[PATCH v4 13/16] accel/tcg: Extract store_atom_insert_al16 to host header

[Richard Henderson <richard.henderson@linaro.org>]

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
 host/include/generic/host/store-insert-al16.h | 50 +++++++++++++++++++
 accel/tcg/ldst_atomicity.c.inc                | 40 +--------------
 2 files changed, 51 insertions(+), 39 deletions(-)
 create mode 100644 host/include/generic/host/store-insert-al16.h

diff --git a/host/include/generic/host/store-insert-al16.h b/host/include/generic/host/store-insert-al16.h
new file mode 100644
index 0000000000..4a1662183d
--- /dev/null
+++ b/host/include/generic/host/store-insert-al16.h
@@ -0,0 +1,50 @@
+/*
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ * Atomic store insert into 128-bit, generic version.
+ *
+ * Copyright (C) 2023 Linaro, Ltd.
+ */
+
+#ifndef HOST_STORE_INSERT_AL16_H
+#define HOST_STORE_INSERT_AL16_H
+
+/**
+ * store_atom_insert_al16:
+ * @p: host address
+ * @val: shifted value to store
+ * @msk: mask for value to store
+ *
+ * Atomically store @val to @p masked by @msk.
+ */
+static inline void ATTRIBUTE_ATOMIC128_OPT
+store_atom_insert_al16(Int128 *ps, Int128 val, Int128 msk)
+{
+#if defined(CONFIG_ATOMIC128)
+    __uint128_t *pu;
+    Int128Alias old, new;
+
+    /* With CONFIG_ATOMIC128, we can avoid the memory barriers. */
+    pu = __builtin_assume_aligned(ps, 16);
+    old.u = *pu;
+    msk = int128_not(msk);
+    do {
+        new.s = int128_and(old.s, msk);
+        new.s = int128_or(new.s, val);
+    } while (!__atomic_compare_exchange_n(pu, &old.u, new.u, true,
+                                          __ATOMIC_RELAXED, __ATOMIC_RELAXED));
+#else
+    Int128 old, new, cmp;
+
+    ps = __builtin_assume_aligned(ps, 16);
+    old = *ps;
+    msk = int128_not(msk);
+    do {
+        cmp = old;
+        new = int128_and(old, msk);
+        new = int128_or(new, val);
+        old = atomic16_cmpxchg(ps, cmp, new);
+    } while (int128_ne(cmp, old));
+#endif
+}
+
+#endif /* HOST_STORE_INSERT_AL16_H */
diff --git a/accel/tcg/ldst_atomicity.c.inc b/accel/tcg/ldst_atomicity.c.inc
index 39ad89800d..6844f85d58 100644
--- a/accel/tcg/ldst_atomicity.c.inc
+++ b/accel/tcg/ldst_atomicity.c.inc
@@ -10,6 +10,7 @@
  */
 
 #include "host/load-extract-al16-al8.h"
+#include "host/store-insert-al16.h"
 
 #ifdef CONFIG_ATOMIC64
 # define HAVE_al8          true
@@ -681,45 +682,6 @@ static void store_atom_insert_al8(uint64_t *p, uint64_t val, uint64_t msk)
                                           __ATOMIC_RELAXED, __ATOMIC_RELAXED));
 }
 
-/**
- * store_atom_insert_al16:
- * @p: host address
- * @val: shifted value to store
- * @msk: mask for value to store
- *
- * Atomically store @val to @p masked by @msk.
- */
-static void ATTRIBUTE_ATOMIC128_OPT
-store_atom_insert_al16(Int128 *ps, Int128Alias val, Int128Alias msk)
-{
-#if defined(CONFIG_ATOMIC128)
-    __uint128_t *pu, old, new;
-
-    /* With CONFIG_ATOMIC128, we can avoid the memory barriers. */
-    pu = __builtin_assume_aligned(ps, 16);
-    old = *pu;
-    do {
-        new = (old & ~msk.u) | val.u;
-    } while (!__atomic_compare_exchange_n(pu, &old, new, true,
-                                          __ATOMIC_RELAXED, __ATOMIC_RELAXED));
-#elif defined(CONFIG_CMPXCHG128)
-    __uint128_t *pu, old, new;
-
-    /*
-     * Without CONFIG_ATOMIC128, __atomic_compare_exchange_n will always
-     * defer to libatomic, so we must use __sync_*_compare_and_swap_16
-     * and accept the sequential consistency that comes with it.
-     */
-    pu = __builtin_assume_aligned(ps, 16);
-    do {
-        old = *pu;
-        new = (old & ~msk.u) | val.u;
-    } while (!__sync_bool_compare_and_swap_16(pu, old, new));
-#else
-    qemu_build_not_reached();
-#endif
-}
-
 /**
  * store_bytes_leN:
  * @pv: host address
-- 
2.34.1