From: kbuild test robot <fengguang.wu@intel.com>
To: Waiman Long <longman@redhat.com>
Cc: kbuild-all@01.org, linux-mm@kvack.org,
Michal Hocko <mhocko@suse.com>,
Andrew Morton <akpm@linux-foundation.org>
Subject: [memcg:since-4.12 539/540] mm/compaction.c:469:8: error: implicit declaration of function 'pageblock_skip_persistent'
Date: Sat, 19 Aug 2017 00:42:40 +0800 [thread overview]
Message-ID: <201708190034.TmrRSDV7%fengguang.wu@intel.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 12421 bytes --]
tree: https://git.kernel.org/pub/scm/linux/kernel/git/mhocko/mm.git since-4.12
head: ba5e8c23db5729ebdbafad983b07434c829cf5b6
commit: 500539d3686a835f6a9740ffc38bed5d74951a64 [539/540] debugobjects: make kmemleak ignore debug objects
config: i386-randconfig-s0-08141822 (attached as .config)
compiler: gcc-6 (Debian 6.2.0-3) 6.2.0 20160901
reproduce:
git checkout 500539d3686a835f6a9740ffc38bed5d74951a64
# save the attached .config to linux build tree
make ARCH=i386
All errors (new ones prefixed by >>):
mm/compaction.c: In function 'isolate_freepages_block':
>> mm/compaction.c:469:8: error: implicit declaration of function 'pageblock_skip_persistent' [-Werror=implicit-function-declaration]
if (pageblock_skip_persistent(page, order)) {
^~~~~~~~~~~~~~~~~~~~~~~~~
>> mm/compaction.c:470:5: error: implicit declaration of function 'set_pageblock_skip' [-Werror=implicit-function-declaration]
set_pageblock_skip(page);
^~~~~~~~~~~~~~~~~~
cc1: some warnings being treated as errors
vim +/pageblock_skip_persistent +469 mm/compaction.c
be976572 Vlastimil Babka 2014-06-04 417
c67fe375 Mel Gorman 2012-08-21 418 /*
9e4be470 Jerome Marchand 2013-11-12 419 * Isolate free pages onto a private freelist. If @strict is true, will abort
9e4be470 Jerome Marchand 2013-11-12 420 * returning 0 on any invalid PFNs or non-free pages inside of the pageblock
9e4be470 Jerome Marchand 2013-11-12 421 * (even though it may still end up isolating some pages).
85aa125f Michal Nazarewicz 2012-01-30 422 */
f40d1e42 Mel Gorman 2012-10-08 423 static unsigned long isolate_freepages_block(struct compact_control *cc,
e14c720e Vlastimil Babka 2014-10-09 424 unsigned long *start_pfn,
85aa125f Michal Nazarewicz 2012-01-30 425 unsigned long end_pfn,
85aa125f Michal Nazarewicz 2012-01-30 426 struct list_head *freelist,
85aa125f Michal Nazarewicz 2012-01-30 427 bool strict)
748446bb Mel Gorman 2010-05-24 428 {
b7aba698 Mel Gorman 2011-01-13 429 int nr_scanned = 0, total_isolated = 0;
bb13ffeb Mel Gorman 2012-10-08 430 struct page *cursor, *valid_page = NULL;
b8b2d825 Xiubo Li 2014-10-09 431 unsigned long flags = 0;
f40d1e42 Mel Gorman 2012-10-08 432 bool locked = false;
e14c720e Vlastimil Babka 2014-10-09 433 unsigned long blockpfn = *start_pfn;
66c64223 Joonsoo Kim 2016-07-26 434 unsigned int order;
748446bb Mel Gorman 2010-05-24 435
748446bb Mel Gorman 2010-05-24 436 cursor = pfn_to_page(blockpfn);
748446bb Mel Gorman 2010-05-24 437
f40d1e42 Mel Gorman 2012-10-08 438 /* Isolate free pages. */
748446bb Mel Gorman 2010-05-24 439 for (; blockpfn < end_pfn; blockpfn++, cursor++) {
66c64223 Joonsoo Kim 2016-07-26 440 int isolated;
748446bb Mel Gorman 2010-05-24 441 struct page *page = cursor;
748446bb Mel Gorman 2010-05-24 442
8b44d279 Vlastimil Babka 2014-10-09 443 /*
8b44d279 Vlastimil Babka 2014-10-09 444 * Periodically drop the lock (if held) regardless of its
8b44d279 Vlastimil Babka 2014-10-09 445 * contention, to give chance to IRQs. Abort if fatal signal
8b44d279 Vlastimil Babka 2014-10-09 446 * pending or async compaction detects need_resched()
8b44d279 Vlastimil Babka 2014-10-09 447 */
8b44d279 Vlastimil Babka 2014-10-09 448 if (!(blockpfn % SWAP_CLUSTER_MAX)
8b44d279 Vlastimil Babka 2014-10-09 449 && compact_unlock_should_abort(&cc->zone->lock, flags,
8b44d279 Vlastimil Babka 2014-10-09 450 &locked, cc))
8b44d279 Vlastimil Babka 2014-10-09 451 break;
8b44d279 Vlastimil Babka 2014-10-09 452
b7aba698 Mel Gorman 2011-01-13 453 nr_scanned++;
f40d1e42 Mel Gorman 2012-10-08 454 if (!pfn_valid_within(blockpfn))
2af120bc Laura Abbott 2014-03-10 455 goto isolate_fail;
2af120bc Laura Abbott 2014-03-10 456
bb13ffeb Mel Gorman 2012-10-08 457 if (!valid_page)
bb13ffeb Mel Gorman 2012-10-08 458 valid_page = page;
9fcd6d2e Vlastimil Babka 2015-09-08 459
9fcd6d2e Vlastimil Babka 2015-09-08 460 /*
9fcd6d2e Vlastimil Babka 2015-09-08 461 * For compound pages such as THP and hugetlbfs, we can save
9fcd6d2e Vlastimil Babka 2015-09-08 462 * potentially a lot of iterations if we skip them at once.
9fcd6d2e Vlastimil Babka 2015-09-08 463 * The check is racy, but we can consider only valid values
9fcd6d2e Vlastimil Babka 2015-09-08 464 * and the only danger is skipping too much.
9fcd6d2e Vlastimil Babka 2015-09-08 465 */
9fcd6d2e Vlastimil Babka 2015-09-08 466 if (PageCompound(page)) {
a93d0214 David Rientjes 2017-08-18 467 const unsigned int order = compound_order(page);
9fcd6d2e Vlastimil Babka 2015-09-08 468
a93d0214 David Rientjes 2017-08-18 @469 if (pageblock_skip_persistent(page, order)) {
a93d0214 David Rientjes 2017-08-18 @470 set_pageblock_skip(page);
a93d0214 David Rientjes 2017-08-18 471 blockpfn = end_pfn;
a93d0214 David Rientjes 2017-08-18 472 } else if (likely(order < MAX_ORDER)) {
a93d0214 David Rientjes 2017-08-18 473 blockpfn += (1UL << order) - 1;
a93d0214 David Rientjes 2017-08-18 474 cursor += (1UL << order) - 1;
9fcd6d2e Vlastimil Babka 2015-09-08 475 }
9fcd6d2e Vlastimil Babka 2015-09-08 476 goto isolate_fail;
9fcd6d2e Vlastimil Babka 2015-09-08 477 }
9fcd6d2e Vlastimil Babka 2015-09-08 478
f40d1e42 Mel Gorman 2012-10-08 479 if (!PageBuddy(page))
2af120bc Laura Abbott 2014-03-10 480 goto isolate_fail;
748446bb Mel Gorman 2010-05-24 481
f40d1e42 Mel Gorman 2012-10-08 482 /*
69b7189f Vlastimil Babka 2014-10-09 483 * If we already hold the lock, we can skip some rechecking.
69b7189f Vlastimil Babka 2014-10-09 484 * Note that if we hold the lock now, checked_pageblock was
69b7189f Vlastimil Babka 2014-10-09 485 * already set in some previous iteration (or strict is true),
69b7189f Vlastimil Babka 2014-10-09 486 * so it is correct to skip the suitable migration target
69b7189f Vlastimil Babka 2014-10-09 487 * recheck as well.
69b7189f Vlastimil Babka 2014-10-09 488 */
69b7189f Vlastimil Babka 2014-10-09 489 if (!locked) {
69b7189f Vlastimil Babka 2014-10-09 490 /*
f40d1e42 Mel Gorman 2012-10-08 491 * The zone lock must be held to isolate freepages.
f40d1e42 Mel Gorman 2012-10-08 492 * Unfortunately this is a very coarse lock and can be
f40d1e42 Mel Gorman 2012-10-08 493 * heavily contended if there are parallel allocations
f40d1e42 Mel Gorman 2012-10-08 494 * or parallel compactions. For async compaction do not
f40d1e42 Mel Gorman 2012-10-08 495 * spin on the lock and we acquire the lock as late as
f40d1e42 Mel Gorman 2012-10-08 496 * possible.
f40d1e42 Mel Gorman 2012-10-08 497 */
8b44d279 Vlastimil Babka 2014-10-09 498 locked = compact_trylock_irqsave(&cc->zone->lock,
8b44d279 Vlastimil Babka 2014-10-09 499 &flags, cc);
f40d1e42 Mel Gorman 2012-10-08 500 if (!locked)
f40d1e42 Mel Gorman 2012-10-08 501 break;
f40d1e42 Mel Gorman 2012-10-08 502
f40d1e42 Mel Gorman 2012-10-08 503 /* Recheck this is a buddy page under lock */
f40d1e42 Mel Gorman 2012-10-08 504 if (!PageBuddy(page))
2af120bc Laura Abbott 2014-03-10 505 goto isolate_fail;
69b7189f Vlastimil Babka 2014-10-09 506 }
748446bb Mel Gorman 2010-05-24 507
66c64223 Joonsoo Kim 2016-07-26 508 /* Found a free page, will break it into order-0 pages */
66c64223 Joonsoo Kim 2016-07-26 509 order = page_order(page);
66c64223 Joonsoo Kim 2016-07-26 510 isolated = __isolate_free_page(page, order);
a4f04f2c David Rientjes 2016-06-24 511 if (!isolated)
a4f04f2c David Rientjes 2016-06-24 512 break;
66c64223 Joonsoo Kim 2016-07-26 513 set_page_private(page, order);
a4f04f2c David Rientjes 2016-06-24 514
748446bb Mel Gorman 2010-05-24 515 total_isolated += isolated;
a4f04f2c David Rientjes 2016-06-24 516 cc->nr_freepages += isolated;
66c64223 Joonsoo Kim 2016-07-26 517 list_add_tail(&page->lru, freelist);
66c64223 Joonsoo Kim 2016-07-26 518
a4f04f2c David Rientjes 2016-06-24 519 if (!strict && cc->nr_migratepages <= cc->nr_freepages) {
932ff6bb Joonsoo Kim 2015-02-12 520 blockpfn += isolated;
932ff6bb Joonsoo Kim 2015-02-12 521 break;
932ff6bb Joonsoo Kim 2015-02-12 522 }
a4f04f2c David Rientjes 2016-06-24 523 /* Advance to the end of split page */
748446bb Mel Gorman 2010-05-24 524 blockpfn += isolated - 1;
748446bb Mel Gorman 2010-05-24 525 cursor += isolated - 1;
2af120bc Laura Abbott 2014-03-10 526 continue;
2af120bc Laura Abbott 2014-03-10 527
2af120bc Laura Abbott 2014-03-10 528 isolate_fail:
2af120bc Laura Abbott 2014-03-10 529 if (strict)
2af120bc Laura Abbott 2014-03-10 530 break;
2af120bc Laura Abbott 2014-03-10 531 else
2af120bc Laura Abbott 2014-03-10 532 continue;
2af120bc Laura Abbott 2014-03-10 533
748446bb Mel Gorman 2010-05-24 534 }
748446bb Mel Gorman 2010-05-24 535
a4f04f2c David Rientjes 2016-06-24 536 if (locked)
a4f04f2c David Rientjes 2016-06-24 537 spin_unlock_irqrestore(&cc->zone->lock, flags);
a4f04f2c David Rientjes 2016-06-24 538
9fcd6d2e Vlastimil Babka 2015-09-08 539 /*
9fcd6d2e Vlastimil Babka 2015-09-08 540 * There is a tiny chance that we have read bogus compound_order(),
9fcd6d2e Vlastimil Babka 2015-09-08 541 * so be careful to not go outside of the pageblock.
9fcd6d2e Vlastimil Babka 2015-09-08 542 */
9fcd6d2e Vlastimil Babka 2015-09-08 543 if (unlikely(blockpfn > end_pfn))
9fcd6d2e Vlastimil Babka 2015-09-08 544 blockpfn = end_pfn;
9fcd6d2e Vlastimil Babka 2015-09-08 545
e34d85f0 Joonsoo Kim 2015-02-11 546 trace_mm_compaction_isolate_freepages(*start_pfn, blockpfn,
e34d85f0 Joonsoo Kim 2015-02-11 547 nr_scanned, total_isolated);
e34d85f0 Joonsoo Kim 2015-02-11 548
e14c720e Vlastimil Babka 2014-10-09 549 /* Record how far we have got within the block */
e14c720e Vlastimil Babka 2014-10-09 550 *start_pfn = blockpfn;
e14c720e Vlastimil Babka 2014-10-09 551
f40d1e42 Mel Gorman 2012-10-08 552 /*
f40d1e42 Mel Gorman 2012-10-08 553 * If strict isolation is requested by CMA then check that all the
f40d1e42 Mel Gorman 2012-10-08 554 * pages requested were isolated. If there were any failures, 0 is
f40d1e42 Mel Gorman 2012-10-08 555 * returned and CMA will fail.
f40d1e42 Mel Gorman 2012-10-08 556 */
2af120bc Laura Abbott 2014-03-10 557 if (strict && blockpfn < end_pfn)
f40d1e42 Mel Gorman 2012-10-08 558 total_isolated = 0;
f40d1e42 Mel Gorman 2012-10-08 559
bb13ffeb Mel Gorman 2012-10-08 560 /* Update the pageblock-skip if the whole pageblock was scanned */
bb13ffeb Mel Gorman 2012-10-08 561 if (blockpfn == end_pfn)
edc2ca61 Vlastimil Babka 2014-10-09 562 update_pageblock_skip(cc, valid_page, total_isolated, false);
bb13ffeb Mel Gorman 2012-10-08 563
7f354a54 David Rientjes 2017-02-22 564 cc->total_free_scanned += nr_scanned;
397487db Mel Gorman 2012-10-19 565 if (total_isolated)
010fc29a Minchan Kim 2012-12-20 566 count_compact_events(COMPACTISOLATED, total_isolated);
748446bb Mel Gorman 2010-05-24 567 return total_isolated;
748446bb Mel Gorman 2010-05-24 568 }
748446bb Mel Gorman 2010-05-24 569
:::::: The code at line 469 was first introduced by commit
:::::: a93d021466d5ab3b2598e59e3a055c577a3f120b mm, compaction: persistently skip hugetlbfs pageblocks
:::::: TO: David Rientjes <rientjes@google.com>
:::::: CC: Michal Hocko <mhocko@suse.com>
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 27288 bytes --]
next reply other threads:[~2017-08-18 16:43 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-18 16:42 kbuild test robot [this message]
2017-08-18 16:57 ` Waiman Long
2017-08-18 21:03 ` Andrew Morton
2017-08-21 6:38 ` Michal Hocko
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=201708190034.TmrRSDV7%fengguang.wu@intel.com \
--to=fengguang.wu@intel.com \
--cc=akpm@linux-foundation.org \
--cc=kbuild-all@01.org \
--cc=linux-mm@kvack.org \
--cc=longman@redhat.com \
--cc=mhocko@suse.com \
/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