Merge tag 's390-6.14-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux

Pull s390 updates from Alexander Gordeev:

 - Select config option KASAN_VMALLOC if KASAN is enabled

 - Select config option VMAP_STACK unconditionally

 - Implement arch_atomic_inc() / arch_atomic_dec() functions which
   result in a single instruction if compiled for z196 or newer
   architectures

 - Make layering between atomic.h and atomic_ops.h consistent

 - Comment s390 preempt_count implementation

 - Remove pre MARCH_HAS_Z196_FEATURES preempt count implementation

 - GCC uses the number of lines of an inline assembly to calculate
   number of instructions and decide on inlining. Therefore remove
   superfluous new lines from a couple of inline assemblies.

 - Provide arch_atomic_*_and_test() implementations that allow the
   compiler to generate slightly better code.

 - Optimize __preempt_count_dec_and_test()

 - Remove __bootdata annotations from declarations in header files

 - Add missing include of <linux/smp.h> in abs_lowcore.h to provide
   declarations for get_cpu() and put_cpu() used in the code

 - Fix suboptimal kernel image base when running make kasan.config

 - Remove huge_pte_none() and huge_pte_none_mostly() as are identical to
   the generic variants

 - Remove unused PAGE_KERNEL_EXEC, SEGMENT_KERNEL_EXEC, and
   REGION3_KERNEL_EXEC defines

 - Simplify noexec page protection handling and change the page, segment
   and region3 protection definitions automatically if the instruction
   execution-protection facility is not available

 - Save one instruction and prefer EXRL instruction over EX in string,
   xor_*(), amode31 and other functions

 - Create /dev/diag misc device to fetch diagnose specific information
   from the kernel and provide it to userspace

 - Retrieve electrical power readings using DIAGNOSE 0x324 ioctl

 - Make ccw_device_get_ciw() consistent and use array indices instead of
   pointer arithmetic

 - s390/qdio: Move memory alloc/pointer arithmetic for slib and sl into
   one place

 - The sysfs core now allows instances of 'struct bin_attribute' to be
   moved into read-only memory. Make use of that in s390 code

 - Add missing TLB range adjustment in pud_free_tlb()

 - Improve topology setup by adding early polarization detection

 - Fix length checks in codepage_convert() function

 - The generic bitops implementation is nearly identical to the s390
   one. Switch to the generic variant and decrease a bit the kernel
   image size

 - Provide an optimized arch_test_bit() implementation which makes use
   of flag output constraint. This generates slightly better code

 - Provide memory topology information obtanied with DIAGNOSE 0x310
   using ioctl.

 - Various other small improvements, fixes, and cleanups

Also, some changes came in through a merge of 'pci-device-recovery'
branch:

 - Add PCI error recovery status mechanism

 - Simplify and document debug_next_entry() logic

 - Split private data allocation and freeing out of debug file open()
   and close() operations

 - Add debug_dump() function that gets a textual representation of a
   debug info (e.g. PCI recovery hardware error logs)

 - Add formatted content of pci_debug_msg_id to the PCI report

* tag 's390-6.14-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (48 commits)
  s390/futex: Fix FUTEX_OP_ANDN implementation
  s390/diag: Add memory topology information via diag310
  s390/bitops: Provide optimized arch_test_bit()
  s390/bitops: Switch to generic bitops
  s390/ebcdic: Fix length decrement in codepage_convert()
  s390/ebcdic: Fix length check in codepage_convert()
  s390/ebcdic: Use exrl instead of ex
  s390/amode31: Use exrl instead of ex
  s390/stackleak: Use exrl instead of ex in __stackleak_poison()
  s390/lib: Use exrl instead of ex in xor functions
  s390/topology: Improve topology detection
  s390/tlb: Add missing TLB range adjustment
  s390/pkey: Constify 'struct bin_attribute'
  s390/sclp: Constify 'struct bin_attribute'
  s390/pci: Constify 'struct bin_attribute'
  s390/ipl: Constify 'struct bin_attribute'
  s390/crypto/cpacf: Constify 'struct bin_attribute'
  s390/qdio: Move memory alloc/pointer arithmetic for slib and sl into one place
  s390/cio: Use array indices instead of pointer arithmetic
  s390/qdio: Rename feature flag aif_osa to aif_qdio
  ...

56 files changed, 1577 insertions, 681 deletions

diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
index 0077969170e8..e4931a32b896 100644
--- a/arch/s390/Kconfig
+++ b/arch/s390/Kconfig
@@ -233,6 +233,7 @@ config S390
 	select HAVE_VIRT_CPU_ACCOUNTING_IDLE
 	select IOMMU_HELPER		if PCI
 	select IOMMU_SUPPORT		if PCI
+	select KASAN_VMALLOC if KASAN
 	select LOCK_MM_AND_FIND_VMA
 	select MMU_GATHER_MERGE_VMAS
 	select MMU_GATHER_NO_GATHER
@@ -255,6 +256,7 @@ config S390
 	select USER_STACKTRACE_SUPPORT
 	select VDSO_GETRANDOM
 	select VIRT_CPU_ACCOUNTING
+	select VMAP_STACK
 	select ZONE_DMA
 	# Note: keep the above list sorted alphabetically
 
@@ -688,32 +690,6 @@ config MAX_PHYSMEM_BITS
 	  Increasing the number of bits also increases the kernel image size.
 	  By default 46 bits (64TB) are supported.
 
-config CHECK_STACK
-	def_bool y
-	depends on !VMAP_STACK
-	prompt "Detect kernel stack overflow"
-	help
-	  This option enables the compiler option -mstack-guard and
-	  -mstack-size if they are available. If the compiler supports them
-	  it will emit additional code to each function prolog to trigger
-	  an illegal operation if the kernel stack is about to overflow.
-
-	  Say N if you are unsure.
-
-config STACK_GUARD
-	int "Size of the guard area (128-1024)"
-	range 128 1024
-	depends on CHECK_STACK
-	default "256"
-	help
-	  This allows you to specify the size of the guard area at the lower
-	  end of the kernel stack. If the kernel stack points into the guard
-	  area on function entry an illegal operation is triggered. The size
-	  needs to be a power of 2. Please keep in mind that the size of an
-	  interrupt frame is 184 bytes for 31 bit and 328 bytes on 64 bit.
-	  The minimum size for the stack guard should be 256 for 31 bit and
-	  512 for 64 bit.
-
 endmenu
 
 menu "I/O subsystem"
diff --git a/arch/s390/Makefile b/arch/s390/Makefile
index 7fd57398221e..3f25498dac65 100644
--- a/arch/s390/Makefile
+++ b/arch/s390/Makefile
@@ -72,15 +72,6 @@ cflags-$(CONFIG_FRAME_POINTER) += -fno-optimize-sibling-calls
 KBUILD_AFLAGS_DECOMPRESSOR += $(aflags-y)
 KBUILD_CFLAGS_DECOMPRESSOR += $(cflags-y)
 
-ifneq ($(call cc-option,-mstack-size=8192 -mstack-guard=128),)
-  CC_FLAGS_CHECK_STACK := -mstack-size=$(STACK_SIZE)
-  ifeq ($(call cc-option,-mstack-size=8192),)
-    CC_FLAGS_CHECK_STACK += -mstack-guard=$(CONFIG_STACK_GUARD)
-  endif
-  export CC_FLAGS_CHECK_STACK
-  cflags-$(CONFIG_CHECK_STACK) += $(CC_FLAGS_CHECK_STACK)
-endif
-
 ifdef CONFIG_EXPOLINE
   ifdef CONFIG_EXPOLINE_EXTERN
     CC_FLAGS_EXPOLINE := -mindirect-branch=thunk-extern
diff --git a/arch/s390/boot/boot.h b/arch/s390/boot/boot.h
index 7521a9d75fa2..56244fe78182 100644
--- a/arch/s390/boot/boot.h
+++ b/arch/s390/boot/boot.h
@@ -13,7 +13,6 @@
 struct machine_info {
 	unsigned char has_edat1 : 1;
 	unsigned char has_edat2 : 1;
-	unsigned char has_nx : 1;
 };
 
 struct vmlinux_info {
diff --git a/arch/s390/boot/startup.c b/arch/s390/boot/startup.c
index 6087d38c7235..e6b06692ddc8 100644
--- a/arch/s390/boot/startup.c
+++ b/arch/s390/boot/startup.c
@@ -30,6 +30,9 @@ unsigned long __bootdata_preserved(vmemmap_size);
 unsigned long __bootdata_preserved(MODULES_VADDR);
 unsigned long __bootdata_preserved(MODULES_END);
 unsigned long __bootdata_preserved(max_mappable);
+unsigned long __bootdata_preserved(page_noexec_mask);
+unsigned long __bootdata_preserved(segment_noexec_mask);
+unsigned long __bootdata_preserved(region_noexec_mask);
 int __bootdata_preserved(relocate_lowcore);
 
 u64 __bootdata_preserved(stfle_fac_list[16]);
@@ -51,8 +54,14 @@ static void detect_facilities(void)
 	}
 	if (test_facility(78))
 		machine.has_edat2 = 1;
-	if (test_facility(130))
-		machine.has_nx = 1;
+	page_noexec_mask = -1UL;
+	segment_noexec_mask = -1UL;
+	region_noexec_mask = -1UL;
+	if (!test_facility(130)) {
+		page_noexec_mask &= ~_PAGE_NOEXEC;
+		segment_noexec_mask &= ~_SEGMENT_ENTRY_NOEXEC;
+		region_noexec_mask &= ~_REGION_ENTRY_NOEXEC;
+	}
 }
 
 static int cmma_test_essa(void)
diff --git a/arch/s390/boot/vmem.c b/arch/s390/boot/vmem.c
index 3fa28db2fe59..881a1ece422f 100644
--- a/arch/s390/boot/vmem.c
+++ b/arch/s390/boot/vmem.c
@@ -63,13 +63,10 @@ static void kasan_populate_shadow(unsigned long kernel_start, unsigned long kern
 	pud_t pud_z = __pud(__pa(kasan_early_shadow_pmd) | _REGION3_ENTRY);
 	p4d_t p4d_z = __p4d(__pa(kasan_early_shadow_pud) | _REGION2_ENTRY);
 	unsigned long memgap_start = 0;
-	unsigned long untracked_end;
 	unsigned long start, end;
 	int i;
 
 	pte_z = __pte(__pa(kasan_early_shadow_page) | pgprot_val(PAGE_KERNEL_RO));
-	if (!machine.has_nx)
-		pte_z = clear_pte_bit(pte_z, __pgprot(_PAGE_NOEXEC));
 	crst_table_init((unsigned long *)kasan_early_shadow_p4d, p4d_val(p4d_z));
 	crst_table_init((unsigned long *)kasan_early_shadow_pud, pud_val(pud_z));
 	crst_table_init((unsigned long *)kasan_early_shadow_pmd, pmd_val(pmd_z));
@@ -93,15 +90,10 @@ static void kasan_populate_shadow(unsigned long kernel_start, unsigned long kern
 	kasan_populate(kernel_start + TEXT_OFFSET, kernel_end, POPULATE_KASAN_MAP_SHADOW);
 	kasan_populate(0, (unsigned long)__identity_va(0), POPULATE_KASAN_ZERO_SHADOW);
 	kasan_populate(AMODE31_START, AMODE31_END, POPULATE_KASAN_ZERO_SHADOW);
-	if (IS_ENABLED(CONFIG_KASAN_VMALLOC)) {
-		untracked_end = VMALLOC_START;
-		/* shallowly populate kasan shadow for vmalloc and modules */
-		kasan_populate(VMALLOC_START, MODULES_END, POPULATE_KASAN_SHALLOW);
-	} else {
-		untracked_end = MODULES_VADDR;
-	}
+	/* shallowly populate kasan shadow for vmalloc and modules */
+	kasan_populate(VMALLOC_START, MODULES_END, POPULATE_KASAN_SHALLOW);
 	/* populate kasan shadow for untracked memory */
-	kasan_populate((unsigned long)__identity_va(ident_map_size), untracked_end,
+	kasan_populate((unsigned long)__identity_va(ident_map_size), VMALLOC_START,
 		       POPULATE_KASAN_ZERO_SHADOW);
 	kasan_populate(kernel_end, _REGION1_SIZE, POPULATE_KASAN_ZERO_SHADOW);
 }
@@ -300,8 +292,6 @@ static void pgtable_pte_populate(pmd_t *pmd, unsigned long addr, unsigned long e
 				continue;
 			entry = __pte(_pa(addr, PAGE_SIZE, mode));
 			entry = set_pte_bit(entry, PAGE_KERNEL);
-			if (!machine.has_nx)
-				entry = clear_pte_bit(entry, __pgprot(_PAGE_NOEXEC));
 			set_pte(pte, entry);
 			pages++;
 		}
@@ -326,8 +316,6 @@ static void pgtable_pmd_populate(pud_t *pud, unsigned long addr, unsigned long e
 			if (can_large_pmd(pmd, addr, next, mode)) {
 				entry = __pmd(_pa(addr, _SEGMENT_SIZE, mode));
 				entry = set_pmd_bit(entry, SEGMENT_KERNEL);
-				if (!machine.has_nx)
-					entry = clear_pmd_bit(entry, __pgprot(_SEGMENT_ENTRY_NOEXEC));
 				set_pmd(pmd, entry);
 				pages++;
 				continue;
@@ -359,8 +347,6 @@ static void pgtable_pud_populate(p4d_t *p4d, unsigned long addr, unsigned long e
 			if (can_large_pud(pud, addr, next, mode)) {
 				entry = __pud(_pa(addr, _REGION3_SIZE, mode));
 				entry = set_pud_bit(entry, REGION3_KERNEL);
-				if (!machine.has_nx)
-					entry = clear_pud_bit(entry, __pgprot(_REGION_ENTRY_NOEXEC));
 				set_pud(pud, entry);
 				pages++;
 				continue;
diff --git a/arch/s390/configs/kasan.config b/arch/s390/configs/kasan.config
index 84c2b551e992..cefbe2ba1228 100644
--- a/arch/s390/configs/kasan.config
+++ b/arch/s390/configs/kasan.config
@@ -1,4 +1,4 @@
 # Help: Enable KASan for debugging
 CONFIG_KASAN=y
 CONFIG_KASAN_INLINE=y
-CONFIG_KASAN_VMALLOC=y
+CONFIG_KERNEL_IMAGE_BASE=0x7FFFE0000000
diff --git a/arch/s390/include/asm/abs_lowcore.h b/arch/s390/include/asm/abs_lowcore.h
index d20df8c923fc..004d17ea05cf 100644
--- a/arch/s390/include/asm/abs_lowcore.h
+++ b/arch/s390/include/asm/abs_lowcore.h
@@ -2,7 +2,7 @@
 #ifndef _ASM_S390_ABS_LOWCORE_H
 #define _ASM_S390_ABS_LOWCORE_H
 
-#include <asm/sections.h>
+#include <linux/smp.h>
 #include <asm/lowcore.h>
 
 #define ABS_LOWCORE_MAP_SIZE	(NR_CPUS * sizeof(struct lowcore))
@@ -25,7 +25,7 @@ static inline void put_abs_lowcore(struct lowcore *lc)
 	put_cpu();
 }
 
-extern int __bootdata_preserved(relocate_lowcore);
+extern int relocate_lowcore;
 
 static inline int have_relocated_lowcore(void)
 {
diff --git a/arch/s390/include/asm/atomic.h b/arch/s390/include/asm/atomic.h
index 6723fca64018..b36dd6a1d652 100644
--- a/arch/s390/include/asm/atomic.h
+++ b/arch/s390/include/asm/atomic.h
@@ -17,13 +17,13 @@
 
 static __always_inline int arch_atomic_read(const atomic_t *v)
 {
-	return __atomic_read(v);
+	return __atomic_read(&v->counter);
 }
 #define arch_atomic_read arch_atomic_read
 
 static __always_inline void arch_atomic_set(atomic_t *v, int i)
 {
-	__atomic_set(v, i);
+	__atomic_set(&v->counter, i);
 }
 #define arch_atomic_set arch_atomic_set
 
@@ -45,6 +45,36 @@ static __always_inline void arch_atomic_add(int i, atomic_t *v)
 }
 #define arch_atomic_add arch_atomic_add
 
+static __always_inline void arch_atomic_inc(atomic_t *v)
+{
+	__atomic_add_const(1, &v->counter);
+}
+#define arch_atomic_inc arch_atomic_inc
+
+static __always_inline void arch_atomic_dec(atomic_t *v)
+{
+	__atomic_add_const(-1, &v->counter);
+}
+#define arch_atomic_dec arch_atomic_dec
+
+static __always_inline bool arch_atomic_sub_and_test(int i, atomic_t *v)
+{
+	return __atomic_add_and_test_barrier(-i, &v->counter);
+}
+#define arch_atomic_sub_and_test arch_atomic_sub_and_test
+
+static __always_inline bool arch_atomic_dec_and_test(atomic_t *v)
+{
+	return __atomic_add_const_and_test_barrier(-1, &v->counter);
+}
+#define arch_atomic_dec_and_test arch_atomic_dec_and_test
+
+static __always_inline bool arch_atomic_inc_and_test(atomic_t *v)
+{
+	return __atomic_add_const_and_test_barrier(1, &v->counter);
+}
+#define arch_atomic_inc_and_test arch_atomic_inc_and_test
+
 #define arch_atomic_sub(_i, _v)		arch_atomic_add(-(int)(_i), _v)
 #define arch_atomic_sub_return(_i, _v)	arch_atomic_add_return(-(int)(_i), _v)
 #define arch_atomic_fetch_sub(_i, _v)	arch_atomic_fetch_add(-(int)(_i), _v)
@@ -94,13 +124,13 @@ static __always_inline bool arch_atomic_try_cmpxchg(atomic_t *v, int *old, int n
 
 static __always_inline s64 arch_atomic64_read(const atomic64_t *v)
 {
-	return __atomic64_read(v);
+	return __atomic64_read((long *)&v->counter);
 }
 #define arch_atomic64_read arch_atomic64_read
 
 static __always_inline void arch_atomic64_set(atomic64_t *v, s64 i)
 {
-	__atomic64_set(v, i);
+	__atomic64_set((long *)&v->counter, i);
 }
 #define arch_atomic64_set arch_atomic64_set
 
@@ -122,6 +152,36 @@ static __always_inline void arch_atomic64_add(s64 i, atomic64_t *v)
 }
 #define arch_atomic64_add arch_atomic64_add
 
+static __always_inline void arch_atomic64_inc(atomic64_t *v)
+{
+	__atomic64_add_const(1, (long *)&v->counter);
+}
+#define arch_atomic64_inc arch_atomic64_inc
+
+static __always_inline void arch_atomic64_dec(atomic64_t *v)
+{
+	__atomic64_add_const(-1, (long *)&v->counter);
+}
+#define arch_atomic64_dec arch_atomic64_dec
+
+static __always_inline bool arch_atomic64_sub_and_test(s64 i, atomic64_t *v)
+{
+	return __atomic64_add_and_test_barrier(-i, (long *)&v->counter);
+}
+#define arch_atomic64_sub_and_test arch_atomic64_sub_and_test
+
+static __always_inline bool arch_atomic64_dec_and_test(atomic64_t *v)
+{
+	return __atomic64_add_const_and_test_barrier(-1, (long *)&v->counter);
+}
+#define arch_atomic64_dec_and_test arch_atomic64_dec_and_test
+
+static __always_inline bool arch_atomic64_inc_and_test(atomic64_t *v)
+{
+	return __atomic64_add_const_and_test_barrier(1, (long *)&v->counter);
+}
+#define arch_atomic64_inc_and_test arch_atomic64_inc_and_test
+
 static __always_inline s64 arch_atomic64_xchg(atomic64_t *v, s64 new)
 {
 	return arch_xchg(&v->counter, new);
diff --git a/arch/s390/include/asm/atomic_ops.h b/arch/s390/include/asm/atomic_ops.h
index 1d6b2056fad8..585678bbcd7a 100644
--- a/arch/s390/include/asm/atomic_ops.h
+++ b/arch/s390/include/asm/atomic_ops.h
@@ -10,50 +10,51 @@
 
 #include <linux/limits.h>
 #include <asm/march.h>
+#include <asm/asm.h>
 
-static __always_inline int __atomic_read(const atomic_t *v)
+static __always_inline int __atomic_read(const int *ptr)
 {
-	int c;
+	int val;
 
 	asm volatile(
-		"	l	%[c],%[counter]\n"
-		: [c] "=d" (c) : [counter] "R" (v->counter));
-	return c;
+		"	l	%[val],%[ptr]\n"
+		: [val] "=d" (val) : [ptr] "R" (*ptr));
+	return val;
 }
 
-static __always_inline void __atomic_set(atomic_t *v, int i)
+static __always_inline void __atomic_set(int *ptr, int val)
 {
-	if (__builtin_constant_p(i) && i >= S16_MIN && i <= S16_MAX) {
+	if (__builtin_constant_p(val) && val >= S16_MIN && val <= S16_MAX) {
 		asm volatile(
-			"	mvhi	%[counter], %[i]\n"
-			: [counter] "=Q" (v->counter) : [i] "K" (i));
+			"	mvhi	%[ptr],%[val]\n"
+			: [ptr] "=Q" (*ptr) : [val] "K" (val));
 	} else {
 		asm volatile(
-			"	st	%[i],%[counter]\n"
-			: [counter] "=R" (v->counter) : [i] "d" (i));
+			"	st	%[val],%[ptr]\n"
+			: [ptr] "=R" (*ptr) : [val] "d" (val));
 	}
 }
 
-static __always_inline s64 __atomic64_read(const atomic64_t *v)
+static __always_inline long __atomic64_read(const long *ptr)
 {
-	s64 c;
+	long val;
 
 	asm volatile(
-		"	lg	%[c],%[counter]\n"
-		: [c] "=d" (c) : [counter] "RT" (v->counter));
-	return c;
+		"	lg	%[val],%[ptr]\n"
+		: [val] "=d" (val) : [ptr] "RT" (*ptr));
+	return val;
 }
 
-static __always_inline void __atomic64_set(atomic64_t *v, s64 i)
+static __always_inline void __atomic64_set(long *ptr, long val)
 {
-	if (__builtin_constant_p(i) && i >= S16_MIN && i <= S16_MAX) {
+	if (__builtin_constant_p(val) && val >= S16_MIN && val <= S16_MAX) {
 		asm volatile(
-			"	mvghi	%[counter], %[i]\n"
-			: [counter] "=Q" (v->counter) : [i] "K" (i));
+			"	mvghi	%[ptr],%[val]\n"
+			: [ptr] "=Q" (*ptr) : [val] "K" (val));
 	} else {
 		asm volatile(
-			"	stg	%[i],%[counter]\n"
-			: [counter] "=RT" (v->counter) : [i] "d" (i));
+			"	stg	%[val],%[ptr]\n"
+			: [ptr] "=RT" (*ptr) : [val] "d" (val));
 	}
 }
 
@@ -73,7 +74,7 @@ static __always_inline op_type op_name(op_type val, op_type *ptr)	\
 }									\
 
 #define __ATOMIC_OPS(op_name, op_type, op_string)			\
-	__ATOMIC_OP(op_name, op_type, op_string, "\n")			\
+	__ATOMIC_OP(op_name, op_type, op_string, "")			\
 	__ATOMIC_OP(op_name##_barrier, op_type, op_string, "bcr 14,0\n")
 
 __ATOMIC_OPS(__atomic_add, int, "laa")
@@ -99,7 +100,7 @@ static __always_inline void op_name(op_type val, op_type *ptr)		\
 }
 
 #define __ATOMIC_CONST_OPS(op_name, op_type, op_string)			\
-	__ATOMIC_CONST_OP(op_name, op_type, op_string, "\n")		\
+	__ATOMIC_CONST_OP(op_name, op_type, op_string, "")		\
 	__ATOMIC_CONST_OP(op_name##_barrier, op_type, op_string, "bcr 14,0\n")
 
 __ATOMIC_CONST_OPS(__atomic_add_const, int, "asi")
@@ -169,4 +170,76 @@ __ATOMIC64_OPS(__atomic64_xor, "xgr")
 
 #endif /* MARCH_HAS_Z196_FEATURES */
 
+#if defined(MARCH_HAS_Z196_FEATURES) && defined(__HAVE_ASM_FLAG_OUTPUTS__)
+
+#define __ATOMIC_TEST_OP(op_name, op_type, op_string, op_barrier)	\
+static __always_inline bool op_name(op_type val, op_type *ptr)		\
+{									\
+	op_type tmp;							\
+	int cc;								\
+									\
+	asm volatile(							\
+		op_string "	%[tmp],%[val],%[ptr]\n"			\
+		op_barrier						\
+		: "=@cc" (cc), [tmp] "=d" (tmp), [ptr] "+QS" (*ptr)	\
+		: [val] "d" (val)					\
+		: "memory");						\
+	return (cc == 0) || (cc == 2);					\
+}									\
+
+#define __ATOMIC_TEST_OPS(op_name, op_type, op_string)			\
+	__ATOMIC_TEST_OP(op_name, op_type, op_string, "")		\
+	__ATOMIC_TEST_OP(op_name##_barrier, op_type, op_string, "bcr 14,0\n")
+
+__ATOMIC_TEST_OPS(__atomic_add_and_test, int, "laal")
+__ATOMIC_TEST_OPS(__atomic64_add_and_test, long, "laalg")
+
+#undef __ATOMIC_TEST_OPS
+#undef __ATOMIC_TEST_OP
+
+#define __ATOMIC_CONST_TEST_OP(op_name, op_type, op_string, op_barrier)	\
+static __always_inline bool op_name(op_type val, op_type *ptr)		\
+{									\
+	int cc;								\
+									\
+	asm volatile(							\
+		op_string "	%[ptr],%[val]\n"			\
+		op_barrier						\
+		: "=@cc" (cc), [ptr] "+QS" (*ptr)			\
+		: [val] "i" (val)					\
+		: "memory");						\
+	return (cc == 0) || (cc == 2);					\
+}
+
+#define __ATOMIC_CONST_TEST_OPS(op_name, op_type, op_string)		\
+	__ATOMIC_CONST_TEST_OP(op_name, op_type, op_string, "")		\
+	__ATOMIC_CONST_TEST_OP(op_name##_barrier, op_type, op_string, "bcr 14,0\n")
+
+__ATOMIC_CONST_TEST_OPS(__atomic_add_const_and_test, int, "alsi")
+__ATOMIC_CONST_TEST_OPS(__atomic64_add_const_and_test, long, "algsi")
+
+#undef __ATOMIC_CONST_TEST_OPS
+#undef __ATOMIC_CONST_TEST_OP
+
+#else /* defined(MARCH_HAS_Z196_FEATURES) && defined(__HAVE_ASM_FLAG_OUTPUTS__) */
+
+#define __ATOMIC_TEST_OP(op_name, op_func, op_type)			\
+static __always_inline bool op_name(op_type val, op_type *ptr)		\
+{									\
+	return op_func(val, ptr) == -val;				\
+}
+
+__ATOMIC_TEST_OP(__atomic_add_and_test,			__atomic_add,		int)
+__ATOMIC_TEST_OP(__atomic_add_and_test_barrier,		__atomic_add_barrier,	int)
+__ATOMIC_TEST_OP(__atomic_add_const_and_test,		__atomic_add,		int)
+__ATOMIC_TEST_OP(__atomic_add_const_and_test_barrier,	__atomic_add_barrier,	int)
+__ATOMIC_TEST_OP(__atomic64_add_and_test,		__atomic64_add,		long)
+__ATOMIC_TEST_OP(__atomic64_add_and_test_barrier,	__atomic64_add_barrier, long)
+__ATOMIC_TEST_OP(__atomic64_add_const_and_test,		__atomic64_add,		long)
+__ATOMIC_TEST_OP(__atomic64_add_const_and_test_barrier,	__atomic64_add_barrier,	long)
+
+#undef __ATOMIC_TEST_OP
+
+#endif /* defined(MARCH_HAS_Z196_FEATURES) && defined(__HAVE_ASM_FLAG_OUTPUTS__) */
+
 #endif /* __ARCH_S390_ATOMIC_OPS__  */
diff --git a/arch/s390/include/asm/bitops.h b/arch/s390/include/asm/bitops.h
index 54a079cd39ed..15aa64e3020e 100644
--- a/arch/s390/include/asm/bitops.h
+++ b/arch/s390/include/asm/bitops.h</