1
0
mirror of https://github.com/Yours3lf/rpi-vk-driver.git synced 2024-11-29 11:24:14 +01:00

assembler now should be fully functional

This commit is contained in:
Unknown 2019-04-14 19:35:21 +01:00
parent 1b402b99bf
commit 061d52907f

View File

@ -301,7 +301,7 @@ uint64_t encode_load_imm(uint8_t pack_unpack_select,
//write per element MS bit and LS bit across simd array //write per element MS bit and LS bit across simd array
uint64_t encode_load_imm_per_elem( uint64_t encode_load_imm_per_elem(
uint8_t signed_or_unsigned, //0 for signed, 1 for unsigned uint8_t signed_or_unsigned, //1 for signed, 0 for unsigned
uint8_t pack_unpack_select, uint8_t pack_unpack_select,
uint8_t pack_mode, uint8_t pack_mode,
qpu_cond cond_add, qpu_cond cond_add,
@ -318,7 +318,7 @@ uint64_t encode_load_imm_per_elem(
uint64_t tmp = 0; uint64_t tmp = 0;
tmp = 0x71; tmp = 0x71;
tmp |= signed_or_unsigned << 1; tmp |= !signed_or_unsigned << 1;
res |= tmp << 57; res |= tmp << 57;
tmp = pack_unpack_select & 1; tmp = pack_unpack_select & 1;
@ -353,23 +353,6 @@ uint64_t encode_load_imm_per_elem(
return res; return res;
} }
/*
Format:
#comment
sig_bit_optional ; dstAdd.pack_mode_optional = add_opcode.sf_optional.condition.unpack_mode_optional(srcA, srcB, imm_optional, raddr_a_optional, raddr_b_optional) ; dstMul.pack_mode_optional = mul_opcode.condition(srcA, srcB) ;
sig_bit_branch ; dstAdd = branch.rel_optional.reg_optional.condition(address, srcA_optional) ; dstMul = branch() ;
sig_bit_none ; dstAdd.pack_mode_optional = sem_inc.sf_optional.condition(sem_number, 27bit_imm_value_optional) ; dstMul.pack_mode_optional = sem_inc.condition() ;
sig_load_imm ; dstAdd.pack_mode_optional = load32.sf_optional.condition(immediate_value) ; dstMul.pack_mode_optional = load32.condition() ;
sig_load_imm ; dstAdd.pack_mode_optional = load16.signed_optional.sf_optional.condition(int16_imm, in16_imm) ; dstMul.pack_mode_optional = load16.condition() ;
Examples:
sig_none ; ra0.nop = add.sf.always(r0, r1, 0) ; rb0.nop = fmul.sf.always(r2, r3) ;
sig_branch ; ra0 = branch.rel.reg.always(0xdeadbeef, ra1) ; rb0 = branch() ;
sig_none ; ra0.nop = sem_inc.sf.always(1, 0x7ffffff) ; rb0.nop = sem_inc.always() ;
sig_load_imm ; ra0.nop = load32.sf.always(0xdeadbeef) ; rb0.nop = load32.always() ;
sig_load_imm ; ra0.nop = load16.sf.signed.always(1, 2) ; rb0.nop = load16.always() ;
*/
qpu_sig_bits parse_sig_bit(char* str) qpu_sig_bits parse_sig_bit(char* str)
{ {
unsigned num_sig_bits = sizeof(qpu_sig_bits_str) / sizeof(const char *); unsigned num_sig_bits = sizeof(qpu_sig_bits_str) / sizeof(const char *);
@ -459,17 +442,24 @@ void parse_dst(char** str, qpu_waddr* waddr, uint8_t* pack_mode, uint8_t* ws, un
*pack_mode = pack_mode_res; *pack_mode = pack_mode_res;
} }
void parse_op_modifiers(char** str, uint8_t* sf, qpu_cond* condition, qpu_unpack* unpack_mode, uint8_t* rel, uint8_t* reg) void parse_op_modifiers(char** str, uint8_t* signed_or_unsigned, uint8_t* pm, uint8_t* sf, qpu_cond* condition, qpu_unpack* unpack_mode, uint8_t* rel, uint8_t* reg)
{ {
char* modifier = strtok(*str, "."); char* modifier = strtok(*str, ".");
//at most 3 modifiers supported //at most 4 modifiers supported
for(int c = 0; c < 3; ++c) for(int c = 0; c < 4; ++c)
{ {
if(modifier) if(modifier)
{ {
*str = modifier; *str = modifier;
if(strcmp(modifier, "pm") == 0)
{
*pm = 1;
modifier = strtok(0, ".");
continue;
}
if(strcmp(modifier, "rel") == 0) if(strcmp(modifier, "rel") == 0)
{ {
*rel = 1; *rel = 1;
@ -491,6 +481,13 @@ void parse_op_modifiers(char** str, uint8_t* sf, qpu_cond* condition, qpu_unpack
continue; continue;
} }
if(strcmp(modifier, "signed") == 0)
{
*signed_or_unsigned = 1;
modifier = strtok(0, ".");
continue;
}
unsigned found = 0; unsigned found = 0;
unsigned num_conds = sizeof(qpu_cond_str) / sizeof(const char *); unsigned num_conds = sizeof(qpu_cond_str) / sizeof(const char *);
@ -730,7 +727,7 @@ void parse_args_sem(char** str, uint8_t* sem, uint32_t* imm32)
*str += 1; *str += 1;
} }
void parse_args_branch(char** str, uint32_t* imm32, uint8_t* raddr_a) void parse_args_branch(char** str, uint32_t* imm32, qpu_branch_cond* branch_cond, uint8_t* raddr_a)
{ {
char* arg = strtok(*str, " \n\v\f\r\t,"); char* arg = strtok(*str, " \n\v\f\r\t,");
@ -742,6 +739,23 @@ void parse_args_branch(char** str, uint32_t* imm32, uint8_t* raddr_a)
arg = strtok(0, " \n\v\f\r\t,"); arg = strtok(0, " \n\v\f\r\t,");
if(arg)
{
unsigned num_branch_conds = sizeof(qpu_branch_cond_str) / sizeof(const char *);
for(unsigned c = 0; c < num_branch_conds && arg; ++c)
{
if(qpu_branch_cond_str[c] && strcmp(arg, qpu_branch_cond_str[c]) == 0)
{
*branch_cond = c;
*str = arg;
break;
}
}
}
arg = strtok(0, " \n\v\f\r\t,");
if(arg) if(arg)
{ {
uint8_t raddr_a_res = 0; uint8_t raddr_a_res = 0;
@ -815,34 +829,26 @@ void parse_args_load(char** str, qpu_load_type load_type, uint32_t* imm32, uint1
} }
uint64_t* assemble_qpu_asm(char* str) void assemble_qpu_asm(char* str, uint64_t* instructions)
{ {
unsigned num_instructions = 0;
char* ptr = str;
while(ptr && *ptr != '\0')
{
ptr = strstr(ptr, ";");
ptr = strstr(ptr+(ptr!=0), ";");
ptr = strstr(ptr+(ptr!=0), ";");
if(ptr)
{
ptr += 1;
num_instructions += 1;
}
}
printf("Num instructions: %i\n", num_instructions);
if(!num_instructions)
{
return 0;
}
uint64_t* instructions = malloc(sizeof(uint64_t)*num_instructions);
unsigned instruction_counter = 0; unsigned instruction_counter = 0;
//delete lines that have comments in them
char* comment_token = strstr(str, "#");
while(comment_token)
{
while(*comment_token != '\n')
{
*comment_token = ' ';
comment_token++;
}
*comment_token = ' ';
comment_token = strstr(comment_token, "#");
}
//parse string token by token
char* token = strtok(str, " \n\v\f\r\t;"); char* token = strtok(str, " \n\v\f\r\t;");
while(token) while(token)
@ -874,7 +880,7 @@ uint64_t* assemble_qpu_asm(char* str)
uint16_t ls_imm16 = 0; uint16_t ls_imm16 = 0;
uint8_t semaphore = 0; uint8_t semaphore = 0;
qpu_load_type load_type = QPU_LOAD32; qpu_load_type load_type = QPU_LOAD32;
uint8_t signed_or_unsigned = 0; uint8_t is_signed = 0;
qpu_branch_cond branch_cond = QPU_COND_BRANCH_ALWAYS; qpu_branch_cond branch_cond = QPU_COND_BRANCH_ALWAYS;
sig_bit = parse_sig_bit(token); sig_bit = parse_sig_bit(token);
@ -893,7 +899,7 @@ uint64_t* assemble_qpu_asm(char* str)
//get modifiers //get modifiers
token = strtok(token, " \n\v\f\r\t("); token = strtok(token, " \n\v\f\r\t(");
parse_op_modifiers(&token, &sf, &cond_add, &unpack_mode, &rel, &reg); parse_op_modifiers(&token, &is_signed, &pack_unpack_select, &sf, &cond_add, &unpack_mode, &rel, &reg);
if(type == QPU_ALU) if(type == QPU_ALU)
{ {
@ -911,7 +917,7 @@ uint64_t* assemble_qpu_asm(char* str)
{ {
//get arguments for branch //get arguments for branch
token = strtok(token, ")"); token = strtok(token, ")");
parse_args_branch(&token, &imm32, &raddr_a); parse_args_branch(&token, &imm32, &branch_cond, &raddr_a);
} }
else if(type == QPU_LOAD_IMM) else if(type == QPU_LOAD_IMM)
{ {
@ -930,7 +936,7 @@ uint64_t* assemble_qpu_asm(char* str)
//get modifiers //get modifiers
token = strtok(token, " \n\v\f\r\t("); token = strtok(token, " \n\v\f\r\t(");
parse_op_modifiers(&token, &sf, &cond_mul, &unpack_mode, &rel, &reg); parse_op_modifiers(&token, &is_signed, &pack_unpack_select, &sf, &cond_mul, &unpack_mode, &rel, &reg);
token = strtok(token, ")"); token = strtok(token, ")");
@ -968,26 +974,99 @@ uint64_t* assemble_qpu_asm(char* str)
} }
else else
{ {
instructions[instruction_counter] = encode_load_imm_per_elem(signed_or_unsigned, pack_unpack_select, pack_mode, cond_add, cond_mul, sf, ws, waddr_add, waddr_mul, ms_imm16, ls_imm16); instructions[instruction_counter] = encode_load_imm_per_elem(is_signed, pack_unpack_select, pack_mode, cond_add, cond_mul, sf, ws, waddr_add, waddr_mul, ms_imm16, ls_imm16);
} }
} }
instruction_counter++; instruction_counter++;
token = strtok(token, " \n\v\f\r\t;"); token = strtok(token, " \n\v\f\r\t;");
} }
return instructions;
} }
void disassemble_qpu_asm(uint64_t instruction)
{
}
/*
Format:
#comment
sig_bit_optional ; dstAdd.pack_mode_optional = add_opcode.pm_optional.sf_optional.condition.unpack_mode_optional(srcA, srcB, imm_optional, raddr_a_optional, raddr_b_optional) ; dstMul.pack_mode_optional = mul_opcode.condition(srcA, srcB) ;
sig_bit_branch ; dstAdd = branch.pm_optional.rel_optional.reg_optional(address, condition, srcA_optional) ; dstMul = branch() ;
sig_bit_none ; dstAdd.pack_mode_optional = sem_inc.pm_optional.sf_optional.condition(sem_number, 27bit_imm_value_optional) ; dstMul.pack_mode_optional = sem_inc.condition() ;
sig_load_imm ; dstAdd.pack_mode_optional = load32.pm_optional.sf_optional.condition(immediate_value) ; dstMul.pack_mode_optional = load32.condition() ;
sig_load_imm ; dstAdd.pack_mode_optional = load16.pm_optional.signed_optional.sf_optional.condition(int16_imm, in16_imm) ; dstMul.pack_mode_optional = load16.condition() ;
Examples:
sig_none ; ra0.nop = add.pm.sf.always(r0, r1, 0) ; rb0.nop = fmul.sf.always(r2, r3) ;
sig_branch ; ra0 = branch.pm.rel.reg.always(0xdeadbeef, ra1) ; rb0 = branch() ;
sig_none ; ra0.nop = sem_inc.pm.sf.always(1, 0x7ffffff) ; rb0.nop = sem_inc.always() ;
sig_load_imm ; ra0.nop = load32.pm.sf.always(0xdeadbeef) ; rb0.nop = load32.always() ;
sig_load_imm ; ra0.nop = load16.pm.sf.signed.always(1, 2) ; rb0.nop = load16.always() ;
#mov
sig_none ; ra0.nop = or(r0, r0) ; rb0 = v8min(r1, r1) ;
*/
int main() int main()
{ {
char asm_code[] = char asm_code[] =
"sig_none ; ra0.nop = add.sf.always.nop(r0, r1, 0) ; rb0.nop = fmul.sf.always(r2, r3) ;" "sig_none ; ra0.nop = add.sf.always.nop(r0, r1, 0) ; rb0.nop = fmul.sf.always(r2, r3) ;"
"sig_branch ; ra0 = branch.rel.reg.always(0xdeadbeef, ra1) ; rb0 = branch() ;" "sig_branch ; ra0 = branch.rel.reg(0xdeadbeef, always, ra1) ; rb0 = branch() ;"
"#hello\n"
"sig_none ; ra0.nop = sem_inc.sf.always(1, 0x7ffffff) ; rb0.nop = sem_inc.always() ;" "sig_none ; ra0.nop = sem_inc.sf.always(1, 0x7ffffff) ; rb0.nop = sem_inc.always() ;"
"sig_load_imm ; ra0.nop = load32.sf.always(0xdeadbeef) ; rb0.nop = load32.always() ;"; "#hello2\n"
"sig_load_imm ; ra0.nop = load32.sf.always(0xdeadbeef) ; rb0.nop = load32.always() ;"
"sig_load_imm ; ra0.nop = load16.sf.signed.always(0xdead, 0xbeef) ; rb0.nop = load16.always() ;";
uint64_t* assembly = assemble_qpu_asm(asm_code); unsigned num_instructions = 0;
char* ptr = asm_code;
while(ptr && *ptr != '\0')
{
ptr = strstr(ptr, ";");
ptr = strstr(ptr+(ptr!=0), ";");
ptr = strstr(ptr+(ptr!=0), ";");
if(ptr)
{
ptr += 1;
num_instructions += 1;
}
}
printf("Num instructions: %i\n", num_instructions);
if(!num_instructions)
{
return 0;
}
uint64_t* instruction_size = sizeof(uint64_t)*num_instructions;
uint64_t* instructions = malloc(instruction_size);
assemble_qpu_asm(asm_code, instructions);
for(int c = 0; c < instruction_size; ++c)
{
unsigned char d = ((char*)instructions)[c];
printf("%#x,\t", d);
if((c+1)%8==0)
{
printf("\n");
}
}
const char asm_instructions[] =
{
0x53, 0x70, 0x9e, 0x2c, 0, 0x60, 0x2, 0x10,
0xef, 0xbe, 0xad, 0xde, 0, 0x20, 0xfc, 0xf0,
0xf1, 0xff, 0xff, 0xff, 0, 0x60, 0x2, 0xe8,
0xef, 0xbe, 0xad, 0xde, 0, 0x60, 0x2, 0xe0,
0xef, 0xbe, 0xad, 0xde, 0, 0x60, 0x2, 0xe2
};
for(int c = 0; c < num_instructions; ++c)
{
disassemble_qpu_asm(asm_instructions);
}
return 0; return 0;
} }