Re: [netfilter] e32a4dc651: BUG:using_smp_processor_id()in_preemptible

[Stefano Brivio <sbrivio@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 1860 bytes --]

On Fri, 5 Jun 2020 08:15:07 +0800
Liu Yiding <yidingx.liu@intel.com> wrote:

> Hi,
> 
> On 6/5/20 6:23 AM, Stefano Brivio wrote:
> > Hi,
> >
> > On Wed, 3 Jun 2020 16:49:34 +0800
> > kernel test robot <lkp@intel.com> wrote:
> >  
> >> Greeting,
> >>
> >> FYI, we noticed the following commit (built with gcc-7):
> >>
> >> commit: e32a4dc6512ce3c1a1920531246e7037896e510a ("netfilter: nf_tables: make sets built-in")
> >> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
> >>
> >> in testcase: kernel-selftests
> >> with following parameters:
> >>
> >> 	group: kselftests-netfilter  
> > I couldn't find this information in the report. Would it be possible to
> > have here an indication of what kselftest specifically is failing?
> > There are a number of tests in that group. I have an obvious suspicion
> > here (nft_concat_range.sh), but it would be nice to know, in general.  
> 
> Yes, it's nft_concat_range.sh caused this error.
> 
> ```
> 
> [  165.249447] # selftests: netfilter: nft_concat_range.sh
> [  165.249453]
> [  165.295336] # TEST: reported issues
> [  165.295342]
> [  165.316525] BUG: using smp_processor_id() in preemptible [00000000] 
> code: nft/6247
> [  165.319547] caller is nft_pipapo_insert+0x464/0x610 [nf_tables]
> [  165.321846] CPU: 1 PID: 6247 Comm: nft Not tainted 
> 5.6.0-rc5-01595-ge32a4dc6512ce3 #1
> 
> ```

Thanks for confirming. I think it would really be helpful to have the
specific test name in the report, in the future.

> And i think this error is caused by some kconfigs.
>
> This test was passed until we enabled all kconfigs defined in 
> netfilter/config. :-(

Then most likely this was CONFIG_NF_TABLES_SET not enabled, which goes
away with Florian's commit e32a4dc6512c as the whole feature is now
built in.

-- 
Stefano