atapi works w/o acpi, fails with

I'm using -current on my shiny new dell inspiron laptop, and I have the
following rather annoying symptom: when using the regular pcibios
configuration, my cd rom drives works without a hitch - but with acpi,
I get missed interrupts.

The most annoying thing is that nothing else seems to be interrupting
on the secondary channel. Any suggestions for what to do? enabling
PCI_{ADDR,BUS,INTR}_FIXUP doesn't seem to matter.

Here's the diff of the dmesgs:

--- dmesg.apm	2006-01-02 21:14:07.000000000 -0800
+++ dmesg.acpi	2006-01-03 02:47:52.000000000 -0800
 <at>  <at>  -3,101 +3,15  <at>  <at>  Copyright (c) 1996, 1997, 1998, 1999, 20
 Copyright (c) 1982, 1986, 1989, 1991, 1993
     The Regents of the University of California.  All rights reserved.

-NetBSD 3.99.15 (APMSHINY) #4: Mon Jan  2 21:09:49 PST 2006
-	dogcow <at> :/usr/obj/obj/usr/src/sys/arch/i386/compile/APMSHINY
+NetBSD 3.99.15 (SHINY) #24: Tue Jan  3 02:43:19 PST 2006
+	dogcow <at> shiny.babymeat.com:/usr/obj/obj/usr/src/sys/arch/i386/compile/SHINY
 total memory = 511 MB
 avail memory = 499 MB
 BIOS32 rev. 0 found at 0xffe90
-PCI BIOS rev. 2.1 found at 0xfbaae
-pcibios: config mechanism [1][x], special cycles [x][x], last bus 10
-PCI IRQ Routing Table rev. 1.0 found at 0xfb7c0, size 176 bytes (9 entries)
-PCI Interrupt Router at 000:31:0 (Intel 82371MX (MPIIX) Mobile PCI I/O IDE Xcelerator compatible)
-PIR Entry 0:
-	Bus: 0  Device: 3

Pentium M 730, possibly correct mV values

Here they are:

Index: est.c
===================================================================
RCS file: /cvsroot/src/sys/arch/i386/i386/est.c,v
retrieving revision 1.14
diff -u -p -r1.14 est.c
--- est.c       31 Dec 2005 17:55:55 -0000      1.14
+++ est.c       3 Jan 2006 14:57:46 -0000
 <at>  <at>  -262,6 +262,15  <at>  <at>  static const struct fq_info pentium_m_n7
        {  600,  988 }
 };

+/* Intel Pentium M processor 730 1.6 GHz */
+static const struct fq_info pentium_m_n730[] = {
+       { 1596, 1308 },
+       { 1330, 1260 },
+       { 1197, 1212 },
+       { 1064, 1180 },
+       {  798,  988 }
+}; 
+
 /* Intel Pentium M processor 735 1.7 GHz */
 static const struct fq_info pentium_m_n735[] = {
        { 1700, 1340 },
 <at>  <at>  -347,7 +356,10  <at>  <at>  static const struct fqlist pentium_m_dot
        ENTRY("1.50", pentium_m_n758),
 #endif
        ENTRY("1.50", pentium_m_n715),
+#if 0

cleandir parallel make race

Cleandir fails for me with current with

--- cleandir-dosboot ---
getcwd() failed: No such file or directory

Essentially, this is because the Makefiles of dosboot, netboot and
pxeboot all include Makefile.booters. Makefile.booters includes
the makefiles for libkern, libz, libsa and libi386, which means
that those last 4 directories get cleaned 3 times (by dosboot,
netboot, pxeboot). When doing a parallel build who knows who will
get there first..

I'm not sure what the best fix is.. .WAIT would mean that building
couldn't happen in parallel, but I think it can, it is just the
cleaning which can't, or rather, the error which should be ignored..

Thoughts?

Cheers,

Patrick

Re: Portege 3000 series IDE support messed up?

On Thu, Dec 29, 2005 at 10:05:44AM +1100, James Haggerty wrote:
> Hi,
> 
> I'm trying to install NetBSD on my Portege 3020ct because I was under the
> impression it was the only (non-Windows) OS to support its weird IDE
> controller. However, as soon as I enable it in the kernel (uncommenting
> the 'pciide* at pnpbios? index ?' and the 'pnpbios* at mainbus?' lines),
> it fails to boot despite detecting all the hardware correctly:
> 
> pciide0 at pnpbios0 index 12 (TOS7300)
> pciide0: io 1f0-1f7 3f6 1800-180f, irq 14
> pciide0: Toshiba Extended IDE Controller
> atabus0 at pciide0 channel 0
> ...
> wd0 at atabus0 drive 0: <TOSHIBA MK6411MAT>
> wd0: drive supports 16-sector PIO transfers, LBA addressing
> wd0: 6194 MB, 13424 cyl, 15 head, 63 sec, 512 bytes/sect x 12685680 sectors
> wd0: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 2 (Ultra/33)
> uvm_fault(0xc03462e0, 0, 0, 1) -> 0xe
> fatal page fault in supervisor mode
> trap type 6 code 0 eip c0295652 cs 8 eflags 10282 cr2 18 ilevel 6
> panic: trap
> syncing disks... done
> rebooting...
> 
> Any help would be much appreciated - I'm quite possibly doing something
> silly, since this is the first time I've tried to compile a NetBSD kernel
> (I can of course get it running if I don't uncomment that line, but just
> with the standard wdc0 ISA controller...)

Re: Portege 3000 series IDE support messed up?

Quoting Manuel Bouyer <bouyer <at> antioche.eu.org>:

> > Any help would be much appreciated - I'm quite possibly doing
> something
> > silly, since this is the first time I've tried to compile a NetBSD
> kernel
> > (I can of course get it running if I don't uncomment that line, but
> just
> > with the standard wdc0 ISA controller...)
>
> See http://www.netbsd.org/cgi-bin/query-pr-single.pl?number=23192 for
> details. I didn't have time to look at a proper fix for this yet.
>

Hmm, as I said in a follow up email, the patch there looks out of date:
current NetBSD code is substantially different, and there appears to be no
equivalent to 'WDC_CAPABILITY_MODE'. Or I just don't know what it is.

Thanks,

James.

----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

Re: atapi works w/o acpi, fails with

On 2006.01.03 03:01:15 -0800, Tom Spindler wrote:
| I'm using -current on my shiny new dell inspiron laptop, and I have the
| following rather annoying symptom: when using the regular pcibios
| configuration, my cd rom drives works without a hitch - but with acpi,
| I get missed interrupts.

Are you using ioapic?

		-- Rui Paulo

Re: atapi works w/o acpi, fails with

I had the same problem when i activated 

options        ACPI_ACTIVATE_DEV

on the kernel, try to desactivate this option and the cdrom should work,
i don't know yet how to make wireless + cdrom working, because if you desactivate this option wireless card
only work on boot time, after some time, it stop to respond, no off no on...

Daniel Bento

--

-- 
// Daniel Bento <danielbento <at> overdestiny.com> http://www.overdestiny.com
// Grupo Astronomia-PT <astronomia.pt <at> gmail.com> http://vibrosoft.org/astronomia-pt
Location  <at>  39.1670° N, -8.8830° E Alt 0 m
-----BEGIN GEEK CODE BLOCK-----
Version: 3.12
GM d- s: a--- C+++ UL++ P+ L++ E--- W++ N+ o+ K- w 
O-- M+ V- PS+++ PE-- Y PGP t 5 X R- tv- b+++ DI- D 
G e- h- !r y? 
------END GEEK CODE BLOCK------

Re: Looping "NetBSD MBR boot"

On Mon, Dec 26, 2005 at 07:22:10PM +0900, Curt Sampson wrote:
> I've got a couple of IBM Intellistation M Pro machines, running NetBSD
> 3.0, one with a pair of SCSI drives, one with a pair of IDE drives. Both
> have run into a strange problem where on boot they just print ""NetBSD
> MBR boot" over and over again. Re-running installboot(8) seems to fix
> the problem.

This is a known problem [1] with disklabel, given a blank drive it has
a habit of copying sector 0 to the first sector of the partition.

This is fixed in current, but disklabel needed a rototil to fix
the problem properly.

2.0 had the same problem - but didn't have the message from the
boot code, so the machine appeared to lock up.

	David

[1] I can't remember the PR number...

--

-- 
David Laight: david <at> l8s.co.uk

Re: Portege 3000 series IDE support messed up?

On Wed, Jan 04, 2006 at 10:00:11AM +1100, James Haggerty wrote:
> > See http://www.netbsd.org/cgi-bin/query-pr-single.pl?number=23192 for
> > details. I didn't have time to look at a proper fix for this yet.
> >
> 
> Hmm, as I said in a follow up email, the patch there looks out of date:
> current NetBSD code is substantially different, and there appears to be no
> equivalent to 'WDC_CAPABILITY_MODE'. Or I just don't know what it is.

Ha yes, lots of things have changed since then. Can you try the attached
patch ? It should definitively fix the issue, but I've no way to test it.

--

-- 
Manuel Bouyer <bouyer <at> antioche.eu.org>
     NetBSD: 26 ans d'experience feront toujours la difference
--

Index: pciide_pnpbios.c
===================================================================
RCS file: /cvsroot/src/sys/arch/i386/pnpbios/pciide_pnpbios.c,v
retrieving revision 1.20
diff -u -r1.20 pciide_pnpbios.c
--- pciide_pnpbios.c	11 Dec 2005 12:17:47 -0000	1.20
+++ pciide_pnpbios.c	4 Jan 2006 22:05:52 -0000
 <at>  <at>  -54,7 +54,6  <at>  <at> 

 static int	pciide_pnpbios_match(struct device *, struct cfdata *, void *);
 static void	pciide_pnpbios_attach(struct device *, struct device *, void *);
-void		pciide_pnpbios_setup_channel(struct ata_channel *);

Re: atapi works w/o acpi, fails with

> I had the same problem when i activated 
> 
> options        ACPI_ACTIVATE_DEV
> 
> on the kernel, try to desactivate this option and the cdrom should work,
> i don't know yet how to make wireless + cdrom working, because if you desactivate this option wireless card
only work on boot time, after some time, it stop to respond, no off no on...

Yup, this was precisely the problem; disabling that and adding
PCI_{ADDR,BUS,INTR}_FIXUP makes everything work properly for me.