LoongArch: BPF: Implement branchless conditional move for TCC

The current implementation handles combined bpf2bpf and tail calls by
checking at runtime whether REG_TCC holds a scalar count or a pointer
address via a conditional jump. This adds branch prediction overhead
in the hot path of tail call execution.

To implement branchless conditional move, use an unsigned comparison
(sltui) combined with mask instructions (maskeqz/masknez) to achieve
branchless classification and blending of incoming scalar counts and
kernel pointers in REG_TCC.

This optimization refactors the inner logic of the helper function,
unifies the offset decrement at the function entry, and removes all
runtime branching from the prologue hot path completely.

Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

0 files changed, 0 insertions, 0 deletions