summaryrefslogtreecommitdiffstats
path: root/src/runtime/internal/sys/intrinsics_common.go
diff options
context:
space:
mode:
Diffstat (limited to 'src/runtime/internal/sys/intrinsics_common.go')
-rw-r--r--src/runtime/internal/sys/intrinsics_common.go158
1 files changed, 158 insertions, 0 deletions
diff --git a/src/runtime/internal/sys/intrinsics_common.go b/src/runtime/internal/sys/intrinsics_common.go
new file mode 100644
index 0000000..48d9759
--- /dev/null
+++ b/src/runtime/internal/sys/intrinsics_common.go
@@ -0,0 +1,158 @@
+// Copyright 2019 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package sys
+
+// Copied from math/bits to avoid dependence.
+
+var len8tab = [256]uint8{
+ 0x00, 0x01, 0x02, 0x02, 0x03, 0x03, 0x03, 0x03, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04, 0x04,
+ 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05, 0x05,
+ 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06,
+ 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06, 0x06,
+ 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07,
+ 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07,
+ 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07,
+ 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+ 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08, 0x08,
+}
+
+var ntz8tab = [256]uint8{
+ 0x08, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x05, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x06, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x05, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x07, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x05, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x06, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x05, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+ 0x04, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00, 0x03, 0x00, 0x01, 0x00, 0x02, 0x00, 0x01, 0x00,
+}
+
+// len64 returns the minimum number of bits required to represent x; the result is 0 for x == 0.
+func Len64(x uint64) (n int) {
+ if x >= 1<<32 {
+ x >>= 32
+ n = 32
+ }
+ if x >= 1<<16 {
+ x >>= 16
+ n += 16
+ }
+ if x >= 1<<8 {
+ x >>= 8
+ n += 8
+ }
+ return n + int(len8tab[x])
+}
+
+// --- OnesCount ---
+
+const m0 = 0x5555555555555555 // 01010101 ...
+const m1 = 0x3333333333333333 // 00110011 ...
+const m2 = 0x0f0f0f0f0f0f0f0f // 00001111 ...
+
+// OnesCount64 returns the number of one bits ("population count") in x.
+func OnesCount64(x uint64) int {
+ // Implementation: Parallel summing of adjacent bits.
+ // See "Hacker's Delight", Chap. 5: Counting Bits.
+ // The following pattern shows the general approach:
+ //
+ // x = x>>1&(m0&m) + x&(m0&m)
+ // x = x>>2&(m1&m) + x&(m1&m)
+ // x = x>>4&(m2&m) + x&(m2&m)
+ // x = x>>8&(m3&m) + x&(m3&m)
+ // x = x>>16&(m4&m) + x&(m4&m)
+ // x = x>>32&(m5&m) + x&(m5&m)
+ // return int(x)
+ //
+ // Masking (& operations) can be left away when there's no
+ // danger that a field's sum will carry over into the next
+ // field: Since the result cannot be > 64, 8 bits is enough
+ // and we can ignore the masks for the shifts by 8 and up.
+ // Per "Hacker's Delight", the first line can be simplified
+ // more, but it saves at best one instruction, so we leave
+ // it alone for clarity.
+ const m = 1<<64 - 1
+ x = x>>1&(m0&m) + x&(m0&m)
+ x = x>>2&(m1&m) + x&(m1&m)
+ x = (x>>4 + x) & (m2 & m)
+ x += x >> 8
+ x += x >> 16
+ x += x >> 32
+ return int(x) & (1<<7 - 1)
+}
+
+var deBruijn64tab = [64]byte{
+ 0, 1, 56, 2, 57, 49, 28, 3, 61, 58, 42, 50, 38, 29, 17, 4,
+ 62, 47, 59, 36, 45, 43, 51, 22, 53, 39, 33, 30, 24, 18, 12, 5,
+ 63, 55, 48, 27, 60, 41, 37, 16, 46, 35, 44, 21, 52, 32, 23, 11,
+ 54, 26, 40, 15, 34, 20, 31, 10, 25, 14, 19, 9, 13, 8, 7, 6,
+}
+
+const deBruijn64 = 0x03f79d71b4ca8b09
+
+// TrailingZeros64 returns the number of trailing zero bits in x; the result is 64 for x == 0.
+func TrailingZeros64(x uint64) int {
+ if x == 0 {
+ return 64
+ }
+ // If popcount is fast, replace code below with return popcount(^x & (x - 1)).
+ //
+ // x & -x leaves only the right-most bit set in the word. Let k be the
+ // index of that bit. Since only a single bit is set, the value is two
+ // to the power of k. Multiplying by a power of two is equivalent to
+ // left shifting, in this case by k bits. The de Bruijn (64 bit) constant
+ // is such that all six bit, consecutive substrings are distinct.
+ // Therefore, if we have a left shifted version of this constant we can
+ // find by how many bits it was shifted by looking at which six bit
+ // substring ended up at the top of the word.
+ // (Knuth, volume 4, section 7.3.1)
+ return int(deBruijn64tab[(x&-x)*deBruijn64>>(64-6)])
+}
+
+// LeadingZeros64 returns the number of leading zero bits in x; the result is 64 for x == 0.
+func LeadingZeros64(x uint64) int { return 64 - Len64(x) }
+
+// LeadingZeros8 returns the number of leading zero bits in x; the result is 8 for x == 0.
+func LeadingZeros8(x uint8) int { return 8 - Len8(x) }
+
+// TrailingZeros8 returns the number of trailing zero bits in x; the result is 8 for x == 0.
+func TrailingZeros8(x uint8) int {
+ return int(ntz8tab[x])
+}
+
+// Len8 returns the minimum number of bits required to represent x; the result is 0 for x == 0.
+func Len8(x uint8) int {
+ return int(len8tab[x])
+}
+
+// Prefetch prefetches data from memory addr to cache
+//
+// AMD64: Produce PREFETCHT0 instruction
+//
+// ARM64: Produce PRFM instruction with PLDL1KEEP option
+func Prefetch(addr uintptr) {}
+
+// PrefetchStreamed prefetches data from memory addr, with a hint that this data is being streamed.
+// That is, it is likely to be accessed very soon, but only once. If possible, this will avoid polluting the cache.
+//
+// AMD64: Produce PREFETCHNTA instruction
+//
+// ARM64: Produce PRFM instruction with PLDL1STRM option
+func PrefetchStreamed(addr uintptr) {}
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-31 21:27:07 +0000