include/exec/memop.h: Expand comment for MO_ATOM_SUBALIGN

Expand the example in the comment documenting MO_ATOM_SUBALIGN,
to be clearer about the atomicity guarantees it represents.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-id: 20250228103222.1838913-1-peter.maydell@linaro.org

diff --git a/include/exec/memop.h b/include/exec/memop.h
index acdb40a9b3b39b80c9d18238bd1c7431e99e5643..407a47d82c7ef4dad2fe2473b5e0b8204abf9db2 100644 (file)
--- a/include/exec/memop.h
+++ b/include/exec/memop.h
@@ -91,8 +91,12 @@ typedef enum MemOp {
      *    Depending on alignment, one or both will be single-copy atomic.
      *    This is the atomicity e.g. of Arm FEAT_LSE2 LDP.
      * MO_ATOM_SUBALIGN: the operation is single-copy atomic by parts
-     *    by the alignment.  E.g. if the address is 0 mod 4, then each
-     *    4-byte subobject is single-copy atomic.
+     *    by the alignment.  E.g. if an 8-byte value is accessed at an
+     *    address which is 0 mod 8, then the whole 8-byte access is
+     *    single-copy atomic; otherwise, if it is accessed at 0 mod 4
+     *    then each 4-byte subobject is single-copy atomic; otherwise
+     *    if it is accessed at 0 mod 2 then the four 2-byte subobjects
+     *    are single-copy atomic.
      *    This is the atomicity e.g. of IBM Power.
      * MO_ATOM_NONE: the operation has no atomicity requirements.
      *