On Fri, 1 Sep 2006, Adam Rosi-Kessel wrote: > Alan Stern wrote: > >> I have a brand new external USB 2.0 250GB hard drive. It shows up like > >> this: > [...] > >> usb 4-1: reset high speed USB device using ehci_hcd and address 2 > >> usb 4-1: device descriptor read/64, error -71 > >> usb 4-1: device descriptor read/64, error -71 > > Does cycling the power on the disk drive (without rebooting) help? > > That does seem to make the drive usable again, albeit at another device node > (e.g., sdb vs. sda). It still fails after a time as above.
This is another indication that the problem lies either in the drive or in some incompatibility that messes up the drive's firmware. The fact that it started working again after you made no change to the computer at all is a good clue that nothing's wrong with the computer. > >> Any ideas how to isolate the problem? I can't figure out if it is a > >> hardware or software problem. I did a large transfer to and from this > >> drive on another system without problem. The results above on the system > >> in question seem reproducible, at least inasmuch as they keep happening > >> after a couple of days of use. > > Your description makes it sound like a hardware or firmware problem in the > > drive or its USB interface. Or just possibly a marginal hardware > > compatibility problem between the USB interface in the drive and the USB > > controller in your computer (or even the USB cable). > > So I tried connecting the drive to a different system -- nothing in common: > motherboard and CPU are different, kernel is 2.6.11 (compiled for AMD by > Debian) rather than 2.6.17.1. On this system it failed even worse, again > after several hours: Yet more proof that the problem is in the drive, not in the computer. > SCSI error : <0 0 0 0> return code = 0x70000 > end_request: I/O error, dev sda, sector 305659975 > EXT2-fs error (device sda1): read_inode_bitmap: Cannot read inode bitmap - > block_group = 1166, inode_bitmap = 38207489 > > And then, a few minutes later, an *unrelated* drive hiccuped in a way it > never has before: > > hdi: dma_timer_expiry: dma status == 0x64 > hdi: DMA interrupt recovery > hdi: lost interrupt > > Followed by: > > [__report_bad_irq+42/144] __report_bad_irq+0x2a/0x90 > [handle_IRQ_event+48/112] handle_IRQ_event+0x30/0x70 > [note_interrupt+108/208] note_interrupt+0x6c/0xd0 > [__do_IRQ+219/224] __do_IRQ+0xdb/0xe0 > [do_IRQ+25/48] do_IRQ+0x19/0x30 > [common_interrupt+26/32] common_interrupt+0x1a/0x20 > [pg0+408983575/1069908992] uhci_free_pending_qhs+0x17/0x60 [uhci_hcd] > [pg0+408984246/1069908992] uhci_irq+0xf6/0x1b0 [uhci_hcd] > [update_process_times+49/256] update_process_times+0x31/0x100 > [pg0+409315053/1069908992] usb_hcd_irq+0x3d/0x70 [usbcore] > [handle_IRQ_event+48/112] handle_IRQ_event+0x30/0x70 > [__do_IRQ+168/224] __do_IRQ+0xa8/0xe0 > [do_IRQ+25/48] do_IRQ+0x19/0x30 > [common_interrupt+26/32] common_interrupt+0x1a/0x20 > [default_idle+35/48] default_idle+0x23/0x30 > [cpu_idle+72/96] cpu_idle+0x48/0x60 > [start_kernel+362/432] start_kernel+0x16a/0x1b0 > [unknown_bootoption+0/432] unknown_bootoption+0x0/0x1b0 > handlers: > [pg0+407672336/1069908992] (ide_intr+0x0/0x130 [ide_core]) > Disabling IRQ #177 > > And then more SCSI I/O errors. It's vaguely possible the two events are related. Note that the error message didn't say there was anything actually wrong with the older drive; it merely said that an interrupt had been lost. Possibly because of interference caused by the new drive. > Does having these two data points help more? It doesn't help _me_... but it should help _you_. There's nothing I can do to fix your broken disk drive. Alan Stern ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Linux-usb-users@lists.sourceforge.net To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-users
