headers
John David Anglin | 4 Sep 00:45
Picon

2.6.23-rc5 warnings

Some section mismatches in 64-bit kernel build:

WARNING: vmlinux.o(.text+0xb0): Section mismatch: reference to .init.text:start_
kernel (between '$pgt_fill_loop' and 'common_stext')
WARNING: vmlinux.o(.text+0xb4): Section mismatch: reference to .init.text:start_
kernel (between '$pgt_fill_loop' and 'common_stext')
WARNING: vmlinux.o(.text+0x290): Section mismatch: reference to .init.text:smp_c
allin (between 'smp_slave_stext' and 'flush_tlb_all_local')
WARNING: vmlinux.o(.text+0x294): Section mismatch: reference to .init.text:smp_c
allin (between 'smp_slave_stext' and 'flush_tlb_all_local')
WARNING: vmlinux.o(.text.write_profile+0x19c): Section mismatch: reference to .i
nit.text:setup_profiling_timer (after 'write_profile')

I think there was a warning about a missing syscall fallocate.

Dave
--

-- 
J. David Anglin                                  dave.anglin <at> nrc-cnrc.gc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6602)
Permalink | Reply | 1 comment |
headers
John David Anglin | 4 Sep 03:19
Picon

Re: 2.6.23-rc5 warnings

Thumbs down on 2.6.23-rc5.  I had segmentation faults in tools like
'make' twice in a row.  Back to 2.6.22.6.

Dave
--

-- 
J. David Anglin                                  dave.anglin <at> nrc-cnrc.gc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6602)
Permalink | Reply | 0 comments |
headers
Jesse Dougherty | 4 Sep 20:27

older HP 9000 Workstations wanted

We are looking to purchase / sorce the following older HP 9000 workstations:

HP 9000 715 Apollo workstations
HP 9000 B180L class workstations
HP 9000 B160L class workstations
HP 9000 B132L class workstations
HP 9000 748i class workstation
HP 9000 745i class workstation
HP 9000 743i class workstation CPU board
HP 9000 744L class workstation CPU board

We pay for shipping. If anyone has any of these surplus HP 9000 workstation
or CPU boards, feel free to contact me for take out pricing.

Thanks
Jesse Dougherty
Cypress Technology, Inc
Re-Sellers of HP 3000/9000 products
8565A Somerset Drive
Largo, FL 33773
727-557-0911 / fax 727-557-0121
jesse <at> cypress-tech.com
www.cypress-tech.com
Permalink | Reply | 0 comments |
headers
Helge Deller | 4 Sep 22:17
X-Face
Picon
Picon

Re: booting a 9000/785/C8000

On Friday 31 August 2007, Rick Vernam wrote:
> I'm new to parisc (hppa) machines - never touched one before.
> I'm trying to boot into linux so that i can do a simple
> 	dd if=/dev/null of=/dev/[hard drive]
> I have about 25-30 machines to work on, so I'm looking for a rather clean 
> solution :-)
> 
> I've tried both Gentoo's and Debian's hppa offerings, but am having the same 
> problems on both (i've used gentoo x86/amd64 for years...):
> 1 - I can boot with 'console=tty0 sti=0/4/0/0' (as added by palo) and the boot 
> process will stop with
> 	Console: colour dummy device 160x64
> at this point, it is not hard locked - I can ctrl-alt-del to reboot.  it would 
> probably sit there indefinitely - I did leave it for about 16-18 hours 
> (running when I left work, still there the next day).
> 2 - I can boot without the 'console=tty0' and the kernel will load, but then 
> fail when it can't find initial console (panic: kernel not syncing).
> 
> I've tried with both the AGP 8x Fire GL X1 that comes in the machine, and a 
> PCI Rage XL of some sort.
> 
> The Gentoo kernel is 2.6.20 (most recent livecd, 2007.0), and debian's is also 
> recent cd with 2.6.18 kernel.
> In both cases, the complete kernel command line were pretty much the same 
> (maybe they varied in looptype? ...and order of arguments)
> 
> TERM=linux root=/dev/ram0 init=/linuxrc cdroot looptype=normal 
> loop=/image.loop hda=scsi hdb=scsi console=tty0 sti=/0/4/0/0 sti_font=VGA8x16 
> palo_kernel=0/vmlinux
>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Helge Deller | 4 Sep 22:21
X-Face
Picon
Picon

Re: 2.6.23-rc5 warnings

On Tuesday 04 September 2007, John David Anglin wrote:
> Some section mismatches in 64-bit kernel build:
> 
> WARNING: vmlinux.o(.text+0xb0): Section mismatch: reference to .init.text:start_
> kernel (between '$pgt_fill_loop' and 'common_stext')
> WARNING: vmlinux.o(.text+0xb4): Section mismatch: reference to .init.text:start_
> kernel (between '$pgt_fill_loop' and 'common_stext')
> WARNING: vmlinux.o(.text+0x290): Section mismatch: reference to .init.text:smp_c
> allin (between 'smp_slave_stext' and 'flush_tlb_all_local')
> WARNING: vmlinux.o(.text+0x294): Section mismatch: reference to .init.text:smp_c
> allin (between 'smp_slave_stext' and 'flush_tlb_all_local')
> WARNING: vmlinux.o(.text.write_profile+0x19c): Section mismatch: reference to .i
> nit.text:setup_profiling_timer (after 'write_profile')
> 
> I think there was a warning about a missing syscall fallocate.

The first two warnings are valid for 32bit as well. Reason is:
[head.S:122]		load32          start_kernel,%r11

fallocate() syscall is missing in 32bit too.

Helge
Permalink | Reply | 0 comments |
headers
Matthias Klose | 4 Sep 22:37
Picon

linux32 personality & config.guess

With recent 64bit kernels linux32 seems to work,

$ linux32 uname -m
parisc

but

$ linux32 /usr/share/misc/config.guess
hppa2.0-unknown-linux-gnu

which seems to break some configury only knowing about hppa and
hppa64. Is config.guess correct, and should configure scripts be
changed?

  Matthias
Permalink | Reply | 0 comments |
headers
Helge Deller | 4 Sep 23:15
X-Face
Picon
Picon

Re: booting a 9000/785/C8000

(please keep the parisc-linux list on CC).

On Tuesday 04 September 2007, Rick Vernam wrote:
> On Tuesday 04 September 2007 03:17:56 pm you wrote:
> > On Friday 31 August 2007, Rick Vernam wrote:
> > > I'm new to parisc (hppa) machines - never touched one before.
> > > I'm trying to boot into linux so that i can do a simple
> > > 	dd if=/dev/null of=/dev/[hard drive]
> > > I have about 25-30 machines to work on, so I'm looking for a rather clean
> > > solution :-)
> > >
> > > I've tried both Gentoo's and Debian's hppa offerings, but am having the
> > > same problems on both (i've used gentoo x86/amd64 for years...):
> > > 1 - I can boot with 'console=tty0 sti=0/4/0/0' (as added by palo) and the
> > > boot process will stop with
> > > 	Console: colour dummy device 160x64
> > > at this point, it is not hard locked - I can ctrl-alt-del to reboot.  it
> > > would probably sit there indefinitely - I did leave it for about 16-18
> > > hours (running when I left work, still there the next day).
> > > 2 - I can boot without the 'console=tty0' and the kernel will load, but
> > > then fail when it can't find initial console (panic: kernel not syncing).
> > >
> > > I've tried with both the AGP 8x Fire GL X1 that comes in the machine, and
> > > a PCI Rage XL of some sort.
> > >
> > > The Gentoo kernel is 2.6.20 (most recent livecd, 2007.0), and debian's is
> > > also recent cd with 2.6.18 kernel.
> > > In both cases, the complete kernel command line were pretty much the same
> > > (maybe they varied in looptype? ...and order of arguments)
> > >
(Continue reading)

Permalink | Reply | 0 comments |
headers
Helge Deller | 4 Sep 23:44
X-Face
Picon
Picon

crash in mux_poll on K200/Kittyhawk (was: Hardware-List (fwd))

Forwarding to parisc-linux mailing list.....

Can someone with MUX knowledge take a look into this problem ?
Additionally, does someone knows what this device is ?
> 12. unknown device at 0xf190c000 [10/4/12] { 2, 0x0, 0x004, 0x0005f }

Thanks,
Helge

On Tuesday 04 September 2007, Rob Sciuk wrote:
> On Tue, 4 Sep 2007, Helge Deller wrote:
> > On Wednesday 29 August 2007, Rob Sciuk wrote:
> >>> I'm having some difficulties with Debian 4.0R1 on an HP9000 Series
> >>> K200, also experienced similar behaviours with Sarge 3.1R6 and Debian
> >>> 4.0R0, where the install freezes/panics at the point at which it
> >>> states:
> >>>
> >>>   	Freeing unused kernel memory: 328K freed.
> >>>
> >>> The Gentoo 2007.0 release on the otherhand goes on to boot the full
> >>> live cd.  I'd very much like to run Debian, though I am just now
> >>> putting Gentoo on a 712/60, and I might resort to that for the K200.
> >>> Can you help me out here or provide pointers which can get me through
> >>> the boot process ??
> >
> > Hi Rob,
> > could you please post the full boot log (e.g. captured with minicom on serial port) ?
> > Helge
> 
> Full log is attached, captured with cu/tip on FreeBSD
(Continue reading)

Permalink | Reply | 0 comments |
headers
John David Anglin | 4 Sep 23:48
Picon

Re: linux32 personality & config.guess

> With recent 64bit kernels linux32 seems to work,
> 
> $ linux32 uname -m
> parisc
> 
> but
> 
> $ linux32 /usr/share/misc/config.guess
> hppa2.0-unknown-linux-gnu
> 
> which seems to break some configury only knowing about hppa and
> hppa64. Is config.guess correct, and should configure scripts be
> changed?

GCC accepts either parisc or hppa, at least in config.gcc.  I tend
to think the parisc form should be deprecated since no other PA-RISC
targets use it.  The others all use hppa.

With respect to hppa variants, this is a hopeless mess and I don't
believe this can be fixed.  For example, libgmp treats hppa2.0w as
indicating a 64-bit runtime.  I'm sure you have hit this.

Nominally, the first part of the target string represents the
architecture of the kernel, and not the userspace architecture.
This is not sufficient to configure runtime applications when multiple
architectures are supported by one kernel.

Dave
--

-- 
J. David Anglin                                  dave.anglin <at> nrc-cnrc.gc.ca
(Continue reading)

Permalink | Reply | 0 comments |
headers
Matthew Wilcox | 5 Sep 00:15

Unknown device 0x0005f

On Tue, Sep 04, 2007 at 11:44:29PM +0200, Helge Deller wrote:
> Additionally, does someone knows what this device is ?
> > 12. unknown device at 0xf190c000 [10/4/12] { 2, 0x0, 0x004, 0x0005f }

I don't.  We don't have anything with the 0x0005f programming model in
the database.  We have:

        {HPHW_A_DMA, 0x004, 0x00050, 0x80, "Lanbrusca 802.3 (36967A)"}, 
        {HPHW_B_DMA, 0x008, 0x00051, 0x80, "HP-PB Transit 802.3"}, 
        {HPHW_B_DMA, 0x004, 0x00052, 0x80, "Miura LAN/Console (J2146A)"}, 
        {HPHW_A_DMA, 0x004, 0x00056, 0x80, "HP-PB LoQuix FDDI"}, 
        {HPHW_A_DMA, 0x004, 0x00057, 0x80, "HP-PB LoQuix FDDI (28670A)"}, 
        {HPHW_B_DMA, 0x008, 0x00058, 0x80, "HP-PB Transit 802.4"}, 
        {HPHW_A_DMA, 0x004, 0x0005E, 0x00, "Gecko Add-on Token Ring"}, 

The owner could probably help us out by taking a look at the card in
slot 10/4/12 and seeing what chips are on it, and what connectors it has.

--

-- 
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours.  We can't possibly take such
a retrograde step."
Permalink | Reply | 0 comments |

Gmane