From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: Richard Henderson <richard.henderson@linaro.org>
Subject: [PULL 33/53] target/i386: implement additional AVX comparison operators
Date: Tue, 18 Oct 2022 15:30:22 +0200 [thread overview]
Message-ID: <20221018133042.856368-34-pbonzini@redhat.com> (raw)
In-Reply-To: <20221018133042.856368-1-pbonzini@redhat.com>
The new implementation of SSE will cover AVX from the get go, so include
the 24 extra comparison operators that are only available with the VEX
prefix.
Based on a patch by Paul Brook <paul@nowt.org>.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
target/i386/ops_sse.h | 38 ++++++++++++++++++++++++++++++++++++
| 27 +++++++++++++++++++++++++
2 files changed, 65 insertions(+)
diff --git a/target/i386/ops_sse.h b/target/i386/ops_sse.h
index 8bb7293975..18d217ebf5 100644
--- a/target/i386/ops_sse.h
+++ b/target/i386/ops_sse.h
@@ -1075,10 +1075,21 @@ void glue(helper_addsubpd, SUFFIX)(CPUX86State *env, Reg *d, Reg *v, Reg *s)
} \
}
+static inline bool FPU_EQU(FloatRelation x)
+{
+ return (x == float_relation_equal || x == float_relation_unordered);
+}
+static inline bool FPU_GE(FloatRelation x)
+{
+ return (x == float_relation_equal || x == float_relation_greater);
+}
#define FPU_EQ(x) (x == float_relation_equal)
#define FPU_LT(x) (x == float_relation_less)
#define FPU_LE(x) (x <= float_relation_equal)
+#define FPU_GT(x) (x == float_relation_greater)
#define FPU_UNORD(x) (x == float_relation_unordered)
+/* We must make sure we evaluate the argument in case it is a signalling NAN */
+#define FPU_FALSE(x) (x == float_relation_equal && 0)
#define FPU_CMPQ(size, a, b) \
float ## size ## _compare_quiet(a, b, &env->sse_status)
@@ -1098,6 +1109,33 @@ SSE_HELPER_CMP(cmpnlt, FPU_CMPS, !FPU_LT)
SSE_HELPER_CMP(cmpnle, FPU_CMPS, !FPU_LE)
SSE_HELPER_CMP(cmpord, FPU_CMPQ, !FPU_UNORD)
+SSE_HELPER_CMP(cmpequ, FPU_CMPQ, FPU_EQU)
+SSE_HELPER_CMP(cmpnge, FPU_CMPS, !FPU_GE)
+SSE_HELPER_CMP(cmpngt, FPU_CMPS, !FPU_GT)
+SSE_HELPER_CMP(cmpfalse, FPU_CMPQ, FPU_FALSE)
+SSE_HELPER_CMP(cmpnequ, FPU_CMPQ, !FPU_EQU)
+SSE_HELPER_CMP(cmpge, FPU_CMPS, FPU_GE)
+SSE_HELPER_CMP(cmpgt, FPU_CMPS, FPU_GT)
+SSE_HELPER_CMP(cmptrue, FPU_CMPQ, !FPU_FALSE)
+
+SSE_HELPER_CMP(cmpeqs, FPU_CMPS, FPU_EQ)
+SSE_HELPER_CMP(cmpltq, FPU_CMPQ, FPU_LT)
+SSE_HELPER_CMP(cmpleq, FPU_CMPQ, FPU_LE)
+SSE_HELPER_CMP(cmpunords, FPU_CMPS, FPU_UNORD)
+SSE_HELPER_CMP(cmpneqq, FPU_CMPS, !FPU_EQ)
+SSE_HELPER_CMP(cmpnltq, FPU_CMPQ, !FPU_LT)
+SSE_HELPER_CMP(cmpnleq, FPU_CMPQ, !FPU_LE)
+SSE_HELPER_CMP(cmpords, FPU_CMPS, !FPU_UNORD)
+
+SSE_HELPER_CMP(cmpequs, FPU_CMPS, FPU_EQU)
+SSE_HELPER_CMP(cmpngeq, FPU_CMPQ, !FPU_GE)
+SSE_HELPER_CMP(cmpngtq, FPU_CMPQ, !FPU_GT)
+SSE_HELPER_CMP(cmpfalses, FPU_CMPS, FPU_FALSE)
+SSE_HELPER_CMP(cmpnequs, FPU_CMPS, !FPU_EQU)
+SSE_HELPER_CMP(cmpgeq, FPU_CMPQ, FPU_GE)
+SSE_HELPER_CMP(cmpgtq, FPU_CMPQ, FPU_GT)
+SSE_HELPER_CMP(cmptrues, FPU_CMPS, !FPU_FALSE)
+
#undef SSE_HELPER_CMP
#if SHIFT == 1
--git a/target/i386/ops_sse_header.h b/target/i386/ops_sse_header.h
index 32ffa8445b..e7866a8395 100644
--- a/target/i386/ops_sse_header.h
+++ b/target/i386/ops_sse_header.h
@@ -237,6 +237,33 @@ SSE_HELPER_CMP(cmpnlt, FPU_CMPS, !FPU_LT)
SSE_HELPER_CMP(cmpnle, FPU_CMPS, !FPU_LE)
SSE_HELPER_CMP(cmpord, FPU_CMPQ, !FPU_UNORD)
+SSE_HELPER_CMP(cmpequ, FPU_CMPQ, FPU_EQU)
+SSE_HELPER_CMP(cmpnge, FPU_CMPS, !FPU_GE)
+SSE_HELPER_CMP(cmpngt, FPU_CMPS, !FPU_GT)
+SSE_HELPER_CMP(cmpfalse, FPU_CMPQ, FPU_FALSE)
+SSE_HELPER_CMP(cmpnequ, FPU_CMPQ, !FPU_EQU)
+SSE_HELPER_CMP(cmpge, FPU_CMPS, FPU_GE)
+SSE_HELPER_CMP(cmpgt, FPU_CMPS, FPU_GT)
+SSE_HELPER_CMP(cmptrue, FPU_CMPQ, !FPU_FALSE)
+
+SSE_HELPER_CMP(cmpeqs, FPU_CMPS, FPU_EQ)
+SSE_HELPER_CMP(cmpltq, FPU_CMPQ, FPU_LT)
+SSE_HELPER_CMP(cmpleq, FPU_CMPQ, FPU_LE)
+SSE_HELPER_CMP(cmpunords, FPU_CMPS, FPU_UNORD)
+SSE_HELPER_CMP(cmpneqq, FPU_CMPS, !FPU_EQ)
+SSE_HELPER_CMP(cmpnltq, FPU_CMPQ, !FPU_LT)
+SSE_HELPER_CMP(cmpnleq, FPU_CMPQ, !FPU_LE)
+SSE_HELPER_CMP(cmpords, FPU_CMPS, !FPU_UNORD)
+
+SSE_HELPER_CMP(cmpequs, FPU_CMPS, FPU_EQU)
+SSE_HELPER_CMP(cmpngeq, FPU_CMPQ, !FPU_GE)
+SSE_HELPER_CMP(cmpngtq, FPU_CMPQ, !FPU_GT)
+SSE_HELPER_CMP(cmpfalses, FPU_CMPS, FPU_FALSE)
+SSE_HELPER_CMP(cmpnequs, FPU_CMPS, !FPU_EQU)
+SSE_HELPER_CMP(cmpgeq, FPU_CMPQ, FPU_GE)
+SSE_HELPER_CMP(cmpgtq, FPU_CMPQ, FPU_GT)
+SSE_HELPER_CMP(cmptrues, FPU_CMPS, !FPU_FALSE)
+
#if SHIFT == 1
DEF_HELPER_3(ucomiss, void, env, Reg, Reg)
DEF_HELPER_3(comiss, void, env, Reg, Reg)
--
2.37.3
next prev parent reply other threads:[~2022-10-18 13:54 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-10-18 13:29 [PULL 00/53] target/i386, scsi, build patches for 2022-10-18 Paolo Bonzini
2022-10-18 13:29 ` [PULL 01/53] virtio-scsi: Send "REPORTED LUNS CHANGED" sense data upon disk hotplug events Paolo Bonzini
2022-10-18 13:29 ` [PULL 02/53] configure: don't enable firmware for targets that are not built Paolo Bonzini
2022-10-18 13:29 ` [PULL 03/53] scsi: Use device_cold_reset() and bus_cold_reset() Paolo Bonzini
2022-10-18 13:29 ` [PULL 04/53] hw/scsi/vmw_pvscsi.c: Use device_cold_reset() to reset SCSI devices Paolo Bonzini
2022-10-18 13:29 ` [PULL 05/53] hyperv: fix SynIC SINT assertion failure on guest reset Paolo Bonzini
2022-10-18 13:29 ` [PULL 06/53] configure: Avoid using strings binary Paolo Bonzini
2022-10-18 13:29 ` [PULL 07/53] target/i386: Use device_cold_reset() to reset the APIC Paolo Bonzini
2022-10-18 13:29 ` [PULL 08/53] target/i386: Save and restore pc_save before tcg_remove_ops_after Paolo Bonzini
2022-10-18 13:29 ` [PULL 09/53] target/i386: Use MMUAccessType across excp_helper.c Paolo Bonzini
2022-10-18 13:29 ` [PULL 10/53] target/i386: Direct call get_hphys from mmu_translate Paolo Bonzini
2022-10-18 13:30 ` [PULL 11/53] target/i386: Introduce structures for mmu_translate Paolo Bonzini
2022-10-18 13:30 ` [PULL 12/53] target/i386: Reorg GET_HPHYS Paolo Bonzini
2022-10-18 13:30 ` [PULL 13/53] target/i386: Add MMU_PHYS_IDX and MMU_NESTED_IDX Paolo Bonzini
2022-10-18 13:30 ` [PULL 14/53] target/i386: Use MMU_NESTED_IDX for vmload/vmsave Paolo Bonzini
2022-10-18 13:30 ` [PULL 15/53] target/i386: Combine 5 sets of variables in mmu_translate Paolo Bonzini
2022-10-18 13:30 ` [PULL 16/53] target/i386: Use atomic operations for pte updates Paolo Bonzini
2022-10-18 13:30 ` [PULL 17/53] target/i386: Use probe_access_full for final stage2 translation Paolo Bonzini
2022-10-18 13:30 ` [PULL 18/53] target/i386: Define XMMReg and access macros, align ZMM registers Paolo Bonzini
2022-10-18 13:30 ` [PULL 19/53] target/i386: make ldo/sto operations consistent with ldq Paolo Bonzini
2022-10-18 13:30 ` [PULL 20/53] target/i386: make rex_w available even in 32-bit mode Paolo Bonzini
2022-10-18 13:30 ` [PULL 21/53] target/i386: add core of new i386 decoder Paolo Bonzini
2022-10-18 13:30 ` [PULL 22/53] target/i386: add ALU load/writeback core Paolo Bonzini
2022-10-18 13:30 ` [PULL 23/53] target/i386: add CPUID[EAX=7,ECX=0].ECX to DisasContext Paolo Bonzini
2022-10-18 13:30 ` [PULL 24/53] target/i386: add CPUID feature checks to new decoder Paolo Bonzini
2022-10-18 13:30 ` [PULL 25/53] target/i386: add AVX_EN hflag Paolo Bonzini
2022-10-18 13:30 ` [PULL 26/53] target/i386: validate VEX prefixes via the instructions' exception classes Paolo Bonzini
2022-10-18 13:30 ` [PULL 27/53] target/i386: validate SSE prefixes directly in the decoding table Paolo Bonzini
2022-10-18 13:30 ` [PULL 28/53] target/i386: move scalar 0F 38 and 0F 3A instruction to new decoder Paolo Bonzini
2022-10-18 13:30 ` [PULL 29/53] target/i386: Prepare ops_sse_header.h for 256 bit AVX Paolo Bonzini
2022-10-18 13:30 ` [PULL 30/53] target/i386: extend helpers to support VEX.V 3- and 4- operand encodings Paolo Bonzini
2022-10-18 13:30 ` [PULL 31/53] target/i386: support operand merging in binary scalar helpers Paolo Bonzini
2022-10-18 13:30 ` [PULL 32/53] target/i386: provide 3-operand versions of unary " Paolo Bonzini
2022-10-18 13:30 ` Paolo Bonzini [this message]
2022-10-18 13:30 ` [PULL 34/53] target/i386: Introduce 256-bit vector helpers Paolo Bonzini
2022-10-18 13:30 ` [PULL 35/53] target/i386: reimplement 0x0f 0x60-0x6f, add AVX Paolo Bonzini
2022-10-18 13:30 ` [PULL 36/53] target/i386: reimplement 0x0f 0xd8-0xdf, 0xe8-0xef, 0xf8-0xff, " Paolo Bonzini
2022-10-18 13:30 ` [PULL 37/53] target/i386: reimplement 0x0f 0x50-0x5f, " Paolo Bonzini
2022-10-18 13:30 ` [PULL 38/53] target/i386: reimplement 0x0f 0x78-0x7f, " Paolo Bonzini
2022-10-18 13:30 ` [PULL 39/53] target/i386: reimplement 0x0f 0x70-0x77, " Paolo Bonzini
2022-10-18 13:30 ` [PULL 40/53] target/i386: reimplement 0x0f 0xd0-0xd7, 0xe0-0xe7, 0xf0-0xf7, " Paolo Bonzini
2022-10-18 13:30 ` [PULL 41/53] target/i386: clarify (un)signedness of immediates from 0F3Ah opcodes Paolo Bonzini
2022-10-18 13:30 ` [PULL 42/53] target/i386: reimplement 0x0f 0x3a, add AVX Paolo Bonzini
2022-10-18 13:30 ` [PULL 43/53] target/i386: Use tcg gvec ops for pmovmskb Paolo Bonzini
2022-10-18 13:30 ` [PULL 44/53] target/i386: reimplement 0x0f 0x38, add AVX Paolo Bonzini
2022-10-18 13:30 ` [PULL 45/53] target/i386: reimplement 0x0f 0xc2, 0xc4-0xc6, " Paolo Bonzini
2022-10-18 13:30 ` [PULL 46/53] target/i386: reimplement 0x0f 0x10-0x17, " Paolo Bonzini
2022-10-18 13:30 ` [PULL 47/53] target/i386: reimplement 0x0f 0x28-0x2f, " Paolo Bonzini
2023-05-09 10:52 ` Peter Maydell
2022-10-18 13:30 ` [PULL 48/53] target/i386: implement XSAVE and XRSTOR of AVX registers Paolo Bonzini
2022-10-18 13:30 ` [PULL 49/53] target/i386: implement VLDMXCSR/VSTMXCSR Paolo Bonzini
2022-10-18 13:30 ` [PULL 50/53] target/i386: Enable AVX cpuid bits when using TCG Paolo Bonzini
2022-10-18 13:30 ` [PULL 51/53] tests/tcg: extend SSE tests to AVX Paolo Bonzini
2022-10-18 13:30 ` [PULL 52/53] target/i386: move 3DNow to the new decoder Paolo Bonzini
2022-10-18 13:30 ` [PULL 53/53] target/i386: remove old SSE decoder Paolo Bonzini
2022-10-18 20:01 ` [PULL 00/53] target/i386, scsi, build patches for 2022-10-18 Stefan Hajnoczi
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=20221018133042.856368-34-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=richard.henderson@linaro.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).