Re: [mm/page_alloc] 2bd8eec68f: BUG:sleeping_function_called_from_invalid_context_at_mm/gup.c

[Mel Gorman <mgorman@techsingularity.net>]

[-- Attachment #1: Type: text/plain, Size: 2000 bytes --]

On Thu, Jul 07, 2022 at 11:55:35PM +0200, Vlastimil Babka wrote:
> On 7/5/22 15:51, Oliver Sang wrote:
> > Hi Andrew Morton,
> > 
> > On Sun, Jul 03, 2022 at 01:22:09PM -0700, Andrew Morton wrote:
> >> On Sun, 3 Jul 2022 17:44:30 +0800 kernel test robot <oliver.sang@intel.com> wrote:
> >> 
> >> > FYI, we noticed the following commit (built with gcc-11):
> >> > 
> >> > commit: 2bd8eec68f740608db5ea58ecff06965228764cb ("[PATCH 7/7] mm/page_alloc: Replace local_lock with normal spinlock")
> >> > url: https://github.com/intel-lab-lkp/linux/commits/Mel-Gorman/Drain-remote-per-cpu-directly/20220613-230139
> >> > base: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git b13baccc3850ca8b8cccbf8ed9912dbaa0fdf7f3
> >> > patch link: https://lore.kernel.org/lkml/20220613125622.18628-8-mgorman(a)techsingularity.net
> >> > 
> >> 
> >> Did this test include the followup patch
> >> mm-page_alloc-replace-local_lock-with-normal-spinlock-fix.patch?
> > 
> > no, we just fetched original patch set and test upon it.
> 
> It appears you fetched v4, not v5. I noticed it from the threading of your
> report that was threaded in the v4 thread, and also the github url: above.
> In v4, pcpu_spin_trylock_irqsave() was missing an unpin, and indeed it's
> missing in the github branch you were testing:
> 

Thanks Vlastimil! This is my fault, I failed to verify that the code in
my tree, Andrew's tree and what Oliver tested were the same so no wonder I
could not find where the missing unpin was. I've gone through mm-unstable
commits be42c869b8e..4143c9b5266 and can confirm that they are now identical
to my own tree which includes Andrew's fix for the smatch warning that
Dan reported.

# git diff HEAD^..mm-pcpspinnoirq-v6r1-mmunstable | wc -l
0

The only difference between my tree and Andrew's is that there is a head
commit for "mm/page_alloc: Do not disable IRQs for per-cpu allocations"
which has been put on hold for now.

-- 
Mel Gorman
SUSE Labs