Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 23 May 2001 13:26:46 -0600
From:      Warner Losh <imp@harmony.village.org>
To:        John Reynolds~ <jreynold@sedona.ch.intel.com>
Cc:        mobile@FreeBSD.ORG
Subject:   Re: TESTIT: Patch to turn pcic_pci into real device 
Message-ID:  <200105231926.f4NJQkE49174@harmony.village.org>
In-Reply-To: Your message of "Wed, 23 May 2001 08:53:04 PDT." <15115.56544.943533.1140@hip186.ch.intel.com> 
References:  <15115.56544.943533.1140@hip186.ch.intel.com>  <200105230335.f4N3ZeE44422@harmony.village.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
In message <15115.56544.943533.1140@hip186.ch.intel.com> John Reynolds~ writes:
: I'll give it a shot. Sorry to ask this if it has been brought up before, but
: what will this patch do? Is it a pre-cursor to something else big or just a
: 'robustness' fix to current code?

Right now we put all PCI chipsets into a legacy mode where they
respond to the old 0x3e0 address and then let the ISA probe do its
magic.  This is broken in several ways, but works well enough to allow
almost all laptops to work correctly.

These patches change that.  They make the pcic driver attach to the
pci bus and talk to the chips via the memory mapped registers rather
than the I/O ports on the ISA bus.  This will make a marginal speed
improvement, but the massive delays in the code likely swamp any speed
increase that you'll see.

There are two benefits from this not completely in the patches yet.
First, is that you can write an interrupt handler that will reset the
itnerrupt assertion at the pci level.  This means potentially fewer
interrupt storms.  It also means that we can use pci cards as well,
but those changes aren't in these patches.

Warner

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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200105231926.f4NJQkE49174>