tests: Migrate to newlib-nano and libgloss-htif

This deprecates the monolithic syscalls.c environment in favor of a more
structured board support package (BSP) based on newlib.
This commit is contained in:
Albert Ou
2019-12-13 15:03:20 -08:00
parent 7059ac3f0f
commit f60e1a9c89
10 changed files with 66 additions and 2225 deletions

1
tests/.gitignore vendored
View File

@@ -1,3 +1,4 @@
*.o
*.riscv
*.dump
libgloss/

View File

@@ -1,12 +1,18 @@
GCC=riscv64-unknown-elf-gcc
OBJDUMP=riscv64-unknown-elf-objdump
CFLAGS=-mcmodel=medany -std=gnu99 -O2 -fno-common -fno-builtin-printf -Wall
LDFLAGS=-static -nostdlib -nostartfiles -lgcc
CFLAGS= -std=gnu99 -O2 -fno-common -fno-builtin-printf -Wall
LDFLAGS= -static
include libgloss.mk
PROGRAMS = pwm blkdev accum charcount nic-loopback big-blkdev pingd
.DEFAULT_GOAL := default
.PHONY: default
default: $(addsuffix .riscv,$(PROGRAMS))
.PHONY: dumps
dumps: $(addsuffix .dump,$(PROGRAMS))
%.o: %.S
@@ -15,11 +21,14 @@ dumps: $(addsuffix .dump,$(PROGRAMS))
%.o: %.c mmio.h
$(GCC) $(CFLAGS) -c $< -o $@
%.riscv: %.o crt.o syscalls.o link.ld
$(GCC) -T link.ld $(LDFLAGS) $< crt.o syscalls.o -o $@
%.riscv: %.o $(libgloss)
$(GCC) $(LDFLAGS) $< -o $@
%.dump: %.riscv
$(OBJDUMP) -D $< > $@
.PHONY: clean
clean:
rm -f *.riscv *.o *.dump
$(if $(libgloss),rm -rf $(libgloss_builddir)/)

View File

@@ -1,237 +0,0 @@
# See LICENSE for license details.
#include "encoding.h"
#if __riscv_xlen == 64
# define LREG ld
# define SREG sd
# define REGBYTES 8
#else
# define LREG lw
# define SREG sw
# define REGBYTES 4
#endif
.section ".text.init"
.globl _start
_start:
li x1, 0
li x2, 0
li x3, 0
li x4, 0
li x5, 0
li x6, 0
li x7, 0
li x8, 0
li x9, 0
li x10,0
li x11,0
li x12,0
li x13,0
li x14,0
li x15,0
li x16,0
li x17,0
li x18,0
li x19,0
li x20,0
li x21,0
li x22,0
li x23,0
li x24,0
li x25,0
li x26,0
li x27,0
li x28,0
li x29,0
li x30,0
li x31,0
# enable FPU and accelerator if present
li t0, MSTATUS_FS | MSTATUS_XS
csrs mstatus, t0
# make sure XLEN agrees with compilation choice
li t0, 1
slli t0, t0, 31
#if __riscv_xlen == 64
bgez t0, 1f
#else
bltz t0, 1f
#endif
2:
li a0, 1
sw a0, tohost, t0
j 2b
1:
#ifdef __riscv_flen
# initialize FPU if we have one
la t0, 1f
csrw mtvec, t0
fssr x0
fmv.s.x f0, x0
fmv.s.x f1, x0
fmv.s.x f2, x0
fmv.s.x f3, x0
fmv.s.x f4, x0
fmv.s.x f5, x0
fmv.s.x f6, x0
fmv.s.x f7, x0
fmv.s.x f8, x0
fmv.s.x f9, x0
fmv.s.x f10,x0
fmv.s.x f11,x0
fmv.s.x f12,x0
fmv.s.x f13,x0
fmv.s.x f14,x0
fmv.s.x f15,x0
fmv.s.x f16,x0
fmv.s.x f17,x0
fmv.s.x f18,x0
fmv.s.x f19,x0
fmv.s.x f20,x0
fmv.s.x f21,x0
fmv.s.x f22,x0
fmv.s.x f23,x0
fmv.s.x f24,x0
fmv.s.x f25,x0
fmv.s.x f26,x0
fmv.s.x f27,x0
fmv.s.x f28,x0
fmv.s.x f29,x0
fmv.s.x f30,x0
fmv.s.x f31,x0
1:
#endif
# initialize trap vector
la t0, trap_entry
csrw mtvec, t0
# initialize global pointer
.option push
.option norelax
la gp, __global_pointer$
.option pop
la tp, _end + 63
and tp, tp, -64
# get core id
csrr a0, mhartid
# for now, assume only 1 core
li a1, 1
1:bgeu a0, a1, 1b
# give each core 128KB of stack + TLS
#define STKSHIFT 17
sll a2, a0, STKSHIFT
add tp, tp, a2
add sp, a0, 1
sll sp, sp, STKSHIFT
add sp, sp, tp
j _init
.align 2
trap_entry:
addi sp, sp, -272
SREG x1, 1*REGBYTES(sp)
SREG x2, 2*REGBYTES(sp)
SREG x3, 3*REGBYTES(sp)
SREG x4, 4*REGBYTES(sp)
SREG x5, 5*REGBYTES(sp)
SREG x6, 6*REGBYTES(sp)
SREG x7, 7*REGBYTES(sp)
SREG x8, 8*REGBYTES(sp)
SREG x9, 9*REGBYTES(sp)
SREG x10, 10*REGBYTES(sp)
SREG x11, 11*REGBYTES(sp)
SREG x12, 12*REGBYTES(sp)
SREG x13, 13*REGBYTES(sp)
SREG x14, 14*REGBYTES(sp)
SREG x15, 15*REGBYTES(sp)
SREG x16, 16*REGBYTES(sp)
SREG x17, 17*REGBYTES(sp)
SREG x18, 18*REGBYTES(sp)
SREG x19, 19*REGBYTES(sp)
SREG x20, 20*REGBYTES(sp)
SREG x21, 21*REGBYTES(sp)
SREG x22, 22*REGBYTES(sp)
SREG x23, 23*REGBYTES(sp)
SREG x24, 24*REGBYTES(sp)
SREG x25, 25*REGBYTES(sp)
SREG x26, 26*REGBYTES(sp)
SREG x27, 27*REGBYTES(sp)
SREG x28, 28*REGBYTES(sp)
SREG x29, 29*REGBYTES(sp)
SREG x30, 30*REGBYTES(sp)
SREG x31, 31*REGBYTES(sp)
csrr a0, mcause
csrr a1, mepc
mv a2, sp
jal handle_trap
csrw mepc, a0
# Remain in M-mode after eret
li t0, MSTATUS_MPP
csrs mstatus, t0
LREG x1, 1*REGBYTES(sp)
LREG x2, 2*REGBYTES(sp)
LREG x3, 3*REGBYTES(sp)
LREG x4, 4*REGBYTES(sp)
LREG x5, 5*REGBYTES(sp)
LREG x6, 6*REGBYTES(sp)
LREG x7, 7*REGBYTES(sp)
LREG x8, 8*REGBYTES(sp)
LREG x9, 9*REGBYTES(sp)
LREG x10, 10*REGBYTES(sp)
LREG x11, 11*REGBYTES(sp)
LREG x12, 12*REGBYTES(sp)
LREG x13, 13*REGBYTES(sp)
LREG x14, 14*REGBYTES(sp)
LREG x15, 15*REGBYTES(sp)
LREG x16, 16*REGBYTES(sp)
LREG x17, 17*REGBYTES(sp)
LREG x18, 18*REGBYTES(sp)
LREG x19, 19*REGBYTES(sp)
LREG x20, 20*REGBYTES(sp)
LREG x21, 21*REGBYTES(sp)
LREG x22, 22*REGBYTES(sp)
LREG x23, 23*REGBYTES(sp)
LREG x24, 24*REGBYTES(sp)
LREG x25, 25*REGBYTES(sp)
LREG x26, 26*REGBYTES(sp)
LREG x27, 27*REGBYTES(sp)
LREG x28, 28*REGBYTES(sp)
LREG x29, 29*REGBYTES(sp)
LREG x30, 30*REGBYTES(sp)
LREG x31, 31*REGBYTES(sp)
addi sp, sp, 272
mret
.section ".tdata.begin"
.globl _tdata_begin
_tdata_begin:
.section ".tdata.end"
.globl _tdata_end
_tdata_end:
.section ".tbss.end"
.globl _tbss_end
_tbss_end:
.section ".tohost","aw",@progbits
.align 6
.globl tohost
tohost: .dword 0
.align 6
.globl fromhost
fromhost: .dword 0

File diff suppressed because it is too large Load Diff

1
tests/htif.ld Symbolic link
View File

@@ -0,0 +1 @@
../toolchains/libgloss/util/htif.ld

50
tests/libgloss.mk Normal file
View File

@@ -0,0 +1,50 @@
# Handle libgloss-htif dependency
ifndef libgloss
ifndef GCC
$(error GCC is not defined)
endif
libgloss_specs := htif_nano.specs
# Test whether libgloss-htif is globally installed and usable
# Define BUILD_LIBGLOSS=1 to unconditionally force a local build
BUILD_LIBGLOSS ?= $(shell { echo 'int main(void) { return 0; }' | \
$(GCC) -xc -specs=$(libgloss_specs) -o /dev/null - 2> /dev/null ; } || \
echo "$$?")
ifneq ($(BUILD_LIBGLOSS),)
$(info libgloss-htif: Using local build)
libgloss_srcdir := ../toolchains/libgloss
libgloss_builddir := libgloss
libgloss_specs := $(libgloss_srcdir)/util/$(libgloss_specs)
libgloss_lib := $(libgloss_builddir)/libgloss_htif.a
libgloss := $(libgloss_lib) $(libgloss_specs) htif.ld
LDFLAGS += -L libgloss
$(libgloss_builddir)/Makefile: $(libgloss_srcdir)/configure
mkdir -p $(dir $@)
cd $(dir $@) && $(realpath $<) \
--prefix=$(shell $(GCC) -print-sysroot) \
--host=$(TARGET) \
--disable-multilib
$(libgloss_lib): $(libgloss_builddir)/Makefile
$(MAKE) -C $(dir $^)
.PHONY: libgloss
libgloss: $(libgloss)
else
$(info libgloss-htif: Using global install)
libgloss := # No additional prerequisites
endif
CFLAGS += -specs=$(libgloss_specs)
LDFLAGS += -specs=$(libgloss_specs)
endif # libgloss

View File

@@ -1,65 +0,0 @@
/*======================================================================*/
/* Proxy kernel linker script */
/*======================================================================*/
/* This is the linker script used when building the proxy kernel. */
/*----------------------------------------------------------------------*/
/* Setup */
/*----------------------------------------------------------------------*/
/* The OUTPUT_ARCH command specifies the machine architecture where the
argument is one of the names used in the BFD library. More
specifically one of the entires in bfd/cpu-mips.c */
OUTPUT_ARCH( "riscv" )
ENTRY(_start)
/*----------------------------------------------------------------------*/
/* Sections */
/*----------------------------------------------------------------------*/
SECTIONS
{
/* text: test code section */
. = 0x80000000;
.text.init : { *(.text.init) }
.tohost ALIGN(0x1000) : { *(.tohost) }
.text : { *(.text) }
/* data segment */
.data ALIGN(0x40) : { *(.data) }
.sdata : {
__global_pointer$ = . + 0x800;
*(.srodata.cst16) *(.srodata.cst8) *(.srodata.cst4) *(.srodata.cst2) *(.srodata*)
*(.sdata .sdata.* .gnu.linkonce.s.*)
}
/* bss segment */
.sbss : {
*(.sbss .sbss.* .gnu.linkonce.sb.*)
*(.scommon)
}
.bss ALIGN(0x40) : { *(.bss) }
/* thread-local data segment */
.tdata :
{
_tls_data = .;
*(.tdata.begin)
*(.tdata)
*(.tdata.end)
}
.tbss :
{
*(.tbss)
*(.tbss.end)
}
/* End of uninitalized data segement */
_end = .;
}

View File

@@ -3,8 +3,8 @@
#include <stdio.h>
#include <string.h>
#include <riscv-pk/encoding.h>
#include "nic.h"
#include "encoding.h"
#define NPACKETS 10
#define TEST_OFFSET 3

View File

@@ -1,470 +0,0 @@
// See LICENSE for license details.
#include <stdint.h>
#include <string.h>
#include <stdarg.h>
#include <stdio.h>
#include <limits.h>
#include <sys/signal.h>
#include "util.h"
#define SYS_write 64
#undef strcmp
extern volatile uint64_t tohost;
extern volatile uint64_t fromhost;
static uintptr_t syscall(uintptr_t which, uint64_t arg0, uint64_t arg1, uint64_t arg2)
{
volatile uint64_t magic_mem[8] __attribute__((aligned(64)));
magic_mem[0] = which;
magic_mem[1] = arg0;
magic_mem[2] = arg1;
magic_mem[3] = arg2;
__sync_synchronize();
tohost = (uintptr_t)magic_mem;
while (fromhost == 0)
;
fromhost = 0;
__sync_synchronize();
return magic_mem[0];
}
#define NUM_COUNTERS 2
static uintptr_t counters[NUM_COUNTERS];
static char* counter_names[NUM_COUNTERS];
void setStats(int enable)
{
int i = 0;
#define READ_CTR(name) do { \
while (i >= NUM_COUNTERS) ; \
uintptr_t csr = read_csr(name); \
if (!enable) { csr -= counters[i]; counter_names[i] = #name; } \
counters[i++] = csr; \
} while (0)
READ_CTR(mcycle);
READ_CTR(minstret);
#undef READ_CTR
}
void __attribute__((noreturn)) tohost_exit(uintptr_t code)
{
tohost = (code << 1) | 1;
while (1);
}
uintptr_t __attribute__((weak)) handle_trap(uintptr_t cause, uintptr_t epc, uintptr_t regs[32])
{
tohost_exit(1337);
}
void exit(int code)
{
tohost_exit(code);
}
void abort()
{
exit(128 + SIGABRT);
}
void printstr(const char* s)
{
syscall(SYS_write, 1, (uintptr_t)s, strlen(s));
}
void __attribute__((weak)) thread_entry(int cid, int nc)
{
// multi-threaded programs override this function.
// for the case of single-threaded programs, only let core 0 proceed.
while (cid != 0);
}
int __attribute__((weak)) main(int argc, char** argv)
{
// single-threaded programs override this function.
printstr("Implement main(), foo!\n");
return -1;
}
static void init_tls()
{
register void* thread_pointer asm("tp");
extern char _tls_data;
extern __thread char _tdata_begin, _tdata_end, _tbss_end;
size_t tdata_size = &_tdata_end - &_tdata_begin;
memcpy(thread_pointer, &_tls_data, tdata_size);
size_t tbss_size = &_tbss_end - &_tdata_end;
memset(thread_pointer + tdata_size, 0, tbss_size);
}
void _init(int cid, int nc)
{
init_tls();
thread_entry(cid, nc);
// only single-threaded programs should ever get here.
int ret = main(0, 0);
char buf[NUM_COUNTERS * 32] __attribute__((aligned(64)));
char* pbuf = buf;
for (int i = 0; i < NUM_COUNTERS; i++)
if (counters[i])
pbuf += sprintf(pbuf, "%s = %d\n", counter_names[i], counters[i]);
if (pbuf != buf)
printstr(buf);
exit(ret);
}
#undef putchar
int putchar(int ch)
{
static __thread char buf[64] __attribute__((aligned(64)));
static __thread int buflen = 0;
buf[buflen++] = ch;
if (ch == '\n' || buflen == sizeof(buf))
{
syscall(SYS_write, 1, (uintptr_t)buf, buflen);
buflen = 0;
}
return 0;
}
void printhex(uint64_t x)
{
char str[17];
int i;
for (i = 0; i < 16; i++)
{
str[15-i] = (x & 0xF) + ((x & 0xF) < 10 ? '0' : 'a'-10);
x >>= 4;
}
str[16] = 0;
printstr(str);
}
static inline void printnum(void (*putch)(int, void**), void **putdat,
unsigned long long num, unsigned base, int width, int padc)
{
unsigned digs[sizeof(num)*CHAR_BIT];
int pos = 0;
while (1)
{
digs[pos++] = num % base;
if (num < base)
break;
num /= base;
}
while (width-- > pos)
putch(padc, putdat);
while (pos-- > 0)
putch(digs[pos] + (digs[pos] >= 10 ? 'a' - 10 : '0'), putdat);
}
static unsigned long long getuint(va_list *ap, int lflag)
{
if (lflag >= 2)
return va_arg(*ap, unsigned long long);
else if (lflag)
return va_arg(*ap, unsigned long);
else
return va_arg(*ap, unsigned int);
}
static long long getint(va_list *ap, int lflag)
{
if (lflag >= 2)
return va_arg(*ap, long long);
else if (lflag)
return va_arg(*ap, long);
else
return va_arg(*ap, int);
}
static void vprintfmt(void (*putch)(int, void**), void **putdat, const char *fmt, va_list ap)
{
register const char* p;
const char* last_fmt;
register int ch, err;
unsigned long long num;
int base, lflag, width, precision, altflag;
char padc;
while (1) {
while ((ch = *(unsigned char *) fmt) != '%') {
if (ch == '\0')
return;
fmt++;
putch(ch, putdat);
}
fmt++;
// Process a %-escape sequence
last_fmt = fmt;
padc = ' ';
width = -1;
precision = -1;
lflag = 0;
altflag = 0;
reswitch:
switch (ch = *(unsigned char *) fmt++) {
// flag to pad on the right
case '-':
padc = '-';
goto reswitch;
// flag to pad with 0's instead of spaces
case '0':
padc = '0';
goto reswitch;
// width field
case '1':
case '2':
case '3':
case '4':
case '5':
case '6':
case '7':
case '8':
case '9':
for (precision = 0; ; ++fmt) {
precision = precision * 10 + ch - '0';
ch = *fmt;
if (ch < '0' || ch > '9')
break;
}
goto process_precision;
case '*':
precision = va_arg(ap, int);
goto process_precision;
case '.':
if (width < 0)
width = 0;
goto reswitch;
case '#':
altflag = 1;
goto reswitch;
process_precision:
if (width < 0)
width = precision, precision = -1;
goto reswitch;
// long flag (doubled for long long)
case 'l':
lflag++;
goto reswitch;
// character
case 'c':
putch(va_arg(ap, int), putdat);
break;
// string
case 's':
if ((p = va_arg(ap, char *)) == NULL)
p = "(null)";
if (width > 0 && padc != '-')
for (width -= strnlen(p, precision); width > 0; width--)
putch(padc, putdat);
for (; (ch = *p) != '\0' && (precision < 0 || --precision >= 0); width--) {
putch(ch, putdat);
p++;
}
for (; width > 0; width--)
putch(' ', putdat);
break;
// (signed) decimal
case 'd':
num = getint(&ap, lflag);
if ((long long) num < 0) {
putch('-', putdat);
num = -(long long) num;
}
base = 10;
goto signed_number;
// unsigned decimal
case 'u':
base = 10;
goto unsigned_number;
// (unsigned) octal
case 'o':
// should do something with padding so it's always 3 octits
base = 8;
goto unsigned_number;
// pointer
case 'p':
static_assert(sizeof(long) == sizeof(void*));
lflag = 1;
putch('0', putdat);
putch('x', putdat);
/* fall through to 'x' */
// (unsigned) hexadecimal
case 'x':
base = 16;
unsigned_number:
num = getuint(&ap, lflag);
signed_number:
printnum(putch, putdat, num, base, width, padc);
break;
// escaped '%' character
case '%':
putch(ch, putdat);
break;
// unrecognized escape sequence - just print it literally
default:
putch('%', putdat);
fmt = last_fmt;
break;
}
}
}
int printf(const char* fmt, ...)
{
va_list ap;
va_start(ap, fmt);
vprintfmt((void*)putchar, 0, fmt, ap);
va_end(ap);
return 0; // incorrect return value, but who cares, anyway?
}
int sprintf(char* str, const char* fmt, ...)
{
va_list ap;
char* str0 = str;
va_start(ap, fmt);
void sprintf_putch(int ch, void** data)
{
char** pstr = (char**)data;
**pstr = ch;
(*pstr)++;
}
vprintfmt(sprintf_putch, (void**)&str, fmt, ap);
*str = 0;
va_end(ap);
return str - str0;
}
void* memcpy(void* dest, const void* src, size_t len)
{
if ((((uintptr_t)dest | (uintptr_t)src | len) & (sizeof(uintptr_t)-1)) == 0) {
const uintptr_t* s = src;
uintptr_t *d = dest;
while (d < (uintptr_t*)(dest + len))
*d++ = *s++;
} else {
const char* s = src;
char *d = dest;
while (d < (char*)(dest + len))
*d++ = *s++;
}
return dest;
}
void* memset(void* dest, int byte, size_t len)
{
if ((((uintptr_t)dest | len) & (sizeof(uintptr_t)-1)) == 0) {
uintptr_t word = byte & 0xFF;
word |= word << 8;
word |= word << 16;
word |= word << 16 << 16;
uintptr_t *d = dest;
while (d < (uintptr_t*)(dest + len))
*d++ = word;
} else {
char *d = dest;
while (d < (char*)(dest + len))
*d++ = byte;
}
return dest;
}
size_t strlen(const char *s)
{
const char *p = s;
while (*p)
p++;
return p - s;
}
size_t strnlen(const char *s, size_t n)
{
const char *p = s;
while (n-- && *p)
p++;
return p - s;
}
int strcmp(const char* s1, const char* s2)
{
unsigned char c1, c2;
do {
c1 = *s1++;
c2 = *s2++;
} while (c1 != 0 && c1 == c2);
return c1 - c2;
}
char* strcpy(char* dest, const char* src)
{
char* d = dest;
while ((*d++ = *src++))
;
return dest;
}
long atol(const char* str)
{
long res = 0;
int sign = 0;
while (*str == ' ')
str++;
if (*str == '-' || *str == '+') {
sign = *str == '-';
str++;
}
while (*str) {
res *= 10;
res += *str++ - '0';
}
return sign ? -res : res;
}

View File

@@ -1,135 +0,0 @@
// See LICENSE for license details.
#ifndef __UTIL_H
#define __UTIL_H
//--------------------------------------------------------------------------
// Macros
// Set HOST_DEBUG to 1 if you are going to compile this for a host
// machine (ie Athena/Linux) for debug purposes and set HOST_DEBUG
// to 0 if you are compiling with the smips-gcc toolchain.
#ifndef HOST_DEBUG
#define HOST_DEBUG 0
#endif
// Set PREALLOCATE to 1 if you want to preallocate the benchmark
// function before starting stats. If you have instruction/data
// caches and you don't want to count the overhead of misses, then
// you will need to use preallocation.
#ifndef PREALLOCATE
#define PREALLOCATE 0
#endif
// Set SET_STATS to 1 if you want to carve out the piece that actually
// does the computation.
#if HOST_DEBUG
#include <stdio.h>
static void setStats(int enable) {}
#else
extern void setStats(int enable);
#endif
#include <stdint.h>
#define static_assert(cond) switch(0) { case 0: case !!(long)(cond): ; }
static void printArray(const char name[], int n, const int arr[])
{
#if HOST_DEBUG
int i;
printf( " %10s :", name );
for ( i = 0; i < n; i++ )
printf( " %3d ", arr[i] );
printf( "\n" );
#endif
}
static void printDoubleArray(const char name[], int n, const double arr[])
{
#if HOST_DEBUG
int i;
printf( " %10s :", name );
for ( i = 0; i < n; i++ )
printf( " %g ", arr[i] );
printf( "\n" );
#endif
}
static int verify(int n, const volatile int* test, const int* verify)
{
int i;
// Unrolled for faster verification
for (i = 0; i < n/2*2; i+=2)
{
int t0 = test[i], t1 = test[i+1];
int v0 = verify[i], v1 = verify[i+1];
if (t0 != v0) return i+1;
if (t1 != v1) return i+2;
}
if (n % 2 != 0 && test[n-1] != verify[n-1])
return n;
return 0;
}
static int verifyDouble(int n, const volatile double* test, const double* verify)
{
int i;
// Unrolled for faster verification
for (i = 0; i < n/2*2; i+=2)
{
double t0 = test[i], t1 = test[i+1];
double v0 = verify[i], v1 = verify[i+1];
int eq1 = t0 == v0, eq2 = t1 == v1;
if (!(eq1 & eq2)) return i+1+eq1;
}
if (n % 2 != 0 && test[n-1] != verify[n-1])
return n;
return 0;
}
static void __attribute__((noinline)) barrier(int ncores)
{
static volatile int sense;
static volatile int count;
static __thread int threadsense;
__sync_synchronize();
threadsense = !threadsense;
if (__sync_fetch_and_add(&count, 1) == ncores-1)
{
count = 0;
sense = threadsense;
}
else while(sense != threadsense)
;
__sync_synchronize();
}
static uint64_t lfsr(uint64_t x)
{
uint64_t bit = (x ^ (x >> 1)) & 1;
return (x >> 1) | (bit << 62);
}
#ifdef __riscv
#include "encoding.h"
#endif
#define stringify_1(s) #s
#define stringify(s) stringify_1(s)
#define stats(code, iter) do { \
unsigned long _c = -read_csr(mcycle), _i = -read_csr(minstret); \
code; \
_c += read_csr(mcycle), _i += read_csr(minstret); \
if (cid == 0) \
printf("\n%s: %ld cycles, %ld.%ld cycles/iter, %ld.%ld CPI\n", \
stringify(code), _c, _c/iter, 10*_c/iter%10, _c/_i, 10*_c/_i%10); \
} while(0)
#endif //__UTIL_H