ALI 15X3 DMA Freeze
I'm using kernel 2.4.18-pre9, which has a problem when I switch DMA on the IDE hard drive. With DMA enabled, I cannot complete a kernel compile without the machine locking up. The motherboard is an ASUS board, with 512MB of DDR RAM and an Athlon 1800 XP. If I turn the ide DMA off, then I can compile as many linux kernels as I like with no lockup. This DMA lockup also occurs on other 2.4.x kernels. Is anyone interested in getting me to test a fix for this problem?
Feb 10 10:44:35 solaris kernel: PCI: Using IRQ router ALI [10b9/1533] at 00:07.0 Feb 10 10:44:35 solaris kernel: ALI15X3: IDE controller on PCI bus 00 dev 20 Feb 10 10:44:35 solaris kernel: ALI15X3: chipset revision 196 Feb 10 10:44:35 solaris kernel: ALI15X3: not 100%% native mode: will probe irqs later Feb 10 10:44:35 solaris kernel: ALI15X3: ATA-66/100 forced bit set (WARNING)!!
with dma off hdparm -Tt /dev/hda gives:
/dev/hda:
Timing buffer-cache reads: 128 MB in 0.56 seconds =228.57 MB/sec
Timing buffered disk reads: 64 MB in 14.85 seconds = 4.31 MB/sec
with dma off hdparm -Tt /dev/hda gives:
/dev/hda:
Timing buffer-cache reads: 128 MB in 0.49 seconds =261.22 MB/sec
Timing buffered disk reads: 64 MB in 1.70 seconds = 37.65 MB/sec
Cheers. Paul
Quoted 3 lines
I'm using kernel 2.4.18-pre9, which has a problem when I switch DMA on the IDE hard drive. With DMA enabled, I cannot complete a kernel compile without the machine locking up. The motherboard is an
Start by picking up the ide updates either from the -ac kernel or from Andre's web site.
I've picked up ide.2.4.17.02072002.patch from www.linux-ide.org and compiled it with 2.4.17. I presume that this was the patch you were referring to?
With Andre's IDE patch and 2.4.17, my machine locked up when compiling 2.4.17 with both the DMA on and the DMA off.
With patch-2.4.18-pre9, the machine only locked up when DMA was set on, with DMA off, I could recompile the kernel as many times as I wanted.
Following are kernel messages from the kernel boot with Andre's IDE patch.
Feb 10 15:25:10 solaris kernel: Calibrating delay loop... 3158.83 BogoMIPS Feb 10 15:25:10 solaris kernel: PCI: Using IRQ router ALI [10b9/1533] at 00:07.0 Feb 10 15:25:10 solaris kernel: ALI15X3: IDE controller on PCI bus 00 dev 20 Feb 10 15:25:10 solaris kernel: ALI15X3: chipset revision 196 Feb 10 15:25:10 solaris kernel: ALI15X3: not 100%% native mode: will probe irqs later Feb 10 15:25:10 solaris kernel: ALI15X3: ATA-66/100 forced bit set (WARNING)!! Feb 10 15:25:10 solaris kernel: ALI15X3: simplex device: DMA disabled Feb 10 15:25:10 solaris kernel: ide0: ALI15X3 Bus-Master DMA disabled (BIOS) Feb 10 15:25:10 solaris kernel: ALI15X3: simplex device: DMA disabled Feb 10 15:25:10 solaris kernel: ide1: ALI15X3 Bus-Master DMA disabled (BIOS) Feb 10 15:25:11 solaris random: Initializing random number generator: succeeded
Feb 10 15:25:10 solaris kernel: ide_setup: ide0=ata66 Feb 10 15:25:10 solaris kernel: Uniform Multi-Platform E-IDE driver Revision: 6.31 Feb 10 15:25:10 solaris kernel: ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx Feb 10 15:25:10 solaris kernel: ALI15X3: IDE controller on PCI bus 00 dev 20 Feb 10 15:25:10 solaris kernel: ide0: ALI15X3 Bus-Master DMA disabled (BIOS) Feb 10 15:25:10 solaris kernel: ide1: ALI15X3 Bus-Master DMA disabled (BIOS) Feb 10 15:25:10 solaris kernel: ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 Feb 10 15:25:10 solaris kernel: ide1 at 0x170-0x177,0x376 on irq 15
Please let me know what to try to test next.
Regards. Paul
Quoted 6 lines · Alan Cox
Quoted 3 lines
I'm using kernel 2.4.18-pre9, which has a problem when I switch DMA on the IDE hard drive. With DMA enabled, I cannot complete a kernel compile without the machine locking up. The motherboard is an
Start by picking up the ide updates either from the -ac kernel or from Andre's web site.
Quoted 3 lines
Feb 10 15:25:10 solaris kernel: ide0: ALI15X3 Bus-Master DMA disabled (BIOS) Feb 10 15:25:10 solaris kernel: ALI15X3: simplex device: DMA disabled Feb 10 15:25:10 solaris kernel: ide1: ALI15X3 Bus-Master DMA disabled (BIOS)
Why do you have the IDE DMA disabled in the BIOS ? Is that a setting you made or something your bios knows about your hardware ?
I dont have IDE DMA discabled in the BIOS. That is just something that the ide patch (ide.2.4.17.02072002.patch) printed out. How can I prove that the reporting of IDE DMA disabled is a bug in the patch?
As you can see below, hdparm lets me turn DMA on,
its just that the machine looks up before I
can compile a single Linux kernel.
Here are the boot messages from a vanilla 2.4.18-pre9 boot
[root@solaris root]# dmesg|more
Linux version 2.4.18-pre9 ([email protected]) (gcc version 2.96 20000731 (Red Hat Linux 7.1 2.96-98)) #1 Sun Feb 10 16:59:03 EST 2002
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
ALI15X3: IDE controller on PCI bus 00 dev 20
PCI: No IRQ known for interrupt pin A of device 00:04.0.
ALI15X3: chipset revision 196
ALI15X3: not 100% native mode: will probe irqs later
ALI15X3: ATA-66/100 forced bit set (WARNING)!!
ide0: BM-DMA at 0xd400-0xd407, BIOS settings: hda:DMA, hdb:DMA
ide1: BM-DMA at 0xd408-0xd40f, BIOS settings: hdc:pio, hdd:DMAhda: IC35L040AVER07-0, ATA DISK drive
hdb: CDU5211, ATAPI CD/DVD-ROM drive
hdd: IOMEGA ZIP 250 ATAPI, ATAPI FLOPPY drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide1 at 0x170-0x177,0x376 on irq 15
hda: 80418240 sectors (41174 MB) w/1916KiB Cache, CHS=5005/255/63
hdb: ATAPI 52X CD-ROM drive, 120kB Cache, UDMA(33) Uniform CD-ROM driver Revision: 3.12
Partition check:
hda: hda1 hda2
[root@solaris root]# cat /etc/redhat-release
Red Hat Linux release 7.2 (Enigma)
[root@solaris root]# hdparm -I /dev/hda/dev/hda:
Model=CI530L04VARE700- , FwRev=REO44AA6, SerialNo= S PXXTRT3479
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs } RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=40 BuffType=DualPortCache, BuffSize=1916kB, MaxMultSect=16, MultSect=off CurCHS=16383/16/63, CurSects=-66060037, LBA=yes, LBAsects=80418240 IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120} PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5 AdvancedPM=yes: disabled (255)
Drive Supports : ATA/ATAPI-5 T13 1321D revision 1 : ATA-2 ATA-3 ATA-4 ATA-5
[root@solaris root]# hdparm /dev/hda
/dev/hda:
multcount = 0 (off) I/O support = 0 (default 16-bit) unmaskirq = 0 (off) using_dma = 0 (off) keepsettings = 0 (off) nowerr = 0 (off) readonly = 0 (off) readahead = 8 (on)
geometry = 5005/255/63, sectors = 80418240, start = 0
[root@solaris root]# hdparm -d1 /dev/hda/dev/hda:
setting using_dma to 1 (on)
using_dma = 1 (on)
[root@solaris root]# hdparm /dev/hda/dev/hda:
multcount = 0 (off) I/O support = 0 (default 16-bit) unmaskirq = 0 (off) using_dma = 1 (on) keepsettings = 0 (off) nowerr = 0 (off) readonly = 0 (off) readahead = 8 (on)
geometry = 5005/255/63, sectors = 80418240, start = 0
Let me know what to check next.
Cheers. Paul
Quoted 6 lines · Alan Cox
Quoted 3 lines
Feb 10 15:25:10 solaris kernel: ide0: ALI15X3 Bus-Master DMA disabled (BIOS) Feb 10 15:25:10 solaris kernel: ALI15X3: simplex device: DMA disabled Feb 10 15:25:10 solaris kernel: ide1: ALI15X3 Bus-Master DMA disabled (BIOS)Why do you have the IDE DMA disabled in the BIOS ? Is that a setting you made or something your bios knows about your hardware ?
Hello Mark,
Thats an interesting point. I do. I always thought that Linux supported PNP. Could you tell me a bit more about what setting "PNP OS" actually does?
Cheers. Paul
Quoted 3 lines · Mark Hahn
Quoted 1 line
PCI: No IRQ known for interrupt pin A of device 00:04.0.
are you sure you don't have the bios set for "PNP OS"?
Now I have turned off "PNP OS" and fiddled a little with the boot options. Same basic problem remains, the machine compiles the kernel fine with DMA off but freezes when compiling the kernel with DMA on.
Here are the new boot messages:
Regards. Paul
Kernel command line: auto BOOT_IMAGE=2.4.17.9 ro root=301 BOOT_FILE=/boot/bzImage-2.4.17.9 idebus=66 ide0=ata66 pci=biosirqide_setup: idebus=66ide_setup: ide0=ata66
PCI: PCI BIOS revision 2.10 entry at 0xf1170, last bus=1
PCI: Using configuration type 1
PCI: Probing PCI hardware
Unknown bridge resource 0: assuming transparent
PCI: Using IRQ router ALI [10b9/1533] at 00:07.0
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 66MHz system bus speed for PIO modes
ALI15X3: IDE controller on PCI bus 00 dev 20
PCI: No IRQ known for interrupt pin A of device 00:04.0.
ALI15X3: chipset revision 196
ALI15X3: not 100% native mode: will probe irqs later
ALI15X3: ATA-66/100 forced bit set (WARNING)!!
ide0: BM-DMA at 0xd400-0xd407, BIOS settings: hda:DMA, hdb:DMA
ide1: BM-DMA at 0xd408-0xd40f, BIOS settings: hdc:pio, hdd:DMAhda: IC35L040AVER07-0, ATA DISK drive
hdb: CDU5211, ATAPI CD/DVD-ROM drive
hdd: IOMEGA ZIP 250 ATAPI, ATAPI FLOPPY drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide1 at 0x170-0x177,0x376 on irq 15
hda: 80418240 sectors (41174 MB) w/1916KiB Cache, CHS=5005/255/63
hdb: ATAPI 52X CD-ROM drive, 120kB Cache, UDMA(33) Uniform CD-ROM driver Revision: 3.12
Partition check:
hda: hda1 hda2
Quoted 3 lines · Mark Hahn
Quoted 1 line
PCI: No IRQ known for interrupt pin A of device 00:04.0.
are you sure you don't have the bios set for "PNP OS"?