[v2] target/arm: Scalable Vector Extension

[Qemu-devel] [PATCH v2 54/67] target/arm: Implement SVE prefetches

Posted by Richard Henderson 7 years, 8 months ago

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/arm/translate-sve.c |  9 +++++++++
 target/arm/sve.decode      | 23 +++++++++++++++++++++++
 2 files changed, 32 insertions(+)

diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c
index ca49b94924..63c7a0e8d8 100644
--- a/target/arm/translate-sve.c
+++ b/target/arm/translate-sve.c
@@ -3958,3 +3958,12 @@ static void trans_ST1_zprz(DisasContext *s, arg_ST1_zprz *a, uint32_t insn)
     do_mem_zpz(s, a->rd, a->pg, a->rm, a->scale * a->msz,
                cpu_reg_sp(s, a->rn), fn);
 }
+
+/*
+ * Prefetches
+ */
+
+static void trans_PRF(DisasContext *s, arg_PRF *a, uint32_t insn)
+{
+    /* Prefetch is a nop within QEMU.  */
+}
diff --git a/target/arm/sve.decode b/target/arm/sve.decode
index edd9340c02..f0144aa2d0 100644
--- a/target/arm/sve.decode
+++ b/target/arm/sve.decode
@@ -801,6 +801,29 @@ LD1RQ_zprr	1010010 .. 00 ..... 000 ... ..... ..... \
 LD1RQ_zpri	1010010 .. 00 0.... 001 ... ..... ..... \
 		@rpri_load_msz nreg=0
 
+# SVE 32-bit gather prefetch (scalar plus 32-bit scaled offsets)
+PRF		1000010 00 -1 ----- 0-- --- ----- 0 ----
+
+# SVE 32-bit gather prefetch (vector plus immediate)
+PRF		1000010 -- 00 ----- 111 --- ----- 0 ----
+
+# SVE contiguous prefetch (scalar plus immediate)
+PRF		1000010 11 1- ----- 0-- --- ----- 0 ----
+
+# SVE contiguous prefetch (scalar plus scalar)
+PRF		1000010 -- 00 ----- 110 --- ----- 0 ----
+
+### SVE Memory 64-bit Gather Group
+
+# SVE 64-bit gather prefetch (scalar plus 64-bit scaled offsets)
+PRF		1100010 00 11 ----- 1-- --- ----- 0 ----
+
+# SVE 64-bit gather prefetch (scalar plus unpacked 32-bit scaled offsets)
+PRF		1100010 00 -1 ----- 0-- --- ----- 0 ----
+
+# SVE 64-bit gather prefetch (vector plus immediate)
+PRF		1100010 -- 00 ----- 111 --- ----- 0 ----
+
 ### SVE Memory Store Group
 
 # SVE store predicate register
-- 
2.14.3

Re: [Qemu-devel] [PATCH v2 54/67] target/arm: Implement SVE prefetches

Posted by Peter Maydell 7 years, 7 months ago

On 17 February 2018 at 18:23, Richard Henderson
<richard.henderson@linaro.org> wrote:
> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
> ---
>  target/arm/translate-sve.c |  9 +++++++++
>  target/arm/sve.decode      | 23 +++++++++++++++++++++++
>  2 files changed, 32 insertions(+)
>
> diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c
> index ca49b94924..63c7a0e8d8 100644
> --- a/target/arm/translate-sve.c
> +++ b/target/arm/translate-sve.c
> @@ -3958,3 +3958,12 @@ static void trans_ST1_zprz(DisasContext *s, arg_ST1_zprz *a, uint32_t insn)
>      do_mem_zpz(s, a->rd, a->pg, a->rm, a->scale * a->msz,
>                 cpu_reg_sp(s, a->rn), fn);
>  }
> +
> +/*
> + * Prefetches
> + */
> +
> +static void trans_PRF(DisasContext *s, arg_PRF *a, uint32_t insn)
> +{
> +    /* Prefetch is a nop within QEMU.  */
> +}
> diff --git a/target/arm/sve.decode b/target/arm/sve.decode
> index edd9340c02..f0144aa2d0 100644
> --- a/target/arm/sve.decode
> +++ b/target/arm/sve.decode
> @@ -801,6 +801,29 @@ LD1RQ_zprr 1010010 .. 00 ..... 000 ... ..... ..... \
>  LD1RQ_zpri     1010010 .. 00 0.... 001 ... ..... ..... \
>                 @rpri_load_msz nreg=0
>
> +# SVE 32-bit gather prefetch (scalar plus 32-bit scaled offsets)
> +PRF            1000010 00 -1 ----- 0-- --- ----- 0 ----
> +
> +# SVE 32-bit gather prefetch (vector plus immediate)
> +PRF            1000010 -- 00 ----- 111 --- ----- 0 ----
> +
> +# SVE contiguous prefetch (scalar plus immediate)
> +PRF            1000010 11 1- ----- 0-- --- ----- 0 ----
> +
> +# SVE contiguous prefetch (scalar plus scalar)
> +PRF            1000010 -- 00 ----- 110 --- ----- 0 ----

This one needs something slightly more complicated, because
Rm == 11111 has to be UnallocatedEncoding.

I checked the others and they don't have any unallocated cases
lurking in their decode pseudocode.

thanks
-- PMM