[PATCH 5.4 36/36] compiler.h: fix barrier_data() on clang

From: Greg Kroah-Hartman
Date: Mon Dec 14 2020 - 14:10:40 EST


From: Arvind Sankar <nivedita@xxxxxxxxxxxx>

commit 3347acc6fcd4ee71ad18a9ff9d9dac176b517329 upstream.

Commit 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h
mutually exclusive") neglected to copy barrier_data() from
compiler-gcc.h into compiler-clang.h.

The definition in compiler-gcc.h was really to work around clang's more
aggressive optimization, so this broke barrier_data() on clang, and
consequently memzero_explicit() as well.

For example, this results in at least the memzero_explicit() call in
lib/crypto/sha256.c:sha256_transform() being optimized away by clang.

Fix this by moving the definition of barrier_data() into compiler.h.

Also move the gcc/clang definition of barrier() into compiler.h,
__memory_barrier() is icc-specific (and barrier() is already defined
using it in compiler-intel.h) and doesn't belong in compiler.h.

[rdunlap@xxxxxxxxxxxxx: fix ALPHA builds when SMP is not enabled]

Link: https://lkml.kernel.org/r/20201101231835.4589-1-rdunlap@xxxxxxxxxxxxx
Fixes: 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")
Signed-off-by: Arvind Sankar <nivedita@xxxxxxxxxxxx>
Signed-off-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Tested-by: Nick Desaulniers <ndesaulniers@xxxxxxxxxx>
Reviewed-by: Nick Desaulniers <ndesaulniers@xxxxxxxxxx>
Reviewed-by: Kees Cook <keescook@xxxxxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx>
Link: https://lkml.kernel.org/r/20201014212631.207844-1-nivedita@xxxxxxxxxxxx
Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
[nd: backport to account for missing
commit e506ea451254a ("compiler.h: Split {READ,WRITE}_ONCE definitions out into rwonce.h")
commit d08b9f0ca6605 ("scs: Add support for Clang's Shadow Call Stack (SCS)")]
Signed-off-by: Nick Desaulniers <ndesaulniers@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
include/linux/compiler-clang.h | 6 ------
include/linux/compiler-gcc.h | 19 -------------------
include/linux/compiler.h | 18 ++++++++++++++++--
3 files changed, 16 insertions(+), 27 deletions(-)

--- a/include/linux/compiler-clang.h
+++ b/include/linux/compiler-clang.h
@@ -36,9 +36,3 @@
__has_builtin(__builtin_sub_overflow)
#define COMPILER_HAS_GENERIC_BUILTIN_OVERFLOW 1
#endif
-
-/* The following are for compatibility with GCC, from compiler-gcc.h,
- * and may be redefined here because they should not be shared with other
- * compilers, like ICC.
- */
-#define barrier() __asm__ __volatile__("" : : : "memory")
--- a/include/linux/compiler-gcc.h
+++ b/include/linux/compiler-gcc.h
@@ -14,25 +14,6 @@
# error Sorry, your compiler is too old - please upgrade it.
#endif

-/* Optimization barrier */
-
-/* The "volatile" is due to gcc bugs */
-#define barrier() __asm__ __volatile__("": : :"memory")
-/*
- * This version is i.e. to prevent dead stores elimination on @ptr
- * where gcc and llvm may behave differently when otherwise using
- * normal barrier(): while gcc behavior gets along with a normal
- * barrier(), llvm needs an explicit input variable to be assumed
- * clobbered. The issue is as follows: while the inline asm might
- * access any memory it wants, the compiler could have fit all of
- * @ptr into memory registers instead, and since @ptr never escaped
- * from that, it proved that the inline asm wasn't touching any of
- * it. This version works well with both compilers, i.e. we're telling
- * the compiler that the inline asm absolutely may see the contents
- * of @ptr. See also: https://llvm.org/bugs/show_bug.cgi?id=15495
- */
-#define barrier_data(ptr) __asm__ __volatile__("": :"r"(ptr) :"memory")
-
/*
* This macro obfuscates arithmetic on a variable address so that gcc
* shouldn't recognize the original var, and make assumptions about it.
--- a/include/linux/compiler.h
+++ b/include/linux/compiler.h
@@ -80,11 +80,25 @@ void ftrace_likely_update(struct ftrace_

/* Optimization barrier */
#ifndef barrier
-# define barrier() __memory_barrier()
+/* The "volatile" is due to gcc bugs */
+# define barrier() __asm__ __volatile__("": : :"memory")
#endif

#ifndef barrier_data
-# define barrier_data(ptr) barrier()
+/*
+ * This version is i.e. to prevent dead stores elimination on @ptr
+ * where gcc and llvm may behave differently when otherwise using
+ * normal barrier(): while gcc behavior gets along with a normal
+ * barrier(), llvm needs an explicit input variable to be assumed
+ * clobbered. The issue is as follows: while the inline asm might
+ * access any memory it wants, the compiler could have fit all of
+ * @ptr into memory registers instead, and since @ptr never escaped
+ * from that, it proved that the inline asm wasn't touching any of
+ * it. This version works well with both compilers, i.e. we're telling
+ * the compiler that the inline asm absolutely may see the contents
+ * of @ptr. See also: https://llvm.org/bugs/show_bug.cgi?id=15495
+ */
+# define barrier_data(ptr) __asm__ __volatile__("": :"r"(ptr) :"memory")
#endif

/* workaround for GCC PR82365 if needed */