[Qemu-devel] [PATCH 09/10] cutils: Add generic prefetch

[Paolo Bonzini]

From: Richard Henderson <address@hidden>

There's no real knowledge of the cacheline size,
just prefetching one loop ahead.

Signed-off-by: Richard Henderson <address@hidden>
Message-Id: <address@hidden>
Signed-off-by: Paolo Bonzini <address@hidden>
---
 util/bufferiszero.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/util/bufferiszero.c b/util/bufferiszero.c
index 932b809..4d8a8c8 100644
--- a/util/bufferiszero.c
+++ b/util/bufferiszero.c
@@ -38,6 +38,8 @@ static bool NAME(const void *buf, size_t len)                 
  \
     do {                                                        \
         const VECTYPE *p = buf;                                 \
         VECTYPE t;                                              \
+        __builtin_prefetch(buf + SIZE);                         \
+        barrier();                                              \
         if (SIZE == sizeof(VECTYPE) * 4) {                      \
             t = (p[0] | p[1]) | (p[2] | p[3]);                  \
         } else if (SIZE == sizeof(VECTYPE) * 8) {               \
@@ -210,6 +212,9 @@ bool buffer_is_zero(const void *buf, size_t len)
         return true;
     }
 
+    /* Fetch the beginning of the buffer while we select the accelerator.  */
+    __builtin_prefetch(buf);
+
     /* Use an optimized zero check if possible.  Note that this also
        includes a check for an unrolled loop over 64-bit integers.  */
     return select_accel_fn(buf, len);
-- 
1.8.3.1