Messages in this thread | | | Date | Sat, 30 Oct 1999 04:10:21 -0700 (PDT) | From | Andre Hedrick <> | Subject | Re: 2.2.x: "hda: no DRQ after issuing WRITE", but not on 2.0.x kernels |
| |
It appears to be a leading interrupt that does not get cleared in the driver. Since this is an evolving issue as the hardware changes, it is painful to attempt to wrap a headache around this new issue.
First thing to try is a chipset tuning "monster" patch on your local kernel mirror.
Since you are poking an ATA-66 device with an ATA-33 host adapter, you need to verfiy that the "controller is in the correct mode and the drive is back-speed to the correct transfer rate.
What does "hdparm -iI /dev/hda" return?
Andre Hedrick The Linux IDE guy
On Thu, 21 Oct 1999, Dan Browning wrote:
> Linux hackers: > > I recieve the following kernel error message at random times throughout the > day (approx 10/hour): > > hda: status timeout: status=0xd0 { Busy } > hda: no DRQ after issuing WRITE > ide0: reset: success > > I have gotten this message on 2.2.5, 2.2.9, 2.2.12, and 2.2.13, but not any > of the 2.0.x kernels. I have reproduced the error on 4 identical computers, > and 2 that are similarly configured (give or take a few megahertz and > motherboards). My current debugging machine is: > > PII-400, 128mb RAM, 6.4GB IDE Seagate ST36421A in udma33 mode, Intel > etherexpress pro/100+ management > adapter on an intel se440bx-2 mb (also reproduced on asus p2b mb). I've > used all manners of IDE cables. > > Boot up (on 2.2.13) says: > > PIIX4: IDE controller on PCI bus 00 dev 39 > PIIX4: not 100% native mode: will probe irqs later > ide0: BM-DMA at 0x1060-0x1067, BIOS settings hda:DMA, hdb:pio > ide1: BM-DMA at 0x1068-0x106f, BIOS settings hdc:DMA, hdd:pio > hda: ST36421A, ATA DISK drive > hdc: TOSHIBA CD-ROM XM-6502B, ATAPI CDROM drive > ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 > ide1 at 0x170-0x177,0x376 on irq 15 > hda: ST36421A, 6150MB w/256kB Cache, CHS=784/255/63 > hdc: ATAPI 40X CD-ROM drive, 256kB Cache > Uniform CDROM driver Revision: 2.56 > > hdparm -t says about 2.69 to 3.97 MB/sec > hdparm -T says about 105 to 107 MB/sec > > I'm on the most updated BIOS update from Intel (P15), and I've tried > changing the mode in the BIOS to NORMAL, LBA, etc. but it makes no > difference. It doesn't seem to make a difference if I run a lot of > processes or none. Even disk-intensive commands like $ls -R / run fine, and > don't cause any errors. > > Attached are the only two messages that I could find regarding my problem in > the archives. The replies are: "your hdd is probably dieing," and "are your > cables short enough?". So I'm assuming that the problem has not been solved > yet. I wanted to run a 2.3 kernel to test it, but 2.3.19 wouldn't compile > for me. I'll try again with 2.3.22 if you guys think it will get me > anywhere (at at least help debug). Though I don't think it's the hdd's > fault, I pulled the following info on the drive I'm using: > > ST36421A > U4 Family > Cylinders: 833 > Heads: 240 > Sectors: 63 > Capacity: 6.45 GB > Speed: 5400 rpm > Seek time: 10.5 ms avg > SeaFAX#: 364212 > Ultra ATA/66 (Ultra DMA/66) > > Would it have anything to do with the fact that the drive is UDMA/66? The > motherboard only supports umda33 tho. Here's some info from intel on the > se-440bx-2 motherboard: > > Intel 82440BX AGPset and PCI/IDE Interface > Intel 82443BX PCI/AGP controller (PAC) > Integrated PCI bus mastering controller > Integrated Accelerated Graphics Port (AGP) controller > Intel 82371EB PCI ISA IDE Xcelerator (PIIX4E) > Multifunction PCI-to-ISA bridge > Universal Serial Bus (USB) and DMA controllers > Two fast IDE interfaces, that support up to four IDE drives or devices > Power management logic > Real-time clock > I/O Features > SMC FDC37M707 Super I/O controller > > Intel also has about 16 un-fixed errata on this board. The only > semi-relevant errata I could find is: "BIOS does not impliment S4BIOS Power > state"--and says it will not be fixed. > > Also, who is the IDE maintainer(s)? If this does turn out to be a kernel > problem, direct contact with him or her would be helpful. Of course, it > will probably be one of my retarded mistakes (as a nerd, I make many). > > > Re: ide drive busy - is this "dangerous"? > Philip Blundell (Philip.Blundell@pobox.com) > Sat, 27 Dec 1997 23:00:04 +0000 > > Messages sorted by: [ date ][ thread ][ subject ][ author ] > Next message: Philip Blundell: "Re: PPP broken sometime after 2.1.66?" > Previous message: Bruce Elliott: "Alpha osf_sys.c problems" > Maybe in reply to: Tuomas Heino: "ide drive busy - is this "dangerous"?" > > ---------------------------------------------------------------------------- > ---- > >hdc: status timeout: status=0xd0 { Busy } > >hdc: no DRQ after issuing WRITE > >ide1: reset: success > > > >Do that kind of messages mean that something is wrong? > >(besides hardware being slow?) > >(that's too much like the all-too-common-aic7xxx messages back in > the-old-bad- > >days) > > It may mean that your IDE disk is about to fail. Some drives start producing > these symptoms when they have to do retries to read a sector. > p. > > [next message] > > Re: ide drive busy - is this "dangerous"? > mlord (mlord@pobox.com) > Mon, 29 Dec 1997 14:03:02 -0500 > > Messages sorted by: [ date ][ thread ][ subject ][ author ] > Next message: Bryn Paul Arnold Jones: "Re: hardlinks.... sucks... ;-(" > Previous message: Yuri Kuzmenko: "Re: hardlinks.... sucks... ;-(" > Maybe in reply to: Tuomas Heino: "ide drive busy - is this "dangerous"?" > > ---------------------------------------------------------------------------- > ---- > Tuomas Heino wrote: > > > > hdc: status timeout: status=0xd0 { Busy } > > hdc: no DRQ after issuing WRITE > > ide1: reset: success > > > > Do that kind of messages mean that something is wrong? > > If you see a lot of them, then something is wrong somewhere. > Double-check the connectors and ensure the cable lengths > are 18" or less. > > Some drives seem to generate these errors once in a while, > and as you can see, Linux notices and corrects for the problem. > -- > mlord@pobox.com > The Linux IDE guy > > > I hope this helps you guys, > > Dan Browning > Network Administrator > Cyclone Computer Systems > > > - > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.rutgers.edu > Please read the FAQ at http://www.tux.org/lkml/ >
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |