Warning: the dma on your hard drive is turned off
Hello!
I am running Slackware Linux-12.1 with kernel-generic-smp-2.6.24.5-smp, a fresh new installation. Recently I started getting the following messaged during booting. Earlier they appeared at some boots only but now they appear whenever I boot up the machine. Here they are: Warning: The dma on your hard drive is turned off. This may really slow down the fsck process. I never got these messages in Slackware-12.0. Here is some info: dmesg | grep hda: ide0: BM-DMA at 0xffa0-0xffa7, BIOS settings: hda:DMA, hdb:DMA hda: Maxtor 6E040L0, ATA DISK drive hda: host max PIO4 wanted PIO255(auto-tune) selected PIO4 hda: UDMA/100 mode selected hda: max request size: 128KiB hda: 80293248 sectors (41110 MB) w/2048KiB Cache, CHS=65535/16/63 hda: cache flushes supported hda: hda1 hda2 < hda5 > hda: DMA disabled dmesg | grep hdb ide0: BM-DMA at 0xffa0-0xffa7, BIOS settings: hda:DMA, hdb:DMA hdb: ST3160812A, ATA DISK drive hdb: host max PIO4 wanted PIO255(auto-tune) selected PIO4 hdb: UDMA/100 mode selected hdb: max request size: 512KiB hdb: 312581808 sectors (160041 MB) w/8192KiB Cache, CHS=19457/255/63 hdb: cache flushes supported hdb: dma_intr: status=0x7f { DriveReady DeviceFault SeekComplete DataRequest CorrectedError Index Error } hdb: dma_intr: error=0x7f { DriveStatusError UncorrectableError SectorIdNotFound TrackZeroNotFound AddrMarkNotFound }, LBAsect=1647103278975, high=98175, low=98175, sector=15230687 hdb: DMA disabled dmesg | grep dma: hdb: dma_intr: status=0x7f { DriveReady DeviceFault SeekComplete DataRequest CorrectedError Index Error } hdb: dma_intr: error=0x7f { DriveStatusError UncorrectableError SectorIdNotFound TrackZeroNotFound AddrMarkNotFound }, LBAsect=1647103278975, high=98175, low=98175, sector=15230687 dmesg | grep ide ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx ide0: BM-DMA at 0xffa0-0xffa7, BIOS settings: hda:DMA, hdb:DMA ide1: BM-DMA at 0xffa8-0xffaf, BIOS settings: hdc:DMA, hdd:DMA Probing IDE interface ide0... ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 Probing IDE interface ide1... hdc: host side 80-wire cable detection failed, limiting max speed to UDMA33 ide1 at 0x170-0x177,0x376 on irq 15 ide-floppy driver 0.99.newide ide: failed opcode was: unknown ide0: reset: success hdparm -I /dev/hda | grep dma: DMA: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5 udma6 hdparm -I /dev/hdb | grep dma DMA: mdma0 mdma1 *mdma2 udma0 udma1 udma2 udma3 udma4 udma5 If I try to manualy hdparm -d1 /dev/hda, the command succeeds and there is dma on both hard drives. My question is whether there is a hardware problem on my system, how to check whether it indeed is a hardware problem. Or what I could do to turn dma on at boot so that these messages do not appear. My system seems to run appreciably slower. Any suggestions will be appreciated. Regards, Martin |
Quote:
|
Quote:
|
What do you see when you run /sbin/hdparm -d /dev/hda?
|
Sorry for my late answer.
I reinstalled the system several times so far to make sure that this is not due to some so. My final conclusion is that probably after a power failure something went wring with my windows drive (/dev/hda). After booting into windows, the message disappeared while booting Linux. Now I no more get these messages during boot. The problem is that "hdparm -d /dev/hda" and "hdparm -d /dev/hdb" both say that dma is off. My system runs fairly slowly until I issue "/hdparm -d 1 /dev/hda" and "hdparm -d 1 /dev/hdb". Then hdparm reports that dma is on (dma=1) for both hard drives. Isn't dma supposed to be enabled by default? Should I add these commands to rc.local or some other more appropriate place? Here is some info from my syslog: May 30 14:55:25 marto kernel: hdb: ST3160812A, ATA DISK drive May 30 14:55:25 marto kernel: hda: Maxtor 6E040L0, ATA DISK drive May 30 14:55:25 marto kernel: ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 May 30 14:55:25 marto kernel: hdd: TEAC CD-W552E, ATAPI CD/DVD-ROM drive May 30 14:55:25 marto kernel: hdc: Optiarc DVD RW AD-7173A, ATAPI CD/DVD-ROM drive May 30 14:55:25 marto kernel: hdc: host side 80-wire cable detection failed, limiting max speed to UDMA33 May 30 14:55:25 marto kernel: ide1 at 0x170-0x177,0x376 on irq 15 May 30 14:55:25 marto kernel: ide-floppy driver 0.99.newide May 30 14:55:25 marto kernel: Driver 'sd' needs updating - please use bus_type methods May 30 14:55:25 marto kernel: Driver 'sr' needs updating - please use bus_type methods May 30 14:55:25 marto kernel: PNP: PS/2 appears to have AUX port disabled, if this is incorrect please boot with i8042.nopnp May 30 14:55:25 marto kernel: raid6: int32x1 473 MB/s May 30 14:55:25 marto kernel: raid6: int32x2 645 MB/s May 30 14:55:25 marto kernel: raid6: int32x4 752 MB/s May 30 14:55:25 marto kernel: raid6: int32x8 455 MB/s May 30 14:55:25 marto kernel: raid6: mmxx1 1665 MB/s May 30 14:55:25 marto kernel: raid6: mmxx2 2109 MB/s May 30 14:55:25 marto kernel: raid6: sse1x1 966 MB/s May 30 14:55:25 marto kernel: raid6: sse1x2 1833 MB/s May 30 14:55:25 marto kernel: raid6: sse2x1 2129 MB/s May 30 14:55:25 marto kernel: raid6: sse2x2 2256 MB/s May 30 14:55:25 marto kernel: raid6: using algorithm sse2x2 (2256 MB/s) May 30 14:55:25 marto kernel: Using IPI No-Shortcut mode May 30 14:55:25 marto kernel: intel_rng: Firmware space is locked read-only. If you can't or May 30 14:55:25 marto kernel: intel_rng: don't want to disable this in firmware setup, and if May 30 14:55:25 marto kernel: intel_rng: you are certain that your system has a functional May 30 14:55:25 marto kernel: intel_rng: RNG, try using the 'no_fwh_detect' option. May 30 14:55:25 marto kernel: rtc_cmos: probe of 00:02 failed with error -16 May 30 14:55:25 marto kernel: ACPI Exception (processor_core-0816): AE_NOT_FOUND, Processor Device is not present [20070126] May 30 14:55:25 marto kernel: videodev: "QCM USB Camera" has no release callback. Please fix your driver for proper sysfs support, see http://lwn.net/Articles/36850/ May 30 14:55:25 marto kernel: fuse init (API version 7.9) May 30 14:55:25 marto kernel: lm85 0-002e: Client (0,0x2e) config is locked. May 30 14:55:48 marto kernel: hdb: dma_intr: status=0x7f { DriveReady DeviceFault SeekComplete DataRequest CorrectedError Index Error } May 30 14:55:48 marto kernel: hdb: dma_intr: error=0x7f { DriveStatusError UncorrectableError SectorIdNotFound TrackZeroNotFound AddrMarkNotFound }, LBAsect=1647103278975, high=98175, low=98175, sector=301759 May 30 14:55:48 marto kernel: ide: failed opcode was: unknown May 30 14:55:50 marto kernel: ide0: reset: success Both drives are correctly recognized. However, there are some error messages that I do not know what to think of. Thank you for your responsiveness. Regards, Martin |
Sorry for cross-posting. I didn't notice that in my previous post the second sentence is unfinished. It should read:"I reinstalled the system several times so far to make sure that this is not due to something that I messed up after the installation."
I would also ask if I should suspect a hardware problem and consult a computer service? What would you recommend? |
Hi,
What does 'hdparm -i /dev/hda' & '/hdparm -i /dev/hdb' show? The 'i' flag is for boot time information. After that issue the 'hdparm -I /dev/hda' & 'hdparm -I /dev/hdb'. The 'I' flag is direct from the HDD. Do you have the newer '80 conductor' cable for your HDD? Edit: It is easier to read a post when long list or data is placed within the vbcode tags. These tags are at the top of the reply window; Quote or Code tag. |
Quote:
Quote:
Code:
root@marto:~# hdparm -i /dev/hda |
As well as the good info posted above by others, some BIOSs have options for turning DMA [on|off|auto|PIO[123]|UDMA[12345]} when booting the system.
I'd guess that an "auto" setting *could* result in random on or off settings at boot, but that's a guess. Do check for that 80-wire cable too, and that the master drive is at the END of the cable, and any slave drive is in the MIDDLE plug. Finally, if you were wondering (I think I saw mention of it above) a great place for hdparm commands to be executed at startup is in the /etc/rc.local script. :) Sasha PS - for the record, AFAIK from reading, SATA drives have DMA enabled automatically by default, I suppose provided that the correct driver is in use too. |
Quote:
|
Quote:
:D Sasha |
Quote:
Thank you very much for your responsiveness. Regards, Martin PS It might be a dull question, but I would also like to ask if there is any sense in reinstalling the operating system not that the hardware setting have changed. |
Well, if all is working, I see no reason at all to reinstall the OS. If anything happens to be slightly misconfigged (which I doubt) by rearranging the drives, I would just watch for that and fix it. Reinstalling seems at this point to be needless.
Also, if you do a bit of research about the ATA/IDE standards over the years, you will read that CS (cable select) is an invention/idea that did not really take off, and that has not been standardized nor implemented as intended by most HW and SW vendors. If anything, it has potential to give you grief at some point, if it does anything at all. If it were me, I would change that jumper from CS to Slave. Happy DMAing :) Sasha |
Quote:
|
OK, gotcha. Check with your friend first, but:
If you look closely at the drive housing, there is often the settings of the jumpers stamped or engraved into the metal. Alternately, many have a sticker on the drive showing the settings. If I remember right, a drive with NO jumper at all will default to Slave, if there is a master present. Don't quote me here though. Above all, if it works, go with it :) and leave it on CS unless you learn otherwise. Cheers! Sasha |
All times are GMT -5. The time now is 11:56 PM. |