CPU Hotplug: Hotplug Script And SIGPWR

From: Rusty Russell (rusty_at_au1.ibm.com)
Date: 01/20/04

  • Next message: Nathan Scott: "Re: page allocation failure" 
    To: vatsa@in.ibm.com
Date:	Tue, 20 Jan 2004 16:44:45 +1100

    In message <20040116174446.A2820@in.ibm.com> you write:
    > Would it make sense if we defer invoking hotplug script _after_
    > the CPU is completely dead (i.e after issuing the CPU_DEAD
    > notification)?

    The original code wanted to block until the hotplug script
    acknowledged the removal before completing it. Greg KH says hotplug
    doesn't work this way, so now it could well be delivered after
    everything is over. If it's simpler, we can just do it after.

    The other issue I wanted to revisit: we currently send SIGPWR to all
    processes which we have to undo the CPU affinity for (with a new
    si_info field containing the cpu going down).

    The main problem is that a process can call sched_setaffinity on
    another (unrelated) task, which might not know about it. One option
    would be to only deliver the signal if it's not SIG_DFL for that
    process. Another would be not to signal, and expect hotplug scripts
    to clean up.

    Thoughts?
    Rusty. 

    --
  Anyone who quotes me in their sig is an idiot. -- Rusty Russell.
-
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: Nathan Scott: "Re: page allocation failure"

    Relevant Pages

    • Re: [patch] genapic: optimize & fix APIC mode setup
      ... the CPU hotplug code, ... Subject: genapic: optimize & fix APIC mode setup ... NETDEV WATCHDOG: eth0: transmit timed out ...
      (Linux-Kernel)
    • Re: [PATCH RFC] Priority boosting for preemptible RCU
      ... int this_cpu, cpu; ... and CPU_DOWN_PREPARE, whichever suits better), in the cpu hotplug ... changes to the hotplug synchronization mechanism. ... of the low frequency of statistics printing -- and of hotplug operations, ...
      (Linux-Kernel)
    • Re: [Patch 4/4] cpusets top mask just online, not all possible
      ... the numa code and other hotplug work might have similar considerations. ... Cpusets and hotplug (CPU or Memory) aren't friends, ... Cpusets builds up additional data structures, used to manage a tasks CPU ...
      (Linux-Kernel)
    • Re: CPU Hotplug: Hotplug Script And SIGPWR
      ... That assumption is not safe. ... >>a hotplug script to fault in and start it up again (which may not ... >CPU A gets yanked out, ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • [GIT PULL][RESEND] KVM cpu hotplug fixes
      ... is that we need a notification about a cpu going away at a point in time ... HOTPLUG: Add CPU_DYING notifier ... +static void hardware_disable ... int cpu = v; ...
      (Linux-Kernel)