On 08/07/2013 02:36 AM, Kees Cook wrote: > On Tue, Aug 6, 2013 at 1:00 AM, Chen Gang <[email protected]> wrote: >> They are 2 related patches for setfsgid(). >> >> Patch 1 for bug fix: return the current gid when error occurs. >> Patch 2 for cleaning code: remove useless variable 'old_fsgid'. >> >> Signed-off-by: Chen Gang <[email protected]> >> -- >> kernel/sys.c | 15 +++++---------- >> 1 files changed, 5 insertions(+), 10 deletions(-) > > Making a change like this might have dramatic effects. So, a few > questions, to help better understand: > > How long as the behavior been this way on Linux? > What is the origin of the documentation that states it differently? > Do existing userspace tools already depend on the current behavior? > What specific problem will be solved by changing this? >
Firstly, your questions are necessary and valuable to get reply. But sorry, I only find it by reading code, and not quite familiar with the details (e.g. how to use it, and its history ...). I only know the original implementation has a kernel API related issue (at least now, it is), and try to send a possible fixing patch to mailing list for getting more discussion and better fixing. So, welcome any members to help reply your valuable questions (and I should also try to answer them, but at least, I really needs time to familiar about it). > Thanks, > > -Kees > Thanks. -- Chen Gang -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
