Revert "ACPI: ignore FADT reset-reg-sup flag"

This reverts commit cf450136bf.

It breaks reboot on at least one Thinkpad T43, as reported by Jörg Otte:
 "On reboot it shuts down as normal.
  The last lines displayed are:

  >Unmounting temporary filesystems.. [OK]
  >Deactivating swap...               [OK]
  >Unmounting local filesystems...    [OK]
  >Will now restart
  >    Restarting system

  Then I hear it accessing the cd-drive, but then it's being stuck."

Jörg bisected the regression to this commit.

That commit fixes another machine (see

  https://bugzilla.kernel.org/show_bug.cgi?id=11533

for details) that has a BIOS bug and doesn't support ACPI reset.
However, at least one of those other reporters no longer even has the
machine in question, and had a different workaround to begin with.
Besides, it clearly was a buggy BIOS.  Let's not break the correct case
to fix that case.

Reported-and-bisected-by: Jörg Otte <jrg.otte@googlemail.com>
Cc: linux-acpi@vger.kernel.org
Cc: Len Brown <lenb@kernel.org>
Cc: Peter Anvin <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

drivers/acpi/acpica/hwxface.c

@@ -74,7 +74,8 @@ acpi_status acpi_reset(void)
 
 	/* Check if the reset register is supported */
 
-	if (!reset_reg->address) {
+	if (!(acpi_gbl_FADT.flags & ACPI_FADT_RESET_REGISTER) ||
+	    !reset_reg->address) {
 		return_ACPI_STATUS(AE_NOT_EXIST);
 	}
 

drivers/acpi/reboot.c

@@ -23,7 +23,8 @@ void acpi_reboot(void)
 	/* Is the reset register supported? The spec says we should be
 	 * checking the bit width and bit offset, but Windows ignores
 	 * these fields */
-	/* Ignore also acpi_gbl_FADT.flags.ACPI_FADT_RESET_REGISTER */
+	if (!(acpi_gbl_FADT.flags & ACPI_FADT_RESET_REGISTER))
+		return;
 
 	reset_value = acpi_gbl_FADT.reset_value;