]> git.karo-electronics.de Git - karo-tx-linux.git/commit
cpuidle: always return with interrupts enabled
authorKevin Hilman <khilman@deeprootsystems.com>
Mon, 26 Oct 2009 23:50:18 +0000 (16:50 -0700)
committerGreg Kroah-Hartman <gregkh@suse.de>
Tue, 10 Nov 2009 00:22:27 +0000 (16:22 -0800)
commitbed202356361d3620a51974eadca78261e049ef0
treecceb72a651e0c952a81e42c1a66c8c019d55875e
parent1133a6e3632b5897f5a25084e38b221cb92f3eb6
cpuidle: always return with interrupts enabled

commit 246eb7f0ed1a8aeddec5313137767658f378949b upstream.

In the case where cpuidle_idle_call() returns before changing state due to
a need_resched(), it was returning with IRQs disabled.

The idle path assumes that the platform specific idle code returns with
interrupts enabled (although this too is undocumented AFAICT) and on ARM
we have a WARN_ON(!(irqs_disabled()) when returning from the idle loop, so
the user-visible effects were only a warning since interrupts were
eventually re-enabled later.

On x86, this same problem exists, but there is no WARN_ON() to detect it.
As on ARM, the interrupts are eventually re-enabled, so I'm not sure of
any actual bugs triggered by this.  It's primarily a
correctness/consistency fix.

This patch ensures IRQs are (re)enabled before returning.

Reported-by: Hemanth V <hemanthv@ti.com>
Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Tested-by: Martin Michlmayr <tbm@cyrius.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
drivers/cpuidle/cpuidle.c