MIPS kernel snapshot testing

[Jimmy Angelakos]

Hi,

I just successfully booted:
http://people.debian.org/~tbm/mips/kernel/r11922/linux-image-2.6.26-1-r5k-ip32_2.6.26-1~experimental.1_mips.deb
and have made some observations.

First and foremost, the IP32 kernel appears to have performance issues.
As the only other OS I've managed to boot on this O2 is the ancient
Gentoo Live CD, I've used it as a measure of comparison and am
attaching below some readouts from both OSs on this machine.

Running hdparm -Tt /dev/sda we can see that it appears the Debian
kernel has about only a third (!) of the memory throughput of the
Gentoo kernel (results attached).

Also, while the dmesg for Gentoo doesn't show anything worrying in the
SCSI subsystem, our dmesg shows:
[17179602.240000] sd 0:0:1:0: [sda] 4197405 512-byte hardware sectors
(2149 MB)
[17179602.280000] sd 0:0:1:0: [sda] Write Protect is off
[17179602.300000] sd 0:0:1:0: [sda] Mode Sense: b3 00 00 08
[17179602.320000] sd 0:0:1:0: [sda] Write cache: disabled, read cache:
enabled, doesn't support DPO or FUA

That is, it disables the HDD write cache for no apparent reason.

What I also find worrying is the usage of 4KB kernel page size on the
r5k kernel, surely r5k should support 16KB and wouldn't that enhance
performance a bit?

Both distros see the USB adapter and in Debian (which we care about) it
works perfectly, udev/HAL handles USB storage devices very well.

Moving on to the sound, the PROM tells us we have Audio : A3 version 1.
When I took it out and had a look at it, it says it's a (fairly
standard I assume) Moosehead AV board and the chip on it is an AD1843.
During bootup, however, the module doesn't get loaded.

A modprobe snd-sgi-o2 yields:
[17179964.716000] ad1843: AD1843 won't initialize
[17179964.716000] sgio2audio: probe of sgio2audio failed with error -5

Please find attached below my test results.

Thank you for your time,
Jimmy

Linux lab-o2 2.6.26-1-r5k-ip32 #1 Wed Jul 30 06:12:17 UTC 2008 mips64 GNU/Linux

Linux livecd 2.6.17.10-mipsgit-20060618-ip32r5k #5 Sun Oct 8 15:35:51 UTC 2006 mips64 R5000 V2.1  FPU V1.0 GNU/Linux


/dev/sda:
 Timing cached reads:    60 MB in  2.03 seconds =  29.50 MB/sec
 Timing buffered disk reads:   24 MB in  3.17 seconds =   7.58 MB/sec


/dev/sda:
 Timing cached reads:   124 MB in  2.00 seconds =  61.94 MB/sec
 Timing buffered disk reads:   24 MB in  3.17 seconds =   7.58 MB/sec

00:01.0 SCSI storage controller: Adaptec AIC-7880U
00:02.0 SCSI storage controller: Adaptec AIC-7880U
00:03.0 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 50)
00:03.1 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 50)
00:03.2 USB Controller: VIA Technologies, Inc. USB 2.0 (rev 51)

00:01.0 SCSI storage controller: Adaptec AIC-7880U
00:02.0 SCSI storage controller: Adaptec AIC-7880U
00:03.0 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 50)
00:03.1 USB Controller: VIA Technologies, Inc. VT82xxxxx UHCI USB 1.1 Controller (rev 50)
00:03.2 USB Controller: VIA Technologies, Inc. USB 2.0 (rev 51)

[    0.000000] Initializing cgroup subsys cpu
[    0.000000] Linux version 2.6.26-1-r5k-ip32 (Debian 2.6.26-1~experimental.1) (cts@...) (gcc version 4.1.3 20080420 (prerelease) (Debian 4.1.2-22)) #1 Wed Jul 30 06:12:17 UTC 2008
[    0.000000] ARCH: SGI-IP32
[    0.000000] PROMLIB: ARC firmware Version 1 Revision 10
[    0.000000] CRIME id a rev 1 at 0x0000000014000000
[    0.000000] CRIME MC: bank 0 base 0x0000000000000000 size 32MiB
[    0.000000] CRIME MC: bank 1 base 0x0000000002000000 size 32MiB
[    0.000000] CRIME MC: bank 2 base 0x0000000004000000 size 32MiB
[    0.000000] CRIME MC: bank 3 base 0x0000000006000000 size 32MiB
[    0.000000] CRIME MC: bank 4 base 0x0000000008000000 size 32MiB
[    0.000000] CRIME MC: bank 5 base 0x000000000a000000 size 32MiB
[    0.000000] CRIME MC: bank 6 base 0x000000000c000000 size 32MiB
[    0.000000] CRIME MC: bank 7 base 0x000000000e000000 size 32MiB
[    0.000000] CPU revision is: 00002321 (R5000)
[    0.000000] FPU revision is: 00002310
[    0.000000] Checking for the multiply/shift bug... no.
[    0.000000] Checking for the daddiu bug... no.
[    0.000000] Determined physical RAM map:
[    0.000000]  memory: 0000000010000000 @ 0000000000000000 (usable)
[    0.000000] Entering add_active_range(0, 0, 65536) 0 entries of 256 used
[    0.000000] Initrd not found or empty - disabling initrd
[    0.000000] Zone PFN ranges:
[    0.000000]   Normal          0 ->    65536
[    0.000000] Movable zone start PFN for each node
[    0.000000] early_node_map[1] active PFN ranges
[    0.000000]     0:        0 ->    65536
[    0.000000] On node 0 totalpages: 65536
[    0.000000]   Normal zone: 896 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 64640 pages, LIFO batch:15
[    0.000000]   Movable zone: 0 pages used for memmap
[17179569.184000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 64640
[17179569.184000] Kernel command line: root=/dev/sda1 video=1280x1024-16@75 auto
[17179569.184000] Primary instruction cache 32kB, VIPT, 2-way, linesize 32 bytes.
[17179569.184000] Primary data cache 32kB, 2-way, VIPT, cache aliases, linesize 32 bytes
[17179569.184000] R5000 SCACHE size 512kB, linesize 32 bytes.
[17179569.184000] PID hash table entries: 1024 (order: 10, 8192 bytes)
[17179569.184000] Calibrating system timer... 180 MHz CPU detected
[17179569.184000] CRIME memory error at 0x113fffe0 ST 0x0c00a828<INV,RE,REID=0x28,NONFATAL>
[17179569.184000] Console: colour dummy device 80x25
[17179569.184000] console [tty0] enabled
[17179569.192000] Dentry cache hash table entries: 32768 (order: 6, 262144 bytes)
[17179569.196000] Inode-cache hash table entries: 16384 (order: 5, 131072 bytes)
[17179569.296000] Memory: 253636k/262144k available (2861k kernel code, 8316k reserved, 940k data, 196k init, 0k highmem)
[17179569.296000] Calibrating delay loop... 179.20 BogoMIPS (lpj=358400)
[17179569.372000] Security Framework initialized
[17179569.372000] SELinux:  Disabled at boot.
[17179569.372000] Capability LSM initialized
[17179569.372000] Mount-cache hash table entries: 256
[17179569.376000] Initializing cgroup subsys ns
[17179569.376000] Initializing cgroup subsys cpuacct
[17179569.376000] Initializing cgroup subsys devices
[17179569.376000] Checking for the daddi bug... no.
[17179569.380000] net_namespace: 1208 bytes
[17179569.384000] NET: Registered protocol family 16
[17179569.388000] MACE PCI rev 1
[17179569.404000] SCSI subsystem initialized
[17179569.432000] Switched to high resolution mode on CPU 0
[17179569.432000] NET: Registered protocol family 2
[17179569.472000] IP route cache hash table entries: 2048 (order: 2, 16384 bytes)
[17179569.476000] TCP established hash table entries: 8192 (order: 5, 131072 bytes)
[17179569.476000] TCP bind hash table entries: 8192 (order: 4, 65536 bytes)
[17179569.480000] TCP: Hash tables configured (established 8192 bind 8192)
[17179569.480000] TCP reno registered
[17179569.492000] NET: Registered protocol family 1
[17179569.500000] audit: initializing netlink socket (disabled)
[17179569.500000] type=2000 audit(1217442741.316:1): initialized
[17179569.504000] VFS: Disk quotas dquot_6.5.1
[17179569.504000] Dquot-cache hash table entries: 512 (order 0, 4096 bytes)
[17179569.504000] msgmni has been set to 495
[17179569.508000] Block layer SCSI generic (bsg) driver version 0.4 loaded (major 254)
[17179569.508000] io scheduler noop registered
[17179569.508000] io scheduler anticipatory registered
[17179569.508000] io scheduler deadline registered
[17179569.508000] io scheduler cfq registered (default)
[17179569.940000] Console: switching to colour frame buffer device 160x64
[17179570.048000] fb0: SGI GBE rev 1 @ 0x16000000 using 4096kB memory
[17179570.092000] Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled
[17179570.100000] serial8250.0: ttyS0 at MMIO 0x1f390000 (irq = 60) is a 16550A
[17179570.100000] serial8250.0: ttyS1 at MMIO 0x1f398000 (irq = 66) is a 16550A
[17179570.128000] brd: module loaded
[17179570.132000] eth0: SGI MACE Ethernet rev. 1
[17179570.136000] PCI: Enabling device 0000:00:01.0 (0046 -> 0047)
[17179571.140000] ahc_pci:0:1:0: Using left over BIOS settings
[17179586.168000] scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 7.0
[17179586.168000]         <Adaptec aic7880 Ultra SCSI adapter>
[17179586.168000]         aic7880: Ultra Wide Channel A, SCSI Id=0, 16/253 SCBs
[17179586.168000]
[17179586.172000] PCI: Enabling device 0000:00:02.0 (0046 -> 0047)
[17179587.176000] ahc_pci:0:2:0: Using left over BIOS settings
[17179587.208000] scsi 0:0:1:0: Direct-Access     SGI      IBM DCAS-32160W  S68A PQ: 0 ANSI: 2
[17179587.212000] scsi0:A:1:0: Tagged Queuing enabled.  Depth 32
[17179587.212000]  target0:0:1: Beginning Domain Validation
[17179587.228000]  target0:0:1: wide asynchronous
[17179587.236000]  target0:0:1: FAST-20 WIDE SCSI 40.0 MB/s ST (50 ns, offset 8)
[17179587.244000]  target0:0:1: Domain Validation skipping write tests
[17179587.244000]  target0:0:1: Ending Domain Validation
[17179587.780000] scsi 0:0:4:0: CD-ROM            TOSHIBA  CD-ROM XM-5701TA 0167 PQ: 0 ANSI: 2
[17179587.780000]  target0:0:4: Beginning Domain Validation
[17179587.784000]  target0:0:4: FAST-10 SCSI 10.0 MB/s ST (100 ns, offset 8)
[17179587.788000]  target0:0:4: Domain Validation skipping write tests
[17179587.788000]  target0:0:4: Ending Domain Validation
[17179602.208000] scsi1 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 7.0
[17179602.208000]         <Adaptec aic7880 Ultra SCSI adapter>
[17179602.208000]         aic7880: Ultra Wide Channel A, SCSI Id=0, 16/253 SCBs
[17179602.208000]
[17179602.212000] Driver 'sd' needs updating - please use bus_type methods
[17179602.220000]  target0:0:1: FAST-20 WIDE SCSI 40.0 MB/s ST (50 ns, offset 8)
[17179602.240000] sd 0:0:1:0: [sda] 4197405 512-byte hardware sectors (2149 MB)
[17179602.280000] sd 0:0:1:0: [sda] Write Protect is off
[17179602.300000] sd 0:0:1:0: [sda] Mode Sense: b3 00 00 08
[17179602.320000] sd 0:0:1:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[17179602.372000] sd 0:0:1:0: [sda] 4197405 512-byte hardware sectors (2149 MB)
[17179602.436000] sd 0:0:1:0: [sda] Write Protect is off
[17179602.460000] sd 0:0:1:0: [sda] Mode Sense: b3 00 00 08
[17179602.476000] sd 0:0:1:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[17179602.504000]  sda: sda1 sda2 sda9 sda11
[17179602.544000] sd 0:0:1:0: [sda] Attached SCSI disk
[17179602.572000] Driver 'sr' needs updating - please use bus_type methods
[17179602.604000] sr0: scsi-1 drive
[17179602.632000] Uniform CD-ROM driver Revision: 3.20
[17179602.664000] sr 0:0:4:0: Attached scsi CD-ROM sr0
[17179602.668000] mice: PS/2 mouse device common for all mice
[17179603.200000] input: AT Raw Set 2 keyboard as /class/input/input0
[17179603.636000] TCP cubic registered
[17179603.664000] NET: Registered protocol family 17
[17179603.700000] registered taskstats version 1
[17179603.728000] scsi: waiting for bus probes to complete ...
[17179604.352000] input: PS/2 Generic Mouse as /class/input/input1
[17179606.200000] kjournald starting.  Commit interval 5 seconds
[17179606.232000] EXT3-fs: mounted filesystem with ordered data mode.
[17179606.264000] VFS: Mounted root (ext3 filesystem) readonly.
[17179606.304000] Freeing unused kernel memory: 196k freed
[17179621.612000] sd 0:0:1:0: Attached scsi generic sg0 type 0
[17179621.648000] sr 0:0:4:0: Attached scsi generic sg1 type 5
[17179624.016000] usbcore: registered new interface driver usbfs
[17179624.052000] usbcore: registered new interface driver hub
[17179624.164000] usbcore: registered new device driver usb
[17179624.244000] PCI: Enabling device 0000:00:03.2 (0000 -> 0002)
[17179624.280000] ehci_hcd 0000:00:03.2: EHCI Host Controller
[17179624.316000] ehci_hcd 0000:00:03.2: new USB bus registered, assigned bus number 1
[17179624.352000] ehci_hcd 0000:00:03.2: irq 22, io mem 0x280022000
[17179624.412000] ehci_hcd 0000:00:03.2: USB 2.0 started, EHCI 0.95, driver 10 Dec 2004
[17179624.448000] usb usb1: configuration #1 chosen from 1 choice
[17179624.484000] hub 1-0:1.0: USB hub found
[17179624.520000] hub 1-0:1.0: 4 ports detected
[17179624.620000] USB Universal Host Controller Interface driver v3.0
[17179624.676000] usb usb1: New USB device found, idVendor=1d6b, idProduct=0002
[17179624.712000] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[17179624.748000] usb usb1: Product: EHCI Host Controller
[17179624.780000] usb usb1: Manufacturer: Linux 2.6.26-1-r5k-ip32 ehci_hcd
[17179624.816000] usb usb1: SerialNumber: 0000:00:03.2
[17179624.852000] PCI: Enabling device 0000:00:03.0 (0000 -> 0001)
[17179624.888000] uhci_hcd 0000:00:03.0: UHCI Host Controller
[17179624.924000] uhci_hcd 0000:00:03.0: new USB bus registered, assigned bus number 2
[17179624.956000] uhci_hcd 0000:00:03.0: irq 18, io base 0x00001800
[17179624.996000] usb usb2: configuration #1 chosen from 1 choice
[17179625.032000] hub 2-0:1.0: USB hub found
[17179625.064000] hub 2-0:1.0: 2 ports detected
[17179625.204000] usb usb2: New USB device found, idVendor=1d6b, idProduct=0001
[17179625.240000] usb usb2: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[17179625.272000] usb usb2: Product: UHCI Host Controller
[17179625.308000] usb usb2: Manufacturer: Linux 2.6.26-1-r5k-ip32 uhci_hcd
[17179625.344000] usb usb2: SerialNumber: 0000:00:03.0
[17179625.376000] PCI: Enabling device 0000:00:03.1 (0000 -> 0001)
[17179625.412000] uhci_hcd 0000:00:03.1: UHCI Host Controller
[17179625.444000] uhci_hcd 0000:00:03.1: new USB bus registered, assigned bus number 3
[17179625.480000] uhci_hcd 0000:00:03.1: irq 21, io base 0x00001820
[17179625.516000] usb usb3: configuration #1 chosen from 1 choice
[17179625.552000] hub 3-0:1.0: USB hub found
[17179625.584000] hub 3-0:1.0: 2 ports detected
[17179626.068000] usb usb3: New USB device found, idVendor=1d6b, idProduct=0001
[17179626.100000] usb usb3: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[17179626.136000] usb usb3: Product: UHCI Host Controller
[17179626.168000] usb usb3: Manufacturer: Linux 2.6.26-1-r5k-ip32 uhci_hcd
[17179626.200000] usb usb3: SerialNumber: 0000:00:03.1
[17179632.112000] Adding 267924k swap on /dev/sda2.  Priority:-1 extents:1 across:267924k
[17179633.280000] EXT3 FS on sda1, internal journal

Linux version 2.6.17.10-mipsgit-20060618-ip32r5k (root@dol-guldur) (gcc version 4.1.1 (Gentoo 4.1.1)) #5 Sun Oct 8 15:35:51 UTC 2006
ARCH: SGI-IP32
PROMLIB: ARC firmware Version 1 Revision 10
CRIME id a rev 1 at 0x0000000014000000
CRIME MC: bank 0 base 0x0000000000000000 size 32MiB
CRIME MC: bank 1 base 0x0000000002000000 size 32MiB
CRIME MC: bank 2 base 0x0000000004000000 size 32MiB
CRIME MC: bank 3 base 0x0000000006000000 size 32MiB
CRIME MC: bank 4 base 0x0000000008000000 size 32MiB
CRIME MC: bank 5 base 0x000000000a000000 size 32MiB
CRIME MC: bank 6 base 0x000000000c000000 size 32MiB
CRIME MC: bank 7 base 0x000000000e000000 size 32MiB
CPU revision is: 00002321
FPU revision is: 00002310
Determined physical RAM map:
 memory: 0000000010000000 @ 0000000000000000 (usable)
On node 0 totalpages: 65536
  DMA zone: 65536 pages, LIFO batch:15
Built 1 zonelists
Kernel command line: real_root=/dev/sr0 cdroot=/dev/loop0 looptype=sgimips nosound console=tty0 video=gbefb:1280x1024-16@75 ip32
Primary instruction cache 32kB, physically tagged, 2-way, linesize 32 bytes.
Primary data cache 32kB, 2-way, linesize 32 bytes.
R5000 SCACHE size 512kB, linesize 32 bytes.
Synthesized TLB refill handler (32 instructions).
Synthesized TLB load handler fastpath (46 instructions).
Synthesized TLB store handler fastpath (46 instructions).
Synthesized TLB modify handler fastpath (45 instructions).
PID hash table entries: 2048 (order: 11, 16384 bytes)
Calibrating system timer... 180 MHz CPU detected
Using 90.017 MHz high precision timer.
Console: colour dummy device 80x25
CRIME memory error at 0x113fffe0 ST 0x0c00a828<INV,RE,REID=0x28,NONFATAL>
Dentry cache hash table entries: 32768 (order: 6, 262144 bytes)
Inode-cache hash table entries: 16384 (order: 5, 131072 bytes)
Memory: 251904k/262144k available (4136k kernel code, 9896k reserved, 831k data, 612k init, 0k highmem)
Calibrating delay loop... 178.68 BogoMIPS (lpj=89344)
Mount-cache hash table entries: 256
Checking for 'wait' instruction...  available.
Checking for the multiply/shift bug... no.
Checking for the daddi bug... no.
Checking for the daddiu bug... no.
NET: Registered protocol family 16
MACE PCI rev 1
SCSI subsystem initialized
usbcore: registered new driver usbfs
usbcore: registered new driver hub
MACEPCI: Master abort at 0x00004300 (C)
MACEPCI: Master abort at 0x00004400 (C)
MACEPCI: Master abort at 0x00004500 (C)
MACEPCI: Master abort at 0x00004600 (C)
MACEPCI: Master abort at 0x00004700 (C)
NET: Registered protocol family 2
IP route cache hash table entries: 2048 (order: 2, 16384 bytes)
TCP established hash table entries: 8192 (order: 4, 65536 bytes)
TCP bind hash table entries: 4096 (order: 3, 32768 bytes)
TCP: Hash tables configured (established 8192 bind 4096)
TCP reno registered
squashfs: version 3.0 (2006/03/15) Phillip Lougher
NTFS driver 2.1.27 [Flags: R/O].
EFS: 1.0a - http://aeschi.ch.eu.org/efs/
JFS: nTxBlock = 1970, nTxLock = 15765
SGI XFS with large block/inode numbers, no debug enabled
io scheduler noop registered
io scheduler anticipatory registered
io scheduler deadline registered (default)
io scheduler cfq registered
pci_hotplug: PCI Hot Plug PCI Core version: 0.5
fakephp: Fake PCI Hot Plug Controller Driver
Console: switching to colour frame buffer device 160x64
fb0: SGI GBE rev 1 @ 0x16000000 using 4096kB memory
rtc: SRM (post-2000) epoch (2000) detected
Real Time Clock Driver v1.12a
Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled
serial8250: ttyS0 at MMIO 0x0 (irq = 53) is a 16550A
serial8250: ttyS1 at MMIO 0x0 (irq = 59) is a 16550A
loop: loaded (max 8 devices)
eth0: SGI MACE Ethernet rev. 1
PCI: Enabling device 0000:00:01.0 (0046 -> 0047)
ahc_pci:0:1:0: Using left over BIOS settings
scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 7.0
        <Adaptec aic7880 Ultra SCSI adapter>
        aic7880: Wide Channel A, SCSI Id=0, 16/253 SCBs

  Vendor: SGI       Model: IBM DCAS-32160W   Rev: S68A
  Type:   Direct-Access                      ANSI SCSI revision: 02
scsi0:A:1:0: Tagged Queuing enabled.  Depth 48
 target0:0:1: Beginning Domain Validation
 target0:0:1: wide asynchronous
 target0:0:1: FAST-10 WIDE SCSI 20.0 MB/s ST (100 ns, offset 8)
 target0:0:1: Domain Validation skipping write tests
 target0:0:1: Ending Domain Validation
  Vendor: TOSHIBA   Model: CD-ROM XM-5701TA  Rev: 0167
  Type:   CD-ROM                             ANSI SCSI revision: 02
 target0:0:4: Beginning Domain Validation
 target0:0:4: FAST-10 SCSI 10.0 MB/s ST (100 ns, offset 8)
 target0:0:4: Domain Validation skipping write tests
 target0:0:4: Ending Domain Validation
PCI: Enabling device 0000:00:02.0 (0046 -> 0047)
ahc_pci:0:2:0: Using left over BIOS settings
scsi1 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 7.0
        <Adaptec aic7880 Ultra SCSI adapter>
        aic7880: Wide Channel A, SCSI Id=0, 16/253 SCBs

st: Version 20050830, fixed bufsize 32768, s/g segs 256
SCSI device sda: 4197405 512-byte hdwr sectors (2149 MB)
sda: Write Protect is off
sda: Mode Sense: b3 00 00 08
SCSI device sda: drive cache: write through
SCSI device sda: 4197405 512-byte hdwr sectors (2149 MB)
sda: Write Protect is off
sda: Mode Sense: b3 00 00 08
SCSI device sda: drive cache: write through
 sda: sda1 sda2 sda9 sda11
sd 0:0:1:0: Attached scsi disk sda
sr0: scsi-1 drive
Uniform CD-ROM driver Revision: 3.20
sr 0:0:4:0: Attached scsi CD-ROM sr0
PCI: Enabling device 0000:00:03.2 (0000 -> 0002)
ehci_hcd 0000:00:03.2: EHCI Host Controller
ehci_hcd 0000:00:03.2: new USB bus registered, assigned bus number 1
ehci_hcd 0000:00:03.2: irq 15, io mem 0x280022000
ehci_hcd 0000:00:03.2: USB 2.0 started, EHCI 0.95, driver 10 Dec 2004
usb usb1: configuration #1 chosen from 1 choice
hub 1-0:1.0: USB hub found
hub 1-0:1.0: 4 ports detected
ohci_hcd: 2005 April 22 USB 1.1 'Open' Host Controller (OHCI) Driver (PCI)
USB Universal Host Controller Interface driver v3.0
PCI: Enabling device 0000:00:03.0 (0000 -> 0001)
uhci_hcd 0000:00:03.0: UHCI Host Controller
uhci_hcd 0000:00:03.0: new USB bus registered, assigned bus number 2
uhci_hcd 0000:00:03.0: irq 11, io base 0x00001800
usb usb2: configuration #1 chosen from 1 choice
hub 2-0:1.0: USB hub found
hub 2-0:1.0: 2 ports detected
PCI: Enabling device 0000:00:03.1 (0000 -> 0001)
uhci_hcd 0000:00:03.1: UHCI Host Controller
uhci_hcd 0000:00:03.1: new USB bus registered, assigned bus number 3
uhci_hcd 0000:00:03.1: irq 14, io base 0x00001820
usb usb3: configuration #1 chosen from 1 choice
hub 3-0:1.0: USB hub found
hub 3-0:1.0: 2 ports detected
Initializing USB Mass Storage driver...
usbcore: registered new driver usb-storage
USB Mass Storage support registered.
usbcore: registered new driver usbserial
drivers/usb/serial/usb-serial.c: USB Serial Driver core
drivers/usb/serial/usb-serial.c: USB Serial support registered for pl2303
usbcore: registered new driver pl2303
drivers/usb/serial/pl2303.c: Prolific PL2303 USB to serial adaptor driver
mice: PS/2 mouse device common for all mice
md: linear personality registered for level -1
md: raid0 personality registered for level 0
md: raid1 personality registered for level 1
md: raid10 personality registered for level 10
md: raid5 personality registered for level 5
md: raid4 personality registered for level 4
raid5: measuring checksumming speed
   8regs     :   308.000 MB/sec
   8regs_prefetch:   308.000 MB/sec
   32regs    :   308.000 MB/sec
   32regs_prefetch:   308.000 MB/sec
raid5: using function: 32regs_prefetch (308.000 MB/sec)
raid6: int64x1     50 MB/s
raid6: int64x2     66 MB/s
raid6: int64x4     82 MB/s
raid6: int64x8     62 MB/s
raid6: using algorithm int64x4 (82 MB/s)
md: raid6 personality registered for level 6
md: multipath personality registered for level -4
md: md driver 0.90.3 MAX_MD_DEVS=256, MD_SB_DISKS=27
md: bitmap version 4.39
TCP bic registered
NET: Registered protocol family 1
NET: Registered protocol family 17
Freeing unused kernel memory: 612k freed
input: AT Raw Set 2 keyboard as /class/input/input0
input: PS/2 Generic Mouse as /class/input/input1
kjournald starting.  Commit interval 5 seconds
EXT3 FS on sda1, internal journal
EXT3-fs: mounted filesystem with ordered data mode.

[Martin Michlmayr]

* Jimmy Angelakos <vyruss@...> [2008-07-31 00:18]:
> I just successfully booted:
> http://people.debian.org/~tbm/mips/kernel/r11922/linux-image-2.6.26-1-r5k-ip32_2.6.26-1~experimental.1_mips.deb
> and have made some observations.

Great report, Jimmy.  I hope Thomas will respond.
--
Martin Michlmayr
http://www.cyrius.com/


--
To UNSUBSCRIBE, email to debian-mips-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

[Jimmy Angelakos]

Martin Michlmayr wrote:
> * Jimmy Angelakos <vyruss@...> [2008-07-31 00:18]:
>> I just successfully booted:
>> http://people.debian.org/~tbm/mips/kernel/r11922/linux-image-2.6.26-1-r5k-ip32_2.6.26-1~experimental.1_mips.deb
>> and have made some observations.
>
> Great report, Jimmy.  I hope Thomas will respond.


Well, it is the middle of the summer and people are going off on
holiday, so it's pretty normal for list traffic to go down :)

I went ahead and compiled a kernel from the linux-next git tree and,
using that kernel, O2 sound works perfectly.

However, the very slow cache performance (RAM throughput issue?) did
not go away. That is clearly a regression as it used to be better in
older kernels.


Jimmy


--
To UNSUBSCRIBE, email to debian-mips-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

[Thomas Bogendoerfer]

On Sun, Aug 03, 2008 at 05:45:31PM +0300, Jimmy Angelakos wrote:
> I went ahead and compiled a kernel from the linux-next git tree and,
> using that kernel, O2 sound works perfectly.

ahh that's good as I can't reproduce that my machine, but I was
pretty sure, that it's a timing issue.

> However, the very slow cache performance (RAM throughput issue?) did
> not go away. That is clearly a regression as it used to be better in
> older kernels.

I'm not really buying into hdparm cached time numbers (-T). -t is
at least usefull to see, that the speed between disk drive and
controller/driver is ok. There could be a regression, but I'd like
to see numbers from more meaningfull benchmarks or even better
real applications.

Thomas.

--
Crap can work. Given enough thrust pigs will fly, but it's not necessary a
good idea.                                                [ RFC1925, 2.3 ]


--
To UNSUBSCRIBE, email to debian-mips-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

[Jimmy Angelakos]

Thomas Bogendoerfer wrote:

> I'm not really buying into hdparm cached time numbers (-T). -t is
> at least usefull to see, that the speed between disk drive and
> controller/driver is ok. There could be a regression, but I'd like
> to see numbers from more meaningfull benchmarks or even better
> real applications.
>
> Thomas.
>

What would you suggest? fio? bonnie++? If you are aware of any
benchmarks that illustrate cache performance specifically, I can try
them out.

Another problem that I see, which looks more serious, is that the SCSI
subsystem has turned off write cache for the hdd, I have no idea why. I
tried using the older Adaptec SCSI driver in the kernel but it made no
difference.


Cheers,
Jimmy


--
To UNSUBSCRIBE, email to debian-mips-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

[Thomas Bogendoerfer]

On Mon, Aug 04, 2008 at 04:44:38AM +0300, Jimmy Angelakos wrote:
> What would you suggest? fio? bonnie++? If you are aware of any
> benchmarks that illustrate cache performance specifically, I can try
> them out.

bonnie++ might be a good start.

> Another problem that I see, which looks more serious, is that the SCSI
> subsystem has turned off write cache for the hdd, I have no idea why. I
> tried using the older Adaptec SCSI driver in the kernel but it made no
> difference.

that's a setting of the harddrive, the kernel just reads it and prints
it. Use sdparm to modify that setting.

   sdparm --set=WCE /dev/sda

Thomas.

--
Crap can work. Given enough thrust pigs will fly, but it's not necessary a
good idea.                                                [ RFC1925, 2.3 ]


--
To UNSUBSCRIBE, email to debian-mips-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

[Jimmy Angelakos]

Thomas Bogendoerfer wrote:
> bonnie++ might be a good start.
...
>    sdparm --set=WCE /dev/sda

Cheers for that, I've enabled it now. Haven't used SCSI in a long time.

Here are attached the results of my testing, all on the same O2
machine. The files are:

bonnie.debian      (Martin's latest snapshot on Debian Lenny)
bonnie.gentoo      (Latest mips-git kernel they have on Gentoo)
bonnie.gentoo-live (Gentoo with the 2006 live CD kernel [2.6.17])

There was no way I could run bonnie on the Live CD as I couldn't
compile/emerge and the already compiled bonnie must have used a
different abi. So I just took out the live CD kernel and booted the
system from that instead.

A reminder: The live CD's hdparm claimed 61 MB/s cache reads but the
same kernel only claimed 29 MB/s on the installed system.

I hope these are useful.

Cheers,
Jimmy

Using uid:0, gid:0.
Writing with putc()...done
Writing intelligently...done
Rewriting...done
Reading with getc()...done
Reading intelligently...done
start 'em...done...done...done...
Create files in sequential order...done.
Stat files in sequential order...done.
Delete files in sequential order...done.
Create files in random order...done.
Stat files in random order...done.
Delete files in random order...done.
Version 1.03d       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
lab-o2         496M  1058  97  5673  41  2734  13  1143  99  5810  16 118.1   5
                    ------Sequential Create------ --------Random Create--------
                    -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16  2338  87 31149  99  3698  82  2471  89 +++++ +++  3881  87
lab-o2,496M,1058,97,5673,41,2734,13,1143,99,5810,16,118.1,5,16,2338,87,31149,99,3698,82,2471,89,+++++,+++,3881,87

Using uid:0, gid:0.
Writing with putc()...done
Writing intelligently...done
Rewriting...done
Reading with getc()...done
Reading intelligently...done
start 'em...done...done...done...
Create files in sequential order...done.
Stat files in sequential order...done.
Delete files in sequential order...done.
Create files in random order...done.
Stat files in random order...done.
Delete files in random order...done.
Version 1.03d       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
lab-o2         488M   999  98  5913  41  2878  12  1049  98  6102  15 116.3   4
                    ------Sequential Create------ --------Random Create--------
                    -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16  2566  93 30313  99  3953  90  2643  94 +++++ +++  4020  91
lab-o2,488M,999,98,5913,41,2878,12,1049,98,6102,15,116.3,4,16,2566,93,30313,99,3953,90,2643,94,+++++,+++,4020,91

Using uid:0, gid:0.
Writing with putc()...done
Writing intelligently...done
Rewriting...done
Reading with getc()...done
Reading intelligently...done
start 'em...done...done...done...
Create files in sequential order...done.
Stat files in sequential order...done.
Delete files in sequential order...done.
Create files in random order...done.
Stat files in random order...done.
Delete files in random order...done.
Version 1.03d       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
lab-o2         488M   439  99  6216  40  2812  12   448  99  6150  14 115.8   4
                    ------Sequential Create------ --------Random Create--------
                    -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16  2867  96 29437 100  3954  90  2721  89 +++++ +++  4592  99
lab-o2,488M,439,99,6216,40,2812,12,448,99,6150,14,115.8,4,16,2867,96,29437,100,3954,90,2721,89,+++++,+++,4592,99

[Thomas Bogendoerfer]

On Wed, Aug 06, 2008 at 05:12:46AM +0300, Jimmy Angelakos wrote:
> Here are attached the results of my testing, all on the same O2
> machine. The files are:

thank you.

> A reminder: The live CD's hdparm claimed 61 MB/s cache reads but the
> same kernel only claimed 29 MB/s on the installed system.

so it's depending on the hdparm binary itself ?

> I hope these are useful.

I don't a real bad spot, just some numbers up, a few others down.

Thomas.

--
Crap can work. Given enough thrust pigs will fly, but it's not necessary a
good idea.                                                [ RFC1925, 2.3 ]


--
To UNSUBSCRIBE, email to debian-mips-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

[Jimmy Angelakos]

Thomas Bogendoerfer wrote:
> On Wed, Aug 06, 2008 at 05:12:46AM +0300, Jimmy Angelakos wrote:
>> A reminder: The live CD's hdparm claimed 61 MB/s cache reads but the
>> same kernel only claimed 29 MB/s on the installed system.
> so it's depending on the hdparm binary itself ?

Perhaps. I used to think that on an idle system the hdparm cached speed
should reflect the RAM's speed, and that's why I found 29 MB/s too
slow, surely the O2's RAM can do better than that.

However in the past few days I ran hdparm -T on my desktop machine and
came up with a cache speed of 6,5 GB/s! My RAM is fast but definitely
not that fast.

If that behaviour is not normal, perhaps we should file a bug with the
hdparm team?

> I don't a real bad spot, just some numbers up, a few others down.

Yes, they're pretty consistent results, considering the kernels are
different too.

> Thomas.

Cheers
Jimmy


--
To UNSUBSCRIBE, email to debian-mips-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...