Messages in this thread | | | Date | Fri, 8 Mar 2013 18:55:14 -0500 | From | Dave Jones <> | Subject | Re: BUG_ON(nd->inode != parent->d_inode); |
| |
On Fri, Mar 08, 2013 at 03:45:37PM -0800, Linus Torvalds wrote: > On Fri, Mar 8, 2013 at 3:30 PM, Dave Jones <davej@redhat.com> wrote: > > [ 100.729401] nd->last.name = (\xffffffd2.W._.N.".\xfffffffe.\xffffff80.^N.?.\xffffffe4.E.8.g.\xffffffd2.N.\xffffffb6.^G.\xfffffff1.\xffffffcc.U.\xffffffda.^_.h.^M.1.\xffffffc5.\xffffff82.%.B.\xffffffe0.\xffffffad.^U.8.^L.c.Z.^K.\xffffffe4.h.J.\xffffffc8.\xffffffad.\xffffff83./.\xffffff80.\xffffffd1.\xffffffe5.\xffffff87.\xffffffc3.\xffffffb2.\xffffffdc.\xffffff9d.\xffffffd1.E.\xffffffab.^B. > > You're printing out a 'char' with %02x, aren't you? And then the sign > extensions gives you six extra 'f' character every time the char is > negative. > > If using %02x, make sure you use 'unsigned char', or "& 255". > > Depending on just what is going on, the "nd->last.name" thing really > is probably a valid string. Strictly speaking it's not necessarily > NUL-terminated, though, and should be printed out to a maximum of > "nd->last.len" characters (but the whole path will be NUL-terminated, > so it's all ok, you just might print out more than one component) > > Linus
That one was printed out with %s
Dave
| |