Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.

Close

#2 Patch for 3.4.1 on 64 bit processors

closed-duplicate
Sachin Mhatre
None
5
2004-04-06
2004-03-17
Dan Bar Dov
No

Modifications were tested both on Opteron (AMD) and
Itanium-2 (Intel).
Most modifications had to do with the management of
the /dev/iscsi directory tree.

The patch is for 3.4.1

Dan
danb@voltaire.com

Discussion

    • assigned_to: nobody --> smhatre
     
  • Logged In: YES
    user_id=752546

    Ok, I've looked at this patch and I've tested it quite a bit.
    Unfortunately, I don't think this is going to work. Sure,
    it works
    on *some* 32/64 bit platforms (SLES8, ia-32, ia-64, amd-64).
    But others, it doesn't work on (i.e. RH8, RH9, RHAS2.1).
    The main reason is that not all kernels export all the needed
    sys_* symbols. You can see if a given kernel has the support
    needed by just grepping /proc/ksyms for all required sys_*
    symbols. It seems to me this is what I ran into before with the
    patch I attempted, and that's why I never pushed for it to be
    submitted into cvs.

    I'm going to attach the patch I currently have against the 3-4
    codeline (it's closer to what's in CVS now, and the original
    patch submitted below contained a garbage file at the end
    of it), but I'm not going to submit it.

    Short of rewriting iscsi-probe.c similar to what was done for
    the 3-5 and 4-0 branches, I would say we're going to have
    to punt on 64-bit support for the 3-4 branch. Someone please
    feel free to disagree with me if you know of a way. However,
    I just don't have the expertise to know how to fix up the
    current
    code in such a way that it's compatible with all reasonable
    kernels out there, won't break if someone removes a
    symbol in a future 2.4 kernel, and works with both 32 and 64
    bit platforms.

    Dan, if you're up to it, maybe you can add to the current
    patch to make it more flexible. Thanks very much for your help.

     
    • status: open --> closed-duplicate
     
  • Logged In: YES
    user_id=210654

    Closing as duplication for RFE #864510 (64 bit support).

    See
    http://sourceforge.net/tracker/index.php?
    func=detail&aid=864510&group_id=26396&atid=387026