Closed RobertCochran closed 5 months ago
It looks like it shows the "code" field of the instruction in the decompilation instead of the syscall number, same as Ghidra's built-in MIPS implementation.
Usually syscalls are called through wrapper functions, so I imagine a good solution would be to give said functions appropriate names. Maybe that could be a feature in the future.
I remember Ghidra added better syscall support a few years back, but I never looked much into it. Maybe I could see if that would help.
Usually syscalls are called through wrapper functions, so I imagine a good solution would be to give said functions appropriate names. Maybe that could be a feature in the future.
In my very limited experience, this has usually been the case, except when it isn't; I have noticed in a couple places in the early bootstrap process of Ratchet 1 where there are direct syscalls. (the helper function in those cases got inlined away?)
This is fixed now. Probably should've done it earlier, but now it's done.
The decompile window always shows the syscall as
syscall(0)
, and not the actual syscall number. Based on various documentation I have found and the code I am working through, the EE usesv1
as the syscall number register, which is apparently a non-standard convention on MIPS.