qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Richard Henderson <richard.henderson@linaro.org>
To: qemu-devel@nongnu.org
Cc: luis.pires@eldorado.org.br, alex.bennee@linaro.org
Subject: [PATCH v3 41/48] tcg/optimize: Sink commutative operand swapping into fold functions
Date: Thu, 21 Oct 2021 14:05:32 -0700	[thread overview]
Message-ID: <20211021210539.825582-42-richard.henderson@linaro.org> (raw)
In-Reply-To: <20211021210539.825582-1-richard.henderson@linaro.org>

Most of these are handled by creating a fold_const2_commutative
to handle all of the binary operators.  The rest were already
handled on a case-by-case basis in the switch, and have their
own fold function in which to place the call.

We now have only one major switch on TCGOpcode.

Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
 tcg/optimize.c | 128 ++++++++++++++++++++++---------------------------
 1 file changed, 56 insertions(+), 72 deletions(-)

diff --git a/tcg/optimize.c b/tcg/optimize.c
index ba068e7d3e..92b35a8c3f 100644
--- a/tcg/optimize.c
+++ b/tcg/optimize.c
@@ -696,6 +696,12 @@ static bool fold_const2(OptContext *ctx, TCGOp *op)
     return false;
 }
 
+static bool fold_const2_commutative(OptContext *ctx, TCGOp *op)
+{
+    swap_commutative(op->args[0], &op->args[1], &op->args[2]);
+    return fold_const2(ctx, op);
+}
+
 static bool fold_masks(OptContext *ctx, TCGOp *op)
 {
     uint64_t a_mask = ctx->a_mask;
@@ -832,7 +838,7 @@ static bool fold_xx_to_x(OptContext *ctx, TCGOp *op)
 
 static bool fold_add(OptContext *ctx, TCGOp *op)
 {
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xi_to_x(ctx, op, 0)) {
         return true;
     }
@@ -891,6 +897,9 @@ static bool fold_addsub2(OptContext *ctx, TCGOp *op, bool add)
 
 static bool fold_add2(OptContext *ctx, TCGOp *op)
 {
+    swap_commutative(op->args[0], &op->args[2], &op->args[4]);
+    swap_commutative(op->args[1], &op->args[3], &op->args[5]);
+
     return fold_addsub2(ctx, op, true);
 }
 
@@ -898,7 +907,7 @@ static bool fold_and(OptContext *ctx, TCGOp *op)
 {
     uint64_t z1, z2;
 
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xi_to_i(ctx, op, 0) ||
         fold_xi_to_x(ctx, op, -1) ||
         fold_xx_to_x(ctx, op)) {
@@ -950,8 +959,13 @@ static bool fold_andc(OptContext *ctx, TCGOp *op)
 static bool fold_brcond(OptContext *ctx, TCGOp *op)
 {
     TCGCond cond = op->args[2];
-    int i = do_constant_folding_cond(ctx->type, op->args[0], op->args[1], cond);
+    int i;
 
+    if (swap_commutative(-1, &op->args[0], &op->args[1])) {
+        op->args[2] = cond = tcg_swap_cond(cond);
+    }
+
+    i = do_constant_folding_cond(ctx->type, op->args[0], op->args[1], cond);
     if (i == 0) {
         tcg_op_remove(ctx->tcg, op);
         return true;
@@ -966,10 +980,14 @@ static bool fold_brcond(OptContext *ctx, TCGOp *op)
 static bool fold_brcond2(OptContext *ctx, TCGOp *op)
 {
     TCGCond cond = op->args[4];
-    int i = do_constant_folding_cond2(&op->args[0], &op->args[2], cond);
     TCGArg label = op->args[5];
-    int inv = 0;
+    int i, inv = 0;
 
+    if (swap_commutative2(&op->args[0], &op->args[2])) {
+        op->args[4] = cond = tcg_swap_cond(cond);
+    }
+
+    i = do_constant_folding_cond2(&op->args[0], &op->args[2], cond);
     if (i >= 0) {
         goto do_brcond_const;
     }
@@ -1214,7 +1232,7 @@ static bool fold_dup2(OptContext *ctx, TCGOp *op)
 
 static bool fold_eqv(OptContext *ctx, TCGOp *op)
 {
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xi_to_x(ctx, op, -1) ||
         fold_xi_to_not(ctx, op, 0)) {
         return true;
@@ -1376,8 +1394,20 @@ static bool fold_mov(OptContext *ctx, TCGOp *op)
 static bool fold_movcond(OptContext *ctx, TCGOp *op)
 {
     TCGCond cond = op->args[5];
-    int i = do_constant_folding_cond(ctx->type, op->args[1], op->args[2], cond);
+    int i;
 
+    if (swap_commutative(-1, &op->args[1], &op->args[2])) {
+        op->args[5] = cond = tcg_swap_cond(cond);
+    }
+    /*
+     * Canonicalize the "false" input reg to match the destination reg so
+     * that the tcg backend can implement a "move if true" operation.
+     */
+    if (swap_commutative(op->args[0], &op->args[4], &op->args[3])) {
+        op->args[5] = cond = tcg_invert_cond(cond);
+    }
+
+    i = do_constant_folding_cond(ctx->type, op->args[1], op->args[2], cond);
     if (i >= 0) {
         return tcg_opt_gen_mov(ctx, op, op->args[0], op->args[4 - i]);
     }
@@ -1414,7 +1444,7 @@ static bool fold_movcond(OptContext *ctx, TCGOp *op)
 
 static bool fold_multiply(OptContext *ctx, TCGOp *op)
 {
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xi_to_i(ctx, op, 0)) {
         return true;
     }
@@ -1423,6 +1453,8 @@ static bool fold_multiply(OptContext *ctx, TCGOp *op)
 
 static bool fold_multiply2(OptContext *ctx, TCGOp *op)
 {
+    swap_commutative(op->args[0], &op->args[2], &op->args[3]);
+
     if (arg_is_const(op->args[2]) && arg_is_const(op->args[3])) {
         uint64_t a = arg_info(op->args[2])->val;
         uint64_t b = arg_info(op->args[3])->val;
@@ -1466,7 +1498,7 @@ static bool fold_multiply2(OptContext *ctx, TCGOp *op)
 
 static bool fold_nand(OptContext *ctx, TCGOp *op)
 {
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xi_to_not(ctx, op, -1)) {
         return true;
     }
@@ -1495,7 +1527,7 @@ static bool fold_neg(OptContext *ctx, TCGOp *op)
 
 static bool fold_nor(OptContext *ctx, TCGOp *op)
 {
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xi_to_not(ctx, op, 0)) {
         return true;
     }
@@ -1515,7 +1547,7 @@ static bool fold_not(OptContext *ctx, TCGOp *op)
 
 static bool fold_or(OptContext *ctx, TCGOp *op)
 {
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xi_to_x(ctx, op, 0) ||
         fold_xx_to_x(ctx, op)) {
         return true;
@@ -1561,8 +1593,13 @@ static bool fold_qemu_st(OptContext *ctx, TCGOp *op)
 static bool fold_setcond(OptContext *ctx, TCGOp *op)
 {
     TCGCond cond = op->args[3];
-    int i = do_constant_folding_cond(ctx->type, op->args[1], op->args[2], cond);
+    int i;
 
+    if (swap_commutative(op->args[0], &op->args[1], &op->args[2])) {
+        op->args[3] = cond = tcg_swap_cond(cond);
+    }
+
+    i = do_constant_folding_cond(ctx->type, op->args[1], op->args[2], cond);
     if (i >= 0) {
         return tcg_opt_gen_movi(ctx, op, op->args[0], i);
     }
@@ -1574,9 +1611,13 @@ static bool fold_setcond(OptContext *ctx, TCGOp *op)
 static bool fold_setcond2(OptContext *ctx, TCGOp *op)
 {
     TCGCond cond = op->args[5];
-    int i = do_constant_folding_cond2(&op->args[1], &op->args[3], cond);
-    int inv = 0;
+    int i, inv = 0;
 
+    if (swap_commutative2(&op->args[1], &op->args[3])) {
+        op->args[5] = cond = tcg_swap_cond(cond);
+    }
+
+    i = do_constant_folding_cond2(&op->args[1], &op->args[3], cond);
     if (i >= 0) {
         goto do_setcond_const;
     }
@@ -1754,7 +1795,7 @@ static bool fold_tcg_ld(OptContext *ctx, TCGOp *op)
 
 static bool fold_xor(OptContext *ctx, TCGOp *op)
 {
-    if (fold_const2(ctx, op) ||
+    if (fold_const2_commutative(ctx, op) ||
         fold_xx_to_i(ctx, op, 0) ||
         fold_xi_to_x(ctx, op, 0) ||
         fold_xi_to_not(ctx, op, -1)) {
@@ -1807,63 +1848,6 @@ void tcg_optimize(TCGContext *s)
             ctx.type = TCG_TYPE_I32;
         }
 
-        /* For commutative operations make constant second argument */
-        switch (opc) {
-        CASE_OP_32_64_VEC(add):
-        CASE_OP_32_64_VEC(mul):
-        CASE_OP_32_64_VEC(and):
-        CASE_OP_32_64_VEC(or):
-        CASE_OP_32_64_VEC(xor):
-        CASE_OP_32_64(eqv):
-        CASE_OP_32_64(nand):
-        CASE_OP_32_64(nor):
-        CASE_OP_32_64(muluh):
-        CASE_OP_32_64(mulsh):
-            swap_commutative(op->args[0], &op->args[1], &op->args[2]);
-            break;
-        CASE_OP_32_64(brcond):
-            if (swap_commutative(-1, &op->args[0], &op->args[1])) {
-                op->args[2] = tcg_swap_cond(op->args[2]);
-            }
-            break;
-        CASE_OP_32_64(setcond):
-            if (swap_commutative(op->args[0], &op->args[1], &op->args[2])) {
-                op->args[3] = tcg_swap_cond(op->args[3]);
-            }
-            break;
-        CASE_OP_32_64(movcond):
-            if (swap_commutative(-1, &op->args[1], &op->args[2])) {
-                op->args[5] = tcg_swap_cond(op->args[5]);
-            }
-            /* For movcond, we canonicalize the "false" input reg to match
-               the destination reg so that the tcg backend can implement
-               a "move if true" operation.  */
-            if (swap_commutative(op->args[0], &op->args[4], &op->args[3])) {
-                op->args[5] = tcg_invert_cond(op->args[5]);
-            }
-            break;
-        CASE_OP_32_64(add2):
-            swap_commutative(op->args[0], &op->args[2], &op->args[4]);
-            swap_commutative(op->args[1], &op->args[3], &op->args[5]);
-            break;
-        CASE_OP_32_64(mulu2):
-        CASE_OP_32_64(muls2):
-            swap_commutative(op->args[0], &op->args[2], &op->args[3]);
-            break;
-        case INDEX_op_brcond2_i32:
-            if (swap_commutative2(&op->args[0], &op->args[2])) {
-                op->args[4] = tcg_swap_cond(op->args[4]);
-            }
-            break;
-        case INDEX_op_setcond2_i32:
-            if (swap_commutative2(&op->args[1], &op->args[3])) {
-                op->args[5] = tcg_swap_cond(op->args[5]);
-            }
-            break;
-        default:
-            break;
-        }
-
         /* Assume all bits affected, and no bits known zero. */
         ctx.a_mask = -1;
         ctx.z_mask = -1;
-- 
2.25.1



  parent reply	other threads:[~2021-10-21 21:48 UTC|newest]

Thread overview: 108+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-10-21 21:04 [PATCH v3 00/48] tcg: optimize redundant sign extensions Richard Henderson
2021-10-21 21:04 ` [PATCH v3 01/48] tcg/optimize: Rename "mask" to "z_mask" Richard Henderson
2021-10-21 21:04 ` [PATCH v3 02/48] tcg/optimize: Split out OptContext Richard Henderson
2021-10-22 14:53   ` Philippe Mathieu-Daudé
2021-10-21 21:04 ` [PATCH v3 03/48] tcg/optimize: Remove do_default label Richard Henderson
2021-10-22 15:12   ` Philippe Mathieu-Daudé
2021-10-21 21:04 ` [PATCH v3 04/48] tcg/optimize: Change tcg_opt_gen_{mov, movi} interface Richard Henderson
2021-10-21 21:04 ` [PATCH v3 05/48] tcg/optimize: Move prev_mb into OptContext Richard Henderson
2021-10-22 15:13   ` Philippe Mathieu-Daudé
2021-10-21 21:04 ` [PATCH v3 06/48] tcg/optimize: Split out init_arguments Richard Henderson
2021-10-21 21:04 ` [PATCH v3 07/48] tcg/optimize: Split out copy_propagate Richard Henderson
2021-10-21 21:04 ` [PATCH v3 08/48] tcg/optimize: Split out fold_call Richard Henderson
2021-10-21 21:05 ` [PATCH v3 09/48] tcg/optimize: Drop nb_oargs, nb_iargs locals Richard Henderson
2021-10-21 21:05 ` [PATCH v3 10/48] tcg/optimize: Change fail return for do_constant_folding_cond* Richard Henderson
2021-10-22 13:53   ` Philippe Mathieu-Daudé
2021-10-21 21:05 ` [PATCH v3 11/48] tcg/optimize: Return true from tcg_opt_gen_{mov, movi} Richard Henderson
2021-10-26 15:13   ` [PATCH v3 11/48] tcg/optimize: Return true from tcg_opt_gen_{mov,movi} Alex Bennée
2021-10-21 21:05 ` [PATCH v3 12/48] tcg/optimize: Split out finish_folding Richard Henderson
2021-10-26 15:16   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 13/48] tcg/optimize: Use a boolean to avoid a mass of continues Richard Henderson
2021-10-26 15:17   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 14/48] tcg/optimize: Split out fold_mb, fold_qemu_{ld,st} Richard Henderson
2021-10-22 13:56   ` Philippe Mathieu-Daudé
2021-10-21 21:05 ` [PATCH v3 15/48] tcg/optimize: Split out fold_const{1,2} Richard Henderson
2021-10-26 15:19   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 16/48] tcg/optimize: Split out fold_setcond2 Richard Henderson
2021-10-26 15:31   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 17/48] tcg/optimize: Split out fold_brcond2 Richard Henderson
2021-10-22 17:26   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 18/48] tcg/optimize: Split out fold_brcond Richard Henderson
2021-10-22 13:58   ` Philippe Mathieu-Daudé
2021-10-21 21:05 ` [PATCH v3 19/48] tcg/optimize: Split out fold_setcond Richard Henderson
2021-10-22 13:59   ` Philippe Mathieu-Daudé
2021-10-21 21:05 ` [PATCH v3 20/48] tcg/optimize: Split out fold_mulu2_i32 Richard Henderson
2021-10-22 14:00   ` Philippe Mathieu-Daudé
2021-10-22 17:32   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 21/48] tcg/optimize: Split out fold_addsub2_i32 Richard Henderson
2021-10-22 14:01   ` Philippe Mathieu-Daudé
2021-10-22 17:33   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 22/48] tcg/optimize: Split out fold_movcond Richard Henderson
2021-10-22 14:05   ` Philippe Mathieu-Daudé
2021-10-22 17:34   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 23/48] tcg/optimize: Split out fold_extract2 Richard Henderson
2021-10-22 14:07   ` Philippe Mathieu-Daudé
2021-10-22 17:34   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 24/48] tcg/optimize: Split out fold_extract, fold_sextract Richard Henderson
2021-10-22 14:08   ` Philippe Mathieu-Daudé
2021-10-22 17:34   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 25/48] tcg/optimize: Split out fold_deposit Richard Henderson
2021-10-22 14:09   ` Philippe Mathieu-Daudé
2021-10-22 17:35   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 26/48] tcg/optimize: Split out fold_count_zeros Richard Henderson
2021-10-22 14:10   ` Philippe Mathieu-Daudé
2021-10-22 17:35   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 27/48] tcg/optimize: Split out fold_bswap Richard Henderson
2021-10-22 17:36   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 28/48] tcg/optimize: Split out fold_dup, fold_dup2 Richard Henderson
2021-10-22 14:12   ` Philippe Mathieu-Daudé
2021-10-22 17:36   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 29/48] tcg/optimize: Split out fold_mov Richard Henderson
2021-10-22 17:39   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 30/48] tcg/optimize: Split out fold_xx_to_i Richard Henderson
2021-10-22 21:04   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 31/48] tcg/optimize: Split out fold_xx_to_x Richard Henderson
2021-10-22 21:04   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 32/48] tcg/optimize: Split out fold_xi_to_i Richard Henderson
2021-10-22 21:17   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 33/48] tcg/optimize: Add type to OptContext Richard Henderson
2021-10-22 22:11   ` Luis Fernando Fujita Pires
2021-10-23 18:25     ` Richard Henderson
2021-10-21 21:05 ` [PATCH v3 34/48] tcg/optimize: Split out fold_to_not Richard Henderson
2021-10-25 14:17   ` Luis Fernando Fujita Pires
2021-10-25 17:31     ` Richard Henderson
2021-10-25 18:13       ` Luis Fernando Fujita Pires
2021-10-25 18:35         ` Richard Henderson
2021-10-21 21:05 ` [PATCH v3 35/48] tcg/optimize: Split out fold_sub_to_neg Richard Henderson
2021-10-25 14:17   ` Luis Fernando Fujita Pires
2021-10-21 21:05 ` [PATCH v3 36/48] tcg/optimize: Split out fold_xi_to_x Richard Henderson
2021-10-25 14:26   ` Luis Fernando Fujita Pires
2021-10-25 17:46     ` Richard Henderson
2021-10-21 21:05 ` [PATCH v3 37/48] tcg/optimize: Split out fold_ix_to_i Richard Henderson
2021-10-25 14:26   ` Luis Fernando Fujita Pires
2021-10-26 19:11   ` Philippe Mathieu-Daudé
2021-10-21 21:05 ` [PATCH v3 38/48] tcg/optimize: Split out fold_masks Richard Henderson
2021-10-26 15:32   ` Alex Bennée
2021-10-26 18:50     ` Richard Henderson
2021-10-21 21:05 ` [PATCH v3 39/48] tcg/optimize: Expand fold_mulu2_i32 to all 4-arg multiplies Richard Henderson
2021-10-22 14:16   ` Philippe Mathieu-Daudé
2021-10-21 21:05 ` [PATCH v3 40/48] tcg/optimize: Expand fold_addsub2_i32 to 64-bit ops Richard Henderson
2021-10-26 16:21   ` Alex Bennée
2021-10-21 21:05 ` Richard Henderson [this message]
2021-10-26 16:27   ` [PATCH v3 41/48] tcg/optimize: Sink commutative operand swapping into fold functions Alex Bennée
2021-10-26 19:33     ` Richard Henderson
2021-10-27 13:22       ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 42/48] tcg/optimize: Add more simplifications for orc Richard Henderson
2021-10-26 16:22   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 43/48] tcg/optimize: Stop forcing z_mask to "garbage" for 32-bit values Richard Henderson
2021-10-26 16:35   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 44/48] tcg/optimize: Optimize sign extensions Richard Henderson
2021-10-26 16:34   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 45/48] tcg/optimize: Propagate sign info for logical operations Richard Henderson
2021-10-26 16:35   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 46/48] tcg/optimize: Propagate sign info for setcond Richard Henderson
2021-10-26 16:36   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 47/48] tcg/optimize: Propagate sign info for bit counting Richard Henderson
2021-10-26 16:36   ` Alex Bennée
2021-10-21 21:05 ` [PATCH v3 48/48] tcg/optimize: Propagate sign info for shifting Richard Henderson
2021-10-26 16:37   ` Alex Bennée

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20211021210539.825582-42-richard.henderson@linaro.org \
    --to=richard.henderson@linaro.org \
    --cc=alex.bennee@linaro.org \
    --cc=luis.pires@eldorado.org.br \
    --cc=qemu-devel@nongnu.org \
    --subject='Re: [PATCH v3 41/48] tcg/optimize: Sink commutative operand swapping into fold functions' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).