Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 27 Feb 2001 17:21:30 +0900
From:      Seigo Tanimura <tanimura@r.dl.itc.u-tokyo.ac.jp>
To:        jhb@FreeBSD.org
Cc:        tanimura@FreeBSD.org, cvs-all@FreeBSD.org, cvs-committers@FreeBSD.org
Subject:   RE: cvs commit: src/sys/dev/sound/isa emu8000.c gusmidi.c mpu.c
Message-ID:  <vmitlwfsdx.wl@rina.r.dl.itc.u-tokyo.ac.jp>
In-Reply-To: In your message of "Mon, 26 Feb 2001 10:50:52 -0800 (PST)" <XFMail.010226105052.jhb@FreeBSD.org>
References:  <200102260736.f1Q7aPd19490@freefall.freebsd.org> <XFMail.010226105052.jhb@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 26 Feb 2001 10:50:52 -0800 (PST),
  John Baldwin <jhb@FreeBSD.org> said:

>> - Mutexify midi(4). The driver runs under the giant lock by default.
>> If you ever want to run midi(4) out of the giant lock, uncomment
>> MIDI_OUTOFGIANT in midi.h. Confirmed to work for csamidi with WITNESS
>> and INVARIANTS.

John> Err, I hope that you do all mutex operations regardless, and only use this

I will do that in my next commit.

John> option to control setting the INTR_MPSAFE flag for any interrupt handlers that
John> midi uses.

Since some interrupt handlers are shared with pcm(4), we have to wait
for pcm(4) to be MP-safe.

A couple of questions to ask:

A. As we might hold a mutex of VM and block, malloc(9) should be
called with holding no mutexes except for Giant. Is that right?

B. Is PI_AV for a multimedia device? If so, I would like to add
INTR_TYPE_AV to enum intr_type and ithread_priority() so that pcm and
midi devices can interrupt at PI_AV.

-- 
Seigo Tanimura <tanimura@r.dl.itc.u-tokyo.ac.jp> <tanimura@FreeBSD.org>

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?vmitlwfsdx.wl>