Re: Removal of POWER4 support

[Benjamin Herrenschmidt]

On Sun, 2018-04-08 at 09:59 +0200, John Paul Adrian Glaubitz wrote:
> Hi Michael!
> 
> I was just pointed at your pull request to remove POWER4 support from the
> kernel and I'm stumped. POWER4 is something that people are still actively
> using.
> 
> I'm not aware of any breakage you are mentioning, we have multiple machines
> running 4.15 kernels on older PPC64 machines without any issues.
> 
> Please keep in mind that IBM isn't the sole manufacturer of POWER machines
> but there are also people running PPC64 Linux on embedded PPC64 machines
> with Freescale CPUs for example.

We aren't dropping support for any Freescale CPUs (BookE). Are you
really aware of anybody using a POWER4 though ? Those were huge
machines using a lot of electricity for little processing power by
today standards...

Note: We aren't removing 970 (aka G5) support, only server POWER4.

Cheers,
Ben.

Re: [PATCH v2 4/5] Patch the CHRP boot script

[Benjamin Herrenschmidt]

On Tue, 2017-11-21 at 09:56 +0100, Mathieu Malaterre wrote:
> Hi Ben,
> 
> Sorry for the direct question, but I assumed you would know the
> answer. We are currently trying to replace yaboot with grub-ieee1275
> in debian-installer, but are facing a very small issue, namely with OF
> path handling. It appears that on Apple hardware only `ofpath` from
> Yaboot returns the correct path. Are you using grub on any of your
> Apple hardware ? Thanks much for comments.

+Nathan

I have used grub on Apple systems in the past but it's quite possible
that I ended up fixing up the path by hand. To be honest I haven't
booted any of my Apple systems in a while, but I still have some I
can dig out to test if needed.

Some more comments below...

> See below for full context.
> 
> On Tue, Nov 21, 2017 at 7:51 AM, Frank Scheiner  wrote:
> > Hi Rick,
> > 
> > On 11/21/2017 02:48 AM, Rick Thomas wrote:
> > > 
> > > If you can tell me *exactly* what to do, and I don’t have to set up an
> > > installation environment to do it, I’ll be happy to test ofpathname vs
> > > ofpath vs devalias on my PowerPC test machine farm.
> > 
> > 
> > Well, to be on the safe side, I think you need to use the latest versions of
> > ofpath and ofpathname, hence running Debian Sid could be the easiest way to
> > make sure this is the case.
> > 
> > You need to have a disk installed, because ofpath and IIC also ofpathname do
> > not or cannot translate non existing device aliases to OF paths.
> > 
> > In general the commands I ran for the p5 520Q below should do. Assuming that
> > the output of ofpathname will always be wrong for Power Macs, it should be
> > sufficient to check only one partition.
> > 
> > I think devalias does only save/return OF paths up to the disk level, but
> > not up to the partition level. But the disk level should already be
> > sufficient to detect differences to ofpath and/or ofpathname.
> > 
> > You have to go to OF to run devalias, but with a glass console you won't be
> > able to copy the output. There are ways to interact with the OF via telnet
> > from another machine (see [1] and possibly [2]), but I haven't tried this
> > yet.
> > 
> > [1]:
> > https://web.archive.org/web/20040202053614/http://developer.apple.com/technotes/tn/tn2004.html
> > 
> > [2]:
> > https://web.archive.org/web/20040202060137/http://developer.apple.com:80/technotes/tn/tn2023.html
> > 
> > > 
> > > I have the following machines:
> > >  Power Mac G5 11,2,
> > >  Power Mac G5  7,2  (I think — it’s turned off right now)
> > >  Mac mini G4 10,1
> > 
> > 
> > I have these types available, too, "except" for the 7,3 which currently
> > won't start up correctly.
> > 
> > Here's the output of ofpath and ofpathname for /dev/sda2 on the my Mac mini
> > (10,1):
> > ```
> > root@mac-mini:~# ofpath /dev/sda2
> > /pci@f400/ata-6@d/@0:2
> > 
> > root@mac-mini:~# ofpathname /dev/sda2
> > /usr/sbin/ofpathname: line 812: warning: command substitution: ignored null
> > byte in input
> > /pci@f400/ata-6@d/scsi@0/sd@0,0

So here, ofpathname gets confused by the fact that Linux shows ATA
devices as scsi (which is a Linux'ism). ofpath has workarounds to deal
with that which should probably be ported over.

> > 
> > Also interesting, it looks like ofpathname cannot even give a "correct"
> > result for an IBM machine, at least not for my p5 520Q:
> > ```
> > root@p5-520q:~# lsblk
> > NAME   MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT
> > sda  8:00 136.7G  0 disk
> > ├─sda1   8:10 7M  0 part
> > ├─sda2   8:20 131.1G  0 part
> > ├─sda3   8:30 1K  0 part
> > └─sda5   8:50   5.6G  0 part
> > sr0 11:01  1024M  0 rom
> > 
> > root@p5-520q:~# ofpath /dev/sda1
> > /pci@8002003/pci@2,4/pci1069,b166@1/scsi@0/@3:1
> > 
> > root@p5-520q:~# ofpathname /dev/sda1
> > /pci@8002003/pci@2,4/pci1069,b166@1/scsi@0/sd@3,0
> > 
> > root@p5-520q:~# ofpath /dev/sda2
> > /pci@8002003/pci@2,4/pci1069,b166@1/scsi@0/@3:2
> > 
> > root@p5-520q:~# ofpathname /dev/sda2
> > /pci@8002003/pci@2,4/pci1069,b166@1/scsi@0/sd@3,0
> > ```
> > 
> > Please notice that ofpathname returns the same result for two different
> > partitions. :-/

Ouch. I think ofpathname is still maintained by some of our IBM guys
isn't it ? Might be worth digging there (or fixing it up !).

Nathan, do you know anything about this ?

> > >  Mac G4 3,4
> > >  Mac G4 dual core 3,6
> > 
> > 
> > Results for these machines would be helpful. It should work to
> > install/upgrade to Debian Sid on one machine only and just move the disk to
> > the other machine later on for testing.

I can test if you are willing to fix the script :)

Cheers,
Ben.

Re: PowerPC agpmode issues

[Benjamin Herrenschmidt]

On Sat, 2016-10-08 at 12:05 -0700, Herminio Hernandez, Jr. wrote:
> Performed an apt dist-upgrade and rebooted with agpmode=4 and still
> the GPU is failing with errors. Should I proceed to place a bug
> report on this?

I wouldn't even try to be honest. The HW issues in Apple AGP
implementation pretty much guarantee that things will crash
with the recent DRIs. Back when AGP memory was just a single
chunk of memory we got it to limp along but since DRI2 it's
pretty much hopeless.

I would stick to PCI GART even if it sucks for performances.

Cheers,
Ben.

> Herminio
> 
> On Tue, Sep 20, 2016 at 11:20 PM, Herminio Hernandez, Jr.  ernande...@gmail.com> wrote:
> > I am still getting errors even with radeon.agpmode=4 is set
> > 
> > rican-linux@Debian-G5:~$ dmesg |grep -e radeom -e drm -e
> > radeon.agpmode
> > [    0.00] Kernel command line: root=UUID=aeca9a67-31d7-4c4b-
> > a0f8-4db328b33305  radeon.agpmode=4
> > [    9.734366] [drm] Initialized drm 1.1.0 20060810
> > [   10.934519] [drm] radeon kernel modesetting enabled.
> > [   10.943743] fb: switching to radeondrmfb from OFfb ATY,Simone
> > [   10.954622] fb: switching to radeondrmfb from OFfb ATY,Simone
> > [   10.956515] [drm] initializing kernel modesetting (RV350
> > 0x1002:0x4150 0x1002:0x4150 0x00).
> > [   10.956552] [drm] register mmio base: 0xA000
> > [   10.956558] [drm] register mmio size: 65536
> > [   11.046327] [drm] Not an x86 BIOS ROM, not using.
> > [   11.046358] [drm] Using device-tree clock info
> > [   11.046369] [drm] AGP mode requested: 4
> > [   11.046528] [drm] Generation 2 PCI interface, using max
> > accessible memory
> > [   11.046582] [drm] Detected VRAM RAM=256M, BAR=256M
> > [   11.046587] [drm] RAM width 128bits DDR
> > [   11.046858] [drm] radeon: 64M of VRAM memory ready
> > [   11.046866] [drm] radeon: 256M of GTT memory ready.
> > [   11.046935] [drm] radeon: 1 quad pipes, 1 Z pipes initialized.
> > [   11.061306] [drm] Supports vblank timestamp caching Rev 2
> > (21.10.2013).
> > [   11.061312] [drm] Driver supports precise vblank timestamp
> > query.
> > [   11.061363] [drm] radeon: irq initialized.
> > [   11.061405] [drm] Loading R300 Microcode
> > [   11.110467] [drm] radeon: ring at 0x0001
> > [   11.257952] [drm:.r100_ring_test [radeon]] *ERROR* radeon: ring
> > test failed (scratch(0x15E4)=0xCAFEDEAD)
> > [   11.258060] [drm:.r100_cp_init [radeon]] *ERROR* radeon: cp
> > isn't working (-22).
> > [   11.404548] [drm:.r100_cp_fini [radeon]] *ERROR* Wait for CP
> > idle timeout, shutting down CP.
> > [   11.551199] [drm] radeon: cp finalized
> > [   12.547388] [drm] radeon: cp finalized
> > [   12.553854] [drm] radeon: ttm finalized
> > [   12.553861] [drm] Forcing AGP to PCI mode
> > [   12.643384] [drm] Not an x86 BIOS ROM, not using.
> > [   12.643406] [drm] Using device-tree clock info
> > [   12.643414] [drm] Generation 2 PCI interface, using max
> > accessible memory
> > [   12.643446] [drm] Detected VRAM RAM=256M, BAR=256M
> > [   12.643452] [drm] RAM width 128bits DDR
> > [   12.643712] [drm] radeon: 64M of VRAM memory ready
> > [   12.643719] [drm] radeon: 512M of GTT memory ready.
> > [   12.643761] [drm] GART: num cpu pages 8192, num gpu pages 131072
> > [   12.645324] [drm] radeon: 1 quad pipes, 1 Z pipes initialized.
> > [   12.645343] [drm] PCI GART of 512M enabled (table at
> > 0x62F8).
> > [   12.645462] [drm] Supports vblank timestamp caching Rev 2
> > (21.10.2013).
> > [   12.645468] [drm] Driver supports precise vblank timestamp
> > query.
> > [   12.645512] [drm] radeon: irq initialized.
> > [   12.645689] [drm] radeon: ring at 0x9001
> > [   12.793131] [drm:.r100_ring_test [radeon]] *ERROR* radeon: ring
> > test failed (scratch(0x15E4)=0xCAFEDEAD)
> > [   12.793235] [drm:.r100_cp_init [radeon]] *ERROR* radeon: cp
> > isn't working (-22).
> > [   12.939961] [drm:.r100_cp_fini [radeon]] *ERROR* Wait for CP
> > idle timeout, shutting down CP.
> > [   12.940193] [drm] radeon: cp finalized
> > [   12.944581] [drm] Connector Table: 12 (mac g5 9600)
> > [   12.944609] [drm] No valid Ext TMDS info found in BIOS
> > [   12.944621] [drm] No TV DAC info found in BIOS
> > [   12.944798] [drm] No TMDS info found in BIOS
> > [   12.944974] [drm] Radeon Display Connectors
> > [   12.944981] [drm] Connector 0:
> > [   12.944985] [drm]   DVI-I-1
> > [   12.944990] [drm]   HPD1
> > [   12.944995] [drm]   DDC: 0x64 0x64 0x64 0x64 0x64 0x64 0x64 0x64
> > [   12.945000] [drm]   Encoders:
> > [   12.945005] [drm] DFP2: INTERNAL_DVO1
> > [   12.945010] [drm] CRT2: INTERNAL_DAC2
> > [   12.945015] [drm] Connector 1:
> > [   12.945019] [drm]   DVI-I-2
> > [   12.945023] [drm]   HPD2
> > [   12.945028] [drm]   DDC: 0x60 0x60 0x60 0x60 0x60 0x60 0x60 0x60
> > [   12.945033] [drm]   Encoders:
> > [   12.945037] [drm] DFP1: INTERNAL_TMDS1
> > [   12.945042] [drm] CRT1: INTERNAL_DAC1
> > [   12.945046] [drm] Connector 2:
> > [   12.945051] [drm]   SVIDEO-1
> > [  

Re: PowerPC agpmode issues

[Benjamin Herrenschmidt]

On Mon, 2016-09-19 at 02:13 -0700, Herminio Hernandez, Jr. wrote:
> 
> I noticed the GPU was set to PCI mode however I did not set
> radeon.agpmode=-1 in yaboot. I then installed Lubuntu 16.10 along
> side Debian. I noticed that GPU was working and it was not set to
> force PCI mode. Has there been a change to kernel in Debian to force
> this by default? If so it looks like it is breaking on certain G5
> machines. I am attaching bothe dmesg files for review.
> 
> Thanks and sorry for the previous garbled email.

Hrm... this is odd. I would expect PCI mode to work and AGP to be
busted to be honest. Can you provide the complete log ? I wonder
if it's trying to enable the PCIe bypass high aperture and hits the
Radeon address limit...

Ben.

Re: iBook G3 stopped booting after last testing update

[Benjamin Herrenschmidt]

On Wed, 2013-01-16 at 10:29 +0100, Michel Dänzer wrote:
> On Die, 2013-01-15 at 22:33 +0100, Martin Kuball wrote: 
> > Am Tuesday 15 January 2013 schrieb Michel Dänzer:
> > > On Mon, 2013-01-14 at 19:47 +0100, Martin Kuball wrote: 
> > > > 
> > > > After spending to much time figuring out what's going on, I have at
> > > > least a working system. The error is triggered when laptop-mode is
> > > > setting the auto negotiation attribute. So it may be a hardware error
> > > > after all. Because if I bring up the network manually (ifup -a) it
> > > > works.
> > > 
> > > It might still be a sungem driver bug, or the driver might at least be
> > > able to work around the hardware problem. Consider reporting a bug
> > > against the kernel, preferably upstream. 
> > 
> > Can you tell me how to do this or point me to a howto?
> 
> Actually, it looks like there is already a bug report for this,
> including a bisected commit introducing the problem:
> 
> https://bugzilla.kernel.org/show_bug.cgi?id=42992
> 
> 
> Ben, any ideas? 

Looks like it's accessing the chip as a result of an ioctl while the
chip is powered down. I'll have a look.

Ben.



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/1358376244.2782.14.camel@pasglop

Re: iMac-G3 fails to boot with 2.6.37-1-powerpc (Bug#614221)

[Benjamin Herrenschmidt]

On Tue, 2011-03-08 at 13:00 +0100, Michel Dänzer wrote:
> On Die, 2011-03-08 at 07:30 +1100, Benjamin Herrenschmidt wrote: 
> > On Mon, 2011-03-07 at 10:45 +0100, Michel Dänzer wrote:
> > > 
> > > I've been using radeon KMS on my PowerBook ever since I got it working
> > > initially about 1.5 years ago and fixing issues as time permits.
> > > 
> > > The major outstanding issues I'm aware of are:
> > > 
> > >   * The uninorth AGP driver doesn't allow AGP transfer rates beyond
> > > 1x to work reliably with KMS. Benjamin Herrenschmidt (CC'd) was
> > > working on a fix for this, any progress Ben? 
> > 
> > I have hacks.
> 
> FWIW, I've been using the hack at the end of this mail for AGP >1x.

 .../...

Ok I'll try to give it another go here one of these days, I think I
still have some of my old hacks too.

Cheers,
Ben.

> > But I never got it working reliably. In fact, on the laptop I have
> > here, even with PCI GART, it's still unstable if I use KMS/DRI2.
> 
> I know your PowerBook was affected by the problem fixed by commit
> b7d8cce5b558e0c0aa6898c9865356481598b46d ('drm/radeon/kms: MC vram map
> needs to be >= pci aperture size'), which went into 2.6.37. Still no
> better with that?
> 
> 
> > >   * Come to think of it, the OFfb handover probably only works for
> > > me thanks to a patch "drm/radeon: Add early unregister of
> > > firmware fb's" by BenH. Ben/Dave, what's the status of that? 
> > 
> > Dave ? Was this ever merged ?
> 
> Apparently it went into 2.6.37, I didn't notice because it went into
> different places in the code compared to the patch I had from you.
> 
> 
> > >   * Various endianness issues in the Mesa drivers.
> > 
> > Right, and the later aren't getting any better :-( It's busted even
> > without KMS nowadays.
> > 
> > Unfortunately, I have about 0 time to spend on that at the moment.
> > 
> > I recommend that distros stick to radeonfb + UMS for the time being on
> > ppc32.
> 
> One problem being that the classic Mesa r300 driver is essentially
> unmaintained, and evidently rotting
> 
> 
> commit 9a86d7fa5bb0b4fe228becf9ed9831bac985702c
> Author: Michel Dänzer 
> Date:   Thu Jan 6 18:34:28 2011 +0100
> 
> agp/uninorth: Fix lockups with radeon KMS and >1x.
> 
> diff --git a/drivers/char/agp/uninorth-agp.c b/drivers/char/agp/uninorth-agp.c
> index f845a8f..a32c492 100644
> --- a/drivers/char/agp/uninorth-agp.c
> +++ b/drivers/char/agp/uninorth-agp.c
> @@ -80,7 +80,7 @@ static void uninorth_tlbflush(struct agp_memory *mem)
>  ctrl | UNI_N_CFG_GART_INVAL);
>   pci_write_config_dword(agp_bridge->dev, UNI_N_CFG_GART_CTRL, ctrl);
>  
> - if (uninorth_rev <= 0x30) {
> + if (!mem && uninorth_rev <= 0x30) {
>   pci_write_config_dword(agp_bridge->dev, UNI_N_CFG_GART_CTRL,
>  ctrl | UNI_N_CFG_GART_2xRESET);
>   pci_write_config_dword(agp_bridge->dev, UNI_N_CFG_GART_CTRL,
> 
> 



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/1299623992.22236.40.camel@pasglop

Re: iMac-G3 fails to boot with 2.6.37-1-powerpc (Bug#614221)

[Benjamin Herrenschmidt]

On Mon, 2011-03-07 at 10:45 +0100, Michel Dänzer wrote:
> 
> I've been using radeon KMS on my PowerBook ever since I got it working
> initially about 1.5 years ago and fixing issues as time permits.
> 
> The major outstanding issues I'm aware of are:
> 
>   * The uninorth AGP driver doesn't allow AGP transfer rates beyond
> 1x to work reliably with KMS. Benjamin Herrenschmidt (CC'd) was
> working on a fix for this, any progress Ben? 

I have hacks. But I never got it working reliably. In fact, on the
laptop I have here, even with PCI GART, it's still unstable if I use
KMS/DRI2.

>   * Come to think of it, the OFfb handover probably only works for
> me thanks to a patch "drm/radeon: Add early unregister of
> firmware fb's" by BenH. Ben/Dave, what's the status of that? 

Dave ? Was this ever merged ?

>   * LVDS backlight control was only added for 2.6.38, and some
> concerns were raised about the prerequisite backlight
> infrastructure changes, so I'm not sure if it's staying in for
> 2.6.38 final. 
>   * No suspend to RAM support yet. Suspend to disk seems to work
> though without AGP, I have some hackish fixes for it to work
> with AGP as well that I need to clean up and submit. 
>   * Various endianness issues in the Mesa drivers.

Right, and the later aren't getting any better :-( It's busted even
without KMS nowadays.

Unfortunately, I have about 0 time to spend on that at the moment.

I recommend that distros stick to radeonfb + UMS for the time being on
ppc32.

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/1299529827.9759.19.camel@pasglop

Re: no sound on powermac3,3 (powerbook g4 550) debian502 gnome

[Benjamin Herrenschmidt]

On Sat, 2009-09-12 at 19:33 -0300, Otavio Salvador wrote:
> Do you know a way for us to know if aoa or powerpc is the right
> choice? Or the oposite, a way to know that aoa is known to be broken
> in a set of machine ids or like?
> 
Well, generally speaking, aoa only picks up machines it knows about no ?

Ben.



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: PowerPC 8500??

[Benjamin Herrenschmidt]

On Wed, 2009-09-09 at 05:52 -0400, Frank Rubinsky wrote:
> On Sep 9, 2009, at 2:09 AM, Brad Boyer wrote:
> 
> > I went and looked on the Apple support downloads web site (which is
> > at http://support.apple.com/downloads/) and saw a link to "Older
> > Software Downloads" (http://support.apple.com/kb/TA48312? 
> > viewlocale=en_US)
> > which appears to have page after page of old stuff, including System 6
> > and most of the versions of System 7 (looks like all but 7.6).
> 
> 
> Good catch - I missed that page, but I just looked closely at it, and  
> didn't see any base OS downloads.  Everything seems to be a point  
> release update/upgrade - which requires the major release - which  
> isn't on that site.   It looks like this page has everything Apple  
> released for free but nothing (OS at least) that was sold (or  
> included with new machines).

One option with those old machines is booting from OF. It's tricky,
requires a network with a BOOTP server and a lot of praying but ...

I have an old dual 604 SMP 8500 on my desk taking the dust :-) I've been
working on reviving it recently. There's various issues with recent
kernels in trying to boot them from OF, in large part due to bugs in
those old Apple OF.

If you don't find a solution involving MacOS or miboot, I'd suggest you
ping me every now and then to see whether I got a chance to sort out
some of those issues and get it back into a shape where it can be
netbooted with a debian install ramdisk entirely from OF.

Cheers,
Ben.


-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: no sound on powermac3,3 (powerbook g4 550) debian502 gnome

[Benjamin Herrenschmidt]

On Mon, 2009-09-07 at 12:12 -0300, Rogério Brito wrote:
> Right, but now the snd-powermac driver is what is being used instead
> of
> the snd-aoa* driver.
> 
> Perhaps BenH (one of the principal developers of Linux on PowerPC)
> could
> shed some comments here with the final word so that we can settle this
> snd-aoa vs. snd-powermac for the time being.
> 
Ouch :-)

I've been avoiding the sound stuff like a plague since the Apple stuff
is a total mess :-)

If aoa works generally better on this machine then the distro should use
it by default. I was somewhat hoping somebody would take over those
sound drivers once for all but they seem to systematically scare off all
the volunteers we can find :-)

Sadly, I don't have the time nowadays to do much work on them myself.

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

On Sat, 2009-06-27 at 19:46 +0200, Laszlo Fekete wrote:
> Hello!
> 
> Thank you very much, this patch works me too.
> 
> Maybe this patch will be in the debian kernel someday?

The patch isn't actually correct just yet :-) Michael will
be posting a proper one next week. It should be possible to
request its inclusion into debian separately, we'll probably
send it to sta...@kernel.org as well.

Cheers,
Ben.

> Thank you: blackluck
> 
> Michael Ellerman wrote: 
> > On Wed, 2009-06-24 at 15:56 +1000, Benjamin Herrenschmidt wrote:
> >   
> > > On Wed, 2009-06-24 at 15:53 +1000, Michael Ellerman wrote:
> > > 
> > > > Doesn't fix my machine :/
> > > > 
> > > >   
> > > That doesn't make sense ... What if you remove the bit inside the ifdef
> > > CONFIG_MPIC_BROKEN_REGREAD in _mpic_read() ?
> > > 
> > > If that makes a difference, then it would be interesting to add a printk
> > > in there that prints what the original value "val" is and what we have
> > > in the shadow...
> > > 
> > 
> > With this patch it boots:
> > 
> > diff --git a/arch/powerpc/sysdev/mpic.c b/arch/powerpc/sysdev/mpic.c
> > index 2353adc..fc17289 100644
> > --- a/arch/powerpc/sysdev/mpic.c
> > +++ b/arch/powerpc/sysdev/mpic.c
> > @@ -231,13 +231,16 @@ static inline u32 _mpic_irq_read(struct mpic *mpic, 
> > unsign
> > unsigned intisu = src_no >> mpic->isu_shift;
> > unsigned intidx = src_no & mpic->isu_mask;
> > unsigned intval;
> > +   unsigned intshadow;
> >  
> > val = _mpic_read(mpic->reg_type, &mpic->isus[isu],
> >  reg + (idx * MPIC_INFO(IRQ_STRIDE)));  
> >  #ifdef CONFIG_MPIC_BROKEN_REGREAD
> > -   if (reg == 0)
> > -   val = (val & (MPIC_VECPRI_MASK | MPIC_VECPRI_ACTIVITY)) |
> > +   if (reg == 0) {
> > +   shadow = (val & (MPIC_VECPRI_MASK | MPIC_VECPRI_ACTIVITY)) |
> > mpic->isu_reg0_shadow[idx];
> > +   printk("%s: val 0x%x shadow 0x%x\n", __func__, val, shadow);
> > +   }
> >  #endif
> > return val;
> >  }
> > 
> > 
> > And I see:
> > 
> > sym53c8xx :00:0c.0: enabling device (0140 -> 0143)
> > sym0: <896> rev 0x7 at pci :00:0c.0 irq 17
> > sym0: No NVRAM, ID 7, Fast-40, SE, parity checking
> > _mpic_irq_read: val 0x80480004 shadow 0x80080014
> > _mpic_irq_read: val 0x480004 shadow 0x480004
> > 
> > 
> > 
> > cheers
> >   


-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

On Wed, 2009-06-24 at 16:19 +1000, Michael Ellerman wrote:

> sym53c8xx :00:0c.0: enabling device (0140 -> 0143)
> sym0: <896> rev 0x7 at pci :00:0c.0 irq 17
> sym0: No NVRAM, ID 7, Fast-40, SE, parity checking
> _mpic_irq_read: val 0x80480004 shadow 0x80080014
> _mpic_irq_read: val 0x480004 shadow 0x480004

Strange.. it's like we read a different vector and polarity than
what is in the cache. Oh well, I'll have a closer look tomorrow
in the office.

Ben.



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

On Wed, 2009-06-24 at 15:53 +1000, Michael Ellerman wrote:
> Doesn't fix my machine :/
> 
That doesn't make sense ... What if you remove the bit inside the ifdef
CONFIG_MPIC_BROKEN_REGREAD in _mpic_read() ?

If that makes a difference, then it would be interesting to add a printk
in there that prints what the original value "val" is and what we have
in the shadow...

Cheers,
Ben.


-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

On Wed, 2009-06-24 at 07:57 +1000, Benjamin Herrenschmidt wrote:
> > I tried on a POWER3 box I have here "IBM,7044-170" and things work fine
> > here with current upstream. (I suspect a much smaller machine).
> > 
> > I will really need an actual bisection here... In the meantime, I'll see
> > if I can get my hand on one of these machines here.
> 
> Ok so I think we may have found it... looks like
> CONFIG_MPIC_BROKEN_REGREAD is what breaks it which is enabled by PA-Semi
> support in the .config.
> 
> Can you verify that disabling PA-Semi support removes that option from
> your .config and that once removed, it works again ?
> 
> We don't know yet -why- it breaks it, still investigating.

Do the following patch also fix it ?

Index: linux-work/arch/powerpc/sysdev/mpic.c
===
--- linux-work.orig/arch/powerpc/sysdev/mpic.c  2009-06-24 09:24:51.0 
+1000
+++ linux-work/arch/powerpc/sysdev/mpic.c   2009-06-24 09:26:45.0 
+1000
@@ -230,14 +230,16 @@ static inline u32 _mpic_irq_read(struct 
 {
unsigned intisu = src_no >> mpic->isu_shift;
unsigned intidx = src_no & mpic->isu_mask;
+   unsigned intval;
 
+   val = _mpic_read(mpic->reg_type, &mpic->isus[isu],
+reg + (idx * MPIC_INFO(IRQ_STRIDE)));  
 #ifdef CONFIG_MPIC_BROKEN_REGREAD
if (reg == 0)
-   return mpic->isu_reg0_shadow[idx];
-   else
+   val = (val & (MPIC_VECPRI_MASK | MPIC_VECPRI_ACTIVITY)) |
+   mpic->isu_reg0_shadow[idx];
 #endif
-   return _mpic_read(mpic->reg_type, &mpic->isus[isu],
- reg + (idx * MPIC_INFO(IRQ_STRIDE)));
+   return val;
 }
 
 static inline void _mpic_irq_write(struct mpic *mpic, unsigned int src_no,
@@ -251,7 +253,8 @@ static inline void _mpic_irq_write(struc
 
 #ifdef CONFIG_MPIC_BROKEN_REGREAD
if (reg == 0)
-   mpic->isu_reg0_shadow[idx] = value;
+   mpic->isu_reg0_shadow[idx] =
+   value & ~(MPIC_VECPRI_MASK | MPIC_VECPRI_ACTIVITY);
 #endif
 }
 



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

> I tried on a POWER3 box I have here "IBM,7044-170" and things work fine
> here with current upstream. (I suspect a much smaller machine).
> 
> I will really need an actual bisection here... In the meantime, I'll see
> if I can get my hand on one of these machines here.

Ok so I think we may have found it... looks like
CONFIG_MPIC_BROKEN_REGREAD is what breaks it which is enabled by PA-Semi
support in the .config.

Can you verify that disabling PA-Semi support removes that option from
your .config and that once removed, it works again ?

We don't know yet -why- it breaks it, still investigating.

Cheers,
Ben.

> Cheers,
> Ben.
> 
> > Cheers,
> > Ben.
> > 
> > > Thanks: blackluck
> > > 
> > > Laszlo Fekete wrote:
> > > > Hello!
> > > >
> > > > I'm sorry about the annoyances, but I'd welcome all ideas, suggestions 
> > > > to see what needs to be done or should be tested for the solution.
> > > >
> > > > Thank you very much!
> > > >
> > > > Guennadi Liakhovetski wrote:
> > > >> Ok, first attempt to forward this to scsi was wrong, as pointed out 
> > > >> by Matthew Wilcox this does indeed look like an interrupt problem - 
> > > >> no interrupts drom SCSI, IDE, keyboar. Might be a known problem, I 
> > > >> guess. In any case, I think, the OP would be grateful for any hints.
> > > >>
> > > >> Thanks
> > > >> Guennadi
> > > >> ---
> > > >> Guennadi Liakhovetski, Ph.D.
> > > >> Freelance Open-Source Software Developer
> > > >> http://www.open-technology.de/
> > > >>
> > > >> -- Forwarded message --
> > > >> Date: Sat, 13 Jun 2009 16:22:07 +0200
> > > >> From: Laszlo Fekete 
> > > >> To: debian-powerpc@lists.debian.org
> > > >> Subject: sym scsi driver problem with 2.6.26 or newer debian kernel 
> > > >> on p610
> > > >> Resent-Date: Sat, 13 Jun 2009 14:29:55 + (UTC)
> > > >> Resent-From: debian-powerpc@lists.debian.org
> > > >>
> > > >> This is a multi-part message in MIME format.
> > > >> 
> > > >>
> > > >> Hello!
> > > >>
> > > >>
> > > >>
> > > >>
> > > >>
> > > >> Pls help me with sym scsi driver problem.
> > > >>
> > > >>
> > > >>
> > > >> I have Ibm P610 (and tested it on P630 and P640 too), installed debian
> > > >>
> > > >> etch and upgraded to lenny.
> > > >>
> > > >>
> > > >>
> > > >> But with 2.6.26 or newer kernel it's not booting, it's hang on sym scsi
> > > >>
> > > >> bus scan.
> > > >>
> > > >>
> > > >>
> > > >>
> > > >>
> > > >> Whats the problem with it, or how can I fix this?
> > > >>
> > > >>
> > > >>
> > > >>
> > > >>
> > > >> I attach the output from minicom with 2.6.29, 2.6.26, and the working
> > > >>
> > > >> 2.6.24 kernel booting.
> > > >>
> > > >>
> > > >>
> > > >>
> > > >>
> > > >> Thank you very much!
> > > >>
> > > >>
> > > >>   
> > > >
> > > >
> > > 
> > > ___
> > > Linuxppc-dev mailing list
> > > linuxppc-...@lists.ozlabs.org
> > > https://lists.ozlabs.org/listinfo/linuxppc-dev
> > 
> > ___
> > Linuxppc-dev mailing list
> > linuxppc-...@lists.ozlabs.org
> > https://lists.ozlabs.org/listinfo/linuxppc-dev
> 
> ___
> Linuxppc-dev mailing list
> linuxppc-...@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev


-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

On Sun, 2009-06-14 at 11:10 +0200, Guennadi Liakhovetski wrote:
> P610 (and tested it on P630 and P640 too)

Also, send me a tarball of /proc/device-tree please.

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

On Tue, 2009-06-23 at 11:18 +1000, Benjamin Herrenschmidt wrote:
> On Mon, 2009-06-22 at 11:10 +0200, Laszlo Fekete wrote:
> > Hello!
> > 
> > Is there any idea for the solution?
> 
> Hard to tell yet. Looks indeed like something is wrong with the
> interrupt controller.
> 
> Any chance you can bisect that ? I'll also have a look on my side,
> it's definitely not something obvious.

I tried on a POWER3 box I have here "IBM,7044-170" and things work fine
here with current upstream. (I suspect a much smaller machine).

I will really need an actual bisection here... In the meantime, I'll see
if I can get my hand on one of these machines here.

Cheers,
Ben.

> Cheers,
> Ben.
> 
> > Thanks: blackluck
> > 
> > Laszlo Fekete wrote:
> > > Hello!
> > >
> > > I'm sorry about the annoyances, but I'd welcome all ideas, suggestions 
> > > to see what needs to be done or should be tested for the solution.
> > >
> > > Thank you very much!
> > >
> > > Guennadi Liakhovetski wrote:
> > >> Ok, first attempt to forward this to scsi was wrong, as pointed out 
> > >> by Matthew Wilcox this does indeed look like an interrupt problem - 
> > >> no interrupts drom SCSI, IDE, keyboar. Might be a known problem, I 
> > >> guess. In any case, I think, the OP would be grateful for any hints.
> > >>
> > >> Thanks
> > >> Guennadi
> > >> ---
> > >> Guennadi Liakhovetski, Ph.D.
> > >> Freelance Open-Source Software Developer
> > >> http://www.open-technology.de/
> > >>
> > >> -- Forwarded message --
> > >> Date: Sat, 13 Jun 2009 16:22:07 +0200
> > >> From: Laszlo Fekete 
> > >> To: debian-powerpc@lists.debian.org
> > >> Subject: sym scsi driver problem with 2.6.26 or newer debian kernel 
> > >> on p610
> > >> Resent-Date: Sat, 13 Jun 2009 14:29:55 + (UTC)
> > >> Resent-From: debian-powerpc@lists.debian.org
> > >>
> > >> This is a multi-part message in MIME format.
> > >> 
> > >>
> > >> Hello!
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> Pls help me with sym scsi driver problem.
> > >>
> > >>
> > >>
> > >> I have Ibm P610 (and tested it on P630 and P640 too), installed debian
> > >>
> > >> etch and upgraded to lenny.
> > >>
> > >>
> > >>
> > >> But with 2.6.26 or newer kernel it's not booting, it's hang on sym scsi
> > >>
> > >> bus scan.
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> Whats the problem with it, or how can I fix this?
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> I attach the output from minicom with 2.6.29, 2.6.26, and the working
> > >>
> > >> 2.6.24 kernel booting.
> > >>
> > >>
> > >>
> > >>
> > >>
> > >> Thank you very much!
> > >>
> > >>
> > >>   
> > >
> > >
> > 
> > ___
> > Linuxppc-dev mailing list
> > linuxppc-...@lists.ozlabs.org
> > https://lists.ozlabs.org/listinfo/linuxppc-dev
> 
> ___
> Linuxppc-dev mailing list
> linuxppc-...@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev


-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: sym scsi driver problem with 2.6.26 or newer debian kernel on p610 (fwd)

[Benjamin Herrenschmidt]

On Mon, 2009-06-22 at 11:10 +0200, Laszlo Fekete wrote:
> Hello!
> 
> Is there any idea for the solution?

Hard to tell yet. Looks indeed like something is wrong with the
interrupt controller.

Any chance you can bisect that ? I'll also have a look on my side,
it's definitely not something obvious.

Cheers,
Ben.

> Thanks: blackluck
> 
> Laszlo Fekete wrote:
> > Hello!
> >
> > I'm sorry about the annoyances, but I'd welcome all ideas, suggestions 
> > to see what needs to be done or should be tested for the solution.
> >
> > Thank you very much!
> >
> > Guennadi Liakhovetski wrote:
> >> Ok, first attempt to forward this to scsi was wrong, as pointed out 
> >> by Matthew Wilcox this does indeed look like an interrupt problem - 
> >> no interrupts drom SCSI, IDE, keyboar. Might be a known problem, I 
> >> guess. In any case, I think, the OP would be grateful for any hints.
> >>
> >> Thanks
> >> Guennadi
> >> ---
> >> Guennadi Liakhovetski, Ph.D.
> >> Freelance Open-Source Software Developer
> >> http://www.open-technology.de/
> >>
> >> -- Forwarded message --
> >> Date: Sat, 13 Jun 2009 16:22:07 +0200
> >> From: Laszlo Fekete 
> >> To: debian-powerpc@lists.debian.org
> >> Subject: sym scsi driver problem with 2.6.26 or newer debian kernel 
> >> on p610
> >> Resent-Date: Sat, 13 Jun 2009 14:29:55 + (UTC)
> >> Resent-From: debian-powerpc@lists.debian.org
> >>
> >> This is a multi-part message in MIME format.
> >> 
> >>
> >> Hello!
> >>
> >>
> >>
> >>
> >>
> >> Pls help me with sym scsi driver problem.
> >>
> >>
> >>
> >> I have Ibm P610 (and tested it on P630 and P640 too), installed debian
> >>
> >> etch and upgraded to lenny.
> >>
> >>
> >>
> >> But with 2.6.26 or newer kernel it's not booting, it's hang on sym scsi
> >>
> >> bus scan.
> >>
> >>
> >>
> >>
> >>
> >> Whats the problem with it, or how can I fix this?
> >>
> >>
> >>
> >>
> >>
> >> I attach the output from minicom with 2.6.29, 2.6.26, and the working
> >>
> >> 2.6.24 kernel booting.
> >>
> >>
> >>
> >>
> >>
> >> Thank you very much!
> >>
> >>
> >>   
> >
> >
> 
> ___
> Linuxppc-dev mailing list
> linuxppc-...@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev


-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Re: Ibook G4 freezes

[Benjamin Herrenschmidt]

On Sat, 2009-03-14 at 17:45 +, Madeleine Paterson wrote:
> Hi,
> I've an ibook that freezes randomly, and I just found your reply to  
> someone else with the same problem. I'd like to give it a go. Thing  
> is, I don't understand the tech-speak. could you help me out with an  
> idiot's guide to the procedure?
> Thanks
> Peter Benson
> Your page is  http://lists.debian.org/debian-powerpc/2005/12/msg00275.html

Are you using MacOS or Linux ? :-)

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to debian-powerpc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: backlight module for nvidia cards -- control backlight even with offb

[Benjamin Herrenschmidt]

No time to look at this right now, please ping me if you have no news in
a week or so.

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Re(2): [SOLVED] "AIGLX: Screen 0 is not dri capable"

[Benjamin Herrenschmidt]

On Tue, 2007-05-29 at 17:19 +0200, Georg Heinrich wrote:
> Yes, that did it.
> Strange enough, the touchpad still works!
> Thank you very much,

You probably need to rebuild synaptics from source (and file a debian
bug report). The touchpad is ADB or USB ? The ADB one just works fine
without, but the USB one works too but pretty badly without synaptics.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Again: "AIGLX: Screen 0 is not dri capable"

[Benjamin Herrenschmidt]

> Backtrace:
> 0: /usr/bin/X(xf86SigHandler+0x94) [0x1008eb98]
> 1: [0x100374]
> 2: [0x7fcac80c]
> 3: /usr/lib/xorg/modules/libxaa.so(XAAFillMono8x8PatternRectsScreenOrigin
> +0xdc) [0xf75301c]
> 4: /usr/lib/xorg/modules/libxaa.so(XAAPaintWindow+0x52c) [0xf761090]
> 5: /usr/bin/X [0x10138684]
> 6: /usr/bin/X(miWindowExposures+0x328) [0x100e057c]
> 7: /usr/bin/X [0x100a966c]
> 8: /usr/bin/X(MapWindow+0x3e4) [0x1002b6a8]
> 9: /usr/bin/X(InitRootWindow+0xfc) [0x1002b860]
> 10: /usr/bin/X(main+0x41c) [0x10025460]
> 11: /lib/tls/libc.so.6 [0xfc7e994]
> 12: /lib/tls/libc.so.6(__libc_start_main+0xb0) [0xfc7ead0]
> 
> Fatal server error:
> Caught signal 11.  Server aborting

That looks like the server is crashing which is weird. Could be a bug in
whatever version of X is currently in debian (is that etch ?).

X looks like release 7.1.1 and all the modules that matter seem properly
compiled against that one... except maybe synaptics. Can you try
disabling it see if it makes a difference ?

Also, try removing the UseFBDev option and tell us if it makes any
difference.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: PowerPC buildd is looking for a new home

[Benjamin Herrenschmidt]

> That's all fine and dandy, but for your personal pet project you don't
> need machines sponsored to the Debian project, maintained by the
> Debian project and stuff.  As you said, you own a G5 your own, so you
> can work on it.
> 
> As long as there's no perspective for powerpc64 to be included in the
> archive, there is no need for a buildd running on .debian.org hardware
> maintained by debian.org.
> 
> If it's your pet project, feel free to work on it, maintain as many
> machines as you like, talk to sponsors and bandwidth donators as you
> like.  This is not the business of the Debian project.

Note that, FYI, a ppc64 machine such as a Power5 or G5 will perfectly
fine run a 32 bits powerpc buildd, and will do it much faster than any
G4 you can find. Also, IBM Power5 machines are probably more reliably,
remotely admistrable, etc...

So it -does- make some sense to use these especially if they were
donated with the intend of those going to Debian.

Ben.
 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Beryl - do I dare?

[Benjamin Herrenschmidt]

> Been there done that. Add the key this message is signed with to your
> APT keyring and add the following to /etc/apt/sources.list:
> 
> deb
> http://www.oakcourt.dyndns.org/~andrew/linux/powerbook/aiglx-zero-copy-tfp
>  ./
> # optionally
> deb-src
> http://www.oakcourt.dyndns.org/~andrew/linux/powerbook/aiglx-zero-copy-tfp
> ./
> 
> 512Kbps bandwidth max, however, so if it gets slow someone might have to
> mirror them.
> 
> Minor caveat: If you want things WRT to upgrades to work normally, pin
> that repository somehow. I haven't figured out an easy and sane way to
> keep packages I've made personal changes to higher up on APT's list than
> the official repos, and I followed policy with regards to non-maintainer
> uploads (6.5.2-3 to 6.5.2-3.1) and now aptitude/update-notifier wants to
> "upgrade" mesa to 6.5.2-4 in unstable. The only thing I have figured is
> to bump the epoch (2:6.5.2-3) which is a bit messy and won't always
> yield desired results, especially if I end up distributing the packages.

Thanks ! Well... actually I'm running ubuntu on that laptop so I won't
even try installing debian X packages but I though it would be useful to
others. I'll build my own X with patches one of these days.

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Beryl - do I dare?

[Benjamin Herrenschmidt]

On Thu, 2007-04-26 at 15:59 -0400, Andrew J. Barr wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
> Taylor Oliphant wrote:
> > Anyone have Beryl running? I have a powerbook g4 and
> > would love to get it up and running.
> 
> What graphics card? Compiz runs fine on my ATI RV350 with the free
> radeon driver (well, what else?) and these patches:
> 
> http://people.freedesktop.org/~daenzer/aiglx-zero-copy-tfp/
> 
> You _need_ these for AIGLX compositing and a discrete graphics card
> (e.g. it's own memory). You will want to throw your nice shiny Powerbook
> across the room if you don't have them and you try to run Compiz or
> Beryl. Trust me.

Care to make .debs  ? :-)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: changing mac address of Sun GEM

[Benjamin Herrenschmidt]

On Sun, 2007-03-25 at 13:22 +0200, Wolfgang Pfeiffer wrote:
> On Fri, Mar 23 2007, at 00:06 +0100, Yves-Alexis Perez wrote:
> > On jeu, 2007-03-22 at 07:29 +1100, Benjamin Herrenschmidt wrote:
> > > No, the driver doesn't implement it. Should be trivial to add though.
> > 
> > MHmh, I use a powerbook g4, my network card is a sungem too and I can
> > do:
> > 
> > sudo ip l set eth0 address 00:11:24:xx:xx:xx
> > 
> > and it works. I'm using a 2.6.20 kernel but I remember it working since
> > a long time so I'm wondering what doesn't the driver implement.
> > 
> > Regards,
> 
> 
> Same here on a alubook G4: works like a charm to change the Ethernet
> MAC address as I like, as it seems. Provided the interface is
> down. Even on this old 2.6.18:

Yes, it works when the if is down when you don't have the
set_mac_address callback afaik. Having it allows changing when the
interface is up which is useful with nowadays userland thingies that
tend to keep interfaces up to poll their link (argh ! sungem is smart
enough to continue managing the link when the interface is down, but not
all drivers are)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: changing mac address of Sun GEM

[Benjamin Herrenschmidt]

On Sat, 2007-03-24 at 05:08 +0800, Ruben Vandeginste wrote:
> At Fri, 23 Mar 2007 09:09:37 +1100, Benjamin Herrenschmidt wrote:
> > 
> > the patch is whitespace damaged though... 
> 
> Ok, that should be fixed now.
> 
> > Appart from that, what's the output of strace macchanger ? Might give a
> > clue why it complains...
> 
> To be honest, this is the first time I try to change anything in the
> kernel... What exactly is the meaning of "netif_running"?  I thought
> it indicated whether the interface is up or down, but that seems not
> to be true.  What happened, is that the data structure of the net
> device was updated with the new mac address, but it was not written to
> the hardware.

netif_running() means the device is open() (which generally means the
interface is up).

> > I'll give it a try myself as soon as I find some time and will submit
> > the patch upstream with fixups if needed.
> 
> In this patch, the new mac address is written to the hardware
> regardless of "netif_running".  It seems to work, but probably needs
> fixups...

You need to be a bit careful here. The chip might be off (power
managed).

I would suggest you do something a bit like change_mtu there, though you
don't actually need to take the locks, only the mutex, and please do
test gp->running before touching the hardware (with the lock held).

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: changing mac address of Sun GEM

[Benjamin Herrenschmidt]

On Fri, 2007-03-23 at 03:53 +0800, Ruben Vandeginste wrote:
> At Thu, 22 Mar 2007 07:29:34 +1100, Benjamin Herrenschmidt wrote:
> > 
> > No, the driver doesn't implement it. Should be trivial to add though.
> 
> Ok.  I gave this a try by looking at the initialisation code for the
> card and at the code for setting the mac address in other drivers.  I
> have something that seems to work, but that is most likely not 'good',
> since I basically just write the values in the correct registers and
> don't care about anything else.  In any case, after changing it with
> "macchanger", "ifconfig" shows the new mac address.  "macchanger" did
> however complain that it could not change the mac address.
> 
> Comments?

Thanks !

the patch is whitespace damaged though... 

Appart from that, what's the output of strace macchanger ? Might give a
clue why it complains...

I'll give it a try myself as soon as I find some time and will submit
the patch upstream with fixups if needed.

Ben.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: changing mac address of Sun GEM

[Benjamin Herrenschmidt]

On Wed, 2007-03-21 at 19:59 +0100, Ruben Vandeginste wrote:
> Hello,
> 
> Is it possible to change the MAC address of the ethernet card in an
> iBook G4 ?
> 
> I tried to use "macchanger" like this (obviously with an other mac
> address):
> 
> guava ~ # macchanger --mac 00:00:00:00:00:00 eth0
> Current MAC: 00:00:00:00:00:00 (unknown)
> ERROR: Can't change MAC: interface up or not permission: Device or resource 
> busy
> 
> The eth0 interface is down and I am executing this as root.
> 
> Someone knows why this is not working?

No, the driver doesn't implement it. Should be trivial to add though.


Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Mac mini: serious clock drift

[Benjamin Herrenschmidt]

On Thu, 2007-03-08 at 16:04 +0100, Bram Senders wrote:
> Hi there,
> 
> I am having issues with clock drift on my G4 Mac mini.  The clock drifts
> so much that even ntpd can't keep up with it.  I keep getting messages
> like "time reset +2.404131 s" from ntpd approximately every 15 to 20
> minutes.

I think the calibration OF puts in the device-tree is crap or we are
using the VIA to calibrate and we should not ...

>From what I remember, Apple has additional calibration code in Darwin
based on the KeyLargo timers (for machines with a KeyLargo). Might be
worth implementing something similar.

Look at pmac_calibrate_decr() in arch/powerpc/platform/powermac/time.c,
and see wether it's calling via_calibrate_decr() or not on your machine.
If not, try calling it and tell us if it improves calibration.

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: G3 B/W fiber card blues

[Benjamin Herrenschmidt]

On Mon, 2007-02-12 at 15:30 -0500, Rick Thomas wrote:

> Maybe I should try the whole thing again and write down the details.
> 
> Can you give me a URL for the linux ieee1394 mailing list?

https://lists.sourceforge.net/lists/listinfo/linux1394-devel

Ben.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: G3 B/W fiber card blues

[Benjamin Herrenschmidt]

> I've got a couple of B&W G3 machines.  I've tried every Debian kernel  
> up thru the latest in SID.  The pcilynx driver crashes on bootup  
> every time.  I can disable it by putting it in the blacklist file,  
> which gives me a clean boot but no FireWire.
> 
> You're right about USB.  It works well enough for keyboard and mouse  
> on the B&W G3.
> 
> Should I file a bugreport?  Is so, against which package?

Best might be to send a mail to the linux ieee1394 mailing list though I
can foresee people telling you to get your self a cheap OCHI1394 PCI
card instead :-)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: powerbook wakeup/boot display issue

[Benjamin Herrenschmidt]

On Tue, 2007-02-06 at 15:20 +0100, Michel Dänzer wrote:
> On Tue, 2007-02-06 at 17:00 +1100, Benjamin Herrenschmidt wrote:
> > > 
> > > There seems to be a race condition somewhere which sometimes causes this
> > > when setting a video mode. I can usually get back the correct display by
> > > forcing a mode set, e.g. by switching between console and X.
> > 
> > The problem seems to be due to the panel powerup sequence. It looks like
> > you need to do a well timed sequence of events when toggling the various
> > LVDS related control bits for the panel to come reliably up. However,
> > from trial & error, I've never been able to come up with a sequence that
> > works perfectly for everybody. I suspect differnent panel models require
> > different sequences.
> 
> Maybe this could be improved in radeonfb though, given that it doesn't
> seem to happen (or at least much less frequently) with the X radeon
> driver without UseFBDev?

Happens for me with it every now and then too.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: powerbook wakeup/boot display issue

[Benjamin Herrenschmidt]

> FWIW, I haven't heard of this blooming damaging hardware. Still it's
> probably a good idea not to leave it indefinitely. :)
> 
> There seems to be a race condition somewhere which sometimes causes this
> when setting a video mode. I can usually get back the correct display by
> forcing a mode set, e.g. by switching between console and X.

The problem seems to be due to the panel powerup sequence. It looks like
you need to do a well timed sequence of events when toggling the various
LVDS related control bits for the panel to come reliably up. However,
from trial & error, I've never been able to come up with a sequence that
works perfectly for everybody. I suspect differnent panel models require
different sequences.

Unfortunately, ATI never answered my various questions about this.

On the other hand, I seriously doubt that this can damage hardware. The
fact is that if a flawky chip ends up dying while in linux, people tend
to blame linux :-) If that happens in MacOS, they rightfully blame the
hardware... go figure.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: G3 B/W fiber card blues

[Benjamin Herrenschmidt]

On Wed, 2007-01-31 at 10:57 -0600, Dan Page wrote:
> Hello.  I've just joined this list so I apologize if this issue has been
> addressed:
> 
> I've installed Debian sarge on a Blue and white G3.  This box is planned
> to be used as a DNS resolver for our network. I got DJB DNS set up and
> working like a champ, my problem lies in the network card.  We use
> gigabit fiber for our core network and I chose this model because it has
> the 64 bit pci bus that most giga fiber cards use.  I installed a Intel
> gigabit card installed.  The e1000.0 module that this card requires  is
> present in the kernel but for some reason it doesn't load.  I don't know
> if its related but the drivers for the USB and Firewire don't load
> properly either.   here is an except from the dmesg log

USB has issues on the B&W G3 due to Apple using an old fairly broken
chip though I though it worked well enough for keyboard/mouse at
least ... 

FireWire is a pcilynx chip (pretty much one of the only ones to use that
chip) and I don't think it's ever been properly tested on big endian.

In any case, you should try a newer kernel

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Debian has failed us

[Benjamin Herrenschmidt]

On Thu, 2007-01-11 at 11:40 -0600, Mathew Binkley wrote:
> Greetings.  I am the senior system administrator at Vanderbilt
> University's supercomputing center.  We operate a 1500 processor cluster
> for researchers at Vanderbilt.
> 
> Our current cluster is divided between 840 Intel/AMD x86 processors, and
> 672 IBM PowerPC 970FX processors.  To date, we have required different
> operating systems on each architecture because of poor OS support for
> the PowerPC's.  SuSE has been our only option so far.

How so ? I'm routinely running debian/ppc on all sort of PowerPC
machines here, including js20's, 21's POWER5's etc...

You might need a bit of hacking to get the initial install though, and I
agree that should be improved in debian (though do they have access to
hardware for that ?)

Also, there are other options, like Fedora/RHEL, YDL, ...

> Etch supports PowerPC's running in 64-bit mode

Running in full 64 bits mode isn't necessarily the best idea on PowerPC
performance wise. It's generally recommended to run a 32 bits distro
with only the minimum set of 64 bits libraries so that applications that
have to be 64 bits can run. It's very different from AMD64 where 64 bits
brings all sorts of improvement starting with much more registers. On
PowerPC, we have many registers in the first place, thus 64 bits mode is
almost purely overhead unless you application has good use of 64 bits
pointers or arithmetic.

>  so I was eager to try it
> on our cluster several months ago.  I uncovered a small bug in the
> installer (the AMD 74xx driver was not compiled in the debian-installer
> kernel) which prevented me from installing Debian, and reported it to
> the PowerPC list in September 2006.  I was promised that the driver
> would be included shortly.

Ah yes, I remember that.

> Unfortunately, some sort of personal issues between developers in the
> PowerPC group and the debian-installer group has prevented this minor
> fix from occuring.  I do not understand what that problem *is*, only
> that personal problems are getting in the way of real-world results.

Yes, I've seen various mails/rants etc... and haven't wanted to take
part of it neither in the past, but it looks like this has gone too far,
and it's more and more looking like even perfectly good bug fixes that
are needed for most users are being rejected on the sole basis of the
person they originate from. Either that or somebody has an agenda of
fucking up powerpc support in debian...

> I had hoped to be able to install Debian on our cluster during our
> refresh this spring in order to have a consistent OS across all
> architectures.  Since I have not been able to get this minor bug
> resolved in three months, we will be installing Centos 5 instead.

Can't you get a custom build of the installer ? I mean, I agree the
situation sucks, but it's not like you can't find somebody willing to
provide you with a custom build if you ask around ...

> I have personally been using Debian since slink, and I would like to
> believe that Debian can sort these issues out before it becomes a
> permanent handicap to the organization.
> 
> Finally, I would like to extend thanks to Sven Luthor for tracking down
> the bug and trying to get me some assistance.

Luther, not Luthor, he's not that evil yet :-)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Wakeup on mouse movement

[Benjamin Herrenschmidt]

On Wed, 2007-01-10 at 11:02 +0100, Matthias Grimm wrote:
> Hi,
> 
> I just upgraded to the debian kernel 2.6.18-3 (I used a self-configured
> kernel before) and found an odd behaviour:
> 
> When the machines was suspended to RAM it woke up again after I
> moved the mouse. In the past The machine woke only up after triggering
> a key on the keyboard but never on mouse movement.
> 
> Is this a feature or a bug? How could I disable this "feature" in case
> it is one?
> 
> I used the 2.6.18 kernel before with my own configuration and it never
> woke up on mouse movement. 

I suspect it's a change in the USB code to enable remote wakeup by USB
devices. It's a good idea in general but does have the side effect that
a lot of mice are stupid enough to trigger a wakeup when they are
moved...


Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [FYI] debian-installer for SONY PS3

[Benjamin Herrenschmidt]

On Thu, 2007-01-04 at 09:48 -0500, Joey Hess wrote:
> It occurs to me that this might be of interest here. The rest of the
> thread is in debian-boot.

As noticed in his message, there are 2 major issues:

 - Kernel. Right now, we really only have this 2.6.16 port done by Sony,
which is somewhat limitative ;-) There is some work in progress to
rebase the port and merge it upstream properly, but that takes time and
will not be ready for a couple of monthes yet at best I reckon (stuff
happens on #cell on oftc)

 - Bootloader. Kboot is really not a terrific option at this point,
especially since every distro ends up doing it's own hacked version of
it incompatible with each other that needs to be flashed in the ps3.
That's why I've started working with jk on a proper graphical
bootloader, though it will require to sync with distros as we'll need
something better than the current kboot-type config files (though I'll
try to make it compatible with existing stuff).

Nothing ready to show there but it's being worked on.

Ben.



> Kenshi Muto wrote:
> > -BEGIN PGP SIGNED MESSAGE-
> > Hash: SHA1
> > 
> > Hi,
> > 
> > Takeshi Yaegashi, who is my friend, succeeded to install Debian
> > on SONY PlayStation 3 (PS3). PS3 is PPC64 architecture and
> > some Linux distributions and NetBSD are already ported for PS3.
> > 
> > His wiki describes how you can install Debian on PS3 by using his
> > customized Debian-Installer:
> > http://www.keshi.org/moin/moin.cgi/PS3/Debian/Installer
> > SVN diff against current d-i trunk (still under development,
> > he is carefully changing d-i to keep small):
> > http://www.keshi.org/moin/moin.cgi/PS3/Debian/Installer/Porting
> > 
> > Although I haven't PS3, it's very interesting that d-i merges
> > his PS3 PPC specific patch for post-Etch.
> > 
> > I talked with him for a while and he pointed some TODO.
> > 
> > 1. We need newer kernel or patched kernel to support PS3 architecture.
> > 2. We have to ITP ps3pf-utils package to use video on X and some
> >features.
> > 3. We have to use new kboot.udeb instead of other PPC installer.
> >kboot is Linux based boot loader and can be installed from
> >SONY official kit. It finds a boot configuration from the 
> >partition has specific label and loads the file. So we don't
> >have to install something to install boot sector. kboot.udeb
> >just locates the configuration file kboot.conf to /etc directory.
> > 
> > Thanks,
> > - -- 
> > Kenshi Muto
> > [EMAIL PROTECTED]
> > -BEGIN PGP SIGNATURE-
> > Version: GnuPG v1.4.6 (GNU/Linux)
> > Comment: Processed by Mailcrypt 3.5.8+ 
> > 
> > iEYEARECAAYFAkWaQRIACgkQQKW+7XLQPLEjWgCfe43uecBQxrPbWQ6NeR/7S51b
> > XacAoMPWBWn8mA+IBpEVBVwU0fKTXLWJ
> > =9QPL
> > -END PGP SIGNATURE-
> > 
> > 
> > -- 
> > To UNSUBSCRIBE, email to [EMAIL PROTECTED]
> > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
> > 
> 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: no way to boot from mobile firewire hdd?

[Benjamin Herrenschmidt]

On Thu, 2007-01-04 at 13:33 +0900, 2g wrote:
> since my iBookG4's internal hdd got dead  
> 
> bought
> 
> http://www.lacie.com/products/product.htm?pid=10558   
> 
> could happily restore osx and boot from it
> 
> but installing debian nor ubuntu seem to fail at  
> 
> installing yaboot process 
> 
>   
> 
> well trying to install the bootstrap to the external hdd  
> 
> is stupid anyway? 
> 
> weird thing is that   
> 
> option boot doesn't show neither osx nor debian on the external   
> 
> but osx on it boots anyway
> 
>   
> 
> then there's no way to boot debain from this external HDD?
> 
> even if i could   
> 
> nothing like yaboot oriented dual(choosing with only one key) thing   
> 
> available?

It can be made to work though I suspect none of the installers will do
it properly by default, that is you'll have to hack a bit around.

Basically, if you set it up so that it boots with an initrd that can
mount root by label, it should work. Then, you have to hack around the
yaboot config files to get to the right OF path.

Now, another option is if the external drive you just bought is an IDE
2.5" laptop drive, just to swap it with your internal one :-)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: iMac G5 Bugs and new wiki

[Benjamin Herrenschmidt]

On Tue, 2006-12-12 at 21:29 +1300, Torrance wrote:
> Hi Michel,
> 
> We've discovered the problem lies with the kernel configuration: the
> standard Debian 2.6.18 kernel Switches Users fine. Our custom kernels
> however, do not.
> 
> Unfortunately, I'm not sure what exactly isn't configured properly. My
> kernel is essentially a standard g5_defconfig, plus firewall support.
> I've attached the config file, and would be most interested if anyone
> can spot the problem. 

Could be a problem with nvidiafb ? What if you disable it ?

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: A few questions from an iMac G5 user...

[Benjamin Herrenschmidt]

> 1) Is it possible to enable either sleep or suspend functions for the
> iMac G5? I am currently running the 2.6.19 kernel, but I don't have
> the slightest clue how to enable suspend or sleep. These functions
> never worked in Ubuntu. Is this possible? 

There is currently no sleep support for G5s Suspend To Disk should
be doable without too much pain though I didn't have time to do anything
about it, I remember some patches were posted at one point doing at
least part of the job.

As for Suspend To RAM... well... the iMac G5 is an SMU based machines. I
_think_ I know what command can be sent to the SMU to instruct it to put
the machine to sleep, though I haven't really tried doing it. The main
problem is that sleep will shut down the video chip and we currently
can't bring an nVIdia chip back.

It might be possiblea after a good deal of reverse engineering, or help
from nVidia (don't hold your breath on that), but not at the moment.

> 2) Is it a good idea to enable ESD? I was having a few system freezes
> when using multiple multimedia players (Totem Xine, Rhythmbox and
> VLC), and have since enabled esd (and installed gstreamer-esd and
> vlc-esd). Things seem to be fine, but I'm not really sure what is an
> ideal sound setup. 

System freeze ? Or just application blocking because another one is
using the driver ?

> 3) How do I install java for running, eg. internet java applets? In
> Ubuntu we had to download from an IBM site. Currently, that site
> requires registration and even then told me that export laws (or
> something) would mean the download would be reviewed and I'd be
> notified when it was available. Of course, this never happened. So how
> else do ppc users get java installed? Alternatively, how long will it
> be before we see Adobe's recently released source code making its way
> into sid? 

Weird, they might have changes something... I had no problem last time I
d/l the IBM Java. That's the best option at the moment. Now that Sun
Java is GPL, there is hope to have more options available though.

You can also use gij/gcj but they can't run quite as much things as the
IBM JVM can (at least IBM own internal expense report system :-) Though
I've been told they can run Eclipse.

There's also Kaffe but that's pretty outdated... 

> 4) Is it possible to brighten the screen on an iMac G5. It looks
> darker and duller than OS X, but besides altering the gamma settings,
> there seems no way to brighten the screen.

Yes, there is a magic register and indeed, it looks like the value set
in there by nvidiafb might be slightly lower than what OS X does on this
machine

The problem is that there are several machine models with similar nVidia
chips and they seem to require differnent "max" values there, though we
haven't quite figured how to properly identify which model needs what,
so we are stuck right now with what we think is a "safe" value (ie.
won't damage the display).

I have some contacts at nVidia that were supposed to enquire about that
but I had no reply over the past few monthes, I'll try to ask again.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Bug#397942: g5 imacs now silent?

[Benjamin Herrenschmidt]

On Thu, 2006-11-30 at 23:04 +0100, Étienne Bersac wrote:
> Hi,
> 
> Beware not to confuse iMac G5 rev A&B and rev C (among PowerMac G5).
> iMac G5 rev C (chip PowerMac 12,1) do not have yet their windfarm
> driver. Benh told that he began the work on this driver but has
> difficulties to get access to de device. Benh, can you confirm ?

Yes, I did some initial work to dig as much info as I could from darwin,
but it seems that apple, once again, changed everything and thus I'm not
confident writing and releasing a driver without physical access to the
HW to verify by myself that the fans are behaving properly.

I haven't managed to get access to that machine yet.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: PowerMac G4 : how to enable the server mode ?

[Benjamin Herrenschmidt]

On Tue, 2006-11-28 at 14:27 -0500, Stefan Bruda wrote:
> Michael Schmitz wrote:
> >
> > I'm pretty sure the decision to disable autoboot on shutdown was a design
> > decision. If you want a shutdown followed by automatic powerup, you should
> > have used reboot.
> 
> Except on power failures...

Well, on power failures it works. On explicit manual shutdown, I disable
it or the machine would just come back (actually some machines would,
some won't, some might even silently disable it in that case, I think it
depends on the generation of PMU firmware).

> By the way, has anybody tried this trick of the PowerMacs G5?  I have a
> first generation (PowerMac7,2) and as far as I figured out I am able to
> enable server_mode but this does not seem to make any difference.  Any
> (other) experience on the matter?

It should work on a PMU based G5 like this one... I'm pretty sure I had
it working. For SMU based machines, you need a different mecanism to
send the SMU command. I reverse engineered the command the other day on
irc and I think somebody wrote a tool though I can't remember who :-0

Ben.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: PowerMac G4 : how to enable the server mode ?

[Benjamin Herrenschmidt]

> I found this through Rich Johnson, the maintainer of the "autoboot"
> utility.. and I've just pulled plug on one of our G4 servers and lo!
> the server_mode=1 survived the reboot.  It also remains after a
> conventional reboot.

Yes, it's disabled if you do a shutdown tho.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: PowerMac G4 : how to enable the server mode ?

[Benjamin Herrenschmidt]

> > Well, by the way, where has BenH gone? I haven't seen him here since the
> > end of october.
> 
> bought an macintel?  ;)

No way :-)

Just busy as usual...

Ben.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ftp.debian.org: Please remove numactl for alpha/powerpc from unstable

[Benjamin Herrenschmidt]

On Mon, 2006-10-30 at 15:40 +0100, Andi Kleen wrote:
> > 
> > Andi, can you fix libnuma ? :)
> 
> I fixed libnuma to return ENOSYS on ppc, still waiting for an official
> syscall number

>From a pending patch by Stephen:

--- a/include/asm-powerpc/unistd.h
+++ b/include/asm-powerpc/unistd.h
@@ -276,7 +276,7 @@ #endif
 #define __NR_rtas  255
 #define __NR_sys_debug_setcontext 256
 /* Number 257 is reserved for vserver */
-/* 258 currently unused */
+#define __NR_migrate_pages 258
 #define __NR_mbind 259
 #define __NR_get_mempolicy 260
 #define __NR_set_mempolicy 261

Paul conformed the number is final.

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ftp.debian.org: Please remove numactl for alpha/powerpc from unstable

[Benjamin Herrenschmidt]

On Mon, 2006-10-30 at 17:13 +1100, Ian Wienand wrote:
> On Mon, Oct 30, 2006 at 04:56:09PM +1100, Benjamin Herrenschmidt wrote:
> > Can't you just apply a debian patch fixing the syscall number and do a
> > build with that ? It doesn't need the syscall to work, and as I said, we
> > do use it regulary, I would be fairly annoyed if it disappeared
> > completely even temporarily.
> 
> Ok, the last thing I want to do is annoy people!
> 
> I just uploaded 0.9.10-2 with the syscall patched to 258 as you
> mentioned.  When a new release comes out I'll deal with that, and keep
> an eye on Power now I know it has some users ;)

Thanks !

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ftp.debian.org: Please remove numactl for alpha/powerpc from unstable

[Benjamin Herrenschmidt]

On Mon, 2006-10-30 at 16:29 +1100, Ian Wienand wrote:
> On Mon, Oct 30, 2006 at 04:06:31PM +1100, Benjamin Herrenschmidt wrote:
> > Or if not ,what is the exact missing bit ? numactl should work on
> > powerpc and is useful on it. It shouldn't be dropped just bcs there is a
> > temporary build failure.
> 
> I wish to drop it for now firstly because (and this is my fault) I
> ignored for it a bit too long and the testing version is *way* out of
> date.  I'd like .10 in testing at least for those architectures that
> support it out-of-the-box.  I am happy to (and intended to) put
> powerpc back in when it can work.
> 
> > Is it the new sys_migrate_pages that isn't hooked (yet) on powerpc ?
> 
> Yes; but worse at the moment because upstream manually calls syscall
> 280, which is just wrong (this is what the current version in unstable
> is doing, which is another reason to get rid of it for now).
> 
> If you get it wired up in the kernel, and Andi fixes up libnuma I'm
> happy to put it all back together again.

Can't you just apply a debian patch fixing the syscall number and do a
build with that ? It doesn't need the syscall to work, and as I said, we
do use it regulary, I would be fairly annoyed if it disappeared
completely even temporarily.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ftp.debian.org: Please remove numactl for alpha/powerpc from unstable

[Benjamin Herrenschmidt]

On Mon, 2006-10-30 at 16:29 +1100, Ian Wienand wrote:
> On Mon, Oct 30, 2006 at 04:06:31PM +1100, Benjamin Herrenschmidt wrote:
> > Or if not ,what is the exact missing bit ? numactl should work on
> > powerpc and is useful on it. It shouldn't be dropped just bcs there is a
> > temporary build failure.
> 
> I wish to drop it for now firstly because (and this is my fault) I
> ignored for it a bit too long and the testing version is *way* out of
> date.  I'd like .10 in testing at least for those architectures that
> support it out-of-the-box.  I am happy to (and intended to) put
> powerpc back in when it can work.
> 
> > Is it the new sys_migrate_pages that isn't hooked (yet) on powerpc ?
> 
> Yes; but worse at the moment because upstream manually calls syscall
> 280, which is just wrong (this is what the current version in unstable
> is doing, which is another reason to get rid of it for now).
> 
> If you get it wired up in the kernel, and Andi fixes up libnuma I'm
> happy to put it all back together again.

Andi, can you fix libnuma ? :)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ftp.debian.org: Please remove numactl for alpha/powerpc from unstable

[Benjamin Herrenschmidt]

On Mon, 2006-10-30 at 15:57 +1100, Ian Wienand wrote:
> On Mon, Oct 30, 2006 at 03:47:25PM +1100, Benjamin Herrenschmidt wrote:
> > Ugh ? Is there a replacement ? numactl is useful on 64 bits powerpc
> > machines. Or is numactl itself being replaced by something else ?
> 
> The latest libnuma depends on, in particular, the migrate_pages system
> call, which isn't wired up for power/ppc (at least in 2.6.18).  Since
> that call went in a long time ago by kernel standards and nobody did
> this yet, I figured nobody was testing/using it.
> 
> Originally libnuma had a hard-coded syscall number for power, but that
> hardcoded number has been replaced in the kernel by something totally
> unrelated.  Thus I removed it because at this point, it's just wrong.

Looks like there is some problems with the ABI of the syscall which
makes things complicated, and is one of the reasons it wasn't hooked up
yet. There's a patch coming soon that should hook it up but I don't know
if it will make 2.6.19. In the meantime, we'll make sure the syscall
number is reserved. It will be 258, next to mbind and currently not
used.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ftp.debian.org: Please remove numactl for alpha/powerpc from unstable

[Benjamin Herrenschmidt]

On Mon, 2006-10-30 at 15:47 +1100, Benjamin Herrenschmidt wrote:
> On Mon, 2006-10-30 at 09:40 +1100, Ian Wienand wrote:
> > Package: ftp.debian.org
> > Severity: normal
> > 
> > Hi,
> > 
> > I have dropped Alpha and PowerPC from numactl, as they don't have the
> > kernel support to build the latest versions.  It's very unlikely
> > anyone cares about this package on those architectures.
> > 
> > Could you please remove them from unstable so the package can make it
> > into testing.
> 
> Ugh ? Is there a replacement ? numactl is useful on 64 bits powerpc
> machines. Or is numactl itself being replaced by something else ?

Or if not ,what is the exact missing bit ? numactl should work on
powerpc and is useful on it. It shouldn't be dropped just bcs there is a
temporary build failure. Is it the new sys_migrate_pages that isn't
hooked (yet) on powerpc ?

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: ftp.debian.org: Please remove numactl for alpha/powerpc from unstable

[Benjamin Herrenschmidt]

On Mon, 2006-10-30 at 09:40 +1100, Ian Wienand wrote:
> Package: ftp.debian.org
> Severity: normal
> 
> Hi,
> 
> I have dropped Alpha and PowerPC from numactl, as they don't have the
> kernel support to build the latest versions.  It's very unlikely
> anyone cares about this package on those architectures.
> 
> Could you please remove them from unstable so the package can make it
> into testing.

Ugh ? Is there a replacement ? numactl is useful on 64 bits powerpc
machines. Or is numactl itself being replaced by something else ?

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: bug#390432 - still no boot 2.6.18 here

[Benjamin Herrenschmidt]

On Tue, 2006-10-24 at 15:45 +1000, Benjamin Herrenschmidt wrote:
> On Mon, 2006-10-23 at 22:00 -0700, Brian Morris wrote:
> 
> > > What model of 'book is it ? A wallstreet ?
> > this is powerbook3400, ppc603ev, 80mbRAM,
> 
> I have access to one of these, I'll do some tests.

Hrm... hard disk dead... No time to find a replacement right now though,
I'll investigate some other time.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: bug#390432 - still no boot 2.6.18 here

[Benjamin Herrenschmidt]

On Mon, 2006-10-23 at 22:00 -0700, Brian Morris wrote:

> > What model of 'book is it ? A wallstreet ?
> this is powerbook3400, ppc603ev, 80mbRAM,

I have access to one of these, I'll do some tests.

> > What image file are you trying to boot from BootX ? Have you tried the
> i am trying to boot the latest version from debian unstable, that is
>  2.6.18-1-powerpc, that is debian version (dpkg list) 2.6.18-3
> > "no video driver" option ? have you waited long enough ? (it might be
> > booting but failing to display...)
> 
> well i don't have a black screen, i get the first line and it hangs.
> on 15 there is a pause where the screen clears and that was
> i think the "black out" place before on g3s but not here. that is
> what you called the trampoline ? i don't think this gets that far.
> 
> this powerbook has chips65550 (chips is now subdiv of intel),
> not ati. that may explain why quik works on 15 and almost
> works on 18. actually i needed quik with 2.6.8 on sarge it turned
> out i was getting the random scsci hangs with bootx, but that was
> further on in the boot, no evidence of that here.
> 
> quik is quik, it boots faster and runs faster, it would be much
> better if i could use it. but it acts like it can't see the ram disk initrd.

You might be able to load the zImage.initrd ... I'll have a look

> i would be more willing to try building my own kernel if i knew
> better what i could take out (config=n). my only guess that quik
> will not boot because some people reported earlier with monolithic
> kernel it would not boot a kernel bigger than 4mb.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: bug#390432 - still no boot 2.6.18 here

[Benjamin Herrenschmidt]

On Mon, 2006-10-23 at 03:01 -0700, Brian Morris wrote:
> well i tried again,  still same result as ever trying
> to boot 2.6.18 on my old powerbook.
> 
> synopsis: bootx hangs after first line "welcome to linux",
> quik goes all the way to VFS: ... unknown block(0,0)
> (as if there is no initramfs). no floppy drive, so can't
> try miboot.

Quik can load an initrd at all ? I don't remember...

What model of 'book is it ? A wallstreet ?

What image file are you trying to boot from BootX ? Have you tried the
"no video driver" option ? have you waited long enough ? (it might be
booting but failing to display...)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: old world nonsense, please rectify

[Benjamin Herrenschmidt]

On Sun, 2006-10-22 at 13:54 +0200, Sven Luther wrote:
> On Sun, Oct 22, 2006 at 10:41:58AM +0200, Holger Levsen wrote:
> > Hi,
> > 
> > On Sunday 22 October 2006 10:20, Benjamin Herrenschmidt wrote:
> > > What is the exact problem ? I'm pretty sure I had 2.6.18 booting fine on
> > > a couple of oldworld boxes here before relase. (A fix for IDE on some
> > > performa went in later on, I think in 2.6.18.1)
> > 
> > The problem I see here (even on newworld), is that 32mb is too little 
> > memory 
> > for kernel (4.3mb) plus initrd (5mb, both compressed sizes).

4Mb compressed ? wow that's a pretty loaded config... the one I use on
my laptop is only 2Mb compressed.

> Ah, ...
> 
> Can you try :
> 
>   1) using yaird instead of initramfs-tools, or initramfs-tools with the
>   modules=dep option.
> 
>   2) use the powerpc-miboot flavour instead of the powerpc one.
> 
> And inform us if any of those help.
> 
> Friendly,
> 
> Sven Luther
> 
> 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: old world nonsense, please rectify

[Benjamin Herrenschmidt]

On Sun, 2006-10-22 at 10:13 +0200, Sven Luther wrote:
> On Sat, Oct 21, 2006 at 08:38:06PM -0700, brian wrote:
> > i was figuring there was something wrong with
> > the 2.6.15 kernel. why it was deleted from 
> > the debian archives when some old world power
> > macs still need it, and nobody responding to
> > any of my postings here. 
> 
> This is a known problem, please help get it fixed in time for the etch
> release.
> 
> > but today i find this:
> > http://packages.ubuntulinux.org/dapper/base/linux-image-2.6.15-23-powerpc
> > 
> > so why am i still booting 2.6.15-1 in debian ?
> 
> Debian/etch will ship with 2.6.18, on all arches. This currently has some
> issues on oldworld powermac, something probably minor to fix, but major in
> that actually booting does not work.
> 
> What we need is someone interested enough in this issue to follow it,
> summarize the actual solution, and investigate with upstream how to fix this.

What is the exact problem ? I'm pretty sure I had 2.6.18 booting fine on
a couple of oldworld boxes here before relase. (A fix for IDE on some
performa went in later on, I think in 2.6.18.1)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: synaptics PowerBook ?

[Benjamin Herrenschmidt]

On Tue, 2006-10-17 at 18:25 +0300, Eddy Petrișor wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
> 
> Matteo Bigoi - Bigo! wrote:
> >>> patch - working also on 19-rc1 - for the synaptics driver on a lot of
> >>> *books:
> >>>
> >>> http://www.cs.unibo.it/~bigoi/
> >>>
> >>> you will find all the infos and latest patches there.
> >>> Try it.
> >> Any reason why this patch isn't submited upstream ? :)
> >  
> > This patch does not use sysfs and dynamic values and LKML users said
> > it can't be submitted upstream. Anyway the original idea of Luca Bigliardi 
> > was to submit this patch in the mainstream and if there is someone who would
> > help me,that would be really appreciated.
> 
> I am no expert, but is this applicable to the patch? What should be
> exposed in sysfs?

The switch between raw and normal mode.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Daystar Genesis MP Quad 604e mac clone with debian

[Benjamin Herrenschmidt]

On Tue, 2006-10-17 at 00:14 +1000, Dana Sibera wrote:
> I've just come across one of these beasties - 4x180MHz 604e CPUs -  
> and it's just not taking to the kernels supplied with debian very  
> well. I've tried with bootx 1.2.2 using the kernels on the netinst cd  
> - both 2.4 and 2.6 for PowerPC. 2.4 doesn't get to the first text  
> screen but just locks up the machine, and 2.6 only gets the first  
> text screen - without getting to the stage of showing a Penguin up  
> top of the screen.
> 
> It has 896MB RAM, a single internal SCSI drive and I've tried with  
> both an ixmicro twinturbo 128 card, and a rage pro, but neither make  
> a difference.
> 
> Any clues on which way to turn now?

Weird We do have support for the 4 CPU card in there. I'm not sure
what is wrong at this point.

What precise kernel version have you tried ? Also, what graphics card
are you using in this machine ? Have you tried BootX option "no video
driver" ?

It might be useful to try to netboot from OF instead. Might be more
reliable, and using OF over a serial port will probably provide better
diagnostics.

You can do that by setting up a bootp/tftp server (enable the bootp
option in dhcp for the former, and install tftpd for the later) and
doing "boot enet". With a little bit of luck it will work on your OF
version (it works on most oldworld machines), at least provided you use
the on-board ethernet.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: powermac fan control modules detection ...

[Benjamin Herrenschmidt]

On Sun, 2006-10-15 at 12:16 +0200, Sven Luther wrote:
> Hi, ...
> 
> I am working at integrating fan control into d-i and initramfs-tools.
> 
> For this, we currently use the following list of modules :
> 
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/therm_pm72.ko
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_core.ko
>   
> /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_cpufreq_clamp.ko
>   
> /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_lm75_sensor.ko
>   
> /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_max6690_sensor.ko
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_pid.ko
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_pm112.ko
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_pm81.ko
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_pm91.ko
>   
> /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_smu_controls.ko
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_smu_sat.ko
>   
> /lib/modules/2.6.17-2-powerpc64/kernel/drivers/macintosh/windfarm_smu_sensors.ko
>   /lib/modules/2.6.17-2-powerpc64/kernel/drivers/i2c/busses/i2c-powermac.ko
> 
> I wonder if anyone is missing.
> 
> Benh, am i right in thinking that those modules are only needed for 64bit
> powermac hardware ? 

i2c-powermac.ko is also useful for various other things on 32 bits. All
the other ones in your list are indeed 64 bits only.

You also need to make sure cpufreq_64 is enabled for them to work.

> Also, i wonder if there is some way for detecting which machine needs which
> module ? Some /proc/device-tree parsing ? Or maybe simply a /proc/cpuinfo
> model matrix ? 

i2c-powermac.ko : should pretty much be built-in

Then, it depends on the machine. The "main" modules tend to auto-load
the other ones, but there is a bug where it misses
windfarm_cpufreq_clamp. and windfarm_pm112 blatantly "forgets" to
autoload anything. So the rule is:

Machine type:

 - PowerMac7,2, PowerMac7,3, RackMac3,1: therm_pm72

 - PowerMac8,1, PowerMac8,2: windfarm_pm81 + windfarm_cpufreq_clamp

 - PowerMac9,1: windfarm_pm91 + windfarm_cpufreq_clamp

 - PowerMac11,2: windfarm_pm112 + windfarm_cpufreq_clamp +
windfarm_smu_sat + windfarm_smu_sensors + windfarm_smu_controls +
windfarm_max6690_sensor.ko + windfarm_lm75_sensor.ko

 - PowerMac12,1:  (iMac G5 + iSight, I welcome
somebody in australia giving me hardware access to one for a week or two
to fix that).

I'll do some patches to fix autoloading of sub-modules so that only the
"main" one is needed.

Ben.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: 2.6.18 on sid boots with BootX ?

[Benjamin Herrenschmidt]

On Wed, 2006-10-11 at 19:58 -0300, Fábio Rabelo wrote:
> Hi all, just checking .
> The current kernel 2.6.18 present in Debian Unstable ( sid ) is able to 
> boot a Beige G3 with BootX ?
> I have one of this, and is working fine as gateway/firewall of one of my 
> costumers, but I will personally in his place to upgrade some other 
> things, so I could take the chance and update this machine too ?!?

I think I've fixed all of the known problems with BootX booting in
2.6.18 but best is you try and report to us :)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Powermac G5 won't boot headless?

[Benjamin Herrenschmidt]

On Wed, 2006-10-11 at 10:59 -0400, Alexandre Gauthier wrote:
> Le mercredi 11 octobre 2006 02:11, Benjamin Herrenschmidt a écrit :
> > On Wed, 2006-10-11 at 00:29 -0400, Alexandre Gauthier wrote:
> > > Hi!
> > >
> > > I am planning on using a Dual PowerMac G5 as a server. I installed Debian
> > > Stable and apt-pinned a few things, such as the kernel from Testing.
> > >
> > > All is well, until I realized that somehow, the machine apparently will
> > > NOT boot headless, that is, if the monitor is not present... this is
> > > slightly inconvenient in the case of a mail server that will almost never
> > > have a monitor plugged in, unless something is broken.
> > >
> > > I was wondering if anyone had an idea about this, or maybe a workaround?
> >
> > No monitor plugged or no video card ? I don't have a problem with the
> > first case though I haven't tested the second case.
> >
> > Ben.
> 
> No monitor. The machine does have a Video Card.
> 
> What happens when I yank the DVI cable from the machine, and reboot is:
> 
> 1) The chime is heard
> 2) Fans kick in gradually.
> 
> Since I can't see where it is in the boot process at all, I assume it is 
> somewhere in yaboot, wich would indeed cause the fans to kick in to full 
> speed if left on for a while...
> 
> But it never boots completely. Add a monitor, and it's fine...

interesting... I have a dual with a serial port, I'll do some tests. It
could be a yaboot issue.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Powermac G5 won't boot headless?

[Benjamin Herrenschmidt]

On Wed, 2006-10-11 at 00:29 -0400, Alexandre Gauthier wrote:
> Hi!
> 
> I am planning on using a Dual PowerMac G5 as a server. I installed Debian 
> Stable and apt-pinned a few things, such as the kernel from Testing.
> 
> All is well, until I realized that somehow, the machine apparently will NOT 
> boot headless, that is, if the monitor is not present... this is slightly 
> inconvenient in the case of a mail server that will almost never have a 
> monitor plugged in, unless something is broken.
> 
> I was wondering if anyone had an idea about this, or maybe a workaround?

No monitor plugged or no video card ? I don't have a problem with the
first case though I haven't tested the second case. 

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: zImage.prep on 2.6.18 & 2.6.19-rc1-git5

[Benjamin Herrenschmidt]

On Tue, 2006-10-10 at 09:42 +0200, Alexander Baldeck wrote:
> Hey all,
> 
> I've been trying to get an old Motorola Powerstack II working on Linux 
> for the past few days. Everything is working fine except for the fact 
> that I have to use a kernel 2.6.8 or earlier. As I'd like to run newer 
> udevs this is unfortunately an inacceptible state, no uevents and all.
> 
> Now then, to the actual problem which occurs while building the kernel 
> like so:
> 
>   $ make zImage
> 
> I was used to do that when building kernels and then dd 
> arch/ppc/boot/images/zImage.prep to my PreP partition. Now with 2.6.18 
> and 2.6.19-rc-git5 the following error pops up somewhere near the end of 
> my build:

Paul Mackerras is working on moving PReP over to arch/powerpc along with
various things. I'm sure he would be interested in a tester with a
PowerStack II. I would suggest that you continue that discussion with
him on [EMAIL PROTECTED]

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: synaptics PowerBook ?

[Benjamin Herrenschmidt]

On Tue, 2006-10-10 at 09:11 +0200, Andrea Lusuardi - UoVoBW wrote:
> On Tue, 10 Oct 2006 15:51:18 +1000
> Benjamin Herrenschmidt <[EMAIL PROTECTED]> wrote:
> 
> > 
> > Any reason why this patch isn't submited upstream ? :)
> > 
> > Ben.
> 
> the patch actually isn't mine.
> It has first been written by Luca Bigliardi but he could not find
> anyone willing to submit it because there was a flame - i do not
> remeber exactly between him and who - because it did not use sysfs.

I remember doing a comment about that but not flaming anybody :) It
might have degenrated into a flamewar later on. I think the patch could
use being modified to have a way to "switch" between the 2 modes at
runtime though, or at the very least with a kernel command line option.

> Then Bigliardi stopped mantaining it and Matteo Bigoi took over, but
> he always made minor modifications so it could just work with newer
> kernels.

Ok.

> I don't know if the patch is actually ready for upstream, needs
> cleaning or any of the such becaus i always just used it - both autors
> are friends and study colleagues.
> I suggest you - or someone willing to take it upstream - contact Matteo
> Bigoi for all the informations.

It certainly could go upstream with a bit of work, I was just suprised
there was no apperent effort in that direction.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: synaptics PowerBook ?

[Benjamin Herrenschmidt]

On Mon, 2006-10-09 at 23:47 +0200, [EMAIL PROTECTED] wrote:
> On Mon, 09 Oct 2006 1:44:23 +0300
> Eddy Petrișor <[EMAIL PROTECTED]> wrote:
> 
> > Hello,
> > 
> > Some time ago I have followed some threads on different lists/forums
> > about setting up the synaptics driver so you could scroll if the right
> > side of the touchpad was used or if the touchpad was touched with two
> > fingers at the same time.
> > 
> > I tried to configure the touchpad of my powerbook, but I didn't
> > managed to. Is it possible to do that on a PowerBook (5,2)? If it can
> > be done, how?
> 
> i really do not know if the powerbook is adb-based, but there is a
> 
> patch - working also on 19-rc1 - for the synaptics driver on a lot of
> *books:
> 
> http://www.cs.unibo.it/~bigoi/
> 
> you will find all the infos and latest patches there.
> Try it.

Any reason why this patch isn't submited upstream ? :)

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [Fwd: hdc: irq timeout: status=0xc0 { Busy }]

[Benjamin Herrenschmidt]

On Thu, 2006-10-05 at 17:02 -0700, Jack Bates wrote:
> I continue to experience this problem with 2.6.18-mm3. I am about to try
> building a kernel without ide-cd support. I wonder if anyone else is
> experiencing system hangs on PowerBooks associated with ide-cd?

Could be useful if you:

1- tested 2.6.18 not -mm
2- told me what precise machine model that is
3- sent me a complete dmesg log
4- sent me a tarball of /proc/device-tree

thanks.

Ben.

> Thanks, Jack
> email message attachment, "Forwarded message - system hang: hdc: irq
> timeout: status=0xc0 { Busy }"
> >  Forwarded Message 
> > From: Jack Bates <[EMAIL PROTECTED]>
> > To: linux-ide@vger.kernel.org
> > Subject: system hang: hdc: irq timeout: status=0xc0 { Busy }
> > Date: Sun, 17 Sep 2006 21:06:09 -0700
> > 
> > Since upgrading to 2.6.18-rc6-mm1 from 2.6.13-rc6-mm2 my system started
> > hanging. I copied the below error from /var/log/messages.
> > 
> > When my system hangs, it is unresponsive for about 10 seconds, then
> > responsive for 1 second, unresponsive for 10 seconds, then responsive
> > for 1 second, etc. I need to reboot to use my system again.
> > 
> > I am not sleeping or waking, suspending or resuming my system. I am only
> > powering down to recover from this error. It happens roughly every 24
> > hours. I usually wake up to find my system hung, reboot, use it in the
> > morning and again in the evening without problem, but wake to find it
> > hung again next morning.
> > 
> > /dev/hdc is my PowerBook's cd/dvd-rom drive. I don't have a disc in the
> > drive when the error occurs, however I've checked the drive and I'm able
> > to read cds.
> > 
> > tj recommended using libata/PATA, however I can't find an
> > apple/uninorth/powermac PATA controller config option.
> > 
> > Any ideas what is causing this error, why the IDE error handler is
> > getting stuck in a loop, or how to use PATA on my PowerBook?
> > 
> > Much thanks - Jack
> > 
> > Sep 16 14:21:43 fis kernel: hdc: irq timeout: status=0xc0 { Busy }
> > Sep 16 14:21:43 fis kernel: ide: failed opcode was: unknown
> > Sep 16 14:22:25 fis kernel: hdc: status timeout: status=0xc0 { Busy }
> > Sep 16 14:22:25 fis kernel: ide: failed opcode was: unknown
> > Sep 16 14:22:25 fis kernel: Call Trace:
> > Sep 16 14:22:25 fis kernel: [D95656E0] [C00085A4] show_stack+0x3c/0x194
> > (unreliable)
> > Sep 16 14:22:25 fis kernel: [D9565710] [C005452C] softlockup_tick
> > +0x94/0xc4
> > Sep 16 14:22:25 fis kernel: [D9565730] [C0038554] run_local_timers
> > +0x18/0x28
> > Sep 16 14:22:25 fis kernel: [D9565740] [C0038594] update_process_times
> > +0x30/0x7c
> > Sep 16 14:22:25 fis kernel: [D9565750] [C000D30C] timer_interrupt
> > +0x110/0x584
> > Sep 16 14:22:25 fis kernel: [D95657D0] [C0010A64] ret_from_except
> > +0x0/0x14
> > Sep 16 14:22:30 fis kernel: --- Exception: 901 at ide_mm_inb+0x4/0x14
> > Sep 16 14:22:30 fis kernel: LR = ide_wait_stat+0x144/0x1a0
> > Sep 16 14:22:30 fis kernel: [D9565890] [C0184AA0] elv_next_request
> > +0x70/0x1bc (unreliable)
> > Sep 16 14:22:30 fis kernel: [D95658C0] [C02189E4] ide_do_request
> > +0x3f0/0xbc0
> > Sep 16 14:22:30 fis kernel: [D9565940] [C0184CF8] elv_insert+0x10c/0x1f4
> > Sep 16 14:22:30 fis kernel: [D9565960] [C0219288] ide_do_drive_cmd
> > +0xd4/0x12c
> > Sep 16 14:22:30 fis kernel: [D9565990] [C0228CB0]
> > cdrom_queue_packet_command+0x48/0xe8
> > Sep 16 14:22:30 fis kernel: [D95659F0] [C0228FA4] cdrom_check_status
> > +0x70/0x84
> > Sep 16 14:22:30 fis kernel: [D9565AD0] [C022A3F0] cdrom_read_toc
> > +0x3c/0x4a4
> > Sep 16 14:22:30 fis kernel: [D9565B20] [C022B538] idecd_revalidate_disk
> > +0x1c/0x30
> > Sep 16 14:22:30 fis kernel: [D9565B70] [C00A56C0] check_disk_change
> > +0x80/0xc8
> > Sep 16 14:22:30 fis kernel: [D9565B90] [C0265C20] cdrom_open+0x1ac/0xadc
> > Sep 16 14:22:30 fis kernel: [D9565D20] [C0229CAC] idecd_open+0x80/0xf4
> > Sep 16 14:22:30 fis kernel: [D9565D40] [C00A5F84] do_open+0x8c/0x3ec
> > Sep 16 14:22:30 fis kernel: [D9565E80] [C00A651C] blkdev_open+0x34/0x94
> > Sep 16 14:22:30 fis kernel: [D9565EA0] [C0078860] __dentry_open
> > +0xcc/0x254
> > Sep 16 14:22:30 fis kernel: [D9565EC0] [C0078B2C] do_filp_open+0x5c/0x78
> > Sep 16 14:22:30 fis kernel: [D9565F20] [C0078BA0] do_sys_open+0x58/0xf8
> > Sep 16 14:22:30 fis kernel: [D9565F40] [C00103BC] ret_from_syscall
> > +0x0/0x38
> > Sep 16 14:22:30 fis kernel: --- Exception: c01 at 0xfed6c8c
> > Sep 16 14:22:30 fis kernel: LR = 0xff68f54
> > Sep 16 14:22:30 fis kernel: hdc: status timeout: status=0xc0 { Busy }
> > Sep 16 14:22:30 fis kernel: ide: failed opcode was: unknown
> > Sep 16 14:22:30 fis kernel: hdc: status timeout: status=0xc0 { Busy }
> > Sep 16 14:22:30 fis kernel: ide: failed opcode was: unknown
> > Sep 16 14:22:30 fis kernel: hdc: status timeout: status=0xc0 { Busy }
> > Sep 16 14:22:30 fis kernel: ide: failed opcode was: unknown
> > [...]


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact

Re: Debian Etch on IBM JS20 fails on hard drive

[Benjamin Herrenschmidt]

On Tue, 2006-09-26 at 14:51 -0500, Mathew Binkley wrote:
> Hello.  I attemped to install Debian Etch on a IBM JS20 using an iso
> from 09/22/2006.  The install works correctly until it tries (and fails)
> to detect the PATA hard drive.  None of the modules listed will work
> with it.
> 
> When I install SuSE 9.1 on the same blade, it shows the hard drive in
> the logs as:
> 
> Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
> ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
> AMD8111: IDE controller at PCI slot :00:04.1
> AMD8111: chipset revision 3
> AMD8111: :00:04.1 (rev 03) UDMA133 controller
> AMD8111: 100% native mode on irq 32
> ide0: BM-DMA at 0x7c00-0x7c07, BIOS settings: hda:pio, hdb:pio
> ide1: BM-DMA at 0x7c08-0x7c0f, BIOS settings: hdc:pio, hdd:pio
> Probing IDE interface ide0...
> hda: FUJITSU MHT2040AS, ATA DISK drive
> Unhandled interrupt 20, disabled
> ide0 at 0x7400-0x7407,0x6c02 on irq 32
> Probing IDE interface ide1...
> Probing IDE interface ide1...
> hda: max request size: 128KiB
> hda: 78140160 sectors (40007 MB) w/8192KiB Cache, CHS=65535/16/63, UDMA(33)
> hda: cache flushes supported
>  hda: hda1 hda2 hda3
> 
> I'm presuming that the AMD8111 chipset is not currently supported under
> Debian Etch for PPC.  Am I wrong, or is there another explanation?  Thanks!

I suppose the driver isn't enabled in the default config. That should be
easily fixed.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: miboot

[Benjamin Herrenschmidt]

On Mon, 2006-09-25 at 09:18 -0700, Daniel Gimpelevich wrote:
> On Mon, 25 Sep 2006 14:56:50 +0200, Sven Luther wrote:
> 
> > On Sun, Sep 24, 2006 at 03:22:37PM +, rob rob wrote:
> >> Hello all,
> >> I have a powerpc 6500/275 and am trying to install debian etch with the
> >> floppy images from ~wouter, or the beta3 from debian for that matter.
> >> Everything works fine until I insert the net-install.img. Through mail
> >> contact with piotras, after "erroneously" posting on the
> >> debootloaders-miboot list, i found out that all the recent net-install
> >> images are too big for a floppy. I also found the following post which
> >> makes reference to this fact:
> >> http://lists.debian.org/debian-boot/2006/09/msg00823.html Is anybody
> >> looking into this? Is this right list? Should I be posting this on
> >> [EMAIL PROTECTED]
> > 
> > Yes, debian-boot is the right place for that.
> > 
> > Friendly,
> > 
> > Sven Luther
> 
> There is some kind of "debootloaders-miboot" list? Why was I not aware of
> that? Got URL?

It's to be noted that the miboot images generated by the kernel are
busted and have been so for a lng time (possibly most of 2.6.x). We
found that only recently.

That should fix it:

Index: linux-tika/arch/powerpc/boot/Makefile
===
--- linux-tika.orig/arch/powerpc/boot/Makefile  2006-09-26 11:36:30.0 
+1000
+++ linux-tika/arch/powerpc/boot/Makefile   2006-09-26 11:39:51.0 
+1000
@@ -119,7 +119,7 @@
 
 quiet_cmd_gen-miboot = GEN $@
   cmd_gen-miboot = $(OBJCOPY) $(OBJCOPY_MIB_ARGS) \
-  --add-section=$1=$(word 2, $^) $< $@
+  --add-section=$2=$(word 2, $^) $< $@
 
 quiet_cmd_gencoff = COFF$@
   cmd_gencoff = $(OBJCOPY) $(OBJCOPY_COFF_ARGS) $@ && \



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: CPU frequency scaling on G4 systems

[Benjamin Herrenschmidt]

> Maybe the CPU in the mac mini doesn't support frequency scaling ?

Actually, iirc, the CPU does (all 7447A do) but the motherboard doesn't
provide voltage scaling which makes it less interesting. At this point,
the linux driver doesn't support scaling in that case (no voltage
control). It should be relatively trivial to add though.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: iMac G5 fan control in 2.6.16 vs 2.6.17

[Benjamin Herrenschmidt]

On Sat, 2006-09-23 at 15:19 +1000, Alexander McLeay wrote:
> On 23/09/06, Benjamin Herrenschmidt <[EMAIL PROTECTED]> wrote:
> 
> > Make sure the various windfarm_* modules are loaded and that cpufreq
> > support is enabled
> 
> Ah, that is the problem. Do you know why they might load automatically
> under 2.6.16 but not under 2.6.17?

Maybe they were built-in in debian 2.6.16 ?

The windfarm stuff lacks proper auto-load for various reasons (and
always did though I might fix it one day) so it's mostly up to the
distro to properly configure loading of them.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: iMac G5 fan control in 2.6.16 vs 2.6.17

[Benjamin Herrenschmidt]

On Fri, 2006-09-22 at 23:38 +1000, Alexander McLeay wrote:
> I'm running Debian Etch on a rev a iMac G5 (PowerMac8,1) and just
> upgraded today from linux-image-2.6.16-2-powerpc64 (vers 2.6.16-18) to
> linux-image-2.6.17-2-powerpc64 (vers 2.6.17-9). Under 2.6.16, the fan
> control seemed to be working and it was nice and quite. Under 2.6.17,
> the fan is not being slowed down, so the computer is very loud.
> 
> Is this change on purpose/known about? Is there something I can do to
> make it quieter again?

Make sure the various windfarm_* modules are loaded and that cpufreq
support is enabled

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Regression in sleep code for PowerBook5,2

[Benjamin Herrenschmidt]

> OTOH, the old kernel doesn't have the r300 driver compatible with the
> new mesa which I need because I am in the Debian Games Team and need
> it for accelerated grapihcs.

It could be a problem with DRI and r300.. There are some fixes going in
real soon that might help.

Ben.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: dvi output on powerbook g4 15" (5,6)

[Benjamin Herrenschmidt]

On Fri, 2006-09-15 at 15:31 +0200, Michel Dänzer wrote:
> On Fri, 2006-09-15 at 14:51 +0200, Yves-Alexis Perez wrote:
> > 
> > I'm trying to use the dvi output (without vga adaptor) on my mac. It
> > works well if (on the same screen) I use the vga adaptor, but when I use
> > a dvi cable (working on osx), it suddenly doesn't work anymore. Well,
> > Xorg detects it, and adjust the pseudo xinerama mode, but nothing is
> > displayed on it (as if the macintosh detected the screen but the screen
> > didn't detect the mac).
> > 
> > Is anyone had made working the dvi output without vga adaptor ?
> 
> The X.Org radeon driver doesn't support two digital displays at the same
> time yet, see https://bugs.freedesktop.org/show_bug.cgi?id=3621 .

There are ticks to make it work though. Jeremy Kerr collected some tips
at

http://ozlabs.org/~jk/docs/mergefb/

The basic idea is to use a modified radeontool to whack the registers
that X isn't configuring properly.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: iMac G4- etch install iso - xorg

[Benjamin Herrenschmidt]

On Tue, 2006-09-05 at 18:45 +0200, Michel Dänzer wrote:
> On Tue, 2006-09-05 at 16:46 +0200, Ralf Saalmüller wrote:
> > 
> > I think that there's one line missing in the xorg.conf:
> > 
> > Section "Module"
> > Load"fbdevhw"
> > ...
> > EndSection
> > 
> > After I insert the "Load "fbdevhw"" the accelerated nv driver came on 
> > on the whole TFT screen and working.
> 
> If this line is really necessary, that's a bug in the nv driver. It
> should load the fbdevhw module automatically when it needs it.

Most of the time, "nv" shouldn't need fbdev at all anyway...

But from the initial post, it's totally unclear what exact machine model
this user has and which video chip it's using...

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: caution: disc upgrades wallstreet & earlier powerbooks

[Benjamin Herrenschmidt]

> the note on the wall street says it is a hardware
> problem on the motherboard. something to do with udma,
> these were early versions of that aka ata-33. maybe
> newer drives can switch down to older/slower, i don't
> know.

This is weird... first those old motherboards don't support U/DMA at all
(MDMA2 is the best you can get), and the media bay and internal IDE are
separate... I wouldn't have been surprised of such weird bugs in OHare
(the chipset used in the 3400) but Heathrow (used in the wallstreet)
should be much more sane... Oh welll.

> anyway i am trying to live with this after taking 
> my book apart 3 times, returning drive for exchange,
> reformatting, etc leaving it in now, it appears ok. 
> runs great by the way !
> 
> one other thing although appeared ata33 w/ macos8.6,
> if using bootx or any mac stuffs, mac drivers would
> seem to fail with anything less than 9.1. (don't know
> about 10.x..)
> 
> 
> 
> 
> 
> __
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around 
> http://mail.yahoo.com 
> 
> 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Regression in sleep code for PowerBook5,2

[Benjamin Herrenschmidt]

On Sun, 2006-09-03 at 04:02 +0300, Eddy Petrişor wrote:
> On 01/09/06, Eddy Petrişor <[EMAIL PROTECTED]> wrote:
> > > A bit busy with other things At this point, all I can say is that a
> > > bunch of things have been fixed. I would recommend you test linus
> > > current git tree and let me know if it's any better.
> >
> > Ok, I will take the debian configuration and test with linus' current
> > tree (bah, git, again; it was not such a pleasent experience last time
> > I tried it).
> 
> Done, tried it. Almost same result. The screen went blank before
> blocking once and blocked as before another time.
> 
> Do you think I should try to run my old kernel 2.6.14 and see if the
> problem was present back then (compatibility issues with the x.org
> server driver)?

Yes.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Regression in sleep code for PowerBook5,2

[Benjamin Herrenschmidt]

> > >
> > :00:10.0 VGA compatible controller: ATI Technologies Inc RV350
> > [Mobility Radeon 9600 M10] (prog-if 00 [VGA])
> > Subsystem: ATI Technologies Inc RV350 [Mobility Radeon 9600 M10]
> > Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop-
> > ParErr- Stepping- SERR- FastB2B-
> > Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium
> > >TAbort- SERR-  > Latency: 255 (2000ns min), Cache Line Size: 32 bytes
> > Interrupt: pin A routed to IRQ 48
> > Region 0: Memory at b800 (32-bit, prefetchable) [size=128M]
> > Region 1: I/O ports at f400 [size=256]
> > Region 2: Memory at b000 (32-bit, non-prefetchable) [size=64K]
> > Expansion ROM at f100 [disabled] [size=128K]
> > Capabilities: 
> 
> Ben, do you need more info on this? Are you busy with other things or
> di this just slipped? (sorry it sounds demanding, but it seems to be a
> non-native language issue)

A bit busy with other things At this point, all I can say is that a
bunch of things have been fixed. I would recommend you test linus
current git tree and let me know if it's any better.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Regression in sleep code for PowerBook5,2

[Benjamin Herrenschmidt]

On Thu, 2006-08-24 at 16:04 +0300, Eddy Petrişor wrote:
> On 24/08/06, Benjamin Herrenschmidt <[EMAIL PROTECTED]> wrote:
> > On Wed, 2006-08-23 at 11:22 +0300, Eddy Petrişor wrote:
> >
> > > How do I test this?
> >
> > Do a lot of them and see if that locks up :)
> 
> Just after reading you mail I tried to switch to a text console and
> experinced the lock. The laptop was on for about 10 hours and I had
> been browsing, IRC-ing editing  etc. meanwhile.
> 
> What should I do next? :-) Anything I can do to help?

Looks like an X problem to me... what video chip model do you have ?

> > > Does the new kernel needs udev >=0.097-1? I was reluctant to upgrade
> > > to it and stayed at 0.094-? (previous Sid package) because
> > > bluez-bcm203x depends on hotplug, which was provided by the old
> > > version, but not by the new one.
> >
> > No idea.
> 
> It seems not. I have now the new udev in spite of the bluetooth
> regression. Hmm, I could roll-back now, until this issue is fixed.
> 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Regression in sleep code for PowerBook5,2

[Benjamin Herrenschmidt]

On Wed, 2006-08-23 at 11:22 +0300, Eddy Petrişor wrote:

> How do I test this? 

Do a lot of them and see if that locks up :)

> In normal conditions, switching consoles is not a
> problem, or  I haven't observed anything. I was able to reproduce the
> lock yesteday by booting, putting the laptop imediately to sleep,
> working/browsing/chatting in X, then (suspecting the locking will
> appear) closed all apps, tryied to put the laptop to sleep (short
> pressing the power button) and the X session blocked.
> 
> I couldn't swicth to any other console or do anything at that point.
> 
> Does the new kernel needs udev >=0.097-1? I was reluctant to upgrade
> to it and stayed at 0.094-? (previous Sid package) because
> bluez-bcm203x depends on hotplug, which was provided by the old
> version, but not by the new one.

No idea.

> >  - sleep from console mode (no X) is reliable
> I will try that now, after sending this mail (although edited in X, I
> will switch to a console and try to put the laptop to sleep from
> there).
> 
> Thanks.
> 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Regression in sleep code for PowerBook5,2

[Benjamin Herrenschmidt]

On Mon, 2006-08-21 at 01:43 +0300, Eddy Petrişor wrote:
> Hello,
> 
> I have started to experinence lock-ups since the upgrade to the latest
> official Debian Sid kernel for powerpc:
> 
> [EMAIL PROTECTED] ~ $ LANG=C apt-cache policy linux-image-2.6.17-2-powerpc
> linux-image-2.6.17-2-powerpc:
>   Installed: 2.6.17-6
>   Candidate: 2.6.17-6
>   Version table:
>  *** 2.6.17-6 0
> 900 http://ftp.ro.debian.org unstable/main Packages
> 100 /var/lib/dpkg/status
> 
> The lock-ups appear at different times, but I have seen it at the time
> the battery finished and sleep was supposed to be entered to save
> work, I have seen it after trying to close the lid. This did not
> happen with the previous package version which was present in Sid.
> 
> Any help would be appreciated.

I don't know what's going on, as I'm not experiencing that. It should
witch to console mode before doing anything though. If it's not doing
so, it could be that X is the one dying on the console switch... Can you
test if:

 - console switches between X and console are reliable
 - sleep from console mode (no X) is reliable

Thanks,
Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: SCSI/Firewire in alu-book 5,8 not working

[Benjamin Herrenschmidt]

On Thu, 2006-08-17 at 15:27 +0200, Wolfgang Pfeiffer wrote:
> On Thu, Aug 17, 2006 at 07:45:49AM +0200, Benjamin Herrenschmidt wrote:
> > On Sat, 2006-08-12 at 00:26 +0200, Wolfgang Pfeiffer wrote:
> > > On Sat, Aug 12, 2006 at 12:21:53AM +0200, Wolfgang Pfeiffer wrote:
> > > > 
> > > > Sorry: In my first message I made a mistake in the subject line: It's
> > > > corrected in this one ...
> > > > 
> > > 
> > > Still BS: I can only talk for the 5,8 one, *not* for the 5,6 one  ...
> > >  ... :/
> > > 
> > > Sorry #2.
> > 
> > Report this to the linux-1394 mailing list
> 
> Yes, I was planning something like that, i.e. writing to
> [EMAIL PROTECTED] The posting here was meant to
> learn what others experienced with SCSI/FW on newer alubooks ..
> 
> I'll keep [EMAIL PROTECTED] CC-ed when writing to
> linux1394-devel, at least as long as no one votes against it ... :)

Fair enough

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: SCSI/Firewire in alu-book 5,8 not working

[Benjamin Herrenschmidt]

On Sat, 2006-08-12 at 00:26 +0200, Wolfgang Pfeiffer wrote:
> On Sat, Aug 12, 2006 at 12:21:53AM +0200, Wolfgang Pfeiffer wrote:
> > 
> > Sorry: In my first message I made a mistake in the subject line: It's
> > corrected in this one ...
> > 
> 
> Still BS: I can only talk for the 5,8 one, *not* for the 5,6 one  ...
>  ... :/
> 
> Sorry #2.

Report this to the linux-1394 mailing list

Cheers,
Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [PATCH] no-execute -- please test

[Benjamin Herrenschmidt]

> We have a bit per page that says if the page is icache dirty or not.
> On machines with no-execute support, we already avoid flushing the
> page until some process first tries to execute from it.  If we
> extended that to this scheme, when we made a segment executable, we
> would have to find and flush all icache-dirty pages in the segment (up
> to 65536 pages).  We wouldn't want to do that every time we made a
> segment executable - it would need to be optimized (e.g. keep a count
> per segment of icache-dirty pages in the segment).

Note that we need to change the icache flush mecanism anyway as it's
always been racy on ppc32 SMP (though very few people noticed so far :)
and ppc64 SMP with < POWER3 CPUs (without N bit).

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: xorg nv options

[Benjamin Herrenschmidt]

On Sat, 2006-07-15 at 17:03 +0200, Antonio-M. Corbi Bellot wrote:
> El dom, 16-07-2006 a las 00:49 +1000, Benjamin Herrenschmidt escribió:
> 
> > To be more precise, this depends on the type of flat panel that is
> > connected to your machine. Cheap ones (like Apple uses on the iMac G5)
> > have only, I think, 6 bits per components and thus need hardware
> > dithering to avoid uglyness in color gradients.
> 
> I forgot to tell that I only use the internal one in my powerbook-12".
> 
> I haven't been able to use an external monitor/beamer with this laptop
> and xorg. Is it possible now (I have not tried it for a while).

The "nv" driver doesn't support dual head. I think it's possible however
to coerce it into using the external output instead of the internal flat
panel but I don't know the details. Others on the list might.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: xorg nv options

[Benjamin Herrenschmidt]

On Sat, 2006-07-15 at 11:21 +0200, Antonio-M. Corbi Bellot wrote:
> Hi list:
> 
> Under debian/xorg nv driver I had problem showing some graphics,
> specially the ones that had color gradients, the steps of the gradient
> were very noticeable and the same graphic showed much better under OSX.
> 
> Playing with the 'nv' options I found that using 
> 
>   Option  "FPDither" "true"
> 
> in the 'device' section made a big difference.
> 
> If you are having the same 'problem' hope this helps you.

To be more precise, this depends on the type of flat panel that is
connected to your machine. Cheap ones (like Apple uses on the iMac G5)
have only, I think, 6 bits per components and thus need hardware
dithering to avoid uglyness in color gradients.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Need HowTo to solve Ubuntu Breezy PPC system freeze at login on G5.

[Benjamin Herrenschmidt]

On Wed, 2006-07-12 at 11:15 +0200, Johannes Berg wrote:
> On Wed, 2006-07-12 at 10:41 +0200, Michael Schmitz wrote:
> 
> > Search the debian-powerpc list archives for messages relating to sound and
> > G5 - ISTR there is a new sound driver in the works that supports the G5
> > and newer G4 machines.
> 
> It's in Linus current tree, unfortunately the version in -rc1 doesn't
> work. But other than that, snd-aoa works with a PowerMac9,1 iirc. But
> I'm not sure I can recommend running the latest git on a PowerMac9,1
> right now, the irq rework Ben did might have broken something and the
> fixes not in yet. I suggest scouring the linuxppc-dev archives for
> patches from him not in Linus's tree yet, and running latest git + those
> patches.
> 
> > Same thing - work in progress IIRC. BenH and Johannes Berg are the names
> > that come to mind here.
> 
> I don't really have any relation to the thermal control code. :)

All the important bits should be in Linus tree. Thermal should work too
if you load all the windfarm modules.

Ben



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

[PATCH] powerpc: Xserve cpu-meter LEDs

[Benjamin Herrenschmidt]

   "rackmeter: failed to allocate samples page !\n");
+   rc = -ENOMEM;
+   goto bail_release;
+   }
+
+   rm->dma_buf_v = dma_alloc_coherent(&macio_get_pci_dev(mdev)->dev,
+  sizeof(struct rackmeter_dma),
+  &rm->dma_buf_p, GFP_KERNEL);
+   if (rm->dma_buf_v == NULL) {
+   printk(KERN_ERR
+  "rackmeter: failed to allocate dma buffer !\n");
+   rc = -ENOMEM;
+   goto bail_free_samples;
+   }
+#if 0
+   rm->i2s_regs = ioremap(macio_resource_start(mdev, 0), 0x1000);
+#else
+   rm->i2s_regs = ioremap(ri2s.start, 0x1000);
+#endif
+   if (rm->i2s_regs == NULL) {
+   printk(KERN_ERR
+  "rackmeter: failed to map i2s registers !\n");
+   rc = -ENXIO;
+   goto bail_free_dma;
+   }
+#if 0
+   rm->dma_regs = ioremap(macio_resource_start(mdev, 1), 0x100);
+#else
+   rm->dma_regs = ioremap(rdma.start, 0x100);
+#endif
+   if (rm->dma_regs == NULL) {
+   printk(KERN_ERR
+  "rackmeter: failed to map dma registers !\n");
+   rc = -ENXIO;
+   goto bail_unmap_i2s;
+   }
+
+   rc = rackmeter_setup(rm);
+   if (rc) {
+   printk(KERN_ERR
+  "rackmeter: failed to initialize !\n");
+   rc = -ENXIO;
+   goto bail_unmap_dma;
+   }
+
+   rc = request_irq(rm->irq, rackmeter_irq, 0, "rackmeter", rm);
+   if (rc != 0) {
+   printk(KERN_ERR
+  "rackmeter: failed to request interrupt !\n");
+   goto bail_stop_dma;
+   }
+   of_node_put(np);
+   return 0;
+
+ bail_stop_dma:
+   DBDMA_DO_RESET(rm->dma_regs);
+ bail_unmap_dma:
+   iounmap(rm->dma_regs);
+ bail_unmap_i2s:
+   iounmap(rm->i2s_regs);
+ bail_free_dma:
+   dma_free_coherent(&macio_get_pci_dev(mdev)->dev,
+ sizeof(struct rackmeter_dma),
+ rm->dma_buf_v, rm->dma_buf_p);
+ bail_free_samples:
+   free_page((unsigned long)rm->ubuf);
+ bail_release:
+#if 0
+   macio_release_resources(mdev);
+#endif
+ bail_free:
+   kfree(rm);
+ bail:
+   of_node_put(i2s);
+   of_node_put(np);
+   dev_set_drvdata(&mdev->ofdev.dev, NULL);
+   return rc;
+}
+
+static int rackmeter_remove(struct macio_dev* mdev)
+{
+   struct rackmeter *rm = dev_get_drvdata(&mdev->ofdev.dev);
+
+   /* Clear reference to private data */
+   dev_set_drvdata(&mdev->ofdev.dev, NULL);
+
+   /* Stop/reset dbdma */
+   DBDMA_DO_RESET(rm->dma_regs);
+
+   /* Release the IRQ */
+   free_irq(rm->irq, rm);
+
+   /* Unmap registers */
+   iounmap(rm->dma_regs);
+   iounmap(rm->i2s_regs);
+
+   /* Free DMA */
+   dma_free_coherent(&macio_get_pci_dev(mdev)->dev,
+ sizeof(struct rackmeter_dma),
+ rm->dma_buf_v, rm->dma_buf_p);
+
+   /* Free samples */
+   free_page((unsigned long)rm->ubuf);
+
+#if 0
+   /* Release resources */
+   macio_release_resources(mdev);
+#endif
+
+   /* Get rid of me */
+   kfree(rm);
+
+   return 0;
+}
+
+static int rackmeter_shutdown(struct macio_dev* dev)
+{
+   return 0;
+}
+
+static struct of_device_id rackmeter_match[] = {
+   { .name = "i2s" },
+   { }
+};
+
+static struct macio_driver rackmeter_drv = {
+   .name = "rackmeter",
+   .owner = THIS_MODULE,
+   .match_table = rackmeter_match,
+   .probe = rackmeter_probe,
+   .remove = rackmeter_remove,
+   .shutdown = rackmeter_shutdown,
+};
+
+
+static int __init rackmeter_init(void)
+{
+   pr_debug("rackmeter_init()\n");
+
+   return macio_register_driver(&rackmeter_drv);
+}
+
+static void __exit rackmeter_exit(void)
+{
+   pr_debug("rackmeter_exit()\n");
+
+   macio_unregister_driver(&rackmeter_drv);
+}
+
+module_init(rackmeter_init);
+module_exit(rackmeter_exit);
+
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Benjamin Herrenschmidt <[EMAIL PROTECTED]>");
+MODULE_DESCRIPTION("RackMeter: Support vu-meter on XServe front panel");
Index: linux-irq-work/drivers/macintosh/Kconfig
===
--- linux-irq-work.orig/drivers/macintosh/Kconfig   2006-07-10 
16:12:38.0 +1000
+++ linux-irq-work/drivers/macintosh/Kconfig2006-07-10 16:45:50.0 
+1000
@@ -218,4 +218,11 @@
tristate "Support for ANS LCD display"
depends on ADB_CUDA && PPC_PMAC
 
+config PMAC_RACKMETER
+   tristate "

Re: SSA

[Benjamin Herrenschmidt]

On Mon, 2006-07-10 at 11:52 -0400, [EMAIL PROTECTED] wrote:
> Does anyone know if there are plans for linux to support SSA (Serial
> Storage Archetecture), or is it already invisibly there as a low-level
> SCSI driver?  I can get external SSA arays cheaper than I can get the
> mounts for internal SCSI drives for my H50.  
> 
> I will have a 9.? GB drive on internal SCSI for booting.

Not that I know of.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: possibly boot hangs refers to other devices ???

[Benjamin Herrenschmidt]

On Sat, 2006-07-08 at 17:58 -0700, brian wrote:
> sorry, more to tell.
> 
> i just went away for a break and now the disk is making noise.

How so ? While running ? while idle ?

> it is this parking thing, that was like a clunk before i upgraded
> the disk driver of apple from the 8.1 to 9.1 (the 9.2 driver cause
> mac os to hang on restart) - that is now click but it is loud enough
> you notice. also the disk feels rather warm under my palm.
> 
> i guess i will have to go look at hdparm pretty quick at least, or
> go back to bootx for a bit (reference : just zap pram)

if i'ts doing that sort of noise while the machine is ine use, it's time
to think about getting a new disk...

> part of my  urgency on the backups was i  have replacement
> disk for pismo which could then transfer to the 3400 its old one.
> but i need to back up both first. there is not so much hurry,
> except for me to time with etch --> stable, and which machines
> i have would be compatible with that.
> 
> i am a little confuse also about: can use different kernels with
> out reference to version of debian. so for instance 2.4.x with sarge
> or 2.6.15 etc. mixing and matching to me...
> 
> possible i could try one of the old world to netboot from the pismo
> as server to try out some things. but i don't even have LAN working
> yet...
> (just wireless internet - one card shared between powerbooks)
> 
> 
> 
> 
> 
> 
> Benjamin Herrenschmidt <[EMAIL PROTECTED]> wrote:
> On Mon, 2006-07-03 at 22:14 -0700, Daniel Gimpelevich wrote:
> 
> > If you really want to bypass extensions when using BootX,
> boot with the
> > Shift key and use the Control Panel.
> 
> It's not clear to me what machines he's talking about... but I
> think a
> PowerBook G3 model 2000 is a newworld machine which isn't
> supposed to be
> used with BootX.
> 
> 
> In general, I also recommend booting oldworld's via Open
> Firmware when
> possible (it can be made to work with beige G3s for example
> using some
> Apple nvram hacks and netboot works with almost every oldworld
> machine
> out there using bootp/tftp, you can netboot the zImage.coff).
> OF booting
> is always more reliable than BootX. The problem with BootX is
> that it
> "kills" MacOS but that sometimes leaves some hardware devices
> in some
> active state, potentially bus mastering or doing other horrors
> behind
> linux back and corrupting memory. You should use BootX if
> nothing else
> works.
> 
> Ben.
> 
> 
> 
> -- 
> To UNSUBSCRIBE, email to
> [EMAIL PROTECTED]
> with a subject of "unsubscribe". Trouble? Contact
> [EMAIL PROTECTED]
> 
> i am sorry if i have confused people, by mixing topics too much and i
> hope i am
> not doing it again here. 
> 
> i have to say i have put quick on my powerbook3400/Sarge machine now
> and
> it seems to be fine, perhaps a bit better, but not to say all is well
> yet. i took bens
> advice to eliminate possibly bootx quirks with my misc problems. 
>  
> it was sufficient for my boot hangs over scsci "double loading" to
> simply use
> the startup version of bootx for this oldworld machine. that is
> specifically
> to restart rather than using the control panel. but because of my list
> of
> other errors i went ahead and got quik going.
> 
> for reference: i used ofpath command of ybin package to get the path.
> then
> i used "Apple System Disk" to set the open firmware variable including
> that.
> note - it appears, using the command printenv at OF prompt that the
> apple util
> resets the load base and a couple other variables i am not familiar
> with as well.
> i found link to "ASD" at linuxppc/quik page; also note it is not
> necessary, just
> convenenient - if one checks the debian bug reports for quik, there is
> some
> work at automating better on the linix side, as ybin does for yaboot.
> 
> mostly i did not use quik before because i was unsure if i could
> easily still use
> my mac os including conveniently and safely. it took me some while to
> feel assured
> that this was the case. all i do is type bye at the boot prompt rather
> than boot
> (note i told "apple system disk" i wanted a pause). the debian
> installer posts
> a big warning which scared me off and and also some other distros
> advise

Re: possibly boot hangs refers to other devices ???

[Benjamin Herrenschmidt]

> for reference: i used ofpath command of ybin package to get the path.
> then
> i used "Apple System Disk" to set the open firmware variable including
> that.
> note - it appears, using the command printenv at OF prompt that the
> apple util
> resets the load base and a couple other variables i am not familiar
> with as well.
> i found link to "ASD" at linuxppc/quik page; also note it is not
> necessary, just
> convenenient - if one checks the debian bug reports for quik, there is
> some
> work at automating better on the linix side, as ybin does for yaboot.

Note that Apple System Disk will also install some "patches" to OF in
nvram to fix bugs, that are sometimes necessary to reliably boot on some
machines.

> mostly i did not use quik before because i was unsure if i could
> easily still use
> my mac os including conveniently and safely. it took me some while to
> feel assured
> that this was the case. all i do is type bye at the boot prompt rather
> than boot
> (note i told "apple system disk" i wanted a pause). the debian
> installer posts
> a big warning which scared me off and and also some other distros
> advise
> against it. it is not currently set up for beginner to debian even
> transferrees.
> 
> i have my entire linux system as the first partition on the disk, of
> type
> "new world bootblock" w/out a separate partition for quik use ,
> followed by
> swap and mac partitions.

It should not be of type newworld boot. Newworld boot is what yaboot is
generally on for newworld machines and you may end up with problems as
yaboot installer might try to reformat it as HFS, that sort of thing.
 
> more later about my attempts to boot a backup on scsi (internal drive
> here is ata)
> some way some how if not by the powerbook then by a beigeG3/Rev2. (get
> it ?
> back up copy sarge), and perhaps to try the powermanagager stuff again
> on pb3400.
> 
> about the other confusion, sorry also: my testing/etch machine is a
> rather highly
> modified pismo/powerbook-firewire. having used yaboot there also made
> it easier for me to try quik. my problem there is i have no backup,
> yet, but i am
> going to try a different drive/cable.
> 
> backup is important !! computers and disks burn up ! the all do
> eventually,
> often unexpectedly !!
> 
> brian
> 
> 
> 
> 
> 
> 
> __
> How low will we go? Check out Yahoo! Messenger’s low PC-to-Phone call
> rates.


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: TDFX(0): No valid PIO address in PCI config space

[Benjamin Herrenschmidt]

On Sat, 2006-07-08 at 23:10 +0200, [EMAIL PROTECTED] wrote:
> I'm trying to get X-Windows to work but keep running into:
> 
> TDFX(0): No valid PIO address in PCI config space
> 
> What does this Error mean? Any idea what I should try?

Have you tried in one of the other slots ? some of the slots in the 9500
may have issues with PCI IO accesses..

Ben.

> 
> Thanks for your help
> 
> (PowerMac 9500/200, 176MB, Voodoo 3 3000 with 16 MB)
> 
> XFree86.0.log:
> 
> [...]
> (II) Setting vga for screen 0.
> (II) Loading sub module "vgahw"
> (II) LoadModule: "vgahw"
> (II) Loading /usr/X11R6/lib/modules/libvgahw.a
> (II) Module vgahw: vendor="The XFree86 Project"
>   compiled for 4.3.0.1, module version = 0.1.0
>   ABI class: XFree86 Video Driver, version 0.6
> (II) TDFX(0): TDFXFindChips: found 1 chip(s)
> (**) TDFX(0): Depth 24, (--) framebuffer bpp 32
> (==) TDFX(0): RGB weight 888
> (==) TDFX(0): Default visual is TrueColor
> (--) TDFX(0): Chipset: "3dfx Voodoo3"
> (--) TDFX(0): Linear framebuffer at 0xA000
> (--) TDFX(0): MMIO registers at addr 0x9200
> (EE) TDFX(0): No valid PIO address in PCI config space
> (II) UnloadModule: "tdfx"
> (II) UnloadModule: "vgahw"
> (II) Unloading /usr/X11R6/lib/modules/libvgahw.a
> (EE) Screen(s) found, but none have a usable configuration.
> 
> 


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: possibly boot hangs refers to other devices ???

[Benjamin Herrenschmidt]

On Mon, 2006-07-03 at 22:14 -0700, Daniel Gimpelevich wrote:

> If you really want to bypass extensions when using BootX, boot with the
> Shift key and use the Control Panel.

It's not clear to me what machines he's talking about... but I think a
PowerBook G3 model 2000 is a newworld machine which isn't supposed to be
used with BootX.

In general, I also recommend booting oldworld's via Open Firmware when
possible (it can be made to work with beige G3s for example using some
Apple nvram hacks and netboot works with almost every oldworld machine
out there using bootp/tftp, you can netboot the zImage.coff). OF booting
is always more reliable than BootX. The problem with BootX is that it
"kills" MacOS but that sometimes leaves some hardware devices in some
active state, potentially bus mastering or doing other horrors behind
linux back and corrupting memory. You should use BootX if nothing else
works.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Work around for SATA driver bug in Sarge 3.1r2 (how)?

[Benjamin Herrenschmidt]

> I can't tell you where the controller is, but I can tell you that the  
> description matches (i.e. the CD cannot be found). As far as I  
> understand the issue (bug), it isn't that the CD is on the SATA  
> controller, but rather that the SATA controller somehow prevents the  
> loading of the CD drivers. In "etch", the CD loads, with some  
> complaining, by way of some firewire drivers. Here is the info I get  
> from OS X about the CD:

I don't know what debian is mucking around on their kernels. It should
just work. I always do my installs with custom kernels anyway with
everything I need built-in. (a g5_defconfig in your case would work
well)

Ben


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Kernel Oops - Post install on Dual G4 Mac

[Benjamin Herrenschmidt]

On Mon, 2006-07-03 at 12:33 -0500, James Stallings II wrote:
> DOH!!
> 
> Mea Culpa, mea Culpa...
> 
> 
> Not only was the adaptec not integrated on the mainboard, its one I
> put in myself a year ago.
> 
> Oh well, I pulled it but still get the kernel oops in the scsi_mod...
> unfortunately much of the oops output scrolls off the screen. How can
> I page back to the beginning of the crash? ctrl+pgup doesn't seem to
> do it for me - this is with the stock (read default) kernel image.

Hrm.. I have no idea what would be the default Sarge kernel image...
except that it's old ! How are you booting in the first place ? Using a
bootable CD with yaboot on it ? What happens if you try an etch or sid
installer instead  ?

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Work around for SATA driver bug in Sarge 3.1r2 (how)?

[Benjamin Herrenschmidt]

On Wed, 2006-06-28 at 13:02 +0200, Brian Durant wrote:

> I have tried using the above method, but actually can't see how this  
> should work as there doesn't seem to be a way of loading the CD  
> drivers before the SATA driver. A loop develops with the selection of  
> the CD drivers, because the SATA drivers have already loaded and the  
> CD-ROM still cannot be detected, despite only choosing the above  
> quoted CD drivers. Can someone walk me through this, please?
> 
> For those of you that don't understand why I don't just install Etch  
> beta 2, I can say that I have tried that. There is an issue with Etch  
> not being able to detect USB keyboards (and maybe mice as well)  
> during the install. This isn't in the errata and I would consider it  
> a bug. However, as I don't want to run a 'testing" system anyway, it  
> is just another reason to get a Sarge install to work.

The CD on this machine shouldn't be on the SATA controller... it's
supposed to be on the good'old Apple PowerMac IDE controller. I haven't
seen any kind of conflict so far with those...

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Kernel Oops - Post install on Dual G4 Mac

[Benjamin Herrenschmidt]

On Mon, 2006-07-03 at 06:38 -0500, James Stallings II wrote:
> Greetings All,
> 
> I've recently attempted to install debian on my Mac - and it all goes
> quite well until the reboot after installation. It would appear that
> there is an unused adaptec SCSI Host controller integrated into the
> mainboard, which currently has no application on my system (except for
> borking my linux install ;)
> 
> It seems that when the system boots for the first time it probes
> everything in the interest of detecting hardware; unfortunately,
> probing the adaptec throws the kernel into a panic.
> 
> I get the feeling this question is essentially FAQ, but I have been
> unable to find any references to the kernel boot parameters I seem to
> remember for dealing with this scenario; the scsi faqs/doc list
> options for  specifying just about anything in advance except for a
> no-probe or ignore controller option.
> 
> Additionally I tried performing the baseconfig operation from within a
> chroot sheel (expert mode) but this had no effect on the problem.
> 
> Please advise asap, I'd hate to have to give up on this for lack of support.

I don't know of any mac with an on-board adaptec controller... Maybe you
have it as a PCI card in whcih case you can remove it to test if that
helps ? Do you have more details of the actual crash ? Also, what kernel
version is it ?

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: new g5 dual core macs

[Benjamin Herrenschmidt]

On Fri, 2006-06-30 at 18:54 +0200, Thorsten Johannvorderbrueggen wrote:
> hello all,
> 
> it´ s still not working. i downloaded the daily build of etch. the
> installation worked, but all fans are running on maximum speed. after a
> restart, i have the same old problem.

Make sure you have the windfarm_* modules, i2c-powermac and cpufreq
loaded. I don't know if debian properly enables these features. You may
just download a 2.6.17 kernel source tree and build it using the
g5_defconfig that comes with it.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: panics on powerbook3400 & misc problems

[Benjamin Herrenschmidt]

On Thu, 2006-06-29 at 15:47 +0200, Michael Schmitz wrote:
> > > On the topic of boot panics: I just noticed cold booting 2.6.17 (pulled
> > > Tuesday) panics somewhere during knfsd or samba startup. The oops does end
> > > up in the log so it does not seem fatal
> > >
> > > 2.6.17-rc5 does boot fine under the same circumstances.
> >
> > Ouch... looks bad... looks like memory corruption to me.
> 
> I'll try to boil it down to a reasonable test case - in fact, leaving
> samba out of the init scripts it boots fine, and I can hammer it with MOL
> and OO.org. until close to OOM. Ater the system boot has completed, I can
> start samba manually and repeat the stress testing.
> 
> Potential hardware problem that goes away after the machine got used for a
> while?

BTW. Try enabling all the kernel debugging options too, might help spot
the problem sooner.

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: panics on powerbook3400 & misc problems

[Benjamin Herrenschmidt]

On Thu, 2006-06-29 at 15:47 +0200, Michael Schmitz wrote:
> > > On the topic of boot panics: I just noticed cold booting 2.6.17 (pulled
> > > Tuesday) panics somewhere during knfsd or samba startup. The oops does end
> > > up in the log so it does not seem fatal
> > >
> > > 2.6.17-rc5 does boot fine under the same circumstances.
> >
> > Ouch... looks bad... looks like memory corruption to me.
> 
> I'll try to boil it down to a reasonable test case - in fact, leaving
> samba out of the init scripts it boots fine, and I can hammer it with MOL
> and OO.org. until close to OOM. Ater the system boot has completed, I can
> start samba manually and repeat the stress testing.
> 
> Potential hardware problem that goes away after the machine got used for a
> while?

I doubt it

Ben.



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]