Re: Regression in ptrace (Wine) starting with 2.6.33-rc1

[Frederic Weisbecker <fweisbec@gmail.com>]

On Sat, Feb 13, 2010 at 10:29:16PM +0100, Michael Stefaniuc wrote:
> Results 2.6.33-rcX:
> -------------------
> ptrace(PTRACE_ATTACH, 18036, 0, 0)      = 0
> ptrace(PTRACE_POKEUSER, 18036, offsetof(struct user, u_debugreg),  
> 0x42424242) = 0
> ptrace(PTRACE_POKEUSER, 18036, offsetof(struct user, u_debugreg) + 4, 0) = 0
> ptrace(PTRACE_POKEUSER, 18036, offsetof(struct user, u_debugreg) + 8, 0) = 0
> ptrace(PTRACE_POKEUSER, 18036, offsetof(struct user, u_debugreg) + 12,  
> 0) = 0
> ptrace(PTRACE_POKEUSER, 18036, offsetof(struct user, u_debugreg) + 24,  
> 0) = 0
> ptrace(PTRACE_POKEUSER, 18036, offsetof(struct user, u_debugreg) + 28,  
> 0x155) = -1 EINVAL (Invalid argument)
>
> Results 2.6.32:
> ---------------
> trace(PTRACE_ATTACH, 3077, 0, 0)       = 0
> ptrace(PTRACE_POKEUSER, 3077, offsetof(struct user, u_debugreg),  
> 0x42424242) = 0
> ptrace(PTRACE_POKEUSER, 3077, offsetof(struct user, u_debugreg) + 4, 0) = 0
> ptrace(PTRACE_POKEUSER, 3077, offsetof(struct user, u_debugreg) + 8, 0) = 0
> ptrace(PTRACE_POKEUSER, 3077, offsetof(struct user, u_debugreg) + 12, 0) = 0
> ptrace(PTRACE_POKEUSER, 3077, offsetof(struct user, u_debugreg) + 24, 0) = 0
> ptrace(PTRACE_POKEUSER, 3077, offsetof(struct user, u_debugreg) + 28,  
> 0x155) = 0


I see... So this is setting breakpoints on the address 0. The new code
rejects such breakpoints, but the previous one was accepting it.

The point of allowing breakpoints in NULL is discutable. It's not a bug,
neither is it a security hole I think (because if the ptrace breakpoint
triggers from the kernel, it's ignored), it's just pointless, unless
userland map things in 0.

But it's too late to debate this. If the previous code accepted it,
it's an ABI, and we have broken it.

I'm preparing a fix.



> So it looks like something in the setting of DR7 is broken or at least
> changed behavior. The function in Wine that does those calls is
> set_thread_context() from server/ptrace.c .
>
> I'll try to see if I can reproduce the other regression; as it is hidden  
> at the moment by this regression.


Ok.

Thanks.