From: Richard Henderson <richard.henderson@linaro.org>
To: qemu-devel@nongnu.org
Cc: peter.maydell@linaro.org, alex.bennee@linaro.org
Subject: [Qemu-devel] [PATCH 2/9] target/arm: Implement vector shifted FCVT for fp16
Date: Tue, 24 Apr 2018 15:22:53 -1000 [thread overview]
Message-ID: <20180425012300.14698-3-richard.henderson@linaro.org> (raw)
In-Reply-To: <20180425012300.14698-1-richard.henderson@linaro.org>
While we have some of the scalar paths for FCVT for fp16,
we failed to decode the fp16 version of these instructions.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
target/arm/translate-a64.c | 65 ++++++++++++++++++++++++++++++++--------------
1 file changed, 46 insertions(+), 19 deletions(-)
diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index c92e052686..e2d11998bd 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -7120,19 +7120,28 @@ static void handle_simd_shift_fpint_conv(DisasContext *s, bool is_scalar,
bool is_q, bool is_u,
int immh, int immb, int rn, int rd)
{
- bool is_double = extract32(immh, 3, 1);
int immhb = immh << 3 | immb;
- int fracbits = (is_double ? 128 : 64) - immhb;
- int pass;
+ int pass, size, fracbits;
TCGv_ptr tcg_fpstatus;
TCGv_i32 tcg_rmode, tcg_shift;
- if (!extract32(immh, 2, 2)) {
- unallocated_encoding(s);
- return;
- }
-
- if (!is_scalar && !is_q && is_double) {
+ if (immh & 0x8) {
+ size = MO_64;
+ if (!is_scalar && !is_q) {
+ unallocated_encoding(s);
+ return;
+ }
+ } else if (immh & 0x4) {
+ size = MO_32;
+ } else if (immh & 0x2) {
+ size = MO_16;
+ if (!arm_dc_feature(s, ARM_FEATURE_V8_FP16)) {
+ unallocated_encoding(s);
+ return;
+ }
+ } else {
+ /* Should have split out AdvSIMD modified immediate earlier. */
+ assert(immh == 1);
unallocated_encoding(s);
return;
}
@@ -7144,11 +7153,12 @@ static void handle_simd_shift_fpint_conv(DisasContext *s, bool is_scalar,
assert(!(is_scalar && is_q));
tcg_rmode = tcg_const_i32(arm_rmode_to_sf(FPROUNDING_ZERO));
- tcg_fpstatus = get_fpstatus_ptr(false);
+ tcg_fpstatus = get_fpstatus_ptr(size == MO_16);
gen_helper_set_rmode(tcg_rmode, tcg_rmode, tcg_fpstatus);
+ fracbits = (16 << size) - immhb;
tcg_shift = tcg_const_i32(fracbits);
- if (is_double) {
+ if (size == 3) {
int maxpass = is_scalar ? 1 : 2;
for (pass = 0; pass < maxpass; pass++) {
@@ -7165,20 +7175,37 @@ static void handle_simd_shift_fpint_conv(DisasContext *s, bool is_scalar,
}
clear_vec_high(s, is_q, rd);
} else {
- int maxpass = is_scalar ? 1 : is_q ? 4 : 2;
+ void (*fn)(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr);
+ int maxpass = is_scalar ? 1 : (8 << is_q >> size);
+
+ switch (size) {
+ case MO_16:
+ if (is_u) {
+ fn = gen_helper_vfp_toulh;
+ } else {
+ fn = gen_helper_vfp_toslh;
+ }
+ break;
+ case MO_32:
+ if (is_u) {
+ fn = gen_helper_vfp_touls;
+ } else {
+ fn = gen_helper_vfp_tosls;
+ }
+ break;
+ default:
+ g_assert_not_reached();
+ }
+
for (pass = 0; pass < maxpass; pass++) {
TCGv_i32 tcg_op = tcg_temp_new_i32();
- read_vec_element_i32(s, tcg_op, rn, pass, MO_32);
- if (is_u) {
- gen_helper_vfp_touls(tcg_op, tcg_op, tcg_shift, tcg_fpstatus);
- } else {
- gen_helper_vfp_tosls(tcg_op, tcg_op, tcg_shift, tcg_fpstatus);
- }
+ read_vec_element_i32(s, tcg_op, rn, pass, size);
+ fn(tcg_op, tcg_op, tcg_shift, tcg_fpstatus);
if (is_scalar) {
write_fp_sreg(s, rd, tcg_op);
} else {
- write_vec_element_i32(s, tcg_op, rd, pass, MO_32);
+ write_vec_element_i32(s, tcg_op, rd, pass, size);
}
tcg_temp_free_i32(tcg_op);
}
--
2.14.3
next prev parent reply other threads:[~2018-04-25 1:23 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-04-25 1:22 [Qemu-devel] [PATCH 0/9] target/arm: Fixups for ARM_FEATURE_V8_FP16 Richard Henderson
2018-04-25 1:22 ` [Qemu-devel] [PATCH 1/9] target/arm: Implement vector shifted SCVF/UCVF for fp16 Richard Henderson
2018-04-27 16:04 ` Alex Bennée
2018-04-29 14:44 ` Richard Henderson
2018-04-29 15:27 ` Peter Maydell
2018-04-25 1:22 ` Richard Henderson [this message]
2018-04-30 15:55 ` [Qemu-devel] [PATCH 2/9] target/arm: Implement vector shifted FCVT " Alex Bennée
2018-04-25 1:22 ` [Qemu-devel] [PATCH 3/9] target/arm: Fix float16 to/from int16 Richard Henderson
2018-05-01 10:10 ` Alex Bennée
2018-04-25 1:22 ` [Qemu-devel] [PATCH 4/9] target/arm: Clear SVE high bits for FMOV Richard Henderson
2018-05-01 10:44 ` Alex Bennée
2018-04-25 1:22 ` [Qemu-devel] [PATCH 5/9] target/arm: Implement FMOV (general) for fp16 Richard Henderson
2018-04-25 1:31 ` Philippe Mathieu-Daudé
2018-04-25 8:40 ` Richard Henderson
2018-04-25 1:22 ` [Qemu-devel] [PATCH 6/9] target/arm: Implement FCVT (scalar, integer) " Richard Henderson
2018-05-01 10:55 ` Alex Bennée
2018-04-25 1:22 ` [Qemu-devel] [PATCH 7/9] target/arm: Implement FCVT (scalar, fixed-point) " Richard Henderson
2018-05-01 10:57 ` Alex Bennée
2018-04-25 1:22 ` [Qemu-devel] [PATCH 8/9] target/arm: Implement FP data-processing (2 source) " Richard Henderson
2018-05-01 11:13 ` Alex Bennée
2018-05-02 18:28 ` Richard Henderson
2018-05-02 18:47 ` Richard Henderson
2018-04-25 1:23 ` [Qemu-devel] [PATCH 9/9] target/arm: Implement FP data-processing (3 " Richard Henderson
2018-05-01 11:21 ` Alex Bennée
2018-05-02 18:49 ` Richard Henderson
2018-04-25 1:35 ` [Qemu-devel] [PATCH 0/9] target/arm: Fixups for ARM_FEATURE_V8_FP16 no-reply
2018-04-25 9:14 ` Alex Bennée
2018-04-27 17:22 ` Alex Bennée
2018-04-27 18:55 ` Alex Bennée
2018-04-27 19:50 ` Alex Bennée
2018-05-11 2:17 ` Richard Henderson
2018-05-11 21:13 ` Alex Bennée
2018-05-01 15:47 ` Alex Bennée
2018-05-01 18:35 ` 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=20180425012300.14698-3-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).