Lines Matching full:warp

1329   // Get the mask of active threads in the warp.  in emitCriticalRegion()
1372   // Reconverge active threads in the warp.  in emitCriticalRegion()
1436 ///     manner as follows: within OpenMP threads in the same warp,
1476 ///     same warp.  It takes the following arguments as input:
1501 ///     from lane 0 of each warp to a lane in the first warp.
1525 /// Warp Reduction Algorithms
1527 /// On the warp level, we have three algorithms implemented in the
1530 /// Full Warp Reduction
1532 /// The reduce algorithm within a warp where all lanes are active
1547 /// signify full warp reduction.
1556 /// Contiguous Partial Warp Reduction
1558 /// This reduce algorithm is used within a warp where only the first
1591 /// Dispersed Partial Warp Reduction
1593 /// This algorithm is used within a warp when any discontiguous subset of
1630 /// threads in a team.  It first reduces within a warp using the
1632 /// reduced values at the first warp.
1635 /// data from each of the "warp master" (zeroth lane of each warp, where
1636 /// warp-reduced data is held) to the zeroth warp.  This step reduces (in
1637 /// a mathematical sense) the problem of reduction across warp masters in
1638 /// a block to the problem of warp reduction.