While it should be possible to use "normal uncachable" mappings for

write-combining on arm64 as Linux does, this doesn't seem to work on
NXP's LX2160A SoC.  So switch to using "device" mappings for now to
make amdgpu(4) work better.

ok patrick@

diff --git a/sys/arch/arm64/include/pmap.h b/sys/arch/arm64/include/pmap.h
index f3747e6..28f76b5 100644 (file)
--- a/sys/arch/arm64/include/pmap.h
+++ b/sys/arch/arm64/include/pmap.h
@@ -1,4 +1,4 @@
-/* $OpenBSD: pmap.h,v 1.14 2020/10/21 21:53:47 deraadt Exp $ */
+/* $OpenBSD: pmap.h,v 1.15 2021/02/15 20:44:08 kettenis Exp $ */
 /*
  * Copyright (c) 2008,2009,2014 Dale Rahn <drahn@dalerahn.com>
  *
@@ -74,6 +74,7 @@ struct pmap {
 #define PMAP_PA_MASK   ~((paddr_t)PAGE_MASK) /* to remove the flags */
 #define PMAP_NOCACHE   0x1 /* non-cacheable memory */
 #define PMAP_DEVICE    0x2 /* device memory */
+#define PMAP_WC                PMAP_DEVICE
 
 #define PG_PMAP_MOD            PG_PMAP0
 #define PG_PMAP_REF            PG_PMAP1