Hello, I am attempting to fold a cache model into Qemu. For this I need to store the instructions executed between actual lds/sts to Qemu memory, for performance reasons. I figured that a buffer in the CPUState could accommodate this requirement, however tcg only gives the ability to ld/st from host memory at constant offsets. Is there a better solution to this problem than adding new ld/st instructions to tcg that take a computed offset? Thx, James
