[RFC] Detect loops in the solib chain

Daniel Jacobowitz drow@false.org
Thu Jul 17 21:42:00 GMT 2008


On Thu, Jul 17, 2008 at 06:14:59PM -0300, Thiago Jung Bauermann wrote:
> For my own education: it looks to me that this customer won the lottery
> two times here (one for pointing to a link map fake entry, and the other
> for it to be circular). How often can this happen in practice?

Beats me, but my guess is more often than you think - the value you
need for it to be circular is readily available.  For instance
something that shifts the pointer by a word will make us load l_prev
when we wanted l_next.

They definitely did lose the lottery twice though :-)

-- 
Daniel Jacobowitz
CodeSourcery



More information about the Gdb-patches mailing list