Re: Accessing IDE disk with bad sectors freezes the box

From: Søren Schmidt (sos_at_DeepCore.dk)
Date: 05/11/05

  • Next message: Rob: "Re: xl(4) & polling" 
    Date: Wed, 11 May 2005 15:26:02 +0200
To: Ruslan Ermilov <ru@FreeBSD.org>

    Ruslan Ermilov wrote:
    > Hi Soren,
    >
    > I have a disk with lot of bad sectors. When working with it on
    > an AMD64 box running 6-CURRENT, accessing bad areas just freezes
    > the box completely, without any diagnostics. The same disk when
    > plugged into another i386 box running 4-STABLE works properly by
    > issuing errors from the kernel, and reporting EIO to userland.
    >
    > Is there something I can do to try to mitigate this issue?

    Hmm, depending on the problem the disk has and how the controller is
    implemented the behavior on fatal errors can be unpredictable,
    especially when DMA is involved (some controllers just lock the bus
    completely so our timeout code wont get a chance).
    If you switch the disk to PIO mode, does that help any ? 

    -- 
-Søren
_______________________________________________
freebsd-current@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-current-unsubscribe@freebsd.org"
  • Next message: Rob: "Re: xl(4) & polling"

    Relevant Pages

    • PROBLEM: HT1000 drops network packets during disk writes
      ... HT1000 drops network packets during disk writes ... _any_ disk writes are done on onboard SATA controller at the ... All tested kernel versions are affected. ...
      (Linux-Kernel)
    • Re: IO scheduler based IO controller V10
      ... resources like cpu scheduling and memory controllers. ... Also Fernando tells me that memory controller did not support ... Enforcing limits is applicable in the scenario where you are managing ... fairness in terms of disk time can get us better results ...
      (Linux-Kernel)
    • Re: SCO OS 5.0.7 on Qemu
      ... The OSR5 "wd" IDE driver has code in it to ... > controller used in PC/AT class machines), ... > the kernel to look for a SCSI disk on a "wd" HBA. ... Possible UDMA timing mismatch, ...
      (comp.unix.sco.misc)
    • Re: Prime 2550
      ... Isn't sectoring, reverse or forward, just another name for interleave? ... interleave factor of 1 less than the number of sectors on a track. ... optimally from a disk. ... sectoring is faster on non-intelligent controller. ...
      (comp.sys.prime)
    • Re: 3B2 Disks
      ... being able to read the disk in its present format. ... 2 MFM drives on a custom controller. ... SCSI came much later as an add on card. ...
      (comp.sys.3b1)