s390/preempt: move preempt_count to the lowcore

Convert s390 to use a field in the struct lowcore for the CPU
preemption count. It is a bit cheaper to access a lowcore field
compared to a thread_info variable and it removes the depencency
on a task related structure.

bloat-o-meter on the vmlinux image for the default configuration
(CONFIG_PREEMPT_NONE=y) reports a small reduction in text size:

add/remove: 0/0 grow/shrink: 18/578 up/down: 228/-5448 (-5220)

A larger improvement is achieved with the default configuration
but with CONFIG_PREEMPT=y and CONFIG_DEBUG_PREEMPT=n:

add/remove: 2/6 grow/shrink: 59/4477 up/down: 1618/-228762 (-227144)

Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

3 files changed, 139 insertions, 3 deletions

diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index 7b93b78f423c..1df1e6215e85 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -126,7 +126,8 @@ struct lowcore {
 	__u64	percpu_offset;			/* 0x0378 */
 	__u64	vdso_per_cpu_data;		/* 0x0380 */
 	__u64	machine_flags;			/* 0x0388 */
-	__u8	pad_0x0390[0x0398-0x0390];	/* 0x0390 */
+	__u32	preempt_count;			/* 0x0390 */
+	__u8	pad_0x0394[0x0398-0x0394];	/* 0x0394 */
 	__u64	gmap;				/* 0x0398 */
 	__u32	spinlock_lockval;		/* 0x03a0 */
 	__u32	fpu_flags;			/* 0x03a4 */
diff --git a/arch/s390/include/asm/preempt.h b/arch/s390/include/asm/preempt.h
new file mode 100644
index 000000000000..b0776b2c8dcf
--- /dev/null
+++ b/arch/s390/include/asm/preempt.h
@@ -0,0 +1,137 @@
+#ifndef __ASM_PREEMPT_H
+#define __ASM_PREEMPT_H
+
+#include <asm/current.h>
+#include <linux/thread_info.h>
+#include <asm/atomic_ops.h>
+
+#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
+
+#define PREEMPT_ENABLED	(0 + PREEMPT_NEED_RESCHED)
+
+static inline int preempt_count(void)
+{
+	return READ_ONCE(S390_lowcore.preempt_count) & ~PREEMPT_NEED_RESCHED;
+}
+
+static inline void preempt_count_set(int pc)
+{
+	int old, new;
+
+	do {
+		old = READ_ONCE(S390_lowcore.preempt_count);
+		new = (old & PREEMPT_NEED_RESCHED) |
+			(pc & ~PREEMPT_NEED_RESCHED);
+	} while (__atomic_cmpxchg(&S390_lowcore.preempt_count,
+				  old, new) != old);
+}
+
+#define init_task_preempt_count(p)	do { } while (0)
+
+#define init_idle_preempt_count(p, cpu)	do { \
+	S390_lowcore.preempt_count = PREEMPT_ENABLED; \
+} while (0)
+
+static inline void set_preempt_need_resched(void)
+{
+	__atomic_and(~PREEMPT_NEED_RESCHED, &S390_lowcore.preempt_count);
+}
+
+static inline void clear_preempt_need_resched(void)
+{
+	__atomic_or(PREEMPT_NEED_RESCHED, &S390_lowcore.preempt_count);
+}
+
+static inline bool test_preempt_need_resched(void)
+{
+	return !(READ_ONCE(S390_lowcore.preempt_count) & PREEMPT_NEED_RESCHED);
+}
+
+static inline void __preempt_count_add(int val)
+{
+	if (__builtin_constant_p(val) && (val >= -128) && (val <= 127))
+		__atomic_add_const(val, &S390_lowcore.preempt_count);
+	else
+		__atomic_add(val, &S390_lowcore.preempt_count);
+}
+
+static inline void __preempt_count_sub(int val)
+{
+	__preempt_count_add(-val);
+}
+
+static inline bool __preempt_count_dec_and_test(void)
+{
+	return __atomic_add(-1, &S390_lowcore.preempt_count) == 1;
+}
+
+static inline bool should_resched(int preempt_offset)
+{
+	return unlikely(READ_ONCE(S390_lowcore.preempt_count) ==
+			preempt_offset);
+}
+
+#else /* CONFIG_HAVE_MARCH_Z196_FEATURES */
+
+#define PREEMPT_ENABLED	(0)
+
+static inline int preempt_count(void)
+{
+	return READ_ONCE(S390_lowcore.preempt_count);
+}
+
+static inline void preempt_count_set(int pc)
+{
+	S390_lowcore.preempt_count = pc;
+}
+
+#define init_task_preempt_count(p)	do { } while (0)
+
+#define init_idle_preempt_count(p, cpu)	do { \
+	S390_lowcore.preempt_count = PREEMPT_ENABLED; \
+} while (0)
+
+static inline void set_preempt_need_resched(void)
+{
+}
+
+static inline void clear_preempt_need_resched(void)
+{
+}
+
+static inline bool test_preempt_need_resched(void)
+{
+	return false;
+}
+
+static inline void __preempt_count_add(int val)
+{
+	S390_lowcore.preempt_count += val;
+}
+
+static inline void __preempt_count_sub(int val)
+{
+	S390_lowcore.preempt_count -= val;
+}
+
+static inline bool __preempt_count_dec_and_test(void)
+{
+	return !--S390_lowcore.preempt_count && tif_need_resched();
+}
+
+static inline bool should_resched(int preempt_offset)
+{
+	return unlikely(preempt_count() == preempt_offset &&
+			tif_need_resched());
+}
+
+#endif /* CONFIG_HAVE_MARCH_Z196_FEATURES */
+
+#ifdef CONFIG_PREEMPT
+extern asmlinkage void preempt_schedule(void);
+#define __preempt_schedule() preempt_schedule()
+extern asmlinkage void preempt_schedule_notrace(void);
+#define __preempt_schedule_notrace() preempt_schedule_notrace()
+#endif /* CONFIG_PREEMPT */
+
+#endif /* __ASM_PREEMPT_H */
diff --git a/arch/s390/include/asm/thread_info.h b/arch/s390/include/asm/thread_info.h
index f15c0398c363..933794150e5b 100644
--- a/arch/s390/include/asm/thread_info.h
+++ b/arch/s390/include/asm/thread_info.h
@@ -34,7 +34,6 @@ struct thread_info {
 	unsigned long		flags;		/* low level flags */
 	unsigned long		sys_call_table;	/* System call table address */
 	unsigned int		cpu;		/* current CPU */
-	int			preempt_count;	/* 0 => preemptable, <0 => BUG */
 	unsigned int		system_call;
 	__u64			user_timer;
 	__u64			system_timer;
@@ -49,7 +48,6 @@ struct thread_info {
 	.task		= &tsk,			\
 	.flags		= 0,			\
 	.cpu		= 0,			\
-	.preempt_count	= INIT_PREEMPT_COUNT,	\
 }
 
 #define init_thread_info	(init_thread_union.thread_info)