Messages in this thread | | | Date | Wed, 02 Jan 2008 22:20:20 -0500 | From | Masami Hiramatsu <> | Subject | Re: [PATCH] x86: Use is_kprobe_fault to better match usage |
| |
Harvey Harrison wrote: > On Wed, 2008-01-02 at 21:36 -0500, Masami Hiramatsu wrote: >> Hi Harvey, >> >> Harvey Harrison wrote: >>> Currently the notify_page_fault helper is used to test it the page >>> fault was caused by a kprobe causing an early return from do_page_fault. >>> >>> Change the name of the helper to is_kprobe_fault to match the usage and >>> remove the preempt_disable/enable pair around kprobe_running() with an >>> explicit test for preemption. The idea for this comes from a patch >>> by Quentin Barnes to kprobes.c >> Sure, that's right. >> However, since other architectures also have notify_page_fault(), >> I think all of those code might better be changed same time for >> maintainability. >> > > How about a static inline in linux/kprobes.h with a big comment above > about when/why the !preemptible() check is sufficient?
Hmm, fault handling depends on the architecture. But current notify_page_fault()s are very similar. I think unifying it is good idea. We will be happy to review that if you send it.
Many thanks!
> > Harvey > > >
-- Masami Hiramatsu
Software Engineer Hitachi Computer Products (America) Inc. Software Solutions Division
e-mail: mhiramat@redhat.com, masami.hiramatsu.pt@hitachi.com
| |