Available on x86 and target feature
sse
only.Expand description
Converts the lowest 32 bit float in the input vector to a 32 bit integer with truncation.
The result is rounded always using truncation (round towards zero). If the
result cannot be represented as a 32 bit integer the result will be
0x8000_0000
(i32::MIN
) or an invalid operation floating point
exception if unmasked (see _mm_setcsr
).
This corresponds to the CVTTSS2SI
instruction (with 32 bit output).