Re: Trouble with ptrace self-attach rule since kernel > 2.6.14

[Andreas Hobein]

On Friday 01 September 2006 03:39, Andrew Morton wrote:
> I'm unable to identify what patch you're referring to here.  Please be more
> specific so we can ask the person who developed it.

I assume the attached patch from Linus Torvalds causes my problem, since the 
condition was changed from "if (task == current)" to "if (task->tgid == 
current->tgid)" it breaks my application code. There may be other parts of 
the ptrace() kernel code that where changed accordingly that I'm not aware. 

There is also a reply from Roland McGrath (see 
http://lkml.org/lkml/2005/11/9/460) who mentioned that there may occur some 
problems in "some real programs out there". May be I'm the first one who is 
affected by this new behaviour.

To summarise my questions:
- Why should a thread not be allowed to ptrace_attach to a sibling thread
- while a forked child of this thread may do this ?
- Is there any other way to suspend sibling threads at arbitrary points like 
phread_suspend_np() does for example on AIX?

Thanks, Andreas

---------------------------