Re: [PATCH v6 11/16] x86/mtrr: construct a memory map with cache modes

From: Juergen Gross
Date: Tue May 09 2023 - 07:47:53 EST


On 09.05.23 13:36, Borislav Petkov wrote:
On Tue, May 02, 2023 at 02:09:26PM +0200, Juergen Gross wrote:
@@ -841,6 +1137,10 @@ static void generic_set_mtrr(unsigned int reg, unsigned long base,
cache_enable();
local_irq_restore(flags);
+
+ /* On the first CPU rebuild the cache mode memory map. */
+ if (smp_processor_id() == cpumask_first(cpu_online_mask))
+ rebuild_map();

I still don't like this. Anything against doing this instead?

No, Should be fine. Apart from ...


---
diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c
index 163dca53796c..e69c8c56897f 100644
--- a/arch/x86/kernel/cpu/mtrr/generic.c
+++ b/arch/x86/kernel/cpu/mtrr/generic.c
@@ -458,8 +458,11 @@ static void map_add_var(void)
* removals of registers, which are complicated to handle without rebuild of
* the map.
*/
-static void rebuild_map(void)
+void rebuild_map(void)
{
+ if (mtrr_if != &generic_mtrr_ops)
+ return;
+
cache_map_n = cache_map_fixed;
map_add_var();
@@ -1127,10 +1130,6 @@ static void generic_set_mtrr(unsigned int reg, unsigned long base,
cache_enable();
local_irq_restore(flags);
-
- /* On the first CPU rebuild the cache mode memory map. */
- if (smp_processor_id() == cpumask_first(cpu_online_mask))
- rebuild_map();
}
int generic_validate_add_page(unsigned long base, unsigned long size,
diff --git a/arch/x86/kernel/cpu/mtrr/legacy.c b/arch/x86/kernel/cpu/mtrr/legacy.c
index d25882fcf181..30a3d51d06e0 100644
--- a/arch/x86/kernel/cpu/mtrr/legacy.c
+++ b/arch/x86/kernel/cpu/mtrr/legacy.c
@@ -67,6 +67,7 @@ static void mtrr_restore(void)
mtrr_value[i].ltype);
}
}
+ rebuild_map();

... this call not being needed at all. We can reach this point only for
mtrr_if != &generic_mtrr_ops.


Juergen

Attachment: OpenPGP_0xB0DE9DD628BF132F.asc
Description: OpenPGP public key

Attachment: OpenPGP_signature
Description: OpenPGP digital signature