Messages in this thread | | | Date | Tue, 24 Mar 2015 08:59:25 +0100 | From | Petr Mladek <> | Subject | Re: [PATCH v3] kprobes: Disable Kprobe when ftrace arming fails |
| |
On Mon 2015-03-23 23:36:00, Jiri Kosina wrote: > On Mon, 23 Mar 2015, Ingo Molnar wrote: > > > > > > wth is a 'universe' in this context? > > > > > > > > We use the term "universe" to define whether the system or task uses > > > > original or patched functions. It is especially important for patches > > > > that modify semantic of functions. They need more complex consistency > > > > model. It defines when it is safe time for the system or task to start > > > > using the new functions (switch to the new universe). > > > > > > > > In theory, different tasks might be in more universes if more patches are > > > > being applied. In practice, we deal with only two universes. The trick is > > > > that we allow to add new patch only when the whole system has switched > > > > to the previous one. > > > > > > > > > > Is this terminology documented anywhere upstream yet? > > > > Even if it was documented (it isn't), > > There is no point in documenting it upstream, as no upstream code is using > this "universe" notation. It should just be removed from original Petr's > changelog and that's it.
Note that the term "universe" is not used in the commit message => no change is needed.
I used the term to answer the question about interference between Jprobes, ftrace, and live patches. I am sorry again. I should have described it in a broader context.
Anyway, please, let me to repeat. This patch makes sense even without live patching. The ftrace operations were there even before and might have failed.
Best Regards, Petr
| |