qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Alistair Francis <alistair.francis@opensource.wdc.com>
To: qemu-devel@nongnu.org
Cc: alistair23@gmail.com, eopXD <eop.chen@sifive.com>,
	Frank Chang <frank.chang@sifive.com>,
	Alistair Francis <alistair.francis@wdc.com>
Subject: [PULL 02/23] target/riscv: rvv: Fix early exit condition for whole register load/store
Date: Wed, 25 May 2022 08:44:07 +1000	[thread overview]
Message-ID: <20220524224428.552334-3-alistair.francis@opensource.wdc.com> (raw)
In-Reply-To: <20220524224428.552334-1-alistair.francis@opensource.wdc.com>

From: eopXD <eop.chen@sifive.com>

Vector whole register load instructions have EEW encoded in the opcode,
so we shouldn't take SEW here. Vector whole register store instructions
are always EEW=8.

Signed-off-by: eop Chen <eop.chen@sifive.com>
Reviewed-by: Frank Chang <frank.chang@sifive.com>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Message-Id: <165181414065.18540.14828125053334599921-0@git.sr.ht>
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
---
 target/riscv/insn_trans/trans_rvv.c.inc | 58 +++++++++++++------------
 1 file changed, 31 insertions(+), 27 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index 90327509f7..391c61fe93 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -1118,10 +1118,10 @@ GEN_VEXT_TRANS(vle64ff_v, MO_64, r2nfvm, ldff_op, ld_us_check)
 typedef void gen_helper_ldst_whole(TCGv_ptr, TCGv, TCGv_env, TCGv_i32);
 
 static bool ldst_whole_trans(uint32_t vd, uint32_t rs1, uint32_t nf,
-                             gen_helper_ldst_whole *fn, DisasContext *s,
-                             bool is_store)
+                             uint32_t width, gen_helper_ldst_whole *fn,
+                             DisasContext *s, bool is_store)
 {
-    uint32_t evl = (s->cfg_ptr->vlen / 8) * nf / (1 << s->sew);
+    uint32_t evl = (s->cfg_ptr->vlen / 8) * nf / width;
     TCGLabel *over = gen_new_label();
     tcg_gen_brcondi_tl(TCG_COND_GEU, cpu_vstart, evl, over);
 
@@ -1153,38 +1153,42 @@ static bool ldst_whole_trans(uint32_t vd, uint32_t rs1, uint32_t nf,
  * load and store whole register instructions ignore vtype and vl setting.
  * Thus, we don't need to check vill bit. (Section 7.9)
  */
-#define GEN_LDST_WHOLE_TRANS(NAME, ARG_NF, IS_STORE)                      \
+#define GEN_LDST_WHOLE_TRANS(NAME, ARG_NF, WIDTH, IS_STORE)               \
 static bool trans_##NAME(DisasContext *s, arg_##NAME * a)                 \
 {                                                                         \
     if (require_rvv(s) &&                                                 \
         QEMU_IS_ALIGNED(a->rd, ARG_NF)) {                                 \
-        return ldst_whole_trans(a->rd, a->rs1, ARG_NF, gen_helper_##NAME, \
-                                s, IS_STORE);                             \
+        return ldst_whole_trans(a->rd, a->rs1, ARG_NF, WIDTH,             \
+                                gen_helper_##NAME, s, IS_STORE);          \
     }                                                                     \
     return false;                                                         \
 }
 
-GEN_LDST_WHOLE_TRANS(vl1re8_v,  1, false)
-GEN_LDST_WHOLE_TRANS(vl1re16_v, 1, false)
-GEN_LDST_WHOLE_TRANS(vl1re32_v, 1, false)
-GEN_LDST_WHOLE_TRANS(vl1re64_v, 1, false)
-GEN_LDST_WHOLE_TRANS(vl2re8_v,  2, false)
-GEN_LDST_WHOLE_TRANS(vl2re16_v, 2, false)
-GEN_LDST_WHOLE_TRANS(vl2re32_v, 2, false)
-GEN_LDST_WHOLE_TRANS(vl2re64_v, 2, false)
-GEN_LDST_WHOLE_TRANS(vl4re8_v,  4, false)
-GEN_LDST_WHOLE_TRANS(vl4re16_v, 4, false)
-GEN_LDST_WHOLE_TRANS(vl4re32_v, 4, false)
-GEN_LDST_WHOLE_TRANS(vl4re64_v, 4, false)
-GEN_LDST_WHOLE_TRANS(vl8re8_v,  8, false)
-GEN_LDST_WHOLE_TRANS(vl8re16_v, 8, false)
-GEN_LDST_WHOLE_TRANS(vl8re32_v, 8, false)
-GEN_LDST_WHOLE_TRANS(vl8re64_v, 8, false)
-
-GEN_LDST_WHOLE_TRANS(vs1r_v, 1, true)
-GEN_LDST_WHOLE_TRANS(vs2r_v, 2, true)
-GEN_LDST_WHOLE_TRANS(vs4r_v, 4, true)
-GEN_LDST_WHOLE_TRANS(vs8r_v, 8, true)
+GEN_LDST_WHOLE_TRANS(vl1re8_v,  1, 1, false)
+GEN_LDST_WHOLE_TRANS(vl1re16_v, 1, 2, false)
+GEN_LDST_WHOLE_TRANS(vl1re32_v, 1, 4, false)
+GEN_LDST_WHOLE_TRANS(vl1re64_v, 1, 8, false)
+GEN_LDST_WHOLE_TRANS(vl2re8_v,  2, 1, false)
+GEN_LDST_WHOLE_TRANS(vl2re16_v, 2, 2, false)
+GEN_LDST_WHOLE_TRANS(vl2re32_v, 2, 4, false)
+GEN_LDST_WHOLE_TRANS(vl2re64_v, 2, 8, false)
+GEN_LDST_WHOLE_TRANS(vl4re8_v,  4, 1, false)
+GEN_LDST_WHOLE_TRANS(vl4re16_v, 4, 2, false)
+GEN_LDST_WHOLE_TRANS(vl4re32_v, 4, 4, false)
+GEN_LDST_WHOLE_TRANS(vl4re64_v, 4, 8, false)
+GEN_LDST_WHOLE_TRANS(vl8re8_v,  8, 1, false)
+GEN_LDST_WHOLE_TRANS(vl8re16_v, 8, 2, false)
+GEN_LDST_WHOLE_TRANS(vl8re32_v, 8, 4, false)
+GEN_LDST_WHOLE_TRANS(vl8re64_v, 8, 8, false)
+
+/*
+ * The vector whole register store instructions are encoded similar to
+ * unmasked unit-stride store of elements with EEW=8.
+ */
+GEN_LDST_WHOLE_TRANS(vs1r_v, 1, 1, true)
+GEN_LDST_WHOLE_TRANS(vs2r_v, 2, 1, true)
+GEN_LDST_WHOLE_TRANS(vs4r_v, 4, 1, true)
+GEN_LDST_WHOLE_TRANS(vs8r_v, 8, 1, true)
 
 /*
  *** Vector Integer Arithmetic Instructions
-- 
2.35.3



  parent reply	other threads:[~2022-05-24 22:51 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-24 22:44 [PULL 00/23] riscv-to-apply queue Alistair Francis
2022-05-24 22:44 ` [PULL 01/23] target/riscv: Fix VS mode hypervisor CSR access Alistair Francis
2022-05-24 22:44 ` Alistair Francis [this message]
2022-05-24 22:44 ` [PULL 03/23] hw/intc: Pass correct hartid while updating mtimecmp Alistair Francis
2022-05-24 22:44 ` [PULL 04/23] target/riscv: Move Zhinx* extensions on ISA string Alistair Francis
2022-05-24 22:44 ` [PULL 05/23] target/riscv: Add short-isa-string option Alistair Francis
2022-05-24 22:44 ` [PULL 06/23] hw/riscv: Make CPU config error handling generous (virt/spike) Alistair Francis
2022-05-24 22:44 ` [PULL 07/23] hw/riscv: Make CPU config error handling generous (sifive_e/u/opentitan) Alistair Francis
2022-05-24 22:44 ` [PULL 08/23] target/riscv: Fix coding style on "G" expansion Alistair Francis
2022-05-24 22:44 ` [PULL 09/23] target/riscv: Disable "G" by default Alistair Francis
2022-05-24 22:44 ` [PULL 10/23] target/riscv: Change "G" expansion Alistair Francis
2022-05-24 22:44 ` [PULL 11/23] target/riscv: FP extension requirements Alistair Francis
2022-05-24 22:44 ` [PULL 12/23] target/riscv: Move/refactor ISA extension checks Alistair Francis
2022-05-24 22:44 ` [PULL 13/23] hw/vfio/pci-quirks: Resolve redundant property getters Alistair Francis
2022-05-24 22:44 ` [PULL 14/23] hw/riscv/sifive_u: Resolve redundant property accessors Alistair Francis
2022-05-24 22:44 ` [PULL 15/23] target/riscv: check 'I' and 'E' after checking 'G' in riscv_cpu_realize Alistair Francis
2022-05-24 22:44 ` [PULL 16/23] target/riscv: Fix typo of mimpid cpu option Alistair Francis
2022-05-24 22:44 ` [PULL 17/23] target/riscv: Fix csr number based privilege checking Alistair Francis
2022-05-24 22:44 ` [PULL 18/23] target/riscv: Fix hstatus.GVA bit setting for traps taken from HS-mode Alistair Francis
2022-05-24 22:44 ` [PULL 19/23] target/riscv: Set [m|s]tval for both illegal and virtual instruction traps Alistair Francis
2022-05-24 22:44 ` [PULL 20/23] hw/riscv: virt: Fix interrupt parent for dynamic platform devices Alistair Francis
2022-05-24 22:44 ` [PULL 21/23] target/riscv: add zicsr/zifencei to isa_string Alistair Francis
2022-05-24 22:44 ` [PULL 22/23] hw/core: Sync uboot_image.h from U-Boot v2022.01 Alistair Francis
2022-05-24 22:44 ` [PULL 23/23] hw/core: loader: Set is_linux to true for VxWorks uImage Alistair Francis
2022-05-25  3:18 ` [PULL 00/23] riscv-to-apply queue Richard Henderson

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=20220524224428.552334-3-alistair.francis@opensource.wdc.com \
    --to=alistair.francis@opensource.wdc.com \
    --cc=alistair.francis@wdc.com \
    --cc=alistair23@gmail.com \
    --cc=eop.chen@sifive.com \
    --cc=frank.chang@sifive.com \
    --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).