Re: [2.7 "thoughts"] V0.3

From: William Lee Irwin III (wli_at_holomorphy.com)
Date: 10/10/03

  • Next message: Pavel Machek: "Re: -test7: /sys/power/disk not reading right data?" 
    Date:	Fri, 10 Oct 2003 02:11:34 -0700
To: "Frederick, Fabian" <Fabian.Frederick@prov-liege.be>

    On Fri, Oct 10, 2003 at 09:54:12AM +0200, Frederick, Fabian wrote:
    > 2.7 "thoughts"
    > Thanks to Gabor, Stuart, Stephan and others
    > Don't hesitate to send me more or comment.

    Ugh, this is all crackpot wishlist gunk.

    How about some goodies backed with real working code, like:

    * O(1) proc_pid_statm()
            -- originally by bcrl for 2.4, fwd port maintained by wli
    * O(lg(n)) proc_pid_readdir()/proc_task_readdir()
            -- original O(1) proc_pid_readdir() by manfred, rewritten by
            -- wli to use rbtrees for O(lg(n)) seeks into the relevant
            -- lists (walking over empty buckets had overhead)
    * 4KB ia32 kernel stacks + irqstacks
            -- original by bcrl, fwd port maintained by dhansen for a
            -- substantial amount of time, now maintained by wli
    * ia32 leaf pagetable node cache
            -- wli
    * node-local per_cpu areas for ia32 NUMA
            -- wli
    * highpmd, analogue of highpte for pmd's
            -- wli. Gets pmd's on node-local mem on ia32 NUMA, and
            -- alleviates a lot of lowmem pressure under heavy
            -- multiprogramming levels on PAE.

    Some benchmarks of a patchset including these (and several other things)
    are at http://home.earthlink.net/~rwhron/kernel/wli.html, and some ports
    of the patch set are at ftp://ftp.kernel.org/people/wli/kernels/

    Whatever fantasy may be worth, working code is worth a lot more. I'll
    refrain from mentioning prototype-quality patches I'm hacking on atm.

    -- wli
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  • Next message: Pavel Machek: "Re: -test7: /sys/power/disk not reading right data?"

    Relevant Pages

    • Re: reiser4 plugins
      ... > Cycle detection when hard links to directories are allowed. ... > resolved by working code or a formal proof that it is not ... directory inside a file act as a mountpoint. ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: [PATCH][2.6] Allow x86_64 to reenable interrupts on contention
      ... > really worth it? ... text data bss dec hex filename ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • [PATCH 1/2] VmLib wrapped: executable brk
      ... as writable will appear with wrapped VmLib: sorry, but while it's worth ... showing ordinary programs as ordinary, it's not worth much effort to ... avoid showing odd ones as odd. ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: [ANNOUNCE] ktimers subsystem
      ... >> schedule things, need timestamps etc etc. ... it is absolutely and emphatically not worth it. ... to] where a new, clean subsystem introduces significant overhead, but ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: Non-GPL export of invalidate_mmap_range
      ... it's simply the price that has to be paid in order to be able to ... it doesn't seem worth it just to save a stack frame. ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)