x86: c_p_a() fix: reorder TLB / cache flushes to follow Intel recommendation

Intel recommends to first flush the TLBs and then the caches
on caching attribute changes. c_p_a() previously did it the
other way round. Reorder that.

The procedure is still not fully compliant to the Intel documentation
because Intel recommends a all CPU synchronization step between
the TLB flushes and the cache flushes.

However on all new Intel CPUs this is now meaningless anyways
because they support Self-Snoop and can skip the cache flush
step anyway.

[ mingo@elte.hu: decoupled from clflush and ported it to x86.git ]

Signed-off-by: Andi Kleen <ak@suse.de>
Acked-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

2 files changed, 8 insertions, 7 deletions

diff --git a/arch/x86/mm/pageattr_32.c b/arch/x86/mm/pageattr_32.c
index 523fd5b37df9..5cb5c7101f41 100644
--- a/arch/x86/mm/pageattr_32.c
+++ b/arch/x86/mm/pageattr_32.c
@@ -87,6 +87,12 @@ static void flush_kernel_map(void *arg)
         struct list_head *lh = (struct list_head *)arg;
         struct page *p;
 
+        /*
+         * Flush all to work around Errata in early athlons regarding
+         * large page flushing.
+         */
+        __flush_tlb_all();
+
         /* High level code is not ready for clflush yet */
         if (0 && cpu_has_clflush) {
                 list_for_each_entry(p, lh, lru)
@@ -95,12 +101,6 @@ static void flush_kernel_map(void *arg)
                 if (boot_cpu_data.x86_model >= 4)
                         wbinvd();
         }
-
-        /*
-         * Flush all to work around Errata in early athlons regarding
-         * large page flushing.
-         */
-        __flush_tlb_all();
 }
 
 static void set_pmd_pte(pte_t *kpte, unsigned long address, pte_t pte)
diff --git a/arch/x86/mm/pageattr_64.c b/arch/x86/mm/pageattr_64.c
index 4d172881af70..3ccdb1401e67 100644
--- a/arch/x86/mm/pageattr_64.c
+++ b/arch/x86/mm/pageattr_64.c
@@ -82,6 +82,8 @@ static void flush_kernel_map(void *arg)
         struct list_head *l = (struct list_head *)arg;
         struct page *pg;
 
+        __flush_tlb_all();
+
         /* When clflush is available always use it because it is
            much cheaper than WBINVD. */
         /* clflush is still broken. Disable for now. */
@@ -94,7 +96,6 @@ static void flush_kernel_map(void *arg)
                         clflush_cache_range(addr, PAGE_SIZE);
                 }
         }
-        __flush_tlb_all();
 }
 
 static inline void flush_map(struct list_head *l)