Loading ...
Sorry, an error occurred while loading the content.

A bit of info on the "Trailer" data bytes at end of image

Expand Messages
  • paulbart1234
    Not too much info, but I noticed the string eRcOmM as part of the 16-byte Trailer at the end of the Flash image (end of 8MBytes): 00 01 00 00 23 24 90 F7
    Message 1 of 6 , Aug 9, 2004
    • 0 Attachment
      Not too much info, but I noticed the string "eRcOmM" as part of the
      16-byte "Trailer" at the end of the Flash image (end of 8MBytes):

      00 01 00 00 23 24 90 F7 65 52 63 4F 6D 4D 00 6D ....#$÷eRcOmM.m

      I did a case-sensitive search for it, and found the following
      strings in the Redboot code:

      copy kernel code from flash to RAM
      copy ramdisk file from flash to RAM
      eRcOmM
      have eRcOmM
      no node address
      run kernel
      push button
      not have eRcOmM

      (For those who don't know, "Redboot" is a boot loader (from Redhat)
      that can start up in 'serial terminal' mode. You can give it a boot
      script to load images from Flash to RAM, load images from the
      network, etc)

      From the above strings, I'd *guess* that startup code looks for
      the "eRcOmM" string @ the end of the Flash, possibly in order to
      determine whether/not it's got valid code in it to run (as opposed
      to startin up in Redboot serial terminal mode).

      The other hex bytes in the "Trailer" may (for example) contain the
      load address of the kernel image, and a checksum/CRC of it. I have
      no idea.


      BTW, a case-*insensitive* search for "eRcOmM" finds more occurrences
      of it in the Redboot section of Flash ("ErCoMm"). I'm not sure what
      it's used for.

      - Paulb
    • paulbart1234
      I ve now got 2 update images (version 2.3 release 24, and version 2.3 release 25), so can compare. Here s the magic 16 bytes at the end of each: 2.3R24: 00 01
      Message 2 of 6 , Aug 9, 2004
      • 0 Attachment
        I've now got 2 update images (version 2.3 release 24, and version
        2.3 release 25), so can compare.

        Here's the magic 16 bytes at the end of each:

        2.3R24:
        00 01 00 00 23 24 90 F7 65 52 63 4F 6D 4D 00 6D ....#$÷eRcOmM.m

        2.3R25:
        00 01 00 00 23 25 90 F7 65 52 63 4F 6D 4D 00 B9 .
        ...#%÷eRcOmM.¹

        So, as (I think) Brian Lantz pointed out, there's some version info
        here ("23 24" and "23 25" look like the firmware version ("23 =
        2.3") and release.

        Most of the rest of the info is identical between versions: the
        first 4 bytes ("00 01 00 00" (=64K, if that matters)), the "90 F7"
        just after the version/release, and "eRcOmM".

        The last 2 bytes (or, maybe just the last byte) differ: "00 6D" in
        2.3R24, and "00 B9" in 2.3R25. Maybe this is some sort of a
        checksum??? (of this 16-byte block of data?)

        - Paulb
      • joule360
        You will note that strings run on the .bin firmware file does find eCos , btw - I wondered about that because the nslu2 is clearly running a linux kernel
        Message 3 of 6 , Aug 10, 2004
        • 0 Attachment
          You will note that 'strings' run on the .bin firmware file does
          find "eCos", btw - I wondered about that because the nslu2 is
          clearly running a linux kernel and not an ecos kernel. But, redboot
          is part of the eCos project, so maybe that's why.

          nl

          --- In nslu2-linux@yahoogroups.com, "rwhitby" <list.yahoo@r...>
          wrote:
          > I just realised that "eRcOmM" is "ecm" and "ROM" interleaved.
          >
          > A redboot.ecm file is discussed here:
          http://ecos.sourceware.org/docs-
          > 2.0/ref/rebuilding-redboot.html
          >
          > -- Rod
          >
          > --- In nslu2-linux@yahoogroups.com, "paulbart1234" <paulbart@b...>
          > wrote:
          > > I've now got 2 update images (version 2.3 release 24, and
          version
          > > 2.3 release 25), so can compare.
          > >
          > > Here's the magic 16 bytes at the end of each:
          > >
          > > 2.3R24:
          > > 00 01 00 00 23 24 90 F7 65 52 63 4F 6D 4D 00 6D ....#$÷eRcOmM.m
          > >
          > > 2.3R25:
          > > 00 01 00 00 23 25 90 F7 65 52 63 4F 6D 4D 00 B9 .
          > > ...#%÷eRcOmM.¹
          > >
          > > So, as (I think) Brian Lantz pointed out, there's some version
          info
          > > here ("23 24" and "23 25" look like the firmware version ("23 =
          > > 2.3") and release.
          > >
          > > Most of the rest of the info is identical between versions: the
          > > first 4 bytes ("00 01 00 00" (=64K, if that matters)), the "90
          F7"
          > > just after the version/release, and "eRcOmM".
          > >
          > > The last 2 bytes (or, maybe just the last byte) differ: "00 6D"
          in
          > > 2.3R24, and "00 B9" in 2.3R25. Maybe this is some sort of a
          > > checksum??? (of this 16-byte block of data?)
          > >
          > > - Paulb
        • paulbart1234
          Yes, I believe redboot is built on eCos (redboot being a self- contained loader, it uses eCos as its mini-OS for loading Linux) - Paulb ... redboot ...
          Message 4 of 6 , Aug 10, 2004
          • 0 Attachment
            Yes, I believe redboot is built on eCos (redboot being a self-
            contained loader, it uses eCos as its mini-OS for loading Linux)

            - Paulb

            --- In nslu2-linux@yahoogroups.com, "joule360" <joule360@y...> wrote:
            > You will note that 'strings' run on the .bin firmware file does
            > find "eCos", btw - I wondered about that because the nslu2 is
            > clearly running a linux kernel and not an ecos kernel. But,
            redboot
            > is part of the eCos project, so maybe that's why.
            >
            > nl
            >
            > --- In nslu2-linux@yahoogroups.com, "rwhitby" <list.yahoo@r...>
            > wrote:
            > > I just realised that "eRcOmM" is "ecm" and "ROM" interleaved.
            > >
            > > A redboot.ecm file is discussed here:
            > http://ecos.sourceware.org/docs-
            > > 2.0/ref/rebuilding-redboot.html
            > >
            > > -- Rod
            > >
            > > --- In nslu2-linux@yahoogroups.com, "paulbart1234"
            <paulbart@b...>
            > > wrote:
            > > > I've now got 2 update images (version 2.3 release 24, and
            > version
            > > > 2.3 release 25), so can compare.
            > > >
            > > > Here's the magic 16 bytes at the end of each:
            > > >
            > > > 2.3R24:
            > > > 00 01 00 00 23 24 90 F7 65 52 63 4F 6D 4D 00
            6D ....#$÷eRcOmM.m
            > > >
            > > > 2.3R25:
            > > > 00 01 00 00 23 25 90 F7 65 52 63 4F 6D 4D 00 B9 .
            > > > ...#%÷eRcOmM.¹
            > > >
            > > > So, as (I think) Brian Lantz pointed out, there's some version
            > info
            > > > here ("23 24" and "23 25" look like the firmware version ("23
            =
            > > > 2.3") and release.
            > > >
            > > > Most of the rest of the info is identical between versions:
            the
            > > > first 4 bytes ("00 01 00 00" (=64K, if that matters)), the "90
            > F7"
            > > > just after the version/release, and "eRcOmM".
            > > >
            > > > The last 2 bytes (or, maybe just the last byte) differ: "00
            6D"
            > in
            > > > 2.3R24, and "00 B9" in 2.3R25. Maybe this is some sort of a
            > > > checksum??? (of this 16-byte block of data?)
            > > >
            > > > - Paulb
          • Jim Buzbee
            ... Does anyone know if reboot can load a new image via tftp much like the wrt54g? This saved my butt when I loaded a bad firmware on the wrt54g... Jim ...
            Message 5 of 6 , Aug 10, 2004
            • 0 Attachment
              On Tue, 10 Aug 2004, paulbart1234 wrote:

              > Yes, I believe redboot is built on eCos (redboot being a self-
              > contained loader, it uses eCos as its mini-OS for loading Linux)


              Does anyone know if reboot can load a new image via tftp much like the
              wrt54g? This saved my butt when I loaded a bad firmware on the wrt54g...

              Jim


              >
              > - Paulb
              >
              > --- In nslu2-linux@yahoogroups.com, "joule360" <joule360@y...> wrote:
              > > You will note that 'strings' run on the .bin firmware file does
              > > find "eCos", btw - I wondered about that because the nslu2 is
              > > clearly running a linux kernel and not an ecos kernel. But,
              > redboot
              > > is part of the eCos project, so maybe that's why.
              > >
              > > nl
              > >
              > > --- In nslu2-linux@yahoogroups.com, "rwhitby" <list.yahoo@r...>
              > > wrote:
              > > > I just realised that "eRcOmM" is "ecm" and "ROM" interleaved.
              > > >
              > > > A redboot.ecm file is discussed here:
              > > http://ecos.sourceware.org/docs-
              > > > 2.0/ref/rebuilding-redboot.html
              > > >
              > > > -- Rod
              > > >
              > > > --- In nslu2-linux@yahoogroups.com, "paulbart1234"
              > <paulbart@b...>
              > > > wrote:
              > > > > I've now got 2 update images (version 2.3 release 24, and
              > > version
              > > > > 2.3 release 25), so can compare.
              > > > >
              > > > > Here's the magic 16 bytes at the end of each:
              > > > >
              > > > > 2.3R24:
              > > > > 00 01 00 00 23 24 90 F7 65 52 63 4F 6D 4D 00
              > 6D ....#$��eRcOmM.m
              > > > >
              > > > > 2.3R25:
              > > > > 00 01 00 00 23 25 90 F7 65 52 63 4F 6D 4D 00 B9 .
              > > > > ...#%��eRcOmM.�
              > > > >
              > > > > So, as (I think) Brian Lantz pointed out, there's some version
              > > info
              > > > > here ("23 24" and "23 25" look like the firmware version ("23
              > =
              > > > > 2.3") and release.
              > > > >
              > > > > Most of the rest of the info is identical between versions:
              > the
              > > > > first 4 bytes ("00 01 00 00" (=64K, if that matters)), the "90
              > > F7"
              > > > > just after the version/release, and "eRcOmM".
              > > > >
              > > > > The last 2 bytes (or, maybe just the last byte) differ: "00
              > 6D"
              > > in
              > > > > 2.3R24, and "00 B9" in 2.3R25. Maybe this is some sort of a
              > > > > checksum??? (of this 16-byte block of data?)
              > > > >
              > > > > - Paulb
              >
              >
              >
              >
              >
              > Yahoo! Groups Links
              >
              >
              >
              >
              >

              --------------------------------------------------------------------------------
              Jim Buzbee "I was gratified to be able to
              jbuzbee@... answer promptly, and I did. I
              http://batbox.org said I didn't know." Mark Twain
            • paulbart1234
              I know that redboot can load images via tftp, but I think you d need serial port terminal access in order to enable the built-in monitor (which would allow you
              Message 6 of 6 , Aug 10, 2004
              • 0 Attachment
                I know that redboot can load images via tftp, but I think you'd need
                serial port terminal access in order to enable the built-in monitor
                (which would allow you to tftp files)

                - Paulb
                --- In nslu2-linux@yahoogroups.com, Jim Buzbee <jbuzbee@n...> wrote:
                > On Tue, 10 Aug 2004, paulbart1234 wrote:
                >
                > > Yes, I believe redboot is built on eCos (redboot being a self-
                > > contained loader, it uses eCos as its mini-OS for loading Linux)
                >
                >
                > Does anyone know if reboot can load a new image via tftp much like
                the
                > wrt54g? This saved my butt when I loaded a bad firmware on the
                wrt54g...
                >
                > Jim
                >
                >
                > >
                > > - Paulb
                > >
                > > --- In nslu2-linux@yahoogroups.com, "joule360" <joule360@y...>
                wrote:
                > > > You will note that 'strings' run on the .bin firmware file
                does
                > > > find "eCos", btw - I wondered about that because the nslu2 is
                > > > clearly running a linux kernel and not an ecos kernel. But,
                > > redboot
                > > > is part of the eCos project, so maybe that's why.
                > > >
                > > > nl
                > > >
                > > > --- In nslu2-linux@yahoogroups.com, "rwhitby"
                <list.yahoo@r...>
                > > > wrote:
                > > > > I just realised that "eRcOmM" is "ecm" and "ROM" interleaved.
                > > > >
                > > > > A redboot.ecm file is discussed here:
                > > > http://ecos.sourceware.org/docs-
                > > > > 2.0/ref/rebuilding-redboot.html
                > > > >
                > > > > -- Rod
                > > > >
                > > > > --- In nslu2-linux@yahoogroups.com, "paulbart1234"
                > > <paulbart@b...>
                > > > > wrote:
                > > > > > I've now got 2 update images (version 2.3 release 24, and
                > > > version
                > > > > > 2.3 release 25), so can compare.
                > > > > >
                > > > > > Here's the magic 16 bytes at the end of each:
                > > > > >
                > > > > > 2.3R24:
                > > > > > 00 01 00 00 23 24 90 F7 65 52 63 4F 6D 4D 00
                > > 6D ....#$÷eRcOmM.m
                > > > > >
                > > > > > 2.3R25:
                > > > > > 00 01 00 00 23 25 90 F7 65 52 63 4F 6D 4D 00 B9 .
                > > > > > ...#%÷eRcOmM.¹
                > > > > >
                > > > > > So, as (I think) Brian Lantz pointed out, there's some
                version
                > > > info
                > > > > > here ("23 24" and "23 25" look like the firmware version
                ("23
                > > =
                > > > > > 2.3") and release.
                > > > > >
                > > > > > Most of the rest of the info is identical between
                versions:
                > > the
                > > > > > first 4 bytes ("00 01 00 00" (=64K, if that matters)),
                the "90
                > > > F7"
                > > > > > just after the version/release, and "eRcOmM".
                > > > > >
                > > > > > The last 2 bytes (or, maybe just the last byte)
                differ: "00
                > > 6D"
                > > > in
                > > > > > 2.3R24, and "00 B9" in 2.3R25. Maybe this is some sort of
                a
                > > > > > checksum??? (of this 16-byte block of data?)
                > > > > >
                > > > > > - Paulb
                > >
                > >
                > >
                > >
                > >
                > > Yahoo! Groups Links
                > >
                > >
                > >
                > >
                > >
                >
                > -------------------------------------------------------------------
                -------------
                > Jim Buzbee "I was gratified to be able
                to
                > jbuzbee@n... answer promptly, and I did. I
                > http://batbox.org said I didn't know." Mark
                Twain
              Your message has been successfully submitted and would be delivered to recipients shortly.