On Mon, 2023-07-31 at 18:06 +0100, Mark Brown wrote:
Someday when the x86 side is finally upstream I have a manpage for map_shadow_stack. Any differences on the arm side would need to be documented, but I'm not sure why there should be any differences. Like, why not use the same flags? Or have a new flag for token+end marker that x86 can use as well?
Ah, it wasn't clear to me that this was a question rather than just open decisions about the eventual manpage. Looking again I think what you're asking about is that I see that at some point in development I lost the SHADOW_STACK_SET_TOKEN flag which x86 has. I suspect that was a rebasing issue as it wasn't a deliberate decision, there's no reason we couldn't have that. Other than that and the fact that we add both a stack swap token and a top of stack marker I'm not aware of any differences.
The thing I was trying to get at was, we have this shared syscall that means create shadow stack memory and prepopulate it like this flag says. On x86 we optionally support SHADOW_STACK_SET_TOKEN which means put a token right at the end of size. So maybe arm should have a different flag value that includes putting the marker and then the token, and x86 could match it someday if we get markers too.
It could be a different flag, like SHADOW_STACK_SET_TOKEN_MARKER, or it could be SHADOW_STACK_SET_MARKER, and callers could pass (SHADOW_STACK_SET_TOKEN | SHADOW_STACK_SET_MARKER) to get what you have implemented here. What do you think?