8382482: Optimize equals scenario in x86 scalar floating point min/max reduction loops#30806

Open

missa-prime wants to merge 6 commits intoopenjdk:masterfrom

missa-prime:user/missa-prime/avx10_2

Commits on Apr 17, 2026

Adjust instruction sequence used by scalar floating point min/max reduction rules.
missa-prime
committed
Add missing character at the beginning of function definition.
missa-prime
committed

Commits on Apr 21, 2026

Extend dashed line to cover all commented text above emit_fp_min_max.
missa-prime
committed

Commits on Apr 24, 2026

Split AVX10.2 and non-AVX10.2 min/max floating point reduction into separate code paths.
missa-prime
committed

Commits on Apr 26, 2026

Change the composition of the float and double arrays in the FpMinMaxIntrinsics.java file.
missa-prime
committed

Commits on May 4, 2026

Re-combine code paths and use near jumps instead of short jumps to make compiler align loop block address to 32 bytes instead of 16 bytes.
missa-prime
committed