LKML Archive on
help / color / mirror / Atom feed
From: Ingo Molnar <>
To: "Eric W. Biederman" <>
Cc: "H. Peter Anvin" <>, Vivek Goyal <>,
	Neil Horman <>,,,,
Subject: Re: [PATCH], issue EOI to APIC prior to calling crash_kexec in die_nmi path
Date: Thu, 7 Feb 2008 01:39:18 +0100	[thread overview]
Message-ID: <> (raw)
In-Reply-To: <>

* Eric W. Biederman <> wrote:

> Looking at the patch the local_irq_enable() is totally bogus.  As soon 
> was we hit machine_crash_shutdown the first thing we do is disable 
> irqs.


> I'm wondering if someone was using the switch cpus on crash patch that 
> was floating around.  That would require the ipis to work.
> I don't know if nmi_exit makes sense.  There are enough layers of 
> abstraction in that piece of code I can't quickly spot the part that 
> is banging the hardware.
> The location of nmi_exit in the patch is clearly wrong.  crash_kexec 
> is a noop if we don't have a crash kernel loaded (and if we are not 
> the first cpu into it), so if we don't execute the crash code 
> something weird may happen.  Further the code is just more 
> maintainable if that kind of code lives in machine_crash_shutdown.

nmi_exit() has no hw effects - it's just our own bookeeping.

the hw knows that we finished the NMI when we do an iret. Perhaps that's 
the bug or side-effect that made the difference: via enabling irqs we 
get an irq entry, and that does an iret and clears the NMI nested state 
- allowing the kexec context to proceed? I suspect kexec() will do an 
iret eventually (at minimum in the booted up kernel's context) - all 
NMIs are blocked up to that point and maybe the APIC doesnt really like 
being frobbed in that state? In any case, the local_irq_enable() is just 
wrong - it's the worst thing a crashing kernel can do. Perhaps doing an 
intentional iret with a prepared stack-let that just restores to 
still-irqs-off state and jumps to the next instruction could 'exit' the 
NMI context without really having to exit it in the kernel code flow?


  reply	other threads:[~2008-02-07  0:51 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-02-06 19:25 Neil Horman
2008-02-06 19:40 ` Vivek Goyal
2008-02-06 20:12   ` Neil Horman
2008-02-06 20:21     ` H. Peter Anvin
2008-02-06 21:04       ` Neil Horman
2008-02-06 20:35     ` Vivek Goyal
2008-02-06 22:00 ` Ingo Molnar
2008-02-06 22:48   ` Vivek Goyal
2008-02-06 22:53     ` Ingo Molnar
2008-02-06 22:56     ` H. Peter Anvin
2008-02-06 23:36       ` Ingo Molnar
2008-02-06 23:50         ` Vivek Goyal
2008-02-07  0:31         ` Eric W. Biederman
2008-02-07  0:39           ` Ingo Molnar [this message]
2008-02-07  1:30             ` Eric W. Biederman
2008-02-07 12:17           ` Neil Horman
2008-02-07 12:24             ` Ingo Molnar
2008-02-07 20:37               ` Neil Horman
2008-02-08 16:14               ` Neil Horman
2008-02-08 16:45                 ` Vivek Goyal
2008-02-08 17:26                   ` Neil Horman
2008-02-12 21:08                   ` Neil Horman
2008-02-15 14:02                     ` Eric W. Biederman
2008-02-20 14:57                     ` Neil Horman
2008-02-08 16:54               ` Andi Kleen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \ \ \ \ \ \ \ \ \ \ \
    --subject='Re: [PATCH], issue EOI to APIC prior to calling crash_kexec in die_nmi path' \

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).