Origin 200 Status

Origin 200 Status

[Thiemo Seufer]

Hello All,

testing a single-node O200 SMP with current CVS and the patches from
ftp://ftp.linux-mips.org/pub/linux/mips/people/skylark/ , plus the
icache flush cache patch from Ralf, plus the PCI detection workaround
for ip27 gave

- working serial console
- working RTC
- ethernet with ~2.5 MB throughput

and a stable system. The IOC3 detection shows some weirdness:

[4294678.019000] IOC3 part: [], serial: [] => class IP27 BaseIO
[4294678.020000] ioc3_probe : request_irq fails for IRQ 0x4
[4294678.020000]  ttyS0 at IOC3 0x8620178 (irq = 0) is a 16550A
[4294678.027000] ttyS1 at IOC3 0x8620170 (irq = 0) is a 16550A

but this doesn't seem to have consequences for the serial console.


Thiemo


[4294667.296000] Linux version 2.6.12-ip30 (ths@hattusa) (gcc version 4.0.1 20050701 (prerelease) (Debian 4.0.0-12)) #220 SMP Sat Jul 9 22:39:47 CEST 2005
[4294667.296000] ARCH: SGI-IP27
[4294667.296000] PROMLIB: ARC firmware Version 64 Revision 0
[4294667.296000] Discovered 2 cpus on 1 nodes
[4294667.296000] node_distance: router_a NULL
[4294667.296000] ************** Topology ********************
[4294667.296000]     00 
[4294667.296000] 00  255 
[4294667.296000] CPU revision is: 00000e23
[4294667.296000] FPU revision is: 00000900
[4294667.296000] IP27: Running on node 0.
[4294667.296000] Node 0 has a primary CPU, CPU is running.
[4294667.296000] Node 0 has a secondary CPU, CPU is running.
[4294667.296000] Machine is in M mode.
[4294667.296000] Cpu 0, Nasid 0x0: partnum 0xc002 is a bridge
[4294667.296000] CPU 0 clock is 270MHz.
[4294667.296000] Determined physical RAM map:
[4294667.296000] On node 0 totalpages: 262144
[4294667.296000]   DMA zone: 262144 pages, LIFO batch:31
[4294667.296000]   Normal zone: 0 pages, LIFO batch:1
[4294667.296000]   HighMem zone: 0 pages, LIFO batch:1
[4294667.296000] Built 1 zonelists
[4294667.296000] Kernel command line: root=dksc(0,1,0) root=/dev/sda1
[4294667.296000] Primary instruction cache 32kB, physically tagged, 2-way, linesize 64 bytes.
[4294667.296000] Primary data cache 32kB, 2-way, linesize 32 bytes.
[4294667.296000] Unified secondary cache 4096kB 2-way, linesize 128 bytes.
[4294667.296000] Synthesized TLB refill handler (41 instructions).
[4294667.296000] Synthesized TLB load handler fastpath (55 instructions).
[4294667.296000] Synthesized TLB store handler fastpath (55 instructions).
[4294667.296000] Synthesized TLB modify handler fastpath (54 instructions).
[4294667.296000] PID hash table entries: 4096 (order: 12, 131072 bytes)
[4294667.330000] Dentry cache hash table entries: 262144 (order: 9, 2097152 bytes)
[4294667.352000] Inode-cache hash table entries: 131072 (order: 8, 1048576 bytes)
[4294667.895000] Memory: 1023476k/1048576k available (2893k kernel code, 25100k reserved, 1244k data, 232k init, 0k highmem)
[4294667.898000] Calibrating delay loop... 402.43 BogoMIPS (lpj=201216)
[4294667.919000] Mount-cache hash table entries: 256
[4294667.922000] Checking for 'wait' instruction...  unavailable.
[4294667.924000] Checking for the multiply/shift bug... no.
[4294667.926000] Checking for the daddi bug... no.
[4294667.928000] Checking for the daddiu bug... no.
[4294667.930000] REPLICATION: ON nasid 0, ktext from nasid 0, kdata from nasid 0
[4294667.932000] CPU revision is: 00000e23
[4294667.932000] FPU revision is: 00000900
[4294667.932000] Primary instruction cache 32kB, physically tagged, 2-way, linesize 64 bytes.
[4294667.932000] Primary data cache 32kB, 2-way, linesize 32 bytes.
[4294667.933000] Unified secondary cache 4096kB 2-way, linesize 128 bytes.
[4294667.935000] Synthesized TLB refill handler (41 instructions).
[4294667.935000] CPU 1 clock is 270MHz.
[4294667.936000] Calibrating delay loop... 400.38 BogoMIPS (lpj=200192)
[4294667.956000] Brought up 2 CPUs
[4294667.957000] CPU0 attaching sched-domain:
[4294667.957000]  domain 0: span 00000000,00000003
[4294667.957000]   groups: 00000000,00000001 00000000,00000002
[4294667.957000] CPU1 attaching sched-domain:
[4294667.957000]  domain 0: span 00000000,00000003
[4294667.957000]   groups: 00000000,00000002 00000000,00000001
[4294667.960000] NET: Registered protocol family 16
[4294667.966000] SCSI subsystem initialized
[4294667.976000] devfs: 2004-01-31 Richard Gooch (rgooch@atnf.csiro.au)
[4294667.977000] devfs: boot_options: 0x0
[4294667.979000] Initializing Cryptographic API
[4294668.099000] Real Time Clock Driver v1.09b
[4294668.100000] Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled
[4294668.103000] io scheduler noop registered
[4294668.105000] io scheduler anticipatory registered
[4294668.108000] io scheduler deadline registered
[4294668.110000] io scheduler cfq registered
[4294668.114000] loop: loaded (max 8 devices)
[4294668.116000] qla1280: QLA1040 found on PCI bus 0, dev 0
[4294668.117000] PCI: Enabling device 0000:00:00.0 (0006 -> 0007)
[4294668.732000] scsi(0:0): Resetting SCSI BUS
[4294668.736000] scsi0 : QLogic QLA1040 PCI to SCSI Host Adapter
[4294668.736000]        Firmware version:  7.65.00, Driver version 3.25
[4294671.026000]   Vendor: SEAGATE   Model: ST318275LC        Rev: 0001
[4294671.058000]   Type:   Direct-Access                      ANSI SCSI revision: 02
[4294671.061000] scsi(0:0:3:0): Sync: period 10, offset 12, Wide
[4294671.073000]   Vendor: SEAGATE   Model: ST318275LC        Rev: 0001
[4294671.105000]   Type:   Direct-Access                      ANSI SCSI revision: 02
[4294671.109000] scsi(0:0:4:0): Sync: period 10, offset 12, Wide
[4294671.121000]   Vendor: SEAGATE   Model: ST318275LC        Rev: 0001
[4294671.153000]   Type:   Direct-Access                      ANSI SCSI revision: 02
[4294671.157000] scsi(0:0:5:0): Sync: period 10, offset 12, Wide
[4294671.168000]   Vendor: SEAGATE   Model: ST318275LC        Rev: 0001
[4294671.200000]   Type:   Direct-Access                      ANSI SCSI revision: 02
[4294671.204000] scsi(0:0:6:0): Sync: period 10, offset 12, Wide
[4294672.542000] qla1280: QLA1040 found on PCI bus 0, dev 1
[4294672.543000] PCI: Enabling device 0000:00:01.0 (0006 -> 0007)
[4294673.159000] scsi(1:0): Resetting SCSI BUS
[4294673.163000] scsi1 : QLogic QLA1040 PCI to SCSI Host Adapter
[4294673.163000]        Firmware version:  7.65.00, Driver version 3.25
[4294677.449000] SCSI device sda: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.453000] SCSI device sda: drive cache: write back
[4294677.456000] SCSI device sda: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.459000] SCSI device sda: drive cache: write back
[4294677.460000]  /dev/scsi/host0/bus0/target3/lun0: p1 p2 p3 p9 p11
[4294677.484000] Attached scsi disk sda at scsi0, channel 0, id 3, lun 0
[4294677.490000] SCSI device sdb: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.493000] SCSI device sdb: drive cache: write back
[4294677.496000] SCSI device sdb: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.499000] SCSI device sdb: drive cache: write back
[4294677.500000]  /dev/scsi/host0/bus0/target4/lun0: p1 p2 p3 p9 p11
[4294677.523000] Attached scsi disk sdb at scsi0, channel 0, id 4, lun 0
[4294677.529000] SCSI device sdc: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.532000] SCSI device sdc: drive cache: write back
[4294677.534000] SCSI device sdc: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.537000] SCSI device sdc: drive cache: write back
[4294677.538000]  /dev/scsi/host0/bus0/target5/lun0: p1 p2 p3 p9 p11
[4294677.564000] Attached scsi disk sdc at scsi0, channel 0, id 5, lun 0
[4294677.570000] SCSI device sdd: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.573000] SCSI device sdd: drive cache: write back
[4294677.575000] SCSI device sdd: 35566480 512-byte hdwr sectors (18210 MB)
[4294677.578000] SCSI device sdd: drive cache: write back
[4294677.579000]  /dev/scsi/host0/bus0/target6/lun0: p1 p2 p3 p9 p11
[4294677.603000] Attached scsi disk sdd at scsi0, channel 0, id 6, lun 0
[4294677.604000] Attached scsi generic sg0 at scsi0, channel 0, id 3, lun 0,  type 0
[4294677.606000] Attached scsi generic sg1 at scsi0, channel 0, id 4, lun 0,  type 0
[4294677.607000] Attached scsi generic sg2 at scsi0, channel 0, id 5, lun 0,  type 0
[4294677.608000] Attached scsi generic sg3 at scsi0, channel 0, id 6, lun 0,  type 0
[4294677.609000] md: linear personality registered as nr 1
[4294677.610000] md: raid0 personality registered as nr 2
[4294677.611000] md: raid1 personality registered as nr 3
[4294677.612000] md: raid10 personality registered as nr 9
[4294677.613000] md: raid5 personality registered as nr 4
[4294677.614000] raid5: measuring checksumming speed
[4294677.620000]    8regs     :   688.000 MB/sec
[4294677.626000]    8regs_prefetch:   656.000 MB/sec
[4294677.632000]    32regs    :   684.000 MB/sec
[4294677.638000]    32regs_prefetch:   656.000 MB/sec
[4294677.639000] raid5: using function: 8regs (688.000 MB/sec)
[4294677.657000] raid6: int64x1    179 MB/s
[4294677.675000] raid6: int64x2    246 MB/s
[4294677.693000] raid6: int64x4    312 MB/s
[4294677.711000] raid6: int64x8    199 MB/s
[4294677.712000] raid6: using algorithm int64x4 (312 MB/s)
[4294677.713000] md: raid6 personality registered as nr 8
[4294677.714000] md: md driver 0.90.1 MAX_MD_DEVS=256, MD_SB_DISKS=27
[4294677.728000] device-mapper: 4.4.0-ioctl (2005-01-12) initialised: dm-devel@redhat.com
[4294678.019000] IOC3 part: [], serial: [] => class IP27 BaseIO
[4294678.020000] ioc3_probe : request_irq fails for IRQ 0x4
[4294678.020000]  ttyS0 at IOC3 0x8620178 (irq = 0) is a 16550A
[4294678.027000] ttyS1 at IOC3 0x8620170 (irq = 0) is a 16550A
[4294678.033000] Ethernet address is 08:00:69:0d:52:e7.
[4294678.047000] eth0: link up, 100Mbps, half-duplex, lpa 0x40A1
[4294678.048000] eth0: Using PHY 31, vendor 0x20005c0, model 0, rev 1.
[4294678.049000] eth0: IOC3 SSRAM has 128 kbyte.
[4294678.050000] IOC3 Master Driver loaded for 0000:00:02.0
[4294678.051000] NET: Registered protocol family 2
[4294678.065000] IP: routing cache hash table of 4096 buckets, 64Kbytes
[4294678.068000] TCP established hash table entries: 262144 (order: 10, 4194304 bytes)
[4294678.094000] TCP bind hash table entries: 65536 (order: 8, 1048576 bytes)
[4294678.102000] TCP: Hash tables configured (established 262144 bind 65536)
[4294678.103000] NET: Registered protocol family 1
[4294678.104000] NET: Registered protocol family 17
[4294678.105000] NET: Registered protocol family 15
[4294678.109000] devfs_mk_dev: could not append to parent for md/0
[4294678.110000] md: Autodetecting RAID arrays.
[4294678.213000] md: autorun ...
[4294678.214000] md: considering sdd3 ...
[4294678.215000] md:  adding sdd3 ...
[4294678.216000] md: sdd1 has different UUID to sdd3
[4294678.217000] md:  adding sdc3 ...
[4294678.218000] md: sdc1 has different UUID to sdd3
[4294678.219000] md:  adding sdb3 ...
[4294678.220000] md: sdb1 has different UUID to sdd3
[4294678.221000] md: created md0
[4294678.222000] md: bind<sdb3>
[4294678.223000] md: bind<sdc3>
[4294678.224000] md: bind<sdd3>
[4294678.225000] md: running: <sdd3><sdc3><sdb3>
[4294678.231000] raid5: device sdd3 operational as raid disk 2
[4294678.232000] raid5: device sdc3 operational as raid disk 1
[4294678.233000] raid5: device sdb3 operational as raid disk 0
[4294678.237000] raid5: allocated 3218kB for md0
[4294678.238000] raid5: raid level 5 set md0 active with 3 out of 3 devices, algorithm 2
[4294678.239000] RAID5 conf printout:
[4294678.240000]  --- rd:3 wd:3 fd:0
[4294678.241000]  disk 0, o:1, dev:sdb3
[4294678.242000]  disk 1, o:1, dev:sdc3
[4294678.243000]  disk 2, o:1, dev:sdd3
[4294678.244000] md: considering sdd1 ...
[4294678.245000] md:  adding sdd1 ...
[4294678.246000] md:  adding sdc1 ...
[4294678.247000] md:  adding sdb1 ...
[4294678.249000] devfs_mk_dev: could not append to parent for md/1
[4294678.250000] md: created md1
[4294678.251000] md: bind<sdb1>
[4294678.252000] md: bind<sdc1>
[4294678.253000] md: bind<sdd1>
[4294678.255000] md: running: <sdd1><sdc1><sdb1>
[4294678.260000] raid5: device sdd1 operational as raid disk 2
[4294678.261000] raid5: device sdc1 operational as raid disk 1
[4294678.262000] raid5: device sdb1 operational as raid disk 0
[4294678.266000] raid5: allocated 3218kB for md1
[4294678.267000] raid5: raid level 5 set md1 active with 3 out of 3 devices, algorithm 2
[4294678.268000] RAID5 conf printout:
[4294678.269000]  --- rd:3 wd:3 fd:0
[4294678.270000]  disk 0, o:1, dev:sdb1
[4294678.271000]  disk 1, o:1, dev:sdc1
[4294678.272000]  disk 2, o:1, dev:sdd1
[4294678.273000] md: ... autorun DONE.
[4294678.297000] kjournald starting.  Commit interval 5 seconds
[4294678.299000] EXT3-fs: mounted filesystem with ordered data mode.
[4294678.300000] VFS: Mounted root (ext3 filesystem) readonly.
[4294678.301000] Freeing unused kernel memory: 232k freed
[4294681.162000] Adding 506036k swap on /dev/sda2.  Priority:-1 extents:1
[4294681.183000] Adding 506036k swap on /dev/sdb2.  Priority:-2 extents:1
[4294681.205000] Adding 506036k swap on /dev/sdc2.  Priority:-3 extents:1
[4294681.225000] Adding 506036k swap on /dev/sdd2.  Priority:-4 extents:1
[4294681.465000] EXT3 FS on sda1, internal journal
[4294687.981000] kjournald starting.  Commit interval 5 seconds
[4294688.013000] EXT3 FS on md0, internal journal
[4294688.016000] EXT3-fs: mounted filesystem with ordered data mode.

Re: Origin 200 Status

[Stanislaw Skowronek]

> [4294678.019000] IOC3 part: [], serial: [] => class IP27 BaseIO

This is not really weird, the IOC3 on the IP27 board has no NICs. We'd
have to trace the BRIDGE NICs or even HUB NICs to get the serials, so I
decided that it's an overkill. Although it will be required for reliably
detecting MENET (which has a serial# NIC on the BRIDGE).

> [4294678.020000] ioc3_probe : request_irq fails for IRQ 0x4

This is weird. This means the keyboard will not be operational, and I wish
somebody (Ralf) looks into this. The IOC3 on IP27 BaseIO is a dual-slot
device (takes two IRQs, the INTA and INTA+2).

> [4294678.020000]  ttyS0 at IOC3 0x8620178 (irq = 0) is a 16550A
> [4294678.027000] ttyS1 at IOC3 0x8620170 (irq = 0) is a 16550A

Serial ports will not use IRQs on IP27 unless somebody does dynirqs for
this arch. Nevertheless, if you pass 0 to register_serial() it will use
polling. Note that this is exactly the way IP27 always handled the serial
ports in ioc3-eth.c, so shed no tears :)

> [4294678.033000] Ethernet address is 08:00:69:0d:52:e7.

Good news this works. Anyone with Origin 2000 that had problems with MAC
NICs care to test?

Stanislaw

Re: Origin 200 Status

[Ralf Baechle]

On Sun, Jul 10, 2005 at 08:13:56AM +0200, Stanislaw Skowronek wrote:

> > [4294678.019000] IOC3 part: [], serial: [] => class IP27 BaseIO
> 
> This is not really weird, the IOC3 on the IP27 board has no NICs. We'd
> have to trace the BRIDGE NICs or even HUB NICs to get the serials, so I
> decided that it's an overkill. Although it will be required for reliably
> detecting MENET (which has a serial# NIC on the BRIDGE).

Of course there's a NIC.  The driver has no special handling for IP27 and
is capable of reading the NICs on some machines at least.  That would be
a rather suprising ability if there was no NIC associated with that IOC3 ;-)

(A while ago I recall a few second hand Origins were being sold on eBay
with NICs removed ...)

> > [4294678.020000] ioc3_probe : request_irq fails for IRQ 0x4
> 
> This is weird. This means the keyboard will not be operational, and I wish
> somebody (Ralf) looks into this. The IOC3 on IP27 BaseIO is a dual-slot
> device (takes two IRQs, the INTA and INTA+2).

All information that I have says only INTA is being used for IP27.  It
will probably take a bit of testing.

I haven't yet had any complaints about keyboard and mouse being non-
functional on IP27 - even though my Origin has 4 keyboard and 4 mouse
ports ;-)

   Ralf

Re: Origin 200 Status

[Stanislaw Skowronek]

> Of course there's a NIC.  The driver has no special handling for IP27 and
> is capable of reading the NICs on some machines at least.  That would be
> a rather suprising ability if there was no NIC associated with that IOC3 ;-)

There is only a MAC NIC on a machine ths has. Remember that the Octane
does not have the BaseIO NIC at the IOC3; both NICs placed there are
actually hacks because of XBOW MicroLAN bug.

> > This is weird. This means the keyboard will not be operational, and I wish
> > somebody (Ralf) looks into this. The IOC3 on IP27 BaseIO is a dual-slot
> > device (takes two IRQs, the INTA and INTA+2).
> All information that I have says only INTA is being used for IP27.  It
> will probably take a bit of testing.

Yeah, change the interrupt assignments in ioc3.c, plug in a keyboard and
try. If it works, everybody wins :)

Stanislaw

Re: Origin 200 Status

[Ralf Baechle]

On Mon, Jul 11, 2005 at 04:08:00PM +0200, Stanislaw Skowronek wrote:

> > Of course there's a NIC.  The driver has no special handling for IP27 and
> > is capable of reading the NICs on some machines at least.  That would be
> > a rather suprising ability if there was no NIC associated with that IOC3 ;-)
> 
> There is only a MAC NIC on a machine ths has. Remember that the Octane
> does not have the BaseIO NIC at the IOC3; both NICs placed there are
> actually hacks because of XBOW MicroLAN bug.

I was speaking of IP27, obviously.

> > > This is weird. This means the keyboard will not be operational, and I wish
> > > somebody (Ralf) looks into this. The IOC3 on IP27 BaseIO is a dual-slot
> > > device (takes two IRQs, the INTA and INTA+2).
> > All information that I have says only INTA is being used for IP27.  It
> > will probably take a bit of testing.
> 
> Yeah, change the interrupt assignments in ioc3.c, plug in a keyboard and
> try. If it works, everybody wins :)

I could do - but the plane ticket to plug the keyboard is on you ;-)

  Ralf

Re: Origin 200 Status

[Stanislaw Skowronek]

> I was speaking of IP27, obviously.

I was explaining that not every IOC3 will have a Serial# NIC :)

> I could do - but the plane ticket to plug the keyboard is on you ;-)

I'm going to visit Bayern this month anyway :)))

Stanislaw

Re: Origin 200 Status

[Ralf Baechle]

On Tue, Jul 12, 2005 at 09:20:39AM +0200, Stanislaw Skowronek wrote:

> > I was speaking of IP27, obviously.
> 
> I was explaining that not every IOC3 will have a Serial# NIC :)

Well, in a paragraph mentioning IP27 ...

> > I could do - but the plane ticket to plug the keyboard is on you ;-)
> 
> I'm going to visit Bayern this month anyway :)))

Wrong airport then :)

  Ralf

Re: Origin 200 Status

[Geert Uytterhoeven]

On Tue, 12 Jul 2005, Ralf Baechle wrote:
> On Tue, Jul 12, 2005 at 09:20:39AM +0200, Stanislaw Skowronek wrote:
> > > I was speaking of IP27, obviously.
> > 
> > I was explaining that not every IOC3 will have a Serial# NIC :)
> 
> Well, in a paragraph mentioning IP27 ...
> 
> > > I could do - but the plane ticket to plug the keyboard is on you ;-)
> > 
> > I'm going to visit Bayern this month anyway :)))
> 
> Wrong airport then :)

You misunderstood: he's going to pick you up in his private jet, to fly to your
keyboard connector...

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds