >     * libguile/smob.c (clear_smobnum): New helper.
>> >       (finalize_smob): Re-set the smobnum to the "finalized smob" type
>> >       before finalizing.  Fixes #19883.
>> >       (scm_smob_prehistory): Pre-register a "finalized smob" type, which 
>> > has
>> >       no mark procedure.
>> >     * test-suite/standalone/test-smob-mark-race.c: New file.
>> >     * test-suite/standalone/Makefile.am: Arrange to build and run the new
>> >       test.
>>
>> Actually, you may also be missing
>> commit c9910c604279f438728cd268272e1839cbc53835
>> Author: Andy Wingo <address@hidden>
>> Date:   Mon Mar 13 15:47:51 2017 +0100
>
> It must have been another one, v2.2.6 doesn't have the problem anymore, but
>
> [hanwen@localhost guile]$ git describe --contains
> c9910c604279f438728cd268272e1839cbc53835
> v2.2.0~11
> [hanwen@localhost guile]$ git describe --contains
> c01a2a757e3c59727bdfa8d77568bf42525fbe05
> v2.1.4~111

Ok, sorry.  So they managed to get and fix more problems.  Basically
your patch replaces malloc/free with a version that does not bother
complaining about multiple freeing.  But I cannot vouch for our
destructors being executable multiple times, so that "fix" is really
asking for trouble.

> I still get
>
> input/regression/mozart-hrn-3.ly:31:9: error: not a markup
>
>         #(string-append  "It has been typeset and placed in the public "

Frankly, that one looks like a nightmare that should not happen.  Does
it say the same for, say, #"just a string" ?  And "just a string"
(without hash mark)?

> and
>
> $ gs -q -dSAFER -dDEVICEWIDTHPOINTS=595.28 -dDEVICEHEIGHTPOINTS=841.89
> -dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH -r1200 -sDEVICE=pdfwrite
> -dAutoRotatePages=/None -dPrinted=false -sOutputFile=mozart-hrn-3.pdf
> -c.setpdfwrite -f/tmp/lilypond-OCyOzh
> Error: /rangecheck in /--.parsecff--
> Operand stack:
>    false   --nostringval--

Pretty sure this is an encoding problem.

>
> those look fixable, though.

-- 
David Kastrup