pulseaudio server doesn't want to start: # pulseaudio --start E: [(null)] conf-parser.c: [/usr/local/etc/pulse/daemon.conf:19] Missing '='. Line 19 of said file: changequote(`[', `]')dnl Set up m4 quoting There are plenty of these in all configuration files. Note: I am running on FreeBSD/arm.
Seems like an arm-specific bug since there is no problem on my amd64 box.
Is this still relevant.
(In reply to w.schwarzenfeld from comment #2) Yes.
A commit references this bug: Author: tobik Date: Fri Mar 16 11:16:18 UTC 2018 New revision: 464674 URL: https://svnweb.freebsd.org/changeset/ports/464674 Log: audio/pulseaudio-module-sndio: Unbreak on armv6/armv7 `pulseaudio --version` fails to run on armv6/armv7 because etc/pulse/daemon.conf generation fails and still has m4 macros embedded in it. Get the PulseAudio version from pkg-config instead. PR: 208535 Reported by: linimon Changes: head/audio/pulseaudio-module-sndio/Makefile head/audio/pulseaudio-module-sndio/files/patch-Makefile
Forget to close?
(In reply to w.schwarzenfeld from comment #5) Nope. The commit was for pulseaudio-module-sndio not pulseaudio. Last time I checked this problem still exists.
Created attachment 215873 [details] proposed patch I just ran into what I think is basically the same issue with the PulseAudioConfig.cmake building x11-wm/lxqt-panel on aarch64. I think this patch fixes both issues.
The configure script looks for gm4, m4, and assigns the first one found to $M4 .. and then doesn't use that variable when invoking m4. So this fix looks good to me, although it might be even better if it used $M4 instead of hard-coding gm4 there. Then it might be upstreamable. That said, I don't understand why this goes wrong: is there no m4 in base on arm64? Because that's the only way I can think of that this command goes wrong. The m4 constructs in the file are bog-standard and (amd64) base m4 can handle them no problem.
(In reply to Adriaan de Groot from comment #8) There is m4 in base. I think it's due to qemu-user-static magic. This patch is probably not really solving any underlying issue directly but rather avoiding it with code that happens to be correct.
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/ports/commit/?id=3c1f4f5c80627a368ea17fe0918e3623e1b10b84 commit 3c1f4f5c80627a368ea17fe0918e3623e1b10b84 Author: Adriaan de Groot <adridg@FreeBSD.org> AuthorDate: 2021-09-15 12:44:58 +0000 Commit: Adriaan de Groot <adridg@FreeBSD.org> CommitDate: 2021-09-15 15:43:24 +0000 audio/pulseaudio: fix build on ARMv6 in QEMU Switches to using GNU m4 (gm4) for processing PA configuration files; regular m4 behaves oddly in some build situations. PR: 208535 Submitted by: swills audio/pulseaudio/Makefile | 4 ++-- audio/pulseaudio/files/patch-configure (new) | 22 ++++++++++++++++++++++ 2 files changed, 24 insertions(+), 2 deletions(-)
I've applied this under the heading "might as well". Thanks swills.