Hi,
I am working on the old version QEMU which is still using dyngen with my own constraints. I want to know how I can get the guest instruction to intermediate code mapping. Are there any docs?
For example:
For this guest instruction
add [ebp + 0x08], edx
It will be split into multiple micro-ops:
movl A0 ebp
addl A0 0x0c
addl A0 ss.base
movl T1 edx
ldl_raw T0 A0
addl T0 T1
stl_raw T0 A0
update2_cc
My question is how I can get the guest instruction to micron-ops mapping. If I have a "cmpl" guest instruction, how can I find the mapping micro-ops in intermediate code?
Another question is that when I run 32bits Guest VM, I can only allocate up to 2 GB not 4 GB. Are there any special reasons for this design?
Thanks,
- Qiang