Multiple drives on 2 IDE controllers on 2.4.22 kernel

• Previous message: Andrew de Quincey: "Re: [PATCH] deal with lack of acpi prt entries gracefully"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Date:	Tue, 09 Sep 2003 13:53:50 -0700
To: linux-kernel@vger.kernel.org

Background:

I have 4 CD burners, 2 on each channel of the MB's IDE controller:

   VP_IDE: VIA vt8235 (rev 00) IDE UDMA133 controller on pci00:11.1
   ide0: BM-DMA at 0xdc00-0xdc07, BIOS settings: hda:DMA, hdb:DMA
   ide1: BM-DMA at 0xdc08-0xdc0f, BIOS settings: hdc:DMA, hdd:DMA

and a 120 Gig HD (boot drive) on a PCI Promise IDE controller:

   PDC20268: IDE controller at PCI slot 00:09.0
   ide2: BM-DMA at 0xa400-0xa407, BIOS settings: hde:pio, hdf:pio
   ide3: BM-DMA at 0xa408-0xa40f, BIOS settings: hdg:pio, hdh:pio

and have ide-scsi attached to the 4 CD burners via a boot param:

   Kernel command line: auto BOOT_IMAGE=Linux-2.4.22 ro root=210
   ide-scsi hdc=ide-scsi hdd=ide-scsi
   ide_setup: hda=ide-scsi
   ide_setup: hdb=ide-scsi
   ide_setup: hdc=ide-scsi
   ide_setup: hdd=ide-scsi
     ide2: BM-DMA at 0xa400-0xa407, BIOS settings: hde:pio, hd
     ide3: BM-DMA at 0xa408-0xa40f, BIOS settings: hdg:pio, hd
     ide0: BM-DMA at 0xdc00-0xdc07, BIOS settings: hda:DMA, hd
     ide1: BM-DMA at 0xdc08-0xdc0f, BIOS settings: hdc:DMA, hd
   hda: LITE-ON LTR-48246S, ATAPI CD/DVD-ROM drive
   hdb: LITE-ON LTR-48246S, ATAPI CD/DVD-ROM drive
   hdc: LITE-ON LTR-48246S, ATAPI CD/DVD-ROM drive
   hdd: LITE-ON LTR-48246S, ATAPI CD/DVD-ROM drive
   hde: Maxtor 6Y080P0, ATA DISK drive
   hde: attached ide-disk driver.
   hde: host protected area => 1
   hde: 160086528 sectors (81964 MB) w/7936KiB Cache, CHS=158816
    hde: hde1 hde2
   hda: attached ide-scsi driver.
   hdb: attached ide-scsi driver.
   hdc: attached ide-scsi driver.
   hdd: attached ide-scsi driver.

I am using 4 instances of CDRDAO to burn a single image to the 4
different CD burners. When running with 2.4.21-ac1, everything ran
'mostly' fine. Occasionally one of the drives would write a VERY large
lead-out, and the other drive on that channel would not eject until the
lead-out was done on the other drive (eject command wasn't received?).
Burning occurred at full speed in this setup, but the freeze-ups and
delays were frustrating. The firmware on the drives is up to date, and
burning an image to a single drive works flawlessly.

We updated to 2.4.22 to see if something had been fixed, and it looks
like everything is working fine now, but everything is VERY VERY slow.
The drives are burning at about 1/3rd speed as they were. It also
appears as though 1 of the drives will receive a large chunk of data and
spin-up, then another drive receives a large chunk and spins up, etc..
instead of uniformly writing to all 4 drives at the same time.

Running 'top' during the burn shows that kernel time was at 65% and user
time was near zero. It is almost as if the ide driver or the ide scsi
emulation layer was locked up (mutex'd?) on a single command until it
completed (speculation! :-P)

Anyone else see this? Is there an explanation for this? Any pointers to
docs that I should read? I've grep'd the Documentation and searched
elsewhere, but couldn't find any more info on this topic. :-/ Below is
some more information that might be helpful.

Thanks!

-Scott

uname -a:
Linux doop1 2.4.22 #2 Mon Sep 8 14:09:27 CDT 2003 i686 unknown

/proc/cpuinfo:
processor : 0
vendor_id : AuthenticAMD
cpu family : 6
model : 8
model name : AMD Athlon(tm) XP 2200+
stepping : 1
cpu MHz : 1810.027
cache size : 256 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 1
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
mca cmov pat pse36 mmx fxsr sse syscall mmxext 3dnowext 3dnow
bogomips : 3604.48

/proc/ioports:
0000-001f : dma1
0020-003f : pic1
0040-005f : timer
0060-006f : keyboard
0070-007f : rtc
0080-008f : dma page reg
00a0-00bf : pic2
00c0-00df : dma2
00f0-00ff : fpu
0170-0177 : ide1
01f0-01f7 : ide0
02f8-02ff : serial(set)
0376-0376 : ide1
03c0-03df : vga+
03f6-03f6 : ide0
03f8-03ff : serial(set)
0cf8-0cff : PCI conf1
9000-907f : Silicon Integrated Systems [SiS] SiS300/305 PCI/AGP VGA
Display Adapter
9400-9407 : Promise Technology, Inc. 20268
   9400-9407 : ide2
9800-9803 : Promise Technology, Inc. 20268
   9802-9802 : ide2
9c00-9c07 : Promise Technology, Inc. 20268
a000-a003 : Promise Technology, Inc. 20268
a400-a40f : Promise Technology, Inc. 20268
   a400-a407 : ide2
   a408-a40f : ide3
bc00-bc07 : CMD Technology Inc Silicon Image SiI 3112 SATARaid Controller
c000-c003 : CMD Technology Inc Silicon Image SiI 3112 SATARaid Controller
c400-c407 : CMD Technology Inc Silicon Image SiI 3112 SATARaid Controller
c800-c803 : CMD Technology Inc Silicon Image SiI 3112 SATARaid Controller
cc00-cc0f : CMD Technology Inc Silicon Image SiI 3112 SATARaid Controller
d000-d01f : VIA Technologies, Inc. USB
d400-d41f : VIA Technologies, Inc. USB (#2)
d800-d81f : VIA Technologies, Inc. USB (#3)
dc00-dc0f : VIA Technologies, Inc. VT82C586A/B/VT82C686/A/B/VT823x/A/C
PIPC Bus Master IDE
   dc00-dc07 : ide0
   dc08-dc0f : ide1
e000-e0ff : VIA Technologies, Inc. VT8233/A/8235 AC97 Audio Controller
e400-e4ff : VIA Technologies, Inc. VT6102 [Rhine-II]
   e400-e4ff : via-rhine

/proc/iomem:
00000000-0009e7ff : System RAM
0009e800-0009ffff : reserved
000a0000-000bffff : Video RAM area
000c0000-000c7fff : Video ROM
000d0000-000d27ff : Extension ROM
000f0000-000fffff : System ROM
00100000-0ffeffff : System RAM
   00100000-001ee67d : Kernel code
   001ee67e-0022d87f : Kernel data
0fff0000-0fff2fff : ACPI Non-volatile Storage
0fff3000-0fffffff : ACPI Tables
d0000000-d7ffffff : VIA Technologies, Inc. VT8377 [KT400 AGP] Host Bridge
d8000000-dfffffff : Silicon Integrated Systems [SiS] SiS300/305 PCI/AGP
VGA Display Adapter
e0060000-e007ffff : Silicon Integrated Systems [SiS] SiS300/305 PCI/AGP
VGA Display Adapter
e00a0000-e00a3fff : Promise Technology, Inc. 20268
e00a4000-e00a41ff : CMD Technology Inc Silicon Image SiI 3112 SATARaid
Controller
e00a5000-e00a50ff : VIA Technologies, Inc. USB 2.0
e00a6000-e00a60ff : VIA Technologies, Inc. VT6102 [Rhine-II]
   e00a6000-e00a60ff : via-rhine
fec00000-fec00fff : reserved
fee00000-fee00fff : reserved
ffff0000-ffffffff : reserved

lsmod:
Module Size Used by Not tainted
ide-scsi 9152 0
scsi_mod 85512 1 [ide-scsi]
via-rhine 13416 1
mii 2416 0 [via-rhine]
crc32 2816 0 [via-rhine]
rtc 6044 0 (autoclean)

/proc/pci:
PCI devices found:
   Bus 0, device 0, function 0:
     Host bridge: VIA Technologies, Inc. VT8377 [KT400 AGP] Host Bridge
(rev 0).
       Master Capable. Latency=8.
       Prefetchable 32 bit memory at 0xd0000000 [0xd7ffffff].
   Bus 0, device 1, function 0:
     PCI bridge: VIA Technologies, Inc. VT8235 PCI Bridge (rev 0).
       Master Capable. No bursts. Min Gnt=4.
   Bus 0, device 8, function 0:
     VGA compatible controller: Silicon Integrated Systems [SiS]
SiS300/305 PCI/AGP VGA Display Adapter (rev 144).
       Prefetchable 32 bit memory at 0xd8000000 [0xdfffffff].
       Non-prefetchable 32 bit memory at 0xe0060000 [0xe007ffff].
       I/O at 0x9000 [0x907f].
   Bus 0, device 9, function 0:
     Unknown mass storage controller: Promise Technology, Inc. 20268
(rev 2).
       IRQ 10.
       Master Capable. Latency=32. Min Gnt=4.Max Lat=18.
       I/O at 0x9400 [0x9407].
       I/O at 0x9800 [0x9803].
       I/O at 0x9c00 [0x9c07].
       I/O at 0xa000 [0xa003].
       I/O at 0xa400 [0xa40f].
       Non-prefetchable 32 bit memory at 0xe00a0000 [0xe00a3fff].
   Bus 0, device 15, function 0:
     RAID bus controller: CMD Technology Inc Silicon Image SiI 3112
SATARaid Controller (rev 1).
       IRQ 11.
       Master Capable. Latency=32.
       I/O at 0xbc00 [0xbc07].
       I/O at 0xc000 [0xc003].
       I/O at 0xc400 [0xc407].
       I/O at 0xc800 [0xc803].
       I/O at 0xcc00 [0xcc0f].
       Non-prefetchable 32 bit memory at 0xe00a4000 [0xe00a41ff].
   Bus 0, device 16, function 0:
     USB Controller: VIA Technologies, Inc. USB (rev 128).
       IRQ 12.
       Master Capable. Latency=32.
       I/O at 0xd000 [0xd01f].
   Bus 0, device 16, function 1:
     USB Controller: VIA Technologies, Inc. USB (#2) (rev 128).
       IRQ 10.
       Master Capable. Latency=32.
       I/O at 0xd400 [0xd41f].
   Bus 0, device 16, function 2:
     USB Controller: VIA Technologies, Inc. USB (#3) (rev 128).
       IRQ 11.
       Master Capable. Latency=32.
       I/O at 0xd800 [0xd81f].
   Bus 0, device 16, function 3:
     USB Controller: VIA Technologies, Inc. USB 2.0 (rev 130).
       IRQ 5.
       Master Capable. Latency=32.
       Non-prefetchable 32 bit memory at 0xe00a5000 [0xe00a50ff].
   Bus 0, device 17, function 0:
     ISA bridge: VIA Technologies, Inc. VT8235 ISA Bridge (rev 0).
   Bus 0, device 17, function 1:
     IDE interface: VIA Technologies, Inc.
VT82C586A/B/VT82C686/A/B/VT823x/A/C PIPC Bus Master IDE (rev 6).
       Master Capable. Latency=32.
       I/O at 0xdc00 [0xdc0f].
   Bus 0, device 17, function 5:
     Multimedia audio controller: VIA Technologies, Inc. VT8233/A/8235
AC97 Audio Controller (rev 80).
       IRQ 11.
       I/O at 0xe000 [0xe0ff].
   Bus 0, device 18, function 0:
     Ethernet controller: VIA Technologies, Inc. VT6102 [Rhine-II] (rev
116).
       IRQ 12.
       Master Capable. Latency=32. Min Gnt=3.Max Lat=8.
       I/O at 0xe400 [0xe4ff].
       Non-prefetchable 32 bit memory at 0xe00a6000 [0xe00a60ff].

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

• Previous message: Andrew de Quincey: "Re: [PATCH] deal with lack of acpi prt entries gracefully"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Relevant Pages Re: SH 205 ... and what technology is inside: ... Firstly the SH205 uses a 20Mb MFM drive, in general these drives are ... it converts ASCI to SCSI and then ... I heard stories once that it was possible to adapt the controller to ... (comp.sys.atari.st) Re: Looking for SAS/SATA RAID Controller That Supports JBOD ... Right, which is why one speaks of a JBOD cabinet, not a JBOD drive. ... All refer to such a controller as a *RAID* controller supporting JBOD ar- ... -or- an external controller spanning drives in a cabinet. ... (comp.periphs.scsi) Re: How to play an old floppy disk. ... Here are the circuit diagrams for the FDD/HDD controller in the ... Therefore it has the same data rate and should be ... Are you sure there ever were drives for 40 tracks double sided? ... Specifies the size of the floppy disk to format. ... (comp.sys.ibm.pc.hardware.storage) Re: SBS2003 -Limited space, New HD, GHOST ... controller from Dell it'll cost me about $700-$800. ... Purchase two new drives of the size you want. ... support more kinds of raid, but for the OS, RAID1 is a good place to be. ... Regarding the driver issue: ... (microsoft.public.windows.server.sbs) Re: [SLE] Added disk now no boot ... > I have this old Promise Ultra 100 card that supports 4 ide drives. ... > attached, then my RAID controller signs on like it used to, then I get ... The added controller's BIOS loads and prevents your normal boot with a drive ... Is your current RAID controller an actual PCI slot card or is it embedded on ... (SuSE)