aaron/ispc
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Represent MOVMSK'ed masks with int64s rather than int32s.

Browse Source 
This allows us to scale up to 64-wide execution.
This commit is contained in:
Matt Pharr
2012-05-25 11:48:08 -07:00
parent 38cea6dc71
commit 90db01d038
20 changed files with 137 additions and 107 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
stdlib.ispc
View File

@@ -355,7 +355,8 @@ static inline uniform bool all(bool v) {
#else
int32 match = __sext_varying_bool((__sext_varying_bool(v) & __mask) == __mask);
#endif
return __movmsk(match) == (1 << programCount) - 1;
return __movmsk(match) == ((programCount == 64) ? ~0ull :
((1ull << programCount) - 1));
}
__declspec(safe)
@@ -388,14 +389,14 @@ __declspec(safe)
static inline uniform int popcnt(bool v) {
// As with any() and all(), only count across the active lanes
#ifdef ISPC_TARGET_GENERIC
return __popcnt_int32(__movmsk(v & __mask));
return __popcnt_int64(__movmsk(v & __mask));
#else
return __popcnt_int32(__movmsk(__sext_varying_bool(v) & __mask));
return __popcnt_int64(__movmsk(__sext_varying_bool(v) & __mask));
#endif
}
__declspec(safe)
static inline uniform int lanemask() {
static inline uniform unsigned int64 lanemask() {
return __movmsk(__mask);
}
@@ -1615,12 +1616,12 @@ static inline TA atomic_swap_global(uniform TA * uniform ptr, TA value) { \
TA ret[programCount]; \
TA memVal; \
uniform int lastSwap; \
uniform int mask = lanemask(); \
uniform unsigned int64 mask = lanemask(); \
/* First, have the first running program instance (if any) perform \
the swap with memory with its value of "value"; record the \
value returned. */ \
for (; i < programCount; ++i) { \
if ((mask & (1 << i)) == 0) \
if ((mask & (1ull << i)) == 0) \
continue; \
memVal = __atomic_swap_uniform_##TB##_global(ptr, extract(value, i)); \
lastSwap = i; \
@@ -1632,7 +1633,7 @@ static inline TA atomic_swap_global(uniform TA * uniform ptr, TA value) { \
current instance had executed a hardware atomic swap right before \
the last one that did a swap. */ \
for (; i < programCount; ++i) { \
if ((mask & (1 << i)) == 0) \
if ((mask & (1ull << i)) == 0) \
continue; \
ret[lastSwap] = extract(value, i); \
lastSwap = i; \