DBT-RISE/DBT-RISE-TGC
12
0
Fork 0
Code Issues 1 Pull Requests 1 Projects Releases Wiki Activity

general improvements to vector_functions, adds functions to process arithmetic instructions (working add)

Browse Source
This commit is contained in:
Eyck-Alexander Jentzsch 2025-02-12 16:02:53 +01:00
parent 69c8fda5d2
commit 6ce0d97e81
4 changed files with 223 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

51
gen_input/templates/interp/CORENAME.cpp.gtl
View File

@ -164,7 +164,9 @@ if(vector != null) {%>
inline void set_tval(uint64_t new_tval){
tval = new_tval;
}
<%if(vector != null) {%>
<%if(vector != null) {
def xlen = constants.find { it.name == 'XLEN' }?.value ?: 0
def vlen = constants.find { it.name == 'VLEN' }?.value ?: 0 %>
inline void lower(){
this->core.reg.trap_state = 0;
}
@ -174,19 +176,54 @@ if(vector != null) {%>
uint64_t vsseg(uint8_t* V, uint8_t vs3, uint64_t rs1_val, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t elem_byte_size, uint16_t elem_count, int8_t EMUL_pow, uint8_t segment_size){
return softvector::vector_load_store(this->get_arch(), softvector::softvec_write, V, traits::VLEN, vs3, rs1_val, vl, vstart, vtype, vm, elem_byte_size, elem_count, EMUL_pow, segment_size);
}
uint64_t vlsseg(uint8_t* V, uint8_t vd, uint64_t rs1_val, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t elem_byte_size, uint16_t elem_count, int8_t EMUL_pow, uint8_t segment_size, int64_t stride){
uint64_t vlsseg(uint8_t* V, uint8_t vd, uint64_t rs1_val, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t elem_byte_size, uint16_t elem_count, int8_t EMUL_pow, uint8_t segment_size, int${xlen}_t stride){
return softvector::vector_load_store(this->get_arch(), softvector::softvec_read, V, traits::VLEN, vd, rs1_val, vl, vstart, vtype, vm, elem_byte_size, elem_count, EMUL_pow, segment_size, stride);
}
uint64_t vssseg(uint8_t* V, uint8_t vs3, uint64_t rs1_val, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t elem_byte_size, uint16_t elem_count, int8_t EMUL_pow, uint8_t segment_size, int64_t stride){
}
uint64_t vssseg(uint8_t* V, uint8_t vs3, uint64_t rs1_val, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t elem_byte_size, uint16_t elem_count, int8_t EMUL_pow, uint8_t segment_size, int${xlen}_t stride){
return softvector::vector_load_store(this->get_arch(), softvector::softvec_write, V, traits::VLEN, vs3, rs1_val, vl, vstart, vtype, vm, elem_byte_size, elem_count, EMUL_pow, segment_size, stride);
}
}
uint64_t vlxseg(uint8_t* V, uint8_t vd, uint8_t vs2, uint64_t rs1_val, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t elem_byte_size, uint16_t elem_count, uint8_t segment_size, bool ordered){
return softvector::vector_load_store_index(this->get_arch(), softvector::softvec_read, V, traits::VLEN, traits::XLEN, vd, vs2, rs1_val, vl, vstart, vtype, vm, elem_byte_size, elem_count, segment_size, ordered);
}
}
uint64_t vsxseg(uint8_t* V, uint8_t vs3, uint8_t vs2, uint64_t rs1_val, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t elem_byte_size, uint16_t elem_count, uint8_t segment_size, bool ordered){
return softvector::vector_load_store_index(this->get_arch(), softvector::softvec_write, V, traits::VLEN, traits::XLEN, vs3, vs2, rs1_val, vl, vstart, vtype, vm, elem_byte_size, elem_count, segment_size, ordered);
}
void vector_vector_op(uint8_t* V, uint8_t funct, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t vd, uint8_t vs2, uint8_t vs1, uint8_t sew_val){
switch(sew_val){
case 0b000:
softvector::vector_vector_op<${vlen}, uint8_t>(V, funct, vl, vstart, vtype, vm, vs2, vs1, vd);
break;
case 0b001:
softvector::vector_vector_op<${vlen}, uint16_t>(V, funct, vl, vstart, vtype, vm, vs2, vs1, vd);
break;
case 0b010:
softvector::vector_vector_op<${vlen}, uint32_t>(V, funct, vl, vstart, vtype, vm, vs2, vs1, vd);
break;
case 0b011:
softvector::vector_vector_op<${vlen}, uint64_t>(V, funct, vl, vstart, vtype, vm, vs2, vs1, vd);
break;
default:
throw new std::runtime_error("Unsupported sew bit value");
}
}
void vector_imm_op(uint8_t* V, uint8_t funct, uint64_t vl, uint64_t vstart, softvector::vtype_t vtype, bool vm, uint8_t vd, uint8_t vs2, int64_t imm, uint8_t sew_val){
switch(sew_val){
case 0b000:
softvector::vector_imm_op<${vlen}, uint8_t>(V, funct, vl, vstart, vtype, vm, vs2, imm, vd);
break;
case 0b001:
softvector::vector_imm_op<${vlen}, uint16_t>(V, funct, vl, vstart, vtype, vm, vs2, imm, vd);
break;
case 0b010:
softvector::vector_imm_op<${vlen}, uint32_t>(V, funct, vl, vstart, vtype, vm, vs2, imm, vd);
break;
case 0b011:
softvector::vector_imm_op<${vlen}, uint64_t>(V, funct, vl, vstart, vtype, vm, vs2, imm, vd);
break;
default:
throw new std::runtime_error("Unsupported sew bit value");
}
}
<%}%>
uint64_t fetch_count{0};
uint64_t tval{0};

32
src/vm/vector_functions.cpp
View File

@ -44,7 +44,6 @@
#include <stdexcept>
namespace softvector {
unsigned RFS = 32;
bool softvec_read(void* core, uint64_t addr, uint64_t length, uint8_t* data) {
// Read length bytes from addr into *data
@ -59,16 +58,6 @@ bool softvec_write(void* core, uint64_t addr, uint64_t length, uint8_t* data) {
return status == iss::Ok;
}
using vlen_t = uint64_t;
struct vreg_view {
uint8_t* start;
size_t size;
template <typename T> T& get(size_t idx = 0) {
assert((idx * sizeof(T)) <= size);
return *(reinterpret_cast<T*>(start) + idx);
}
};
vtype_t::vtype_t(uint32_t vtype_val) { underlying = (vtype_val & 0x8000) << 32 | (vtype_val & ~0x8000); }
vtype_t::vtype_t(uint64_t vtype_val) { underlying = vtype_val; }
bool vtype_t::vill() { return underlying >> 63; }
@ -85,10 +74,15 @@ double vtype_t::lmul() {
int8_t signed_vlmul = (vlmul >> 2) ? 0b11111000 | vlmul : vlmul;
return pow(2, signed_vlmul);
}
bool vmask_view::operator[](size_t idx) {
assert(idx < elem_count);
return *(start + idx / 8) & (1U << (idx % 8));
}
vreg_view read_vmask(uint8_t* V, uint16_t VLEN, uint16_t elem_count, uint8_t reg_idx) {
vmask_view read_vmask(uint8_t* V, uint16_t VLEN, uint16_t elem_count, uint8_t reg_idx) {
uint8_t* mask_start = V + VLEN / 8 * reg_idx;
return {mask_start, elem_count / 8u}; // this can return size==0 as elem_count can be as low as 1
assert(mask_start + elem_count / 8 <= V + VLEN * RFS / 8);
return {mask_start, elem_count};
}
uint64_t vector_load_store(void* core, std::function<bool(void*, uint64_t, uint64_t, uint8_t*)> load_store_fn, uint8_t* V, uint16_t VLEN,
uint8_t addressed_register, uint64_t base_addr, uint64_t vl, uint64_t vstart, vtype_t vtype, bool vm,
@ -104,13 +98,12 @@ uint64_t vector_load_store(void* core, std::function<bool(void*, uint64_t, uint6
assert(!(addressed_register % emul_stride));
if(!use_stride)
stride = elem_size_byte * segment_size;
vreg_view mask_view = read_vmask(V, VLEN, elem_count, 0);
vmask_view mask_reg = read_vmask(V, VLEN, elem_count);
// elements w/ index smaller than vstart are in the prestart and get skipped
// body is from vstart to min(elem_count, vl)
for(unsigned idx = vstart; idx < std::min(elem_count, vl); idx++) {
unsigned trap_idx = idx;
uint8_t current_mask_byte = mask_view.get<uint8_t>(idx / 8);
bool mask_active = vm ? 1 : current_mask_byte & (1 << idx % 8);
bool mask_active = vm ? 1 : mask_reg[idx];
if(mask_active) {
for(unsigned s_idx = 0; s_idx < segment_size; s_idx++) {
// base + selected register + current_elem + current_segment
@ -172,17 +165,16 @@ uint64_t vector_load_store_index(void* core, std::function<bool(void*, uint64_t,
unsigned data_emul_stride = vtype.lmul() < 1 ? 1 : vtype.lmul();
assert(data_emul_stride * segment_size <= 8);
unsigned data_elem_size_byte = vtype.sew() / 8;
vreg_view mask_view = read_vmask(V, VLEN, elem_count, 0);
vmask_view mask_reg = read_vmask(V, VLEN, elem_count);
// elements w/ index smaller than vstart are in the prestart and get skipped
// body is from vstart to min(elem_count, vl)
for(unsigned idx = vstart; idx < std::min(elem_count, vl); idx++) {
unsigned trap_idx = idx;
uint8_t current_mask_byte = mask_view.get<uint8_t>(idx / 8);
bool mask_active = vm ? 1 : current_mask_byte & (1 << idx % 8);
bool mask_active = vm ? 1 : mask_reg[idx];
if(mask_active) {
uint8_t* offset_elem = V + (index_register * VLEN / 8) + (index_elem_size_byte * idx);
assert(offset_elem <= (V + VLEN * RFS / 8 - index_elem_size_byte)); // ensure reading index_elem_size_bytes is legal
// read sew bits from offset_elem truncate / extend to XLEN bits
// read sew bits from offset_elem, truncate / extend to XLEN bits
uint64_t offset_val = read_n_bits(offset_elem, index_elem_size_byte * 8);
assert(XLEN == 64 | XLEN == 32);
uint64_t mask = XLEN == 64 ? std::numeric_limits<uint64_t>::max() : std::numeric_limits<uint32_t>::max();

13
src/vm/vector_functions.h
View File

@ -41,6 +41,7 @@
#include <functional>
#include <stdint.h>
namespace softvector {
const unsigned RFS = 32;
struct vtype_t {
uint64_t underlying;
@ -52,6 +53,14 @@ struct vtype_t {
bool vma();
bool vta();
};
struct vmask_view {
uint8_t* start;
size_t elem_count;
bool operator[](size_t);
};
vmask_view read_vmask(uint8_t* V, uint16_t VLEN, uint16_t elem_count, uint8_t reg_idx = 0);
template <unsigned VLEN> vmask_view read_vmask(uint8_t* V, uint16_t elem_count, uint8_t reg_idx = 0);
bool softvec_read(void* core, uint64_t addr, uint64_t length, uint8_t* data);
bool softvec_write(void* core, uint64_t addr, uint64_t length, uint8_t* data);
uint64_t vector_load_store(void* core, std::function<bool(void*, uint64_t, uint64_t, uint8_t*)> load_store_fn, uint8_t* V, uint16_t VLEN,
@ -62,5 +71,9 @@ uint64_t vector_load_store_index(void* core, std::function<bool(void*, uint64_t,
uint16_t VLEN, uint8_t XLEN, uint8_t addressed_register, uint8_t index_register, uint64_t base_addr,
uint64_t vl, uint64_t vstart, vtype_t vtype, bool vm, uint8_t elem_size_byte, uint64_t elem_count,
uint8_t segment_size, bool ordered);
template <unsigned VLEN, typename elem_t>
void vector_vector_op(uint8_t* V, unsigned funct6, uint64_t vl, uint64_t vstart, vtype_t vtype, bool vm, unsigned vs2, unsigned vs1,
unsigned vd);
} // namespace softvector
#include "vm/vector_functions.hpp"
#endif /* _VM_VECTOR_FUNCTIONS_H_ */

154
src/vm/vector_functions.hpp Normal file
View File

@ -0,0 +1,154 @@
////////////////////////////////////////////////////////////////////////////////
// Copyright (C) 2025, MINRES Technologies GmbH
// All rights reserved.
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are met:
//
// 1. Redistributions of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// 2. Redistributions in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// 3. Neither the name of the copyright holder nor the names of its contributors
// may be used to endorse or promote products derived from this software
// without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
// AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
// IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
// ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
// LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
// CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
// SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
// INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
// CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
// ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
// POSSIBILITY OF SUCH DAMAGE.
//
// Contributors:
// alex@minres.com - initial API and implementation
////////////////////////////////////////////////////////////////////////////////
#pragma once
#include "vm/vector_functions.h"
#include <functional>
#include <stdexcept>
#include <type_traits>
#ifndef _VM_VECTOR_FUNCTIONS_H_
#error __FILE__ should only be included from vector_functions.h
#endif
#include <math.h>
namespace softvector {
template <typename elem_t> struct vreg_view {
uint8_t* start;
size_t elem_count;
inline elem_t& get(size_t idx = 0) {
assert(idx < elem_count);
return *(reinterpret_cast<elem_t*>(start) + idx);
}
elem_t& operator[](size_t idx) {
assert(idx < elem_count);
return *(reinterpret_cast<elem_t*>(start) + idx);
}
};
template <unsigned VLEN, typename elem_t> vreg_view<elem_t> get_vreg(uint8_t* V, uint8_t reg_idx, uint16_t elem_count) {
assert(V + elem_count * sizeof(elem_t) <= V + VLEN * RFS / 8);
return {V + VLEN / 8 * reg_idx, elem_count};
}
template <unsigned VLEN> vmask_view read_vmask(uint8_t* V, uint16_t elem_count, uint8_t reg_idx) {
uint8_t* mask_start = V + VLEN / 8 * reg_idx;
assert(mask_start + elem_count / 8 <= V + VLEN * RFS / 8);
return {mask_start, elem_count};
}
template <typename elem_t> std::function<elem_t(elem_t, elem_t)> get_funct(unsigned funct) {
switch(funct) {
case 0b000000: {
// VADD
return [](elem_t vs2, elem_t vs1) { return vs2 + vs1; };
}
case 0b000010: {
// VSUB
}
case 0b000100: {
// VMINU
}
case 0b000101: {
// VMIN
}
case 0b000110: {
// VMAXU
}
case 0b000111: {
// VMAX
}
case 0b001001: {
// VAND
}
case 0b001010: {
// VOR
}
case 0b001011: {
// VXOR
}
default:
throw new std::runtime_error("Uknown funct in get_funct");
}
}
template <unsigned VLEN, typename elem_t>
void vector_vector_op(uint8_t* V, unsigned funct6, uint64_t vl, uint64_t vstart, vtype_t vtype, bool vm, unsigned vs2, unsigned vs1,
unsigned vd) {
uint64_t elem_count = VLEN * vtype.lmul() / vtype.sew();
vmask_view mask_reg = read_vmask<VLEN>(V, elem_count);
auto vs1_view = get_vreg<VLEN, elem_t>(V, vs1, elem_count);
auto vs2_view = get_vreg<VLEN, elem_t>(V, vs2, elem_count);
auto vd_view = get_vreg<VLEN, elem_t>(V, vd, elem_count);
auto fn = get_funct<elem_t>(funct6);
// elements w/ index smaller than vstart are in the prestart and get skipped
// body is from vstart to min(elem_count, vl)
for(unsigned idx = vstart; idx < std::min(elem_count, vl); idx++) {
bool mask_active = vm ? 1 : mask_reg[idx];
if(mask_active) {
vd_view[idx] = fn(vs2_view[idx], vs1_view[idx]);
} else {
vd_view[idx] = vtype.vma() ? vd_view[idx] : vd_view[idx];
}
}
// elements w/ index larger than elem_count are in the tail (fractional LMUL)
// elements w/ index larger than vl are in the tail
unsigned maximum_elems = VLEN * vtype.lmul() / (sizeof(elem_t) * 8);
for(unsigned idx = std::min(elem_count, vl); idx < maximum_elems; idx++) {
vd_view[idx] = vtype.vta() ? vd_view[idx] : vd_view[idx];
}
return;
}
template <unsigned VLEN, typename elem_t>
void vector_imm_op(uint8_t* V, unsigned funct6, uint64_t vl, uint64_t vstart, vtype_t vtype, bool vm, unsigned vs2,
typename std::make_signed<elem_t>::type imm, unsigned vd) {
uint64_t elem_count = VLEN * vtype.lmul() / vtype.sew();
vmask_view mask_reg = read_vmask<VLEN>(V, elem_count);
auto vs2_view = get_vreg<VLEN, elem_t>(V, vs2, elem_count);
auto vd_view = get_vreg<VLEN, elem_t>(V, vd, elem_count);
auto fn = get_funct<elem_t>(funct6);
// elements w/ index smaller than vstart are in the prestart and get skipped
// body is from vstart to min(elem_count, vl)
for(unsigned idx = vstart; idx < std::min(elem_count, vl); idx++) {
bool mask_active = vm ? 1 : mask_reg[idx];
if(mask_active) {
vd_view[idx] = fn(vs2_view[idx], imm);
} else {
vd_view[idx] = vtype.vma() ? vd_view[idx] : vd_view[idx];
}
}
// elements w/ index larger than elem_count are in the tail (fractional LMUL)
// elements w/ index larger than vl are in the tail
unsigned maximum_elems = VLEN * vtype.lmul() / (sizeof(elem_t) * 8);
for(unsigned idx = std::min(elem_count, vl); idx < maximum_elems; idx++) {
vd_view[idx] = vtype.vta() ? vd_view[idx] : vd_view[idx];
}
return;
}
} // namespace softvector