NOMMU bogus syscall return values

[Dmitry V. Levin]

On Thu, Mar 03, 2016 at 09:21:48PM -0500, Rich Felker wrote:
> On Fri, Mar 04, 2016 at 02:39:38AM +0300, Dmitry V. Levin wrote:
> > On Wed, Mar 02, 2016 at 09:29:24PM -0500, Rich Felker wrote:
> > > On Thu, Mar 03, 2016 at 05:16:18AM +0300, Dmitry V. Levin wrote:
> > > > On Wed, Mar 02, 2016 at 08:28:30PM -0500, Rich Felker wrote:
> > [...]
> > > > > What should be done about daemonized tracer mode?
> > > > 
> > > > If we could switch stack before clone and then switch stack back in the
> > > > child process, it would be a relatively simple change.
> > > > Otherwise I don't see any simple options.
> > > 
> > > I think it's possible to switch stack in the parent via sigaltstack
> > > and raise, after which the cloned child can longjmp back to the stack
> > > that was the parent's.
> > 
> > This might work.  Whatever signal is chosen, its handler and mask would
> > have to be restored before the final execve call.
> 
> The handler and mask could be restored before even calling clone;
> actually, the whole machinery could be implemented as a wrapper for
> clone that lets you pick whether the child runs a new function on a
> new stack with the parent returning (like normal clone) or the parent
> runs a new function on a new stack (with the child returning).

Sounds good.

> > > Alternatively the child could just re-enter
> > > main() on a new stack with a global var set to indicate that it's the
> > > tracer child.
> > 
> > I think it would be too risky to let the main bulk of strace code run
> > on the new stack.
> 
> Just in terms of stack size available? Or something else?

Just the stack size.


-- 
ldv
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.strace.io/pipermail/strace-devel/attachments/20160304/37a494a1/attachment.bin>