From: kernel test robot <lkp@intel.com>
To: Minchan Kim <minchan@kernel.org>,
Andrew Morton <akpm@linux-foundation.org>
Cc: kbuild-all@lists.01.org,
Linux Memory Management List <linux-mm@kvack.org>,
LKML <linux-kernel@vger.kernel.org>,
joaodias@google.com, surenb@google.com, cgoldswo@codeaurora.org,
willy@infradead.org, mhocko@suse.com, david@redhat.com,
vbabka@suse.cz
Subject: Re: [PATCH 1/2] mm: disable LRU pagevec during the migration temporarily
Date: Wed, 3 Mar 2021 23:11:48 +0800 [thread overview]
Message-ID: <202103032333.bv4y27Ol-lkp@intel.com> (raw)
In-Reply-To: <20210302210949.2440120-1-minchan@kernel.org>
[-- Attachment #1: Type: text/plain, Size: 11653 bytes --]
Hi Minchan,
I love your patch! Yet something to improve:
[auto build test ERROR on block/for-next]
[also build test ERROR on linux/master linus/master v5.12-rc1 next-20210303]
[cannot apply to hnaz-linux-mm/master]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Minchan-Kim/mm-disable-LRU-pagevec-during-the-migration-temporarily/20210303-191809
base: https://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux-block.git for-next
config: nios2-randconfig-r032-20210303 (attached as .config)
compiler: nios2-linux-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/6e669beb75caae92c613a012734b1a2dc9485524
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Minchan-Kim/mm-disable-LRU-pagevec-during-the-migration-temporarily/20210303-191809
git checkout 6e669beb75caae92c613a012734b1a2dc9485524
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=nios2
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
All errors (new ones prefixed by >>):
mm/swap.c:59:6: warning: no previous prototype for 'lru_cache_disabled' [-Wmissing-prototypes]
59 | bool lru_cache_disabled(void)
| ^~~~~~~~~~~~~~~~~~
>> mm/swap.c:871:6: error: conflicting types for 'lru_add_drain_all'
871 | void lru_add_drain_all(void)
| ^~~~~~~~~~~~~~~~~
In file included from mm/swap.c:20:
include/linux/swap.h:347:13: note: previous declaration of 'lru_add_drain_all' was here
347 | extern void lru_add_drain_all(bool force_all_cpus);
| ^~~~~~~~~~~~~~~~~
Kconfig warnings: (for reference only)
WARNING: unmet direct dependencies detected for SERIAL_CORE_CONSOLE
Depends on TTY && HAS_IOMEM
Selected by
- EARLY_PRINTK
vim +/lru_add_drain_all +871 mm/swap.c
053837fce7aa79 Nick Piggin 2006-01-18 768
9852a7212324fd Michal Hocko 2018-01-31 769 /*
9852a7212324fd Michal Hocko 2018-01-31 770 * Doesn't need any cpu hotplug locking because we do rely on per-cpu
9852a7212324fd Michal Hocko 2018-01-31 771 * kworkers being shut down before our page_alloc_cpu_dead callback is
9852a7212324fd Michal Hocko 2018-01-31 772 * executed on the offlined cpu.
9852a7212324fd Michal Hocko 2018-01-31 773 * Calling this function with cpu hotplug locks held can actually lead
9852a7212324fd Michal Hocko 2018-01-31 774 * to obscure indirect dependencies via WQ context.
9852a7212324fd Michal Hocko 2018-01-31 775 */
6e669beb75caae Minchan Kim 2021-03-02 776 void lru_add_drain_all(bool force_all_cpus)
053837fce7aa79 Nick Piggin 2006-01-18 777 {
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 778 /*
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 779 * lru_drain_gen - Global pages generation number
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 780 *
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 781 * (A) Definition: global lru_drain_gen = x implies that all generations
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 782 * 0 < n <= x are already *scheduled* for draining.
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 783 *
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 784 * This is an optimization for the highly-contended use case where a
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 785 * user space workload keeps constantly generating a flow of pages for
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 786 * each CPU.
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 787 */
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 788 static unsigned int lru_drain_gen;
5fbc461636c32e Chris Metcalf 2013-09-12 789 static struct cpumask has_work;
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 790 static DEFINE_MUTEX(lock);
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 791 unsigned cpu, this_gen;
5fbc461636c32e Chris Metcalf 2013-09-12 792
ce612879ddc78e Michal Hocko 2017-04-07 793 /*
ce612879ddc78e Michal Hocko 2017-04-07 794 * Make sure nobody triggers this path before mm_percpu_wq is fully
ce612879ddc78e Michal Hocko 2017-04-07 795 * initialized.
ce612879ddc78e Michal Hocko 2017-04-07 796 */
ce612879ddc78e Michal Hocko 2017-04-07 797 if (WARN_ON(!mm_percpu_wq))
ce612879ddc78e Michal Hocko 2017-04-07 798 return;
ce612879ddc78e Michal Hocko 2017-04-07 799
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 800 /*
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 801 * Guarantee pagevec counter stores visible by this CPU are visible to
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 802 * other CPUs before loading the current drain generation.
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 803 */
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 804 smp_mb();
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 805
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 806 /*
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 807 * (B) Locally cache global LRU draining generation number
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 808 *
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 809 * The read barrier ensures that the counter is loaded before the mutex
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 810 * is taken. It pairs with smp_mb() inside the mutex critical section
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 811 * at (D).
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 812 */
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 813 this_gen = smp_load_acquire(&lru_drain_gen);
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 814
5fbc461636c32e Chris Metcalf 2013-09-12 815 mutex_lock(&lock);
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 816
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 817 /*
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 818 * (C) Exit the draining operation if a newer generation, from another
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 819 * lru_add_drain_all(), was already scheduled for draining. Check (A).
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 820 */
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 821 if (unlikely(this_gen != lru_drain_gen))
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 822 goto done;
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 823
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 824 /*
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 825 * (D) Increment global generation number
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 826 *
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 827 * Pairs with smp_load_acquire() at (B), outside of the critical
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 828 * section. Use a full memory barrier to guarantee that the new global
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 829 * drain generation number is stored before loading pagevec counters.
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 830 *
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 831 * This pairing must be done here, before the for_each_online_cpu loop
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 832 * below which drains the page vectors.
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 833 *
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 834 * Let x, y, and z represent some system CPU numbers, where x < y < z.
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 835 * Assume CPU #z is is in the middle of the for_each_online_cpu loop
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 836 * below and has already reached CPU #y's per-cpu data. CPU #x comes
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 837 * along, adds some pages to its per-cpu vectors, then calls
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 838 * lru_add_drain_all().
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 839 *
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 840 * If the paired barrier is done at any later step, e.g. after the
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 841 * loop, CPU #x will just exit at (C) and miss flushing out all of its
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 842 * added pages.
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 843 */
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 844 WRITE_ONCE(lru_drain_gen, lru_drain_gen + 1);
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 845 smp_mb();
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 846
5fbc461636c32e Chris Metcalf 2013-09-12 847 cpumask_clear(&has_work);
5fbc461636c32e Chris Metcalf 2013-09-12 848 for_each_online_cpu(cpu) {
5fbc461636c32e Chris Metcalf 2013-09-12 849 struct work_struct *work = &per_cpu(lru_add_drain_work, cpu);
5fbc461636c32e Chris Metcalf 2013-09-12 850
6e669beb75caae Minchan Kim 2021-03-02 851 if (force_all_cpus ||
6e669beb75caae Minchan Kim 2021-03-02 852 pagevec_count(&per_cpu(lru_pvecs.lru_add, cpu)) ||
7e0cc01ea181c6 Qian Cai 2020-08-14 853 data_race(pagevec_count(&per_cpu(lru_rotate.pvec, cpu))) ||
b01b2141999936 Ingo Molnar 2020-05-27 854 pagevec_count(&per_cpu(lru_pvecs.lru_deactivate_file, cpu)) ||
b01b2141999936 Ingo Molnar 2020-05-27 855 pagevec_count(&per_cpu(lru_pvecs.lru_deactivate, cpu)) ||
b01b2141999936 Ingo Molnar 2020-05-27 856 pagevec_count(&per_cpu(lru_pvecs.lru_lazyfree, cpu)) ||
5fbc461636c32e Chris Metcalf 2013-09-12 857 need_activate_page_drain(cpu)) {
5fbc461636c32e Chris Metcalf 2013-09-12 858 INIT_WORK(work, lru_add_drain_per_cpu);
ce612879ddc78e Michal Hocko 2017-04-07 859 queue_work_on(cpu, mm_percpu_wq, work);
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 860 __cpumask_set_cpu(cpu, &has_work);
5fbc461636c32e Chris Metcalf 2013-09-12 861 }
5fbc461636c32e Chris Metcalf 2013-09-12 862 }
5fbc461636c32e Chris Metcalf 2013-09-12 863
5fbc461636c32e Chris Metcalf 2013-09-12 864 for_each_cpu(cpu, &has_work)
5fbc461636c32e Chris Metcalf 2013-09-12 865 flush_work(&per_cpu(lru_add_drain_work, cpu));
5fbc461636c32e Chris Metcalf 2013-09-12 866
eef1a429f234f8 Konstantin Khlebnikov 2019-11-30 867 done:
5fbc461636c32e Chris Metcalf 2013-09-12 868 mutex_unlock(&lock);
053837fce7aa79 Nick Piggin 2006-01-18 869 }
6ea183d60c4695 Michal Hocko 2019-02-20 870 #else
6ea183d60c4695 Michal Hocko 2019-02-20 @871 void lru_add_drain_all(void)
6ea183d60c4695 Michal Hocko 2019-02-20 872 {
6ea183d60c4695 Michal Hocko 2019-02-20 873 lru_add_drain();
6ea183d60c4695 Michal Hocko 2019-02-20 874 }
6446a5131e24a8 Ahmed S. Darwish 2020-08-27 875 #endif /* CONFIG_SMP */
053837fce7aa79 Nick Piggin 2006-01-18 876
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 24755 bytes --]
next prev parent reply other threads:[~2021-03-03 15:13 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20210302210949.2440120-1-minchan@kernel.org>
2021-03-03 12:49 ` [PATCH 1/2] mm: disable LRU pagevec during the migration temporarily Michal Hocko
2021-03-03 20:23 ` Minchan Kim
2021-03-04 8:07 ` David Hildenbrand
2021-03-04 15:55 ` Minchan Kim
2021-03-05 16:06 ` Michal Hocko
2021-03-05 20:26 ` Minchan Kim
2021-03-03 13:38 ` kernel test robot
2021-03-03 15:11 ` kernel test robot [this message]
2021-03-03 18:12 ` kernel test robot
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=202103032333.bv4y27Ol-lkp@intel.com \
--to=lkp@intel.com \
--cc=akpm@linux-foundation.org \
--cc=cgoldswo@codeaurora.org \
--cc=david@redhat.com \
--cc=joaodias@google.com \
--cc=kbuild-all@lists.01.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@suse.com \
--cc=minchan@kernel.org \
--cc=surenb@google.com \
--cc=vbabka@suse.cz \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).