powerpc: prepare string/mem functions for KASAN

CONFIG_KASAN implements wrappers for memcpy() memmove() and memset()
Those wrappers are doing the verification then call respectively
__memcpy() __memmove() and __memset(). The arches are therefore
expected to rename their optimised functions that way.

For files on which KASAN is inhibited, #defines are used to allow
them to directly call optimised versions of the functions without
going through the KASAN wrappers.

See commit 393f203f5fd5 ("x86_64: kasan: add interceptors for
memset/memmove/memcpy functions") for details.

Other string / mem functions do not (yet) have kasan wrappers,
we therefore have to fallback to the generic versions when
KASAN is active, otherwise KASAN checks will be skipped.

Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
[mpe: Fixups to keep selftests working]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

4 files changed, 27 insertions, 9 deletions

diff --git a/arch/powerpc/lib/Makefile b/arch/powerpc/lib/Makefile
index 79396e184bca..47a4de434c22 100644
--- a/arch/powerpc/lib/Makefile
+++ b/arch/powerpc/lib/Makefile
@@ -8,9 +8,14 @@ ccflags-$(CONFIG_PPC64)	:= $(NO_MINIMAL_TOC)
 CFLAGS_REMOVE_code-patching.o = $(CC_FLAGS_FTRACE)
 CFLAGS_REMOVE_feature-fixups.o = $(CC_FLAGS_FTRACE)
 
-obj-y += string.o alloc.o code-patching.o feature-fixups.o
+obj-y += alloc.o code-patching.o feature-fixups.o
 
-obj-$(CONFIG_PPC32)	+= div64.o copy_32.o crtsavres.o strlen_32.o
+ifndef CONFIG_KASAN
+obj-y	+=	string.o memcmp_$(BITS).o
+obj-$(CONFIG_PPC32)	+= strlen_32.o
+endif
+
+obj-$(CONFIG_PPC32)	+= div64.o copy_32.o crtsavres.o
 
 obj-$(CONFIG_FUNCTION_ERROR_INJECTION)	+= error-inject.o
 
@@ -34,7 +39,7 @@ obj64-$(CONFIG_KPROBES_SANITY_TEST)	+= test_emulate_step.o \
 					   test_emulate_step_exec_instr.o
 
 obj-y			+= checksum_$(BITS).o checksum_wrappers.o \
-			   string_$(BITS).o memcmp_$(BITS).o
+			   string_$(BITS).o
 
 obj-y			+= sstep.o ldstfp.o quad.o
 obj64-y			+= quad.o
diff --git a/arch/powerpc/lib/copy_32.S b/arch/powerpc/lib/copy_32.S
index ba66846fe973..d5642481fb98 100644
--- a/arch/powerpc/lib/copy_32.S
+++ b/arch/powerpc/lib/copy_32.S
@@ -14,6 +14,7 @@
 #include <asm/ppc_asm.h>
 #include <asm/export.h>
 #include <asm/code-patching-asm.h>
+#include <asm/kasan.h>
 
 #define COPY_16_BYTES		\
 	lwz	r7,4(r4);	\
@@ -68,6 +69,7 @@ CACHELINE_BYTES = L1_CACHE_BYTES
 LG_CACHELINE_BYTES = L1_CACHE_SHIFT
 CACHELINE_MASK = (L1_CACHE_BYTES-1)
 
+#ifndef CONFIG_KASAN
 _GLOBAL(memset16)
 	rlwinm.	r0 ,r5, 31, 1, 31
 	addi	r6, r3, -4
@@ -81,6 +83,7 @@ _GLOBAL(memset16)
 	sth	r4, 4(r6)
 	blr
 EXPORT_SYMBOL(memset16)
+#endif
 
 /*
  * Use dcbz on the complete cache lines in the destination
@@ -91,7 +94,7 @@ EXPORT_SYMBOL(memset16)
  * We therefore skip the optimised bloc that uses dcbz. This jump is
  * replaced by a nop once cache is active. This is done in machine_init()
  */
-_GLOBAL(memset)
+_GLOBAL_KASAN(memset)
 	cmplwi	0,r5,4
 	blt	7f
 
@@ -151,6 +154,7 @@ _GLOBAL(memset)
 	bdnz	9b
 	blr
 EXPORT_SYMBOL(memset)
+EXPORT_SYMBOL_KASAN(memset)
 
 /*
  * This version uses dcbz on the complete cache lines in the
@@ -163,12 +167,12 @@ EXPORT_SYMBOL(memset)
  * We therefore jump to generic_memcpy which doesn't use dcbz. This jump is
  * replaced by a nop once cache is active. This is done in machine_init()
  */
-_GLOBAL(memmove)
+_GLOBAL_KASAN(memmove)
 	cmplw	0,r3,r4
 	bgt	backwards_memcpy
 	/* fall through */
 
-_GLOBAL(memcpy)
+_GLOBAL_KASAN(memcpy)
 1:	b	generic_memcpy
 	patch_site	1b, patch__memcpy_nocache
 
@@ -244,6 +248,8 @@ _GLOBAL(memcpy)
 65:	blr
 EXPORT_SYMBOL(memcpy)
 EXPORT_SYMBOL(memmove)
+EXPORT_SYMBOL_KASAN(memcpy)
+EXPORT_SYMBOL_KASAN(memmove)
 
 generic_memcpy:
 	srwi.	r7,r5,3
diff --git a/arch/powerpc/lib/mem_64.S b/arch/powerpc/lib/mem_64.S
index 3c3be02f33b7..7f6bd031c306 100644
--- a/arch/powerpc/lib/mem_64.S
+++ b/arch/powerpc/lib/mem_64.S
@@ -12,7 +12,9 @@
 #include <asm/errno.h>
 #include <asm/ppc_asm.h>
 #include <asm/export.h>
+#include <asm/kasan.h>
 
+#ifndef CONFIG_KASAN
 _GLOBAL(__memset16)
 	rlwimi	r4,r4,16,0,15
 	/* fall through */
@@ -29,8 +31,9 @@ _GLOBAL(__memset64)
 EXPORT_SYMBOL(__memset16)
 EXPORT_SYMBOL(__memset32)
 EXPORT_SYMBOL(__memset64)
+#endif
 
-_GLOBAL(memset)
+_GLOBAL_KASAN(memset)
 	neg	r0,r3
 	rlwimi	r4,r4,8,16,23
 	andi.	r0,r0,7			/* # bytes to be 8-byte aligned */
@@ -96,8 +99,9 @@ _GLOBAL(memset)
 	stb	r4,0(r6)
 	blr
 EXPORT_SYMBOL(memset)
+EXPORT_SYMBOL_KASAN(memset)
 
-_GLOBAL_TOC(memmove)
+_GLOBAL_TOC_KASAN(memmove)
 	cmplw	0,r3,r4
 	bgt	backwards_memcpy
 	b	memcpy
@@ -139,3 +143,4 @@ _GLOBAL(backwards_memcpy)
 	mtctr	r7
 	b	1b
 EXPORT_SYMBOL(memmove)
+EXPORT_SYMBOL_KASAN(memmove)
diff --git a/arch/powerpc/lib/memcpy_64.S b/arch/powerpc/lib/memcpy_64.S
index 273ea67e60a1..25c3772c1dfb 100644
--- a/arch/powerpc/lib/memcpy_64.S
+++ b/arch/powerpc/lib/memcpy_64.S
@@ -11,6 +11,7 @@
 #include <asm/export.h>
 #include <asm/asm-compat.h>
 #include <asm/feature-fixups.h>
+#include <asm/kasan.h>
 
 #ifndef SELFTEST_CASE
 /* For big-endian, 0 == most CPUs, 1 == POWER6, 2 == Cell */
@@ -18,7 +19,7 @@
 #endif
 
 	.align	7
-_GLOBAL_TOC(memcpy)
+_GLOBAL_TOC_KASAN(memcpy)
 BEGIN_FTR_SECTION
 #ifdef __LITTLE_ENDIAN__
 	cmpdi	cr7,r5,0
@@ -230,3 +231,4 @@ END_FTR_SECTION_IFCLR(CPU_FTR_UNALIGNED_LD_STD)
 	blr
 #endif
 EXPORT_SYMBOL(memcpy)
+EXPORT_SYMBOL_KASAN(memcpy)