[Ksummit-2012-discuss] [ATTEND] writeback and kernel testing
linux at roeck-us.net
Mon Jun 18 20:55:36 UTC 2012
On Mon, Jun 18, 2012 at 03:40:38PM -0400, Steven Rostedt wrote:
> On Sat, 2012-06-16 at 07:07 -0700, Guenter Roeck wrote:
> > I am running a nightly sequence of builds on my tree, for as many targets as
> > possible. allyesconfig, allmodconfig, and a large number of randconfigs. That
> > helps me find most of the problems I had early on, which only show up in some
> > configurations. That combined with a personal rule to only push code upstream
> > which has been in my local tree for at least one test cycle helps a lot to avoid
> > the embarrassment of breaking linux-next or Linus' tree.
> This sounds exactly what ktest does. And you can get it from the local
> kernel tree today. I even added sample conf files in the 3.5 merge
> Which includes a bunch of stuff in the examples/includes directory.
> Which holds a bunch of tests to cross-compile most archs.
I keep saying: One never stops learning.
> I guess I should propose talking about how we can incorporate these
> various utilities such that everyone can benefit. And to spread the word
> about what is out there.
Would be great.
I have another suggestion: Can we get some code analysis tool to run over the
kernel on a regular basis ? Or, if we do, some more ?
I just happened to get a Coverity result from the 3.0 kernel. Most of the bugs
it finds have been corrected in the current kernel, but some are still there.
The most interesting is in ocfs2_validate_and_adjust_move_goal():
bg = NULL;
if (range->me_goal == le64_to_cpu(bg->bg_blkno))
range->me_goal += c_to_b;
No idea how this can work. Either I am missing something, or it proves the point
that it _would_ be helpful to get some better test coverage (and/or maybe that
the above code is never called).
More information about the Ksummit-2012-discuss