[Ksummit-2013-discuss] [ATTEND] static checking; COMPILE_TEST

Mark Brown broonie at kernel.org
Wed Jul 17 09:26:30 UTC 2013


On Tue, Jul 16, 2013 at 09:37:19PM +0200, Jiri Slaby wrote:

> * There are many static analyzers capable to check the kernel. But they
> are seldom used by the kernel developers. Or even reports from such
> tools, when provided, are ignored or even rejected by the maintainers.
> Is it possible to improve this somehow? Or do we intentionally leave

I'm not aware of any checkers that it's particularly easy to run
automatically other than sparse and to a lesser extent coccinelle (it's
a bit harder to install).  I know all my builds have sparse enabled and
people frequently send me patches based on it so it seems to have a good
enough degree of adoption and of course coccinelle seems fairly widely
used too both for checking and things like API transitions.

What are the other checkers that we should be paying attention to?

> this to the downstream (redhat, suse and others who run static analyzers
> and send patches)? But they have only limited resources and knowledge in
> the first place which leads to "fixing" false positives.

Is that really ignoring gings - if people are sending patches based on
the resuls of semantic checkers isn't that the result we're looking for?
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.linuxfoundation.org/pipermail/ksummit-2013-discuss/attachments/20130717/c3f0a685/attachment.sig>


More information about the Ksummit-2013-discuss mailing list