diff mbox series

[11/70] target/arm: Avoid tcg_const_ptr in handle_rev

Message ID 20230227054233.390271-12-richard.henderson@linaro.org
State Superseded
Headers show
Series tcg: Remove tcg_const_* | expand

Commit Message

Richard Henderson Feb. 27, 2023, 5:41 a.m. UTC
Here it is not trivial to notice first initialization, so explicitly
zero the temps.  Use an array for the output, rather than separate
tcg_rd/tcg_rd_hi variables.

Fixes a bug by adding a missing clear_vec_high.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
 target/arm/translate-a64.c | 26 +++++++++++++++-----------
 1 file changed, 15 insertions(+), 11 deletions(-)

Comments

Philippe Mathieu-Daudé March 6, 2023, 3:22 p.m. UTC | #1
On 27/2/23 06:41, Richard Henderson wrote:
> Here it is not trivial to notice first initialization, so explicitly
> zero the temps.  Use an array for the output, rather than separate
> tcg_rd/tcg_rd_hi variables.
> 
> Fixes a bug by adding a missing clear_vec_high.
> 
> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
> ---
>   target/arm/translate-a64.c | 26 +++++++++++++++-----------
>   1 file changed, 15 insertions(+), 11 deletions(-)
> 
> diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
> index 16aa71451c..02a61aa8b7 100644
> --- a/target/arm/translate-a64.c
> +++ b/target/arm/translate-a64.c
> @@ -12003,22 +12003,26 @@ static void handle_rev(DisasContext *s, int opcode, bool u,
>           int esize = 8 << size;
>           int elements = dsize / esize;
>           TCGv_i64 tcg_rn = tcg_temp_new_i64();
> -        TCGv_i64 tcg_rd = tcg_const_i64(0);
> -        TCGv_i64 tcg_rd_hi = tcg_const_i64(0);
> +        TCGv_i64 tcg_rd[2];
> +
> +        for (i = 0; i < 2; i++) {
> +            tcg_rd[i] = tcg_temp_new_i64();
> +            tcg_gen_movi_i64(tcg_rd[i], 0);
> +        }
>   
>           for (i = 0; i < elements; i++) {
>               int e_rev = (i & 0xf) ^ revmask;
> -            int off = e_rev * esize;
> +            int w = (e_rev * esize) / 64;
> +            int o = (e_rev * esize) % 64;
> +
>               read_vec_element(s, tcg_rn, rn, i, size);
> -            if (off >= 64) {
> -                tcg_gen_deposit_i64(tcg_rd_hi, tcg_rd_hi,
> -                                    tcg_rn, off - 64, esize);
> -            } else {
> -                tcg_gen_deposit_i64(tcg_rd, tcg_rd, tcg_rn, off, esize);
> -            }
> +            tcg_gen_deposit_i64(tcg_rd[w], tcg_rd[w], tcg_rn, o, esize);
>           }
> -        write_vec_element(s, tcg_rd, rd, 0, MO_64);
> -        write_vec_element(s, tcg_rd_hi, rd, 1, MO_64);
> +
> +        for (i = 0; i < 2; i++) {
> +            write_vec_element(s, tcg_rd[i], rd, i, MO_64);
> +        }

Preferably splitting the bugfix in a preliminary patch,

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>

> +        clear_vec_high(s, true, rd);
>       }
>   }
>
diff mbox series

Patch

diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index 16aa71451c..02a61aa8b7 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -12003,22 +12003,26 @@  static void handle_rev(DisasContext *s, int opcode, bool u,
         int esize = 8 << size;
         int elements = dsize / esize;
         TCGv_i64 tcg_rn = tcg_temp_new_i64();
-        TCGv_i64 tcg_rd = tcg_const_i64(0);
-        TCGv_i64 tcg_rd_hi = tcg_const_i64(0);
+        TCGv_i64 tcg_rd[2];
+
+        for (i = 0; i < 2; i++) {
+            tcg_rd[i] = tcg_temp_new_i64();
+            tcg_gen_movi_i64(tcg_rd[i], 0);
+        }
 
         for (i = 0; i < elements; i++) {
             int e_rev = (i & 0xf) ^ revmask;
-            int off = e_rev * esize;
+            int w = (e_rev * esize) / 64;
+            int o = (e_rev * esize) % 64;
+
             read_vec_element(s, tcg_rn, rn, i, size);
-            if (off >= 64) {
-                tcg_gen_deposit_i64(tcg_rd_hi, tcg_rd_hi,
-                                    tcg_rn, off - 64, esize);
-            } else {
-                tcg_gen_deposit_i64(tcg_rd, tcg_rd, tcg_rn, off, esize);
-            }
+            tcg_gen_deposit_i64(tcg_rd[w], tcg_rd[w], tcg_rn, o, esize);
         }
-        write_vec_element(s, tcg_rd, rd, 0, MO_64);
-        write_vec_element(s, tcg_rd_hi, rd, 1, MO_64);
+
+        for (i = 0; i < 2; i++) {
+            write_vec_element(s, tcg_rd[i], rd, i, MO_64);
+        }
+        clear_vec_high(s, true, rd);
     }
 }