[PATCH 2 of 3] xen/pat: After suspend re-write PAT if BIOS changed it

[Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>]

# HG changeset patch
# User Simon Graham <simon.graham@virtualcomputer.com>
# Date 1333398413 14400
# Node ID caefa03c38366c3e70d937bc95740c5d33c0892a
# Parent  f1da2ce71ed41d1b74ebe6916ff7710d6579438e
xen/pat: After suspend re-write PAT if BIOS changed it.

Certain AMD machines (this was a MSI or GigaBYTE BIOS) after resume
would reset the PAT MSR causing rather weird issues - where
the pages would (say they would be set to WC) would end up with the
wrong type (as they would use the BIOS PAT instead of the one set by
the hypervisor). And in some cases the PAT was stuck and needed
a couple of WRMSRL to actually take.

Signed-off-by: Simon Graham <simon.graham@virtualcomputer.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>

diff -r f1da2ce71ed4 -r caefa03c3836 xen/arch/x86/acpi/power.c
--- a/xen/arch/x86/acpi/power.c	Mon Apr 02 16:26:48 2012 -0400
+++ b/xen/arch/x86/acpi/power.c	Mon Apr 02 16:26:53 2012 -0400
@@ -41,8 +41,27 @@ static DEFINE_SPINLOCK(pm_lock);
 
 struct acpi_sleep_info acpi_sinfo;
 
+static void pat_resume(void);
 void do_suspend_lowlevel(void);
 
+static void
+pat_resume()
+{
+    u64 pat;
+    if (!cpu_has_pat)
+        return;
+
+    rdmsrl(MSR_IA32_CR_PAT, pat);
+    if (pat != host_pat) {
+	printk(KERN_INFO PREFIX "Found PAT MSR: 0x%"PRIx64"\n", pat);
+	printk(KERN_INFO PREFIX "reseting to 0x%"PRIx64"\n", host_pat);
+	wrmsrl(MSR_IA32_CR_PAT, host_pat);
+	rdmsrl(MSR_IA32_CR_PAT, pat);
+	if (pat != host_pat)
+	    printk(KERN_WARNING PREFIX "PAT MSR stuck on: 0x%"PRIx64"\n", pat);
+    }
+}
+
 static int device_power_down(void)
 {
     console_suspend();
@@ -194,6 +213,7 @@ static int enter_state(u32 state)
     if ( cpu_has_efer )
         write_efer(read_efer());
 
+    pat_resume();
     device_power_up();
 
     mcheck_init(&boot_cpu_data, 0);