[llvm] [llvm][ARM][CodeGen] Disable MEMCPY LDM/STM inlining for v7-m (PR #106378)

[Oliver Stannard via llvm-commits]

ostannard wrote:

Do you have any benchmark numbers for what performance difference this makes? Does this only affect specific v7-M cores, or should it apply to later architecture versions too?

I think this needs a lot more tests, to cover the code paths used for different lengths and alignments. The tests should also check that the correct instructions are emitted, not just that no ldm/stm instructions are used.

https://github.com/llvm/llvm-project/pull/106378