• Bootcode.bin USB storage discovery

    From bob prohaska@3:770/3 to All on Wed Nov 17 16:06:49 2021
    I'm having difficulties getting a Pi3b+ to reliably detect
    a USB hard drive on boot. Getting into the shell on u-boot
    it's possible to repeatedly probe using
    usb reset
    and eventually the drive shows up. At that point
    run bootcmd_usb0
    seems to work, but I so far havent found the "magic words"
    to make the disk show up on the first try. Using the special
    version of bootcode.bin offred at https://www.raspberrypi.com/documentation/computers/raspberry-pi.html#special-bootcode-bin-only-boot-mode

    doesn't seem to help, leaving the machine in an unrecoverable network boot
    loop that can only be stopped by cycling power, even with the serial
    connection turned on.

    The randomness of detection makes it seem like some sort of timing problem. Once the drive is detected it works just fine, so I don't think there's anything intrinsically wrong with it. There is a timeout file on the
    microSD and a config.txt file containing the line
    program_usb_boot_timeout=1
    along with a second timeout file on the target hard disk (not sure
    if that's even involved).

    If anybody has suggestions they'd be much appreciated.

    Thanks for reading,

    bob prohaska
    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From crn@nospam.com@3:770/3 to bob prohaska on Wed Nov 17 17:19:47 2021
    bob prohaska <bp@www.zefox.net> wrote:
    I'm having difficulties getting a Pi3b+ to reliably detect
    a USB hard drive on boot. Getting into the shell on u-boot
    it's possible to repeatedly probe using
    usb reset

    Power the hard drive from its own supply and give it time to
    warm up before starting the pi.
    The pi's ability to supply power to a usb hd is marginal at best.

    --
    http://www.netunix.com/
    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From Lew Pitcher@3:770/3 to crn on Wed Nov 17 17:52:07 2021
    On Wed, 17 Nov 2021 17:19:47 +0000, crn wrote:

    bob prohaska <bp@www.zefox.net> wrote:
    I'm having difficulties getting a Pi3b+ to reliably detect a USB hard
    drive on boot. Getting into the shell on u-boot it's possible to
    repeatedly probe using usb reset

    Power the hard drive from its own supply and give it time to warm up
    before starting the pi.
    The pi's ability to supply power to a usb hd is marginal at best.

    Further: use a powered USB hub between the Pi and the USB HD
    The hub will supply the power for the HD, taking the load off the Pi,
    and you can use the hub to supply power to the Pi.



    --
    Lew Pitcher
    "In Skills, We Trust"
    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)
  • From bob prohaska@3:770/3 to Lew Pitcher on Thu Nov 18 01:39:49 2021
    Lew Pitcher <lew.pitcher@digitalfreehold.ca> wrote:
    On Wed, 17 Nov 2021 17:19:47 +0000, crn wrote:

    bob prohaska <bp@www.zefox.net> wrote:
    I'm having difficulties getting a Pi3b+ to reliably detect a USB hard
    drive on boot. Getting into the shell on u-boot it's possible to
    repeatedly probe using usb reset

    Power the hard drive from its own supply and give it time to warm up
    before starting the pi.
    The pi's ability to supply power to a usb hd is marginal at best.

    Further: use a powered USB hub between the Pi and the USB HD
    The hub will supply the power for the HD, taking the load off the Pi,
    and you can use the hub to supply power to the Pi.

    Understood and done from the beginning. The Pi is not powering the
    disk, but it does appear to be starting and stopping the disk; the power/activity light flashes when the Pi resets.

    The only boot timing parameters I know of are the timeout file on
    the DOS partition of the microSD card and usb_pgood_delay set as
    an environment variable understood to be milliseconds in u-boot.
    I've played with usb_pgood_delay from 1 to 40000 milliseconds and
    it seems to make no consistent difference. I just tried renaming
    the timeout file on the microSD and that didn't cause an obvious
    change of behavior, with the disk discovered on the second usb reset.

    Repeating the usb reset command several times eventually resulted
    in failure, by which time the disk drive light was off, as if the
    drive had gone to sleep.

    I wonder if there might be some parameters on the disk which could
    be set, such as never sleep. I do however have two other examples
    of the same drive/enclosure combo, both on Pi4B machines, and
    neither displays such narcolepsy.

    Thanks for writing,

    bob prohaska
    --- SoupGate-Win32 v1.05
    * Origin: Agency HUB, Dunedin - New Zealand | Fido<>Usenet Gateway (3:770/3)