ardour/livetrax
13
0
Fork 0
Code

possible minor optimizations of atomic loads in int62_t class

Browse Source
This commit is contained in:
Paul Davis 2021-02-10 00:22:45 -07:00
parent 38b1f5eee1
commit 25ecc54208
1 changed files with 16 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
libs/pbd/pbd/int62.h
View File

@ -48,7 +48,7 @@ class alignas(16) int62_t {
/* the "flagbit" follows 2's complement logic. It is "set" if the value is positive and the bit is 1; it is also set if the /* the "flagbit" follows 2's complement logic. It is "set" if the value is positive and the bit is 1; it is also set if the
* value is negative and bit is 0. * value is negative and bit is 0.
*/ */
static int64_t int62 (int64_t v) { if (v >= 0) { return v & ~flagbit_mask; } return (v | flagbit_mask); } static int64_t int62 (int64_t v) { if (v >= 0) { return v & ~flagbit_mask; } return (v | flagbit_mask); }
static bool flagged (int64_t v) { if (v >= 0) { return v & flagbit_mask; } return ((v & flagbit_mask) == 0); } static bool flagged (int64_t v) { if (v >= 0) { return v & flagbit_mask; } return ((v & flagbit_mask) == 0); }
public: public:
@ -85,21 +85,21 @@ class alignas(16) int62_t {
rather than two separate loads for each "part". rather than two separate loads for each "part".
*/ */
int62_t operator- () const { int64_t tmp = v; return int62_t (flagged (tmp), -int62(tmp)); } int62_t operator- () const { const int64_t vv = v.load(); return int62_t (flagged (vv), -int62(vv)); }
int62_t operator+ (int64_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) + n); } int62_t operator+ (int64_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) + n); }
int62_t operator- (int64_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) - n); } int62_t operator- (int64_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) - n); }
int62_t operator* (int64_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) * n); } int62_t operator* (int64_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) * n); }
int62_t operator/ (int64_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) / n); } int62_t operator/ (int64_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) / n); }
int62_t operator% (int64_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) % n); } int62_t operator% (int64_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) % n); }
int62_t operator+ (int62_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) + n.val()); } int62_t operator+ (int62_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) + n.val()); }
int62_t operator- (int62_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) - n.val()); } int62_t operator- (int62_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) - n.val()); }
int62_t operator* (int62_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) * n.val()); } int62_t operator* (int62_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) * n.val()); }
int62_t operator/ (int62_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) / n.val()); } int62_t operator/ (int62_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) / n.val()); }
int62_t operator% (int62_t n) const { int64_t tmp = v; return int62_t (flagged (tmp), int62 (tmp) % n.val()); } int62_t operator% (int62_t n) const { const int64_t vv = v.load(); return int62_t (flagged (vv), int62 (vv) % n.val()); }
/* comparison operators .. will throw if the two objects have different /* comparison operators .. will throw if the two objects have different
* flag settings (which is assumed to indicate that they differ in some * flag settings (which is assumed to indicate that they differ in some
* important respect, and thus should not have their values compared) * important respect, and thus should not have their values compared)
*/ */
@ -118,8 +118,8 @@ class alignas(16) int62_t {
* the semantics are well defined and the computation cost is trivial * the semantics are well defined and the computation cost is trivial
*/ */
bool operator!= (int62_t const & other) const { if (flagged() != other.flagged()) return true; return val() != other.val(); } bool operator!= (int62_t const & other) const { const int64_t vv = v.load(); if (flagged (vv) != other.flagged()) return true; return int62 (vv) != other.val(); }
bool operator== (int62_t const & other) const { if (flagged() != other.flagged()) return false; return val() == other.val(); } bool operator== (int62_t const & other) const { const int64_t vv = v.load(); if (flagged (vv) != other.flagged()) return false; return int62 (vv) == other.val(); }
explicit operator int64_t() const { return int62(v); } explicit operator int64_t() const { return int62(v); }
@ -130,7 +130,7 @@ class alignas(16) int62_t {
bool operator!= (int64_t n) const { return val() != n; } bool operator!= (int64_t n) const { return val() != n; }
bool operator== (int64_t n) const { return val() == n; } bool operator== (int64_t n) const { return val() == n; }
int62_t abs() const { int64_t tmp = v; return int62_t (flagged(tmp), ::abs(int62(tmp))); } int62_t abs() const { const int64_t tmp = v; return int62_t (flagged(tmp), ::abs(int62(tmp))); }
int62_t& operator+= (int64_t n) { int62_t& operator+= (int64_t n) {
while (1) { while (1) {