[dpdk-dev] [RFC PATCH v1 1/5] eal: add new API for wait until scheme

[Feifei Wang <feifei.wang2@arm.com>]

For 'wait until' scheme, add new APIs for more cases:
1. add wait_until_unequal API
2. add wait_until_part_equal API

Signed-off-by: Feifei Wang <feifei.wang2@arm.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
---
 lib/eal/arm/include/rte_pause_64.h  | 271 +++++++++++++++++++-----
 lib/eal/include/generic/rte_pause.h | 309 ++++++++++++++++++++++++++++
 2 files changed, 526 insertions(+), 54 deletions(-)

diff --git a/lib/eal/arm/include/rte_pause_64.h b/lib/eal/arm/include/rte_pause_64.h
index e87d10b8cc..19716276fc 100644
--- a/lib/eal/arm/include/rte_pause_64.h
+++ b/lib/eal/arm/include/rte_pause_64.h
@@ -31,20 +31,12 @@ static inline void rte_pause(void)
 /* Put processor into low power WFE(Wait For Event) state. */
 #define __WFE() { asm volatile("wfe" : : : "memory"); }
 
-static __rte_always_inline void
-rte_wait_until_equal_16(volatile uint16_t *addr, uint16_t expected,
-		int memorder)
-{
-	uint16_t value;
-
-	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
-
-	/*
-	 * Atomic exclusive load from addr, it returns the 16-bit content of
-	 * *addr while making it 'monitored',when it is written by someone
-	 * else, the 'monitored' state is cleared and a event is generated
-	 * implicitly to exit WFE.
-	 */
+/*
+ * Atomic exclusive load from addr, it returns the 16-bit content of
+ * *addr while making it 'monitored', when it is written by someone
+ * else, the 'monitored' state is cleared and a event is generated
+ * implicitly to exit WFE.
+ */
 #define __LOAD_EXC_16(src, dst, memorder) {               \
 	if (memorder == __ATOMIC_RELAXED) {               \
 		asm volatile("ldxrh %w[tmp], [%x[addr]]"  \
@@ -58,6 +50,52 @@ rte_wait_until_equal_16(volatile uint16_t *addr, uint16_t expected,
 			: "memory");                      \
 	} }
 
+/*
+ * Atomic exclusive load from addr, it returns the 32-bit content of
+ * *addr while making it 'monitored', when it is written by someone
+ * else, the 'monitored' state is cleared and a event is generated
+ * implicitly to exit WFE.
+ */
+#define __LOAD_EXC_32(src, dst, memorder) {              \
+	if (memorder == __ATOMIC_RELAXED) {              \
+		asm volatile("ldxr %w[tmp], [%x[addr]]"  \
+			: [tmp] "=&r" (dst)              \
+			: [addr] "r"(src)                \
+			: "memory");                     \
+	} else {                                         \
+		asm volatile("ldaxr %w[tmp], [%x[addr]]" \
+			: [tmp] "=&r" (dst)              \
+			: [addr] "r"(src)                \
+			: "memory");                     \
+	} }
+
+/*
+ * Atomic exclusive load from addr, it returns the 64-bit content of
+ * *addr while making it 'monitored', when it is written by someone
+ * else, the 'monitored' state is cleared and a event is generated
+ * implicitly to exit WFE.
+ */
+#define __LOAD_EXC_64(src, dst, memorder) {              \
+	if (memorder == __ATOMIC_RELAXED) {              \
+		asm volatile("ldxr %x[tmp], [%x[addr]]"  \
+			: [tmp] "=&r" (dst)              \
+			: [addr] "r"(src)                \
+			: "memory");                     \
+	} else {                                         \
+		asm volatile("ldaxr %x[tmp], [%x[addr]]" \
+			: [tmp] "=&r" (dst)              \
+			: [addr] "r"(src)                \
+			: "memory");                     \
+	} }
+
+static __rte_always_inline void
+rte_wait_until_equal_16(volatile uint16_t *addr, uint16_t expected,
+		int memorder)
+{
+	uint16_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
 	__LOAD_EXC_16(addr, value, memorder)
 	if (value != expected) {
 		__SEVL()
@@ -66,7 +104,6 @@ rte_wait_until_equal_16(volatile uint16_t *addr, uint16_t expected,
 			__LOAD_EXC_16(addr, value, memorder)
 		} while (value != expected);
 	}
-#undef __LOAD_EXC_16
 }
 
 static __rte_always_inline void
@@ -77,25 +114,6 @@ rte_wait_until_equal_32(volatile uint32_t *addr, uint32_t expected,
 
 	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
 
-	/*
-	 * Atomic exclusive load from addr, it returns the 32-bit content of
-	 * *addr while making it 'monitored',when it is written by someone
-	 * else, the 'monitored' state is cleared and a event is generated
-	 * implicitly to exit WFE.
-	 */
-#define __LOAD_EXC_32(src, dst, memorder) {              \
-	if (memorder == __ATOMIC_RELAXED) {              \
-		asm volatile("ldxr %w[tmp], [%x[addr]]"  \
-			: [tmp] "=&r" (dst)              \
-			: [addr] "r"(src)                \
-			: "memory");                     \
-	} else {                                         \
-		asm volatile("ldaxr %w[tmp], [%x[addr]]" \
-			: [tmp] "=&r" (dst)              \
-			: [addr] "r"(src)                \
-			: "memory");                     \
-	} }
-
 	__LOAD_EXC_32(addr, value, memorder)
 	if (value != expected) {
 		__SEVL()
@@ -104,7 +122,6 @@ rte_wait_until_equal_32(volatile uint32_t *addr, uint32_t expected,
 			__LOAD_EXC_32(addr, value, memorder)
 		} while (value != expected);
 	}
-#undef __LOAD_EXC_32
 }
 
 static __rte_always_inline void
@@ -115,25 +132,6 @@ rte_wait_until_equal_64(volatile uint64_t *addr, uint64_t expected,
 
 	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
 
-	/*
-	 * Atomic exclusive load from addr, it returns the 64-bit content of
-	 * *addr while making it 'monitored',when it is written by someone
-	 * else, the 'monitored' state is cleared and a event is generated
-	 * implicitly to exit WFE.
-	 */
-#define __LOAD_EXC_64(src, dst, memorder) {              \
-	if (memorder == __ATOMIC_RELAXED) {              \
-		asm volatile("ldxr %x[tmp], [%x[addr]]"  \
-			: [tmp] "=&r" (dst)              \
-			: [addr] "r"(src)                \
-			: "memory");                     \
-	} else {                                         \
-		asm volatile("ldaxr %x[tmp], [%x[addr]]" \
-			: [tmp] "=&r" (dst)              \
-			: [addr] "r"(src)                \
-			: "memory");                     \
-	} }
-
 	__LOAD_EXC_64(addr, value, memorder)
 	if (value != expected) {
 		__SEVL()
@@ -143,6 +141,171 @@ rte_wait_until_equal_64(volatile uint64_t *addr, uint64_t expected,
 		} while (value != expected);
 	}
 }
+
+static __rte_always_inline void
+rte_wait_until_part_equal_16(volatile uint16_t *addr, uint16_t mask,
+		uint16_t expected, int memorder)
+{
+	uint16_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_16(addr, value, memorder)
+	if ((value & mask) != expected) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_16(addr, value, memorder)
+		} while ((value & mask) != expected);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_part_equal_32(volatile uint32_t *addr, uint32_t mask,
+		uint32_t expected, int memorder)
+{
+	uint32_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_32(addr, value, memorder)
+	if ((value & mask) != expected) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_32(addr, value, memorder)
+		} while ((value & mask) != expected);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_part_equal_64(volatile uint64_t *addr, uint64_t mask,
+		uint64_t expected, int memorder)
+{
+	uint64_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_64(addr, value, memorder)
+	if ((value & mask) != expected) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_64(addr, value, memorder)
+		} while ((value & mask) != expected);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_unequal_16(volatile uint16_t *addr, uint16_t original,
+		int memorder)
+{
+	uint16_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_16(addr, value, memorder)
+	if (value == original) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_16(addr, value, memorder)
+		} while (value == original);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_unequal_32(volatile uint32_t *addr, uint32_t original,
+		int memorder)
+{
+	uint32_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_32(addr, value, memorder)
+	if (value == original) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_32(addr, value, memorder)
+		} while (value == original);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_unequal_64(volatile uint64_t *addr, uint64_t original,
+		int memorder)
+{
+	uint64_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_64(addr, value, memorder)
+	if (value == original) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_64(addr, value, memorder)
+		} while (value == original);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_part_unequal_16(volatile uint16_t *addr, uint16_t mask,
+		uint16_t original, int memorder)
+{
+	uint16_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_16(addr, value, memorder)
+	if ((value & mask) == original) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_16(addr, value, memorder)
+		} while ((value & mask) == original);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_part_unequal_32(volatile uint32_t *addr, uint32_t mask,
+		uint32_t original, int memorder)
+{
+	uint32_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_32(addr, value, memorder)
+	if ((value & mask) == original) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_32(addr, value, memorder)
+		} while ((value & mask) == original);
+	}
+}
+
+static __rte_always_inline void
+rte_wait_until_part_unequal_64(volatile uint64_t *addr, uint64_t mask,
+		uint64_t original, int memorder)
+{
+	uint64_t value;
+
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	__LOAD_EXC_64(addr, value, memorder)
+	if ((value & mask) == original) {
+		__SEVL()
+		do {
+			__WFE()
+			__LOAD_EXC_64(addr, value, memorder)
+		} while ((value & mask) == original);
+	}
+}
+
+#undef __LOAD_EXC_16
+#undef __LOAD_EXC_32
 #undef __LOAD_EXC_64
 
 #undef __SEVL
diff --git a/lib/eal/include/generic/rte_pause.h b/lib/eal/include/generic/rte_pause.h
index 668ee4a184..943a886f01 100644
--- a/lib/eal/include/generic/rte_pause.h
+++ b/lib/eal/include/generic/rte_pause.h
@@ -81,6 +81,222 @@ static __rte_always_inline void
 rte_wait_until_equal_64(volatile uint64_t *addr, uint64_t expected,
 		int memorder);
 
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for part bits of *addr to be equal with a 16-bit expected value, with
+ * a relaxed memory ordering model meaning the loads around this API can be
+ * reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param mask
+ * value mask of a specific location
+ * @param expected
+ *  A 16-bit expected value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_part_equal_16(volatile uint16_t *addr, uint16_t mask,
+		uint16_t expected, int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for part bits of *addr to be equal with a 32-bit expected value, with
+ * a relaxed memory ordering model meaning the loads around this API can be
+ * reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param mask
+ * value mask of a specific location
+ * @param expected
+ *  A 32-bit expected value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_part_equal_32(volatile uint32_t *addr, uint32_t mask,
+		uint32_t expected, int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for part bits of *addr to be equal with a 64-bit expected value, with
+ * a relaxed memory ordering model meaning the loads around this API can be
+ * reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param mask
+ * value mask of a specific location
+ * @param expected
+ *  A 64-bit expected value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_part_equal_64(volatile uint64_t *addr, uint64_t mask,
+		uint64_t expected, int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for *addr to be unequal with a 16-bit original value, with a relaxed
+ * memory ordering model meaning the loads around this API can be reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param original
+ *  A 16-bit original value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_unequal_16(volatile uint16_t *addr, uint16_t original,
+		int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for *addr to be unequal with a 32-bit original value, with a relaxed
+ * memory ordering model meaning the loads around this API can be reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param original
+ *  A 32-bit original value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_unequal_32(volatile uint32_t *addr, uint32_t original,
+		int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for *addr to be unequal with a 64-bit original value, with a relaxed
+ * memory ordering model meaning the loads around this API can be reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param original
+ *  A 64-bit original value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_unequal_64(volatile uint64_t *addr, uint64_t original,
+		int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for part bits of *addr to be unequal with a 16-bit original value, with
+ * a relaxed memory ordering model meaning the loads around this API can be
+ * reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param mask
+ * value mask of a specific location
+ * @param original
+ *  A 16-bit original value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_part_unequal_16(volatile uint16_t *addr, uint16_t mask,
+		uint16_t original, int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for part bits of *addr to be unequal with a 32-bit original value, with
+ * a relaxed memory ordering model meaning the loads around this API can be
+ * reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param mask
+ * value mask of a specific location
+ * @param original
+ *  A 32-bit original value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_part_unequal_32(volatile uint32_t *addr, uint32_t mask,
+		uint32_t original, int memorder);
+
+/**
+ * @warning
+ * @b EXPERIMENTAL: this API may change without prior notice
+ *
+ * Wait for part bits of *addr to be unequal with a 64-bit original value, with
+ * a relaxed memory ordering model meaning the loads around this API can be
+ * reordered.
+ *
+ * @param addr
+ *  A pointer to the memory location.
+ * @param mask
+ * value mask of a specific location
+ * @param original
+ *  A 64-bit original value to be in the memory location.
+ * @param memorder
+ *  Two different memory orders that can be specified:
+ *  __ATOMIC_ACQUIRE and __ATOMIC_RELAXED. These map to
+ *  C++11 memory orders with the same names, see the C++11 standard or
+ *  the GCC wiki on atomic synchronization for detailed definition.
+ */
+__rte_experimental
+static __rte_always_inline void
+rte_wait_until_part_unequal_64(volatile uint64_t *addr, uint64_t mask,
+		uint64_t original, int memorder);
+
 #ifndef RTE_WAIT_UNTIL_EQUAL_ARCH_DEFINED
 static __rte_always_inline void
 rte_wait_until_equal_16(volatile uint16_t *addr, uint16_t expected,
@@ -111,6 +327,99 @@ rte_wait_until_equal_64(volatile uint64_t *addr, uint64_t expected,
 	while (__atomic_load_n(addr, memorder) != expected)
 		rte_pause();
 }
+
+static __rte_always_inline void
+rte_wait_until_part_equal_16(volatile uint16_t *addr, uint16_t mask,
+		uint16_t expected, int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while ((__atomic_load_n(addr, memorder) & mask) != expected)
+		rte_pause();
+
+}
+
+static __rte_always_inline void
+rte_wait_until_part_equal_32(volatile uint32_t *addr, uint32_t mask,
+		uint32_t expected, int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while ((__atomic_load_n(addr, memorder) & mask) != expected)
+		rte_pause();
+
+}
+
+static __rte_always_inline void
+rte_wait_until_part_equal_64(volatile uint64_t *addr, uint64_t mask,
+		uint64_t expected, int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while ((__atomic_load_n(addr, memorder) & mask) != expected)
+		rte_pause();
+
+}
+
+static __rte_always_inline void
+rte_wait_until_unequal_16(volatile uint16_t *addr, uint16_t original,
+		int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while (__atomic_load_n(addr, memorder) == original)
+		rte_pause();
+}
+
+static __rte_always_inline void
+rte_wait_until_unequal_32(volatile uint32_t *addr, uint32_t original,
+		int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while (__atomic_load_n(addr, memorder) == original)
+		rte_pause();
+}
+
+static __rte_always_inline void
+rte_wait_until_unequal_64(volatile uint64_t *addr, uint64_t original,
+		int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while (__atomic_load_n(addr, memorder) == original)
+		rte_pause();
+}
+
+static __rte_always_inline void
+rte_wait_until_part_unequal_16(volatile uint16_t *addr, uint16_t mask,
+		uint16_t original, int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while ((__atomic_load_n(addr, memorder) & mask) == original)
+		rte_pause();
+}
+
+static __rte_always_inline void
+rte_wait_until_part_unequal_32(volatile uint32_t *addr, uint32_t mask,
+		uint32_t original, int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while ((__atomic_load_n(addr, memorder) & mask) == original)
+		rte_pause();
+}
+
+static __rte_always_inline void
+rte_wait_until_part_unequal_64(volatile uint64_t *addr, uint64_t mask,
+		uint64_t original, int memorder)
+{
+	assert(memorder == __ATOMIC_ACQUIRE || memorder == __ATOMIC_RELAXED);
+
+	while ((__atomic_load_n(addr, memorder) & mask) == original)
+		rte_pause();
+}
 #endif
 
 #endif /* _RTE_PAUSE_H_ */
-- 
2.25.1