x86/microcode: Do not exit early from __reload_late()
authorBorislav Petkov <bp@suse.de>
Sat, 21 Apr 2018 08:19:30 +0000 (10:19 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Tue, 1 May 2018 19:58:26 +0000 (12:58 -0700)
commit922e5129eb011ebbc62c416bc1c423a20ae359d7
treea9c499d113c1c6e660fd493c927096b18839d823
parent7c6bcaac737fa72dd8aef00cb38b9c96b9b04cd8
x86/microcode: Do not exit early from __reload_late()

commit 09e182d17e8891dd73baba961a0f5a82e9274c97 upstream.

Vitezslav reported a case where the

  "Timeout during microcode update!"

panic would hit. After a deeper look, it turned out that his .config had
CONFIG_HOTPLUG_CPU disabled which practically made save_mc_for_early() a
no-op.

When that happened, the discovered microcode patch wasn't saved into the
cache and the late loading path wouldn't find any.

This, then, lead to early exit from __reload_late() and thus CPUs waiting
until the timeout is reached, leading to the panic.

In hindsight, that function should have been written so it does not return
before the post-synchronization. Oh well, I know better now...

Fixes: bb8c13d61a62 ("x86/microcode: Fix CPU synchronization routine")
Reported-by: Vitezslav Samel <vitezslav@samel.cz>
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Vitezslav Samel <vitezslav@samel.cz>
Tested-by: Ashok Raj <ashok.raj@intel.com>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/20180418081140.GA2439@pc11.op.pod.cz
Link: https://lkml.kernel.org/r/20180421081930.15741-2-bp@alien8.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
arch/x86/kernel/cpu/microcode/core.c