Add vectorized integer division for int32 with AVX512, AVX or SSE.

Reference issue

What does this implement/fix?

This implements vectorized integer division for int32 on AVX512, AVX and SSE

Original author is @chuckyschluz .

This change adds

• Sets HasDiv=1 for Packet4i, Packet8i, and Packet16i
• raises SIGFPE upon division by zero

Additional information

Benchmark data for SSE & AVX here: $2412525