From: Richard Henderson <richard.henderson@linaro.org>
To: qemu-devel@nongnu.org
Cc: peter.maydell@linaro.org, "Alex Bennée" <alex.bennee@linaro.org>
Subject: [PULL 09/10] tcg: Add load_dest parameter to GVecGen2
Date: Wed, 6 May 2020 11:30:01 -0700 [thread overview]
Message-ID: <20200506183002.3192-10-richard.henderson@linaro.org> (raw)
In-Reply-To: <20200506183002.3192-1-richard.henderson@linaro.org>
We have this same parameter for GVecGen2i, GVecGen3,
and GVecGen3i. This will make some SVE2 insns easier
to parameterize.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
include/tcg/tcg-op-gvec.h | 2 ++
tcg/tcg-op-gvec.c | 45 ++++++++++++++++++++++++++++-----------
2 files changed, 34 insertions(+), 13 deletions(-)
diff --git a/include/tcg/tcg-op-gvec.h b/include/tcg/tcg-op-gvec.h
index d89f91f40e..cea6497341 100644
--- a/include/tcg/tcg-op-gvec.h
+++ b/include/tcg/tcg-op-gvec.h
@@ -109,6 +109,8 @@ typedef struct {
uint8_t vece;
/* Prefer i64 to v64. */
bool prefer_i64;
+ /* Load dest as a 2nd source operand. */
+ bool load_dest;
} GVecGen2;
typedef struct {
diff --git a/tcg/tcg-op-gvec.c b/tcg/tcg-op-gvec.c
index 43cac1a0bf..049a55e700 100644
--- a/tcg/tcg-op-gvec.c
+++ b/tcg/tcg-op-gvec.c
@@ -663,17 +663,22 @@ static void expand_clr(uint32_t dofs, uint32_t maxsz)
/* Expand OPSZ bytes worth of two-operand operations using i32 elements. */
static void expand_2_i32(uint32_t dofs, uint32_t aofs, uint32_t oprsz,
- void (*fni)(TCGv_i32, TCGv_i32))
+ bool load_dest, void (*fni)(TCGv_i32, TCGv_i32))
{
TCGv_i32 t0 = tcg_temp_new_i32();
+ TCGv_i32 t1 = tcg_temp_new_i32();
uint32_t i;
for (i = 0; i < oprsz; i += 4) {
tcg_gen_ld_i32(t0, cpu_env, aofs + i);
- fni(t0, t0);
- tcg_gen_st_i32(t0, cpu_env, dofs + i);
+ if (load_dest) {
+ tcg_gen_ld_i32(t1, cpu_env, dofs + i);
+ }
+ fni(t1, t0);
+ tcg_gen_st_i32(t1, cpu_env, dofs + i);
}
tcg_temp_free_i32(t0);
+ tcg_temp_free_i32(t1);
}
static void expand_2i_i32(uint32_t dofs, uint32_t aofs, uint32_t oprsz,
@@ -793,17 +798,22 @@ static void expand_4_i32(uint32_t dofs, uint32_t aofs, uint32_t bofs,
/* Expand OPSZ bytes worth of two-operand operations using i64 elements. */
static void expand_2_i64(uint32_t dofs, uint32_t aofs, uint32_t oprsz,
- void (*fni)(TCGv_i64, TCGv_i64))
+ bool load_dest, void (*fni)(TCGv_i64, TCGv_i64))
{
TCGv_i64 t0 = tcg_temp_new_i64();
+ TCGv_i64 t1 = tcg_temp_new_i64();
uint32_t i;
for (i = 0; i < oprsz; i += 8) {
tcg_gen_ld_i64(t0, cpu_env, aofs + i);
- fni(t0, t0);
- tcg_gen_st_i64(t0, cpu_env, dofs + i);
+ if (load_dest) {
+ tcg_gen_ld_i64(t1, cpu_env, dofs + i);
+ }
+ fni(t1, t0);
+ tcg_gen_st_i64(t1, cpu_env, dofs + i);
}
tcg_temp_free_i64(t0);
+ tcg_temp_free_i64(t1);
}
static void expand_2i_i64(uint32_t dofs, uint32_t aofs, uint32_t oprsz,
@@ -924,17 +934,23 @@ static void expand_4_i64(uint32_t dofs, uint32_t aofs, uint32_t bofs,
/* Expand OPSZ bytes worth of two-operand operations using host vectors. */
static void expand_2_vec(unsigned vece, uint32_t dofs, uint32_t aofs,
uint32_t oprsz, uint32_t tysz, TCGType type,
+ bool load_dest,
void (*fni)(unsigned, TCGv_vec, TCGv_vec))
{
TCGv_vec t0 = tcg_temp_new_vec(type);
+ TCGv_vec t1 = tcg_temp_new_vec(type);
uint32_t i;
for (i = 0; i < oprsz; i += tysz) {
tcg_gen_ld_vec(t0, cpu_env, aofs + i);
- fni(vece, t0, t0);
- tcg_gen_st_vec(t0, cpu_env, dofs + i);
+ if (load_dest) {
+ tcg_gen_ld_vec(t1, cpu_env, dofs + i);
+ }
+ fni(vece, t1, t0);
+ tcg_gen_st_vec(t1, cpu_env, dofs + i);
}
tcg_temp_free_vec(t0);
+ tcg_temp_free_vec(t1);
}
/* Expand OPSZ bytes worth of two-vector operands and an immediate operand
@@ -1088,7 +1104,8 @@ void tcg_gen_gvec_2(uint32_t dofs, uint32_t aofs,
* that e.g. size == 80 would be expanded with 2x32 + 1x16.
*/
some = QEMU_ALIGN_DOWN(oprsz, 32);
- expand_2_vec(g->vece, dofs, aofs, some, 32, TCG_TYPE_V256, g->fniv);
+ expand_2_vec(g->vece, dofs, aofs, some, 32, TCG_TYPE_V256,
+ g->load_dest, g->fniv);
if (some == oprsz) {
break;
}
@@ -1098,17 +1115,19 @@ void tcg_gen_gvec_2(uint32_t dofs, uint32_t aofs,
maxsz -= some;
/* fallthru */
case TCG_TYPE_V128:
- expand_2_vec(g->vece, dofs, aofs, oprsz, 16, TCG_TYPE_V128, g->fniv);
+ expand_2_vec(g->vece, dofs, aofs, oprsz, 16, TCG_TYPE_V128,
+ g->load_dest, g->fniv);
break;
case TCG_TYPE_V64:
- expand_2_vec(g->vece, dofs, aofs, oprsz, 8, TCG_TYPE_V64, g->fniv);
+ expand_2_vec(g->vece, dofs, aofs, oprsz, 8, TCG_TYPE_V64,
+ g->load_dest, g->fniv);
break;
case 0:
if (g->fni8 && check_size_impl(oprsz, 8)) {
- expand_2_i64(dofs, aofs, oprsz, g->fni8);
+ expand_2_i64(dofs, aofs, oprsz, g->load_dest, g->fni8);
} else if (g->fni4 && check_size_impl(oprsz, 4)) {
- expand_2_i32(dofs, aofs, oprsz, g->fni4);
+ expand_2_i32(dofs, aofs, oprsz, g->load_dest, g->fni4);
} else {
assert(g->fno != NULL);
tcg_gen_gvec_2_ool(dofs, aofs, oprsz, maxsz, g->data, g->fno);
--
2.20.1
next prev parent reply other threads:[~2020-05-06 18:34 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-05-06 18:29 [PULL 00/10] tcg patch queue Richard Henderson
2020-05-06 18:29 ` [PULL 01/10] tcg: Add tcg_gen_gvec_dup_imm Richard Henderson
2020-05-06 18:29 ` [PULL 02/10] target/s390x: Use tcg_gen_gvec_dup_imm Richard Henderson
2020-05-06 18:29 ` [PULL 03/10] target/ppc: " Richard Henderson
2020-05-06 18:29 ` [PULL 04/10] target/arm: " Richard Henderson
2020-05-07 14:39 ` Laurent Desnogues
2020-05-07 17:16 ` Richard Henderson
2020-05-06 18:29 ` [PULL 05/10] tcg: Use tcg_gen_gvec_dup_imm in logical simplifications Richard Henderson
2020-05-06 18:29 ` [PULL 06/10] tcg: Remove tcg_gen_gvec_dup{8,16,32,64}i Richard Henderson
2020-05-06 18:29 ` [PULL 07/10] tcg: Add tcg_gen_gvec_dup_tl Richard Henderson
2020-05-06 18:30 ` [PULL 08/10] tcg: Improve vector tail clearing Richard Henderson
2020-05-06 18:30 ` Richard Henderson [this message]
2020-05-06 18:30 ` [PULL 10/10] tcg: Fix integral argument type to tcg_gen_rot[rl]i_i{32, 64} Richard Henderson
2020-05-07 9:54 ` [PULL 00/10] tcg patch queue Peter Maydell
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=20200506183002.3192-10-richard.henderson@linaro.org \
--to=richard.henderson@linaro.org \
--cc=alex.bennee@linaro.org \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).