perf: Optimize NEXT_INST macro by reordering memory loads for better performance #217

Job	Run time
linux-arm-ci-asm	6m 35s
linux-arm-test-suite	6m 7s
linux-native-ci-asm (ubuntu-latest)	1m 19s
linux-native-ci-asm-chaos (ubuntu-latest)	1m 1s
linux-native-ci-asm (ubuntu-24.04-arm)	1m 51s
linux-native-ci-asm-chaos (ubuntu-24.04-arm)	2m 5s
linux-native-test-spawn (ubuntu-latest)	1m 0s
linux-native-test-spawn (ubuntu-24.04-arm)	1m 47s
linux-native-test-suite (ubuntu-latest)	2m 42s
linux-x86-ci (ubuntu-latest)	55s
linux-x86-ci-generated	14s
linux-native-test-suite (ubuntu-24.04-arm)	5m 16s
linux-x86-ci (ubuntu-24.04-arm)	2m 6s
linux-x86-deny	33s
macos-x86-ci-asm	1m 33s
windows-x86-ci-asm	4m 35s
	39m 39s

Provide feedback