Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 11 Aug 2004 15:01:23 -0400
From:      John Baldwin <jhb@FreeBSD.org>
To:        freebsd-current@FreeBSD.org
Cc:        Nate Lawson <nate@root.org>
Subject:   Re: cvs commit: src/sys/dev/acpica acpi_pci_link.c acpi_pcib.c acpi_pcib_acpi.c acpi_pcib_pci.c acpi_pcibvar.h
Message-ID:  <200408111501.23593.jhb@FreeBSD.org>
In-Reply-To: <411A5A12.2070404@root.org>
References:  <!~!UENERkVCMDkAAQACAAAAAAAAAAAAAAAAABgAAAAAAAAA0VcX9IoJqUaXPS8MjT1PdsKAAAAQAAAAThZ4MpwZjU+KndAha//XewEAAAAA@telia.com> <411A5A12.2070404@root.org>

Next in thread | Previous in thread | Raw E-Mail | Index | Archive | Help
On Wednesday 11 August 2004 01:40 pm, Nate Lawson wrote:
> Daniel Eriksson wrote:
> > Nate Lawson wrote:
> >>  Modified files:
> >>    sys/dev/acpica       acpi_pci_link.c acpi_pcib.c
> >>                         acpi_pcib_acpi.c acpi_pcib_pci.c
> >>                         acpi_pcibvar.h
> >>  Log:
> >>  Re-work ACPI PCI IRQ routing (_PRT, link devices).  The old
> >>approach was
> >>  incomplete in that the PRT routing was not aware of link
> >>programming.
> >>  Fix this by doing all routing through the link devices.
> >
> > This(?) breaks interrupt routing pretty badly on ASUS A7V600-X (VIA
> > KT-600 chipset).
> >
> > A kernel/userland from 2004.08.09.13.00.00 works just fine (sorry, no
> > dmesg available right now), but a kernel/userland from
> > 2004.08.11.15.00.00 breaks routing of interrupts to (at least) two
> > HighPoint RocketRAID 454 cards.
> >
> > Below is part of the dmesg from a boot with ACPI on. Does it matter that
> > "PnP OS installed" is turned off in the BIOS? I will run some more tests
> > later tonight and report the results here. This is just a FYI that
> > something might be broken...
>
> That may affect it.  Let me know.
>
> > atapci0: <HighPoint HPT374 (channel 0+1) UDMA133 controller> port
> > 0xb400-0xb4ff,0xb800-0xb803,0xd000-0xd007,0xd400-0xd403,0xd800-0xd807 irq
> > 11 atapci1: <HighPoint HPT374 (channel 2+3) UDMA133 controller> port
> > 0x9800-0x98ff,0xa000-0xa003,0xa400-0xa407,0xa800-0xa803,0xb000-0xb007 irq
> > 11 ahc0: <Adaptec 29160 Ultra160 SCSI adapter> port 0x9400-0x94ff mem
> > 0xed800000-0xed800fff irq 3 at device 12.0 on pci0
> > atapci2: <HighPoint HPT374 (channel 0+1) UDMA133 controller> port
> > 0x7800-0x78ff,0x8000-0x8003,0x8400-0x8407,0x8800-0x8803,0x9000-0x9007 irq
> > 10 atapci3: <HighPoint HPT374 (channel 2+3) UDMA133 controller> port
> > 0x6000-0x60ff,0x6400-0x6403,0x6800-0x6807,0x7000-0x7003,0x7400-0x7407 irq
> > 10 at device 14.1 on pci0
> > atapci4: <VIA 6420 SATA150 controller> port
> > 0x4000-0x40ff,0x4400-0x440f,0x4800-0x4803,0x5000-0x5007,0x5400-0x5403,0x5
> >800 -0x5807 irq 5 at device 15.0 on pci0
> > atapci5: <VIA 8237 UDMA133 controller> port
> > 0x3800-0x380f,0x376,0x170-0x177,0x3f6,0x1f0-0x1f7 irq 14 at device 15.1
> > on pci0
> > ata0: at 0x1f0 irq 14 on atapci5
> > ata1: at 0x170 irq 15 on atapci5
>
> I need the dmesg output from boot -v to see the link priority settings.

He's using an I/O APIC.  These are probably all entries that don't have a link 
device but just a hardwired global interrupt number.  Did you test that case?

-- 
John Baldwin <jhb@FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/



Want to link to this message? Use this URL: <http://docs.FreeBSD.org/cgi/mid.cgi?200408111501.23593.jhb>