Re: hot cache line due to note_interrupt()

From: Andrew Morton (akpm_at_osdl.org)
Date: 11/11/03

  • Next message: Maciej Zenczykowski: "Re: A7N8X (Deluxe) Madness" 
    Date:	Tue, 11 Nov 2003 11:58:04 -0800
To: Ravikiran G Thirumalai <kiran@in.ibm.com>

    Ravikiran G Thirumalai <kiran@in.ibm.com> wrote:
    >
    > On Mon, Nov 10, 2003 at 03:58:44PM -0600, Jack Steiner wrote:
    > >
    > > I dont know the background on note_interrupt() in arch/ia64/kernel/irq.c,
    > > but I had to disable the function on our large systems (IA64).
    > >
    > > The function updates a counter in the irq_desc_t table. An entry in this table
    > > is shared by all cpus that take a specific interrupt #. For most interrupt #'s,
    > > this is a problem but it is prohibitive for the timer tick on big systems.
    > >
    > > Updating the counter causes a cache line to be bounced between
    > > cpus at a rate of at least HZ*active_cpus. (The number of bus transactions
    >
    > The answer to this is probably alloc_percpu for the counters.

    Or just make noirqdebug the default.

    The note_interrupt() stuff is only useful for diagnosing mysterious lockups
    (and hasn't proven useful for that, actually). It should be disabled for
    production use.

    -
    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: Maciej Zenczykowski: "Re: A7N8X (Deluxe) Madness"

    Relevant Pages

    • Re: very strange issue with sata,<4G Ram, and ext3
      ... > kernel has now used (eg the PXE layer or pxe booter forgot to close the ... we tried updating to a newer pxelinux and had the same ... Lead Systems Integrator ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: very strange issue with sata,<4G Ram, and ext3
      ... >> its still DMAing happily into the kernel) ... > into updating the boot rom on the nics. ... Lead Systems Integrator ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: 2.6.2-rc2 nfsd+xfs spins in i_size_read()
      ... this call should not actually be updating the size. ... >>Would be a good starting point, I suspect those calls from the nfs ... >>revalidate call are not really going to change the inode size. ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: [PATCH] fix get_jiffies_64 to work on voyager
      ... >>only unstable time is when we may be updating both halves of it ... >>all x86 arch's. ... 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/ ...
      (Linux-Kernel)
    • Re: [patch] scheduler: rebalance_tick interval update
      ... > 5 jiffies apart: ... same timer tick. ... load_balance code on the same CPU and domain several ticks in a row ... 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/ ...
      (Linux-Kernel)