vector.gather.mask
← vector dialect
Masked vector gather from per-lane signed logical offsets added to the last view axis. True mask lanes read the adjusted coordinate, while false mask lanes do not access memory and take the corresponding passthrough lane.
Operation contract
| Property |
Value |
| Semantic phase |
— |
| Interfaces |
MemoryAccess |
| Memory effects |
read view |
Signature
| Kind |
Name |
Type |
Cardinality |
Description |
| Operand |
view |
view |
required |
Typed source view. |
| Operand |
offsets |
vector |
required |
Per-lane signed element offsets from the logical origin. |
| Operand |
mask |
vector |
required |
i1 vector mask selecting gathered lanes. |
| Operand |
passthrough |
vector |
required |
Value used for masked-off lanes. |
| Operand |
indices |
index |
variadic |
Dynamic logical origin indices. |
| Result |
result |
vector |
required |
Gathered vector value. |
| Attribute |
cache_scope |
enum CacheScope |
optional |
Optional cache/coherency scope required by target lowering. |
| Attribute |
cache_temporal |
enum CacheTemporal |
optional |
Optional temporal cache policy required by target lowering. |
| Attribute |
static_indices |
i64_array |
required |
Static logical origin indices with INT64_MIN sentinels for dynamics. |
Verification constraints
HasI1Element(mask)HasIndexOrNonI1IntegerElement(offsets)SameElementType(view, passthrough, result)SameShape(offsets, mask, passthrough, result)SameType(passthrough, result)
Examples
%v = vector.gather.mask %view[%row, %col][%offsets], %mask, %old : view<[%m]x[%n]xf32, %layout>, vector<4xindex>, vector<4xi1>, vector<4xf32>