x86/efi: Clarify that reset attack mitigation needs appropriate userspace

Some distributions have turned on the reset attack mitigation feature,
which is designed to force the platform to clear the contents of RAM if
the machine is shut down uncleanly. However, in order for the platform
to be able to determine whether the shutdown was clean or not, userspace
has to be configured to clear the MemoryOverwriteRequest flag on
shutdown - otherwise the firmware will end up clearing RAM on every
reboot, which is unnecessarily time consuming. Add some additional
clarity to the kconfig text to reduce the risk of systems being
configured this way.

Signed-off-by: Matthew Garrett <mjg59@google.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-efi@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: stable@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>

diff --git a/drivers/firmware/efi/Kconfig b/drivers/firmware/efi/Kconfig
index aab108e..6047ed4 100644 (file)
--- a/drivers/firmware/efi/Kconfig
+++ b/drivers/firmware/efi/Kconfig
@@ -159,7 +159,10 @@ config RESET_ATTACK_MITIGATION
          using the TCG Platform Reset Attack Mitigation specification. This
          protects against an attacker forcibly rebooting the system while it
          still contains secrets in RAM, booting another OS and extracting the
-         secrets.
+         secrets. This should only be enabled when userland is configured to
+         clear the MemoryOverwriteRequest flag on clean shutdown after secrets
+         have been evicted, since otherwise it will trigger even on clean
+         reboots.
 
 endmenu