Nabble - netbsd-ports

Re: mips64 on port-cobalt

2009-12-18T06:07:06Z

On Fri, Dec 18, 2009 at 2:06 AM, Markus W Kilbinger <mk@...> wrote:

>>>>>> "Andy" == Andy Ruhl <acruhl@...> writes:
>
> >> panic: CPU type (0xffffffff) not supported [... and so on]
>
> Andy> I tried it and got the same thing. Well, I presume I did, it
> Andy> was just scrolling this stuff "forever".
>
> -current's classic/mipsel, e.g. GENERIC kernel seem to have problems,
> too:
>
> When starting the first program / 'init' it displays a message about
> '/usr/lib/O32' path and hangs/stops.
>
> Is this reproducable with gxemul?

I'm not sure which kernel you're referring to. I don't know what
classic/mipsel is? Show me a kernel to boot and I'll try it. I didn't
see this with a GENERIC kernel for Cobalt (which is mipsel) from
http://nyftp.netbsd.org/pub/NetBSD-daily/HEAD/200912140000Z/cobalt/binary/kernel/.

I'm only booting the kernel, not a userland so I don't think I'd see
what you're talking about anyway...

Andy

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-18T06:07:06Z

On Fri, Dec 18, 2009 at 2:06 AM, Markus W Kilbinger <mk@...> wrote:

From forum: port-cobalt

re: mips64 on port-cobalt

2009-12-18T01:10:32Z

>>>>> "Andy" == Andy Ruhl <acruhl@...> writes:

>> panic: CPU type (0xffffffff) not supported [... and so on]

Andy> I tried it and got the same thing. Well, I presume I did, it
Andy> was just scrolling this stuff "forever".

-current's classic/mipsel, e.g. GENERIC kernel seem to have problems,
too:

When starting the first program / 'init' it displays a message about
'/usr/lib/O32' path and hangs/stops.

Is this reproducable with gxemul?

this shoudn't be happening.

can you please describe exactly what you're doing?

thanks.

.mrg.

From forum: port-mips

re: mips64 on port-cobalt

2009-12-18T01:10:32Z

From forum: port-cobalt

Re: mips64 on port-cobalt

2009-12-18T01:06:20Z

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-18T01:06:20Z

From forum: port-cobalt

VMWare virtual machine running NetBSD crashes with "Flexible" network interface

2009-12-17T22:43:19Z

Hello,

I've experienced this problem on both the free version of VMWare
running under Linux and VMWare ESX server. With the Flexible network
driver and a virtual IDE hard disk, a VMWare machine running NetBSD will
just "go away" under a combination of high network and disk load.
There's no crash, the virtual machine just goes "power's off"

The fix for this problem is to use the e1000 driver. The free Linux
version doesn't show this as an option, unless you select "Other
(64-bit)" as the OS type. This works even if you're running the 32-bit
version of VMWare. I figured I should post this here in case someone
else runs in to the same problem. I'm attaching the VMWare crash info below.

Jason M.

Sep 22 04:07:33.342: vcpu-0| Core dump with build build-175625
Sep 22 04:07:33.441: vcpu-0| Writing monitor corefile
"/vmfs/volumes/4a1c6645-8d
a19ecc-0f2a-0022192f8d1d/New Virtual Machine/vmware-core.gz"
Sep 22 04:07:33.441: vcpu-0| CoreDump: dumping core with superuser
privileges
Sep 22 04:07:33.504: vcpu-0| Saving busmem frames
Sep 22 04:07:33.506: vcpu-0| Saving anonymous memory
Sep 22 04:07:33.508: vcpu-0| Beginning monitor coredump
Sep 22 04:07:33.688: vcpu-0| End monitor coredump
Sep 22 04:07:33.688: vcpu-0| Beginning extended monitor coredump
Sep 22 04:07:33.688: vcpu-0| Writing anonymous pages at pos: 401000
Sep 22 04:07:34.264: vcpu-0| Writing monitor corefile
"/vmfs/volumes/4a1c6645-8d
a19ecc-0f2a-0022192f8d1d/New Virtual Machine/vmware64-core.gz"
Sep 22 04:07:34.264: vcpu-0| CoreDump: dumping core with superuser
privileges
Sep 22 04:07:34.289: vcpu-0| VMK Stack for vcpu 0 is at 0x4100c6701000
Sep 22 04:07:34.289: vcpu-0| Saving busmem frames
Sep 22 04:07:34.291: vcpu-0| Saving anonymous memory
Sep 22 04:07:34.293: vcpu-0| Beginning monitor coredump
Sep 22 04:07:35.324: vcpu-0| End monitor coredump
Sep 22 04:07:35.325: vcpu-0| Beginning extended monitor coredump
Sep 22 04:07:35.325: vcpu-0| Writing anonymous pages at pos: 4008000
Sep 22 04:07:37.555: vcpu-0| Msg_Post: Error
Sep 22 04:07:37.555: vcpu-0| [msg.log.monpanic] *** VMware ESX internal
monitor
error ***
Sep 22 04:07:37.555: vcpu-0| vcpu-0:NOT_IMPLEMENTED
devices/net/vlance_shared.c:
705
Sep 22 04:07:37.555: vcpu-0| [msg.log.monpanic.report] Please report
this proble
m by selecting menu item Help > VMware on the Web > Request Support, or
by going
to the Web page
"http://www.vmware.com/info?id=8&logFile=%2fvmfs%2fvolumes%2f4a
1c6645%2d8da19ecc%2d0f2a%2d0022192f8d1d%2fNew%20Virtual%20Machine%2fvmware%2elog

&coreLocation=%2fvmfs%2fvolumes%2f4a1c6645%2d8da19ecc%2d0f2a%2d0022192f8d1d%2fNe

w%20Virtual%20Machine%2fvmware%2dcore%2egz%2c%20%2fvmfs%2fvolumes%2f4a1c6645%2d8

da19ecc%2d0f2a%2d0022192f8d1d%2fNew%20Virtual%20Machine%2fvmware64%2dcore%2egz".

Please provide us with the log file
(/vmfs/volumes/4a1c6645-8da19ecc-0f2a-00221
92f8d1d/New Virtual Machine/vmware.log) and the core file(s)
(/vmfs/volumes/4a1c
6645-8da19ecc-0f2a-0022192f8d1d/New Virtual Machine/vmware-core.gz,
/vmfs/volume
s/4a1c6645-8da19ecc-0f2a-0022192f8d1d/New Virtual
Machine/vmware64-core.gz, /vmf
s/volumes/4a1c6645-8da19ecc-0f2a-0022192f8d1d/New Virtual
Machine/vmware-vmx-zdu
mp.010).
Sep 22 04:07:37.555: vcpu-0| [msg.log.monpanic.serverdebug] If the
problem is re
peatable, please set 'Use Debug Monitor' to 'Yes' in the 'Misc' section
of the C
onfigure Virtual Machine Web page. Then reproduce the incident and file
it accor
ding to the instructions.
Sep 22 04:07:37.556: vcpu-0| [msg.log.monpanic.vmSupport.vmx86] To
collect data
to submit to VMware support, run "vm-support".
Sep 22 04:07:37.556: vcpu-0| [msg.log.monpanic.entitlement] We will
respond on t
he basis of your support entitlement.
Sep 22 04:07:37.556: vcpu-0| [msg.log.monpanic.finish] We appreciate
your feedba
ck,
Sep 22 04:07:37.556: vcpu-0| -- the VMware ESX team.
Sep 22 04:07:37.556: vcpu-0| ----------------------------------------
Sep 22 04:07:37.579: vcpu-0| Exiting vcpu-0
Sep 22 04:07:37.659: vmx| VTHREAD watched thread 4 "vcpu-0" died
Sep 22 04:07:38.559: mks| VTHREAD watched thread 0 "vmx" died

From forum: port-i386

Re: mips64 on port-cobalt

2009-12-17T19:36:12Z

On Thu, Dec 17, 2009 at 2:43 PM, Andy Ruhl <acruhl@...> wrote:
> I tried it and got the same thing. Well, I presume I did, it was just
> scrolling this stuff "forever".
>
> I might give it a try on gxemul, I think they have an emulated Cobalt machine.

Same thing with gxemul as well (get it from pkgsrc). It's worth a shot
to see if it works if you make changes:

gxemul -E Cobalt -x cobalt64

I booted a GENERIC Cobalt kernel OK so it at least does something.

Andy

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T19:36:12Z

From forum: port-cobalt

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T19:32:10Z

for any i386 host which isn't able to load kernel modules, this is a real question.

so, are there that many? and are there drivers in that list, which are not ameanable to being kld-ized?

otherwise, while its an inconvenience, surely this is really just a request to prune GENERIC, because MONOLITHIC is left un-altered, and given a current boot loader, and a pruned GENERIC plus a floppy or other media, the drivers can be (re)loaded as needed.

or have I completely misunderstood?

google, or one of the like waterfall systems would simply fork the prebuilt kernel set into a new subclass, and move on. surely with an automated build, we can leave this one in the space of 'somebody does it for you' and stop worrying...

-G

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T19:32:07Z

On Thu, Dec 17, 2009 at 10:14:54PM -0500, der Mouse wrote:
>
> I know one of the machines I've worked with recently has a BusLogic
> SCSI card in it, though I can't recall whether my own or at work. I
> *think* it's a PCI BusLogic, but would have to track the machine down
> and check to be sure. Would it be of value for me to?

Yes, particularly if it is one of the models which can be configured via
its BIOS to not also show up at the traditional ISA BusLogic/Adaptec
register addresses (I've never understood how that works with a
subtractive-decode PCI-ISA bridge also in the system, either....).

--
Thor Lancelot Simon tls@...
"All of my opinions are consistent, but I cannot present them all
at once." -Jean-Jacques Rousseau, On The Social Contract

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T19:14:54Z

>>> bha_isa
>> No idea, but there is a report about PCI one recently?
> But I think I have let my last BusLogic controller go, so I can't
> test with real hardware any more to see...

I know one of the machines I've worked with recently has a BusLogic
SCSI card in it, though I can't recall whether my own or at work. I
*think* it's a PCI BusLogic, but would have to track the machine down
and check to be sure. Would it be of value for me to?

/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse@...
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

From forum: port-i386

NetBSD on Xserve G5

2009-12-17T16:42:19Z

I am a long-time NetBSD user, albeit on the Alpha platform. We have two Apple Xserve G5 systems that currently have Mac OS X 10.3 Panther installed. That OS is well past end-of-life, so I would like to install a currently-supported open source OS so that we may continue to use this hardware.

My ongoing familiarity with NetBSD (and *BSD in general) naturally inclines me towards installing NetBSD. I checked the supported hardware page of the NetBSD/macppc port Web site and discovered that the Xserve G5 is listed as supported. However, the install guide for NetBSD/macppc 5 lists "On-board SATA controller on G5 systems" under "Unsupported devices." Is this still the case? If so, this would likely be a show-stopper for us. (I am happy to try a -current version on this system if it fixes the lack of onboard SATA support.)

The Xserve G5s we have do not have any installed video cards. They do appear to have a serial port on the back. I'm assuming installing via a serial console is a straightforward supported option.

If anyone has successfully installed NetBSD on an Xserve G5, or is currently using such a setup, I'd be grateful to hear whether you give it a thumbs up or thumbs down; if there are any caveats; and so on. Alternatively, if anyone knows of a better supported open source OS for the Xserve G5, I'd be glad to hear of persuasive suggestions. (I'd much rather stick with *BSD, though.)

Finally, is NetBSD/macppc on the Xserve G5 a 32- or 64-bit system?

Cheers,

Paul.

e-mail: paul@...

"Without music to decorate it, time is just a bunch of boring production
deadlines or dates by which bills must be paid."
--- Frank Vincent Zappa

From forum: port-macppc

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T14:08:36Z

On Fri, Dec 18, 2009 at 04:42:51AM +0900, Izumi Tsutsui wrote:
>
> > bha_isa
>
> No idea, but there is a report about PCI one recently?

Yeah -- I don't have physical hardware, but it fails to attach
with VirtualBox's emulated 946C, which does not present the
ISA view of the registers. Thinking about it I am not sure I
*know* the PCI attachment has worked for quite some time.

But I think I have let my last BusLogic controller go, so I
can't test with real hardware any more to see...

Thor

From forum: port-i386

Re: mips64 on port-cobalt

2009-12-17T13:43:00Z

On Thu, Dec 17, 2009 at 2:05 PM, Markus W Kilbinger <mk@...> wrote:

>>>>>> "Matt" == Matt Thomas <matt@...> writes:
>
> Matt> The kernel didn't exist until a few hours ago.
>
> Matt> I put one at
> Matt> ftp://ftp.netbsd.org/pub/NetBSD/misc/matt/cobalt64.gz
>
> I tried to boot that kernel (uncompressed) on my otherwise unchanged
> Qube 2, but it failed some how:
>
> Cobalt Microserver Diagnostics - 'We serve it, you surf it'
> Built Tue May 25 15:58:41 PDT 1999
>
> 1.LCD Test................................PASS
> 2.Controller Test.........................PASS
> 5.Bank 0:.................................64M
> 6.Bank 1:.................................64M
> 7.Bank 2:.................................64M
> 8.Bank 3:.................................64M
> 9.Serial Test.............................PASS
> 10.PCI Expansion Slot....................**EMPTY**
> 12.IDE Test................................PASS
> 13.Ethernet Test...........................PASS
> 16.RTC Test................................PASS
> BOOTLOADER ramcode: selected partition /dev/hda1
> Decompressing done
> Executing bootloader kernel...
> Decompressing -- done.
>
> >> NetBSD/cobalt 5.99.22 Bootloader, Revision 0.9 [@0x80f00000]
> >> (root@q, Fri Dec 11 12:55:36 MET 2009)
> >> Model: Cobalt Qube 2
> >> Memory: 262144 k
> >> PROM boot string: root=/dev/hda1 ro
> Boot [wd0a:netbsd]: cobalt64
> Loading: wd0a:cobalt64
> 3809296+306360 [220800+211677]=0x456b34
> Starting at 0x80001000
>
> panic: CPU type (0xffffffff) not supported
> Type 1 @0xffffffff80f11220
> Type 2 @0xffffffff80f1122c
> Type 3 @0xffffffff80f11284
> Type 1 @0xffffffff80f11220
> Type 2 @0xffffffff80f1122c
> Type 3 @0xffffffff80f11284
> Type 4 @0xffffffff80f11298
> Type 5 @0xffffffff80f112a4
> invalid bootinfo (standalone boot?)
> panic: CPU type (0xffffffff) not supported
> Type 1 @0xffffffff80f11220
> Type 2 @0xffffffff80f1122c
> Type 3 @0xffffffff80f11284
> Type 1 @0xffffffff80f11220
> Type 2 @0xffffffff80f1122c
> Type 3 @0xffffffff80f11284
> Type 4 @0xffffffff80f11298
> Type 5 @0xffffffff80f112a4
> invalid bootinfo (standalone boot?)
> panic: CPU type (0xffffffff) not supported
> [... and so on]
>
> What can I do/try next?

I tried it and got the same thing. Well, I presume I did, it was just
scrolling this stuff "forever".

I might give it a try on gxemul, I think they have an emulated Cobalt machine.

Andy

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T13:43:00Z

On Thu, Dec 17, 2009 at 2:05 PM, Markus W Kilbinger <mk@...> wrote:

I tried it and got the same thing. Well, I presume I did, it was just
scrolling this stuff "forever".

I might give it a try on gxemul, I think they have an emulated Cobalt machine.

Andy

From forum: port-cobalt

Re: mips64 on port-cobalt

2009-12-17T13:05:55Z

>>>>> "Matt" == Matt Thomas <matt@...> writes:

Matt> The kernel didn't exist until a few hours ago.

Matt> I put one at
Matt> ftp://ftp.netbsd.org/pub/NetBSD/misc/matt/cobalt64.gz

I tried to boot that kernel (uncompressed) on my otherwise unchanged
Qube 2, but it failed some how:

Cobalt Microserver Diagnostics - 'We serve it, you surf it'
Built Tue May 25 15:58:41 PDT 1999

1.LCD Test................................PASS
2.Controller Test.........................PASS
5.Bank 0:.................................64M
6.Bank 1:.................................64M
7.Bank 2:.................................64M
8.Bank 3:.................................64M
9.Serial Test.............................PASS
10.PCI Expansion Slot....................**EMPTY**
12.IDE Test................................PASS
13.Ethernet Test...........................PASS
16.RTC Test................................PASS
BOOTLOADER ramcode: selected partition /dev/hda1
Decompressing done
Executing bootloader kernel...
Decompressing -- done.

>> NetBSD/cobalt 5.99.22 Bootloader, Revision 0.9 [@0x80f00000]
>> (root@q, Fri Dec 11 12:55:36 MET 2009)
>> Model: Cobalt Qube 2
>> Memory: 262144 k
>> PROM boot string: root=/dev/hda1 ro
Boot [wd0a:netbsd]: cobalt64
Loading: wd0a:cobalt64
3809296+306360 [220800+211677]=0x456b34
Starting at 0x80001000

panic: CPU type (0xffffffff) not supported
Type 1 @0xffffffff80f11220
Type 2 @0xffffffff80f1122c
Type 3 @0xffffffff80f11284
Type 1 @0xffffffff80f11220
Type 2 @0xffffffff80f1122c
Type 3 @0xffffffff80f11284
Type 4 @0xffffffff80f11298
Type 5 @0xffffffff80f112a4
invalid bootinfo (standalone boot?)
panic: CPU type (0xffffffff) not supported
Type 1 @0xffffffff80f11220
Type 2 @0xffffffff80f1122c
Type 3 @0xffffffff80f11284
Type 1 @0xffffffff80f11220
Type 2 @0xffffffff80f1122c
Type 3 @0xffffffff80f11284
Type 4 @0xffffffff80f11298
Type 5 @0xffffffff80f112a4
invalid bootinfo (standalone boot?)
panic: CPU type (0xffffffff) not supported
[... and so on]

What can I do/try next?

Markus.

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T13:05:55Z

From forum: port-cobalt

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T13:02:06Z

On Thu, Dec 17, 2009 at 10:46:50AM -0800, Jason Thorpe wrote:
> We dropped support for genuine i386 systems a wile ago now, but we are
> still carrying around a bunch of legacy drivers that should probably
> go... namely ISA SCSI, tape controller, and CD-ROM controller drivers.
>
> Any reason to keep them around?

I agree with the sentiment expressed earlier that ones that are just ISA
attachments for drivers that support other busses should stay.

mcd(4) can probably go. wt(4) probably too.. I think it hasn't worked
right in a decade+, though I no longer remember the issues.

--rafal

--
Time is an illusion; lunchtime, doubly so. |/\/\| Rafal Boni
-- Ford Prefect |\/\/| rafal@...

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T12:11:54Z

Jarle Greipsland <jarle@...> writes:

>> aha_isa
>> aic_isa
> I'm still using these two. Please do not remove them.

Same here!

-tih
--
Self documenting code isn't. User application constraints don't. --Ed Prochak

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T11:42:51Z

> adv_isa

This one also has major PCI variants, so I'm not sure
if it's worth to remove only ISA variant.

> aha_isa

This one is a good example to test MD bounce buffer code
(or DMA address translations) as you wrote in bus_dma paper ;-)

> ahc_isa

This one also have PCI/EISA variants.
(BTW ARC DescStation Tyne here has VL-bus slots so I wondered
if we should move ahc_isa.c from arch/i386/isa to dev/isa ;-)

> aic_isa

There are many users of pcmcia variants.
This dumb (PIO only) ISA one might still be useful on testing
some ancient machines (like some prep) which have ISA
but no IDE?

> bha_isa

No idea, but there is a report about PCI one recently?

> esp_isa

I doubt this ever worked, but this just uses MI ncr53c9x.
(and I refered it when I wrote pcscp(4))

BTW, I still have ESP406 based ISA PnP card,
but it has not worked in my local tree..

> mcd.c
> seagate
> uha_isa
> wt

At least, these ones (and wds(4)) can be omitted from GENERIC.
---
Izumi Tsutsui

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T11:41:28Z

On Thu, Dec 17, 2009 at 10:46:50AM -0800, Jason Thorpe wrote:
> Any reason to keep them around?
>
> seagate

sea at isa? I need that one.

Dave

--
David Young OJC Technologies
dyoung@... Urbana, IL * (217) 278-3933

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T11:28:41Z

Jason Thorpe <thorpej@...> writes:
> We dropped support for genuine i386 systems a wile ago now, but
> we are still carrying around a bunch of legacy drivers that
> should probably go... namely ISA SCSI, tape controller, and
> CD-ROM controller drivers.

> Any reason to keep them around?
>
> aha_isa
> aic_isa
I'm still using these two. Please do not remove them.

-jarle
--
"On a clear disk you can seek forever."
-- Peter O'Toole (Trinity College, Dublin)

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T11:06:22Z

On Thu, 17 Dec 2009 19:01:19 +0000
Matthias Scheler <tron@...> wrote:

> On Thu, Dec 17, 2009 at 10:46:50AM -0800, Jason Thorpe wrote:
> > Any reason to keep them around?
>
> There are i486 and Pentium systems with ISA slots that we still support.

ISA slots can be found all the way up to pentium III motherboards.
Also on quite a few machines supported by NetBSD/alpha.

I think those drivers can be safely commented out in GENERIC though.

-Tobias

From forum: port-i386

Re: Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T11:01:19Z

On Thu, Dec 17, 2009 at 10:46:50AM -0800, Jason Thorpe wrote:
> Any reason to keep them around?

There are i486 and Pentium systems with ISA slots that we still support.

> mcd.c

That one can probably really go.

Kind regards

--
Matthias Scheler http://zhadum.org.uk/

From forum: port-i386

Time to prune some drivers? ISA SCSI, tape, etc.

2009-12-17T10:46:50Z

We dropped support for genuine i386 systems a wile ago now, but we are still carrying around a bunch of legacy drivers that should probably go... namely ISA SCSI, tape controller, and CD-ROM controller drivers.

Any reason to keep them around?

adv_isa
aha_isa
ahc_isa
aic_isa
bha_isa
esp_isa
mcd.c
seagate
uha_isa
wt

-- thorpej

From forum: port-i386

Re: mips64 on port-cobalt

2009-12-17T08:51:25Z

On Dec 17, 2009, at 8:44 AM, Andy Ruhl wrote:

> On Thu, Dec 17, 2009 at 9:28 AM, Matt Thomas <matt@...> wrote:
>>
>> On Dec 17, 2009, at 7:49 AM, Andy Ruhl wrote:
>>
>>> On Thu, Dec 17, 2009 at 8:43 AM, Matt Thomas <matt@...> wrote:
>>>>> I can netboot it if that's good enough for you for now. I have another
>>>>> disk I can install into as well if you need.
>>>>
>>>> Booting with a O32 userland wold be a nice start. Let's see if the kernel even works.
>>>
>>> Where do I get it? I'm not finding it in the places I'm looking.
>>
>> define "it". O32 userland is the standard cobalt userland you are already using.
>
> The kernel. I'm not aware of how to build it if I need to, or if it's
> part of the daily builds or something.
>
> I might have missed something though...

The kernel didn't exist until a few hours ago.

I put one at ftp://ftp.netbsd.org/pub/NetBSD/misc/matt/cobalt64.gz

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T08:44:13Z

On Thu, Dec 17, 2009 at 9:28 AM, Matt Thomas <matt@...> wrote:

>
> On Dec 17, 2009, at 7:49 AM, Andy Ruhl wrote:
>
>> On Thu, Dec 17, 2009 at 8:43 AM, Matt Thomas <matt@...> wrote:
>>>> I can netboot it if that's good enough for you for now. I have another
>>>> disk I can install into as well if you need.
>>>
>>> Booting with a O32 userland wold be a nice start. Let's see if the kernel even works.
>>
>> Where do I get it? I'm not finding it in the places I'm looking.
>
> define "it". O32 userland is the standard cobalt userland you are already using.

The kernel. I'm not aware of how to build it if I need to, or if it's
part of the daily builds or something.

I might have missed something though...

Andy

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T08:28:10Z

On Dec 17, 2009, at 7:49 AM, Andy Ruhl wrote:

> On Thu, Dec 17, 2009 at 8:43 AM, Matt Thomas <matt@...> wrote:
>>> I can netboot it if that's good enough for you for now. I have another
>>> disk I can install into as well if you need.
>>
>> Booting with a O32 userland wold be a nice start. Let's see if the kernel even works.
>
> Where do I get it? I'm not finding it in the places I'm looking.

define "it". O32 userland is the standard cobalt userland you are already using.

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T07:49:38Z

On Thu, Dec 17, 2009 at 8:43 AM, Matt Thomas <matt@...> wrote:
>> I can netboot it if that's good enough for you for now. I have another
>> disk I can install into as well if you need.
>
> Booting with a O32 userland wold be a nice start. Let's see if the kernel even works.

Where do I get it? I'm not finding it in the places I'm looking.

Andy

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T07:44:21Z

On Dec 17, 2009, at 6:52 AM, Markus W Kilbinger wrote:

>>>>>> "Matt" == Matt Thomas <matt@...> writes:
>
> Matt> The algorithmics P5064 already runs a mips64 kernel so the
> Matt> Qube could as well (since they share the same processor). I
> Matt> can make cobalt support mips64 is an hour or two. Is someone
> Matt> willing to test it?
>
> Yes, me! :-)
>
> What about all the compat-lib stuff? Do I have to build and install
> them, too (how, then?)? Or will the new mips64 kernel automatically
> handle the (only) existing old libs correctly?

C++ is still broken.

But a new kernel will run with an O32 userland so you can just upgrade the kernel for now.

But all your existing packages will need to be rebuilt.

From forum: port-cobalt

Re: mips64 on port-cobalt

2009-12-17T07:43:00Z

On Dec 17, 2009, at 6:59 AM, Andy Ruhl wrote:

> On Thu, Dec 17, 2009 at 7:34 AM, Matt Thomas <matt@...> wrote:
>>
>> On Dec 17, 2009, at 3:06 AM, Martin Husemann wrote:
>>
>>> On Thu, Dec 17, 2009 at 12:02:28PM +0100, Markus W Kilbinger wrote:
>>>> 256 MB within the Qube 2.
>>>
>>> Probably not worth running the kernel 64 bit then, but maybe N32 userland.
>>
>> You need a 64 bit kernel to run a N32 userland. Maybe not a N64 kernel but at least a N32 kernel.
>>
>> The algorithmics P5064 already runs a mips64 kernel so the Qube could as well (since they share the same processor). I can make cobalt support mips64 is an hour or two. Is someone willing to test it?
>
> I can netboot it if that's good enough for you for now. I have another
> disk I can install into as well if you need.

Booting with a O32 userland wold be a nice start. Let's see if the kernel even works.

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T06:59:41Z

On Thu, Dec 17, 2009 at 7:34 AM, Matt Thomas <matt@...> wrote:

>
> On Dec 17, 2009, at 3:06 AM, Martin Husemann wrote:
>
>> On Thu, Dec 17, 2009 at 12:02:28PM +0100, Markus W Kilbinger wrote:
>>> 256 MB within the Qube 2.
>>
>> Probably not worth running the kernel 64 bit then, but maybe N32 userland.
>
> You need a 64 bit kernel to run a N32 userland. Maybe not a N64 kernel but at least a N32 kernel.
>
> The algorithmics P5064 already runs a mips64 kernel so the Qube could as well (since they share the same processor). I can make cobalt support mips64 is an hour or two. Is someone willing to test it?

I can netboot it if that's good enough for you for now. I have another
disk I can install into as well if you need.

Andy

From forum: port-mips

Re: mips64 on port-cobalt

2009-12-17T06:59:41Z

On Thu, Dec 17, 2009 at 7:34 AM, Matt Thomas <matt@...> wrote:

I can netboot it if that's good enough for you for now. I have another
disk I can install into as well if you need.

Andy

From forum: port-cobalt

Re: mips64 on port-cobalt

2009-12-17T06:52:46Z

>>>>> "Matt" == Matt Thomas <matt@...> writes:

Matt> The algorithmics P5064 already runs a mips64 kernel so the
Matt> Qube could as well (since they share the same processor). I
Matt> can make cobalt support mips64 is an hour or two. Is someone
Matt> willing to test it?

Yes, me! :-)

What about all the compat-lib stuff? Do I have to build and install
them, too (how, then?)? Or will the new mips64 kernel automatically
handle the (only) existing old libs correctly?

Markus.

From forum: port-mips