Gerd Hoffmann <kra...@redhat.com> writes: > On Mon, May 18, 2020 at 06:59:01AM +0200, Markus Armbruster wrote: >> Gerd Hoffmann <kra...@redhat.com> writes: >> >> > On Fri, May 15, 2020 at 09:06:20AM +0200, Markus Armbruster wrote: >> >> Watch this: >> >> >> >> $ aarch64-softmmu/qemu-system-aarch64 -S -nodefaults -accel qtest >> >> -display none -M vexpress-a15 -audiodev none,id=foo >> >> audio: Device lm4549: audiodev default parameter is deprecated, >> >> please specify audiodev=foo >> >> >> >> I did, didn't I? >> > >> > https://patchwork.ozlabs.org/project/qemu-devel/patch/20200429110214.29037-13-kra...@redhat.com/ >> >> I don't immediately understand how that addresses my issue, but I trust >> it does :) > > Well, not the patch itself, it handles pcspk only. But we could use the > same naming convention for all other onboard audio devices too: Have > them try lookup a "onboard" audio backend and use that if it exists. > Don't initialize if not.
When an onboard device is an integral part of the physical board we're emulating, then it should be connected to a suitable null backend when we don't have a better default. > Not fully sure this is a good plan, thats why the patch is tagged 'RfC'. > Better idea are welcome. In the blockdev world floppy has a similar > problem, is there any way to link floppy and blockdev without -global > isa-fdc.driveA=... meanwhile? Yes, but it won't help with onboard sound: -device floppy just works now. The general problem of configuring onboard devices still needs a solution. This recent thread has some ideas: Subject: Configuring onboard devices (was: Failing property setters + hardwired devices + -global = a bad day) Date: Thu, 30 Apr 2020 12:03:12 +0200 (2 weeks, 3 days, 22 hours ago) Message-ID: <87tv11e1en.fsf...@dusky.pond.sub.org> https://lists.nongnu.org/archive/html/qemu-devel/2020-04/msg05468.html
