* [ammarfaizi2-block:dhowells/linux-fs/netfs-maple 39/44] fs/netfs/buffered_flush.c:545:56: sparse: sparse: Using plain integer as NULL pointer
@ 2022-04-23 13:56 kernel test robot
0 siblings, 0 replies; only message in thread
From: kernel test robot @ 2022-04-23 13:56 UTC (permalink / raw)
To: David Howells; +Cc: kbuild-all, GNU/Weeb Mailing List, linux-kernel
tree: https://github.com/ammarfaizi2/linux-block dhowells/linux-fs/netfs-maple
head: 931e50676c6598d0eda1954ead465519ff91874d
commit: 42a9ba6b52c29bc596f92921805d734c19a34cb3 [39/44] netfs: Generate a write request from ->writepages()
config: arc-randconfig-s032-20220422 (https://download.01.org/0day-ci/archive/20220423/[email protected]/config)
compiler: arc-elf-gcc (GCC) 11.2.0
reproduce:
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# apt-get install sparse
# sparse version: v0.6.4-dirty
# https://github.com/ammarfaizi2/linux-block/commit/42a9ba6b52c29bc596f92921805d734c19a34cb3
git remote add ammarfaizi2-block https://github.com/ammarfaizi2/linux-block
git fetch --no-tags ammarfaizi2-block dhowells/linux-fs/netfs-maple
git checkout 42a9ba6b52c29bc596f92921805d734c19a34cb3
# save the config file
mkdir build_dir && cp config build_dir/.config
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-11.2.0 make.cross C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' O=build_dir ARCH=arc SHELL=/bin/bash fs/netfs/
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <[email protected]>
sparse warnings: (new ones prefixed by >>)
>> fs/netfs/buffered_flush.c:545:56: sparse: sparse: Using plain integer as NULL pointer
fs/netfs/buffered_flush.c:595:48: sparse: sparse: Using plain integer as NULL pointer
vim +545 fs/netfs/buffered_flush.c
437
438 /*
439 * Flush some of the dirty queue, transforming a part of a sequence of dirty
440 * regions into a block we can flush.
441 *
442 * A number of things constrain us:
443 * - The region we write out should not be undergoing modification
444 * - We may need to expand or split the region for a number of reasons:
445 * - Filesystem storage block/object size
446 * - Filesystem RPC size (wsize)
447 * - Cache block size
448 * - Cache DIO block size
449 * - Crypto/compression block size
450 *
451 * This may be entered multiple times simultaneously. Automatic flushing by
452 * the VM is serialised on I_SYNC, but things like fsync() may enter multiple
453 * times simultaneously.
454 */
455 static int netfs_flush_dirty(struct netfs_io_request *wreq,
456 struct writeback_control *wbc,
457 struct netfs_i_context *ctx,
458 struct ma_state *mas,
459 pgoff_t *_first, pgoff_t last,
460 struct netfs_dirty_region *spares[2])
461 {
462 struct netfs_dirty_region *region;
463 struct folio *folio;
464 unsigned long long end;
465 pgoff_t first = *_first;
466 pgoff_t csize = 1UL << ctx->cache_order;
467 long ret;
468
469 XA_STATE(xas, &wreq->mapping->i_pages, 0);
470
471 /* Round out the range we're looking through to accommodate whole cache
472 * blocks. The cache may only be able to store blocks of that size, in
473 * which case we may need to add non-dirty pages to the buffer too.
474 */
475 if (ctx->cache_order) {
476 first = round_down(first, csize);
477 last = round_up_incl(last, csize);
478 }
479
480 _enter("%lx-%lx", first, last);
481
482 rcu_read_lock();
483 mtree_lock(&ctx->dirty_regions);
484
485 /* Find the first dirty region that overlaps the requested range */
486 mas_set(mas, first);
487 do {
488 region = mas_find(mas, last);
489 if (!region)
490 goto found_nothing;
491 } while (netfs_mas_is_flushing(region) ||
492 (netfs_mas_is_valid(region) && region->waiting_on_wb));
493
494 _debug("query D=%x %lx-%lx",
495 netfs_mas_is_valid(region) ? region->debug_id : 0,
496 mas->index, mas->last);
497
498 wreq->first = max(mas->index, first);
499 if (wreq->first > 0) {
500 /* The first folio might extend backwards beyond the start of
501 * the proposed region - in which case we need to include that
502 * also. But at least, in such a case, the folio size has to
503 * be an integer multiple of the cache blocksize.
504 */
505 if (mas->index < wreq->first) {
506 _debug("check folio %lx", wreq->first);
507 xas_set(&xas, wreq->first);
508 do {
509 xas_reset(&xas);
510 folio = xas_load(&xas);
511 } while (xas_retry(&xas, folio));
512
513 if (folio && !xa_is_value(folio)) {
514 /* A region span *should not* end in the middle
515 * of a folio.
516 */
517 BUG_ON(folio->index < mas->index);
518 if (folio->index < wreq->first) {
519 wreq->first = folio->index;
520 mas_set_range(mas, wreq->first, mas->last);
521 }
522 }
523 }
524
525 if (mas->index < wreq->first) {
526 pgoff_t saved_last = mas->last;
527 _debug("splitf %lx-%lx %lx", mas->index, mas->last, first);
528 netfs_split_off_front(ctx, mas, region, &spares[0], first - 1,
529 netfs_dirty_trace_split_off_front);
530 mas_set_range(mas, first, saved_last);
531 }
532
533 wreq->last = mas->last;
534 }
535
536
537 end = wreq->start = wreq->first * PAGE_SIZE;
538 while (mas->last < last) {
539 _debug("flip %lx-%lx", mas->index, mas->last);
540 wreq->last = mas->last;
541 mas_store(mas, netfs_mas_set_flushing(region));
542 if (region != NETFS_COPY_TO_CACHE) {
543 __set_bit(NETFS_RREQ_UPLOAD_TO_SERVER, &wreq->flags);
544 list_add_tail(®ion->flush_link, &wreq->regions);
> 545 trace_netfs_dirty(ctx, region, 0, mas->index, mas->last,
546 netfs_dirty_trace_flush);
547 end = region->to;
548 }
549
550 region = mas_next(mas, mas->last + 1);
551 if (!region || netfs_mas_is_flushing(region) ||
552 region->waiting_on_wb)
553 goto no_more;
554 if (mas->last >= last)
555 break;
556 _debug("query+ D=%x %lx-%lx",
557 netfs_mas_is_valid(region) ? region->debug_id : 0,
558 mas->index, mas->last);
559 }
560
561 /* Deal with the region we're looking at exceeding the specified range.
562 * In such a case, we need to split the region - and the last folio may
563 * extend beyond the end of the proposed region - in which case we need
564 * to include that also. And, again, the folio size has to be an
565 * integer multiple of the cache blocksize.
566 */
567 if (mas->last > last) {
568 xas_set(&xas, last);
569 do {
570 xas_reset(&xas);
571 folio = xas_load(&xas);
572 } while (xas_retry(&xas, folio));
573
574 if (folio && !xa_is_value(folio)) {
575 pgoff_t flast = folio_next_index(folio) - 1;
576
577 _debug("flast %lx %lx %lx", flast, mas->last, last);
578 /* A region span *should not* end in the middle of a folio. */
579 BUG_ON(flast > mas->last);
580 if (flast > last) {
581 last = flast;
582 mas_set_range(mas, mas->index, last);
583 }
584 }
585
586 region = netfs_split_off_front(ctx, mas, region, &spares[1], last,
587 netfs_dirty_trace_split_off_back);
588 }
589
590 wreq->last = mas->last;
591 mas_store(mas, netfs_mas_set_flushing(region));
592 if (region != NETFS_COPY_TO_CACHE) {
593 __set_bit(NETFS_RREQ_UPLOAD_TO_SERVER, &wreq->flags);
594 list_add_tail(®ion->flush_link, &wreq->regions);
595 trace_netfs_dirty(ctx, region, 0, mas->index, mas->last,
596 netfs_dirty_trace_flush2);
597 }
598
599 no_more:
600 /* We've now got a contiguous span. Some of the subspans may only need
601 * writing to the cache, whilst others need writing to both the server
602 * and the cache.
603 */
604 _debug("span %lx-%lx", wreq->first, wreq->last);
605 *_first = last + 1;
606 mtree_unlock(&ctx->dirty_regions);
607 rcu_read_unlock();
608
609 if (wreq->i_size > end)
610 end = min_t(unsigned long long, wreq->i_size, (wreq->last + 1) * PAGE_SIZE);
611 wreq->len = end - wreq->start;
612
613 /* Load the pages into the raw-data buffer and transition them over to
614 * the writeback state.
615 */
616 ret = netfs_flush_get_pages(wreq, ctx);
617 if (ret < 0)
618 goto undo;
619
620 if (wreq->buffering == NETFS_ENC_BUFFER_TO_BOUNCE) {
621 ret = netfs_alloc_buffer(&wreq->bounce, wreq->first,
622 wreq->last - wreq->first + 1);
623 if (ret < 0)
624 goto undo;
625 }
626
627 netfs_writeback_lock(wreq);
628 netfs_writeback_start(wreq);
629
630 wbc->nr_to_write -= wreq->last - wreq->first + 1;
631 *_first = wreq->last + 1;
632 _leave(" = %lx [%lx]", wreq->last - wreq->first + 1, *_first);
633 return 1;
634
635 found_nothing:
636 *_first = last + 1;
637 mtree_unlock(&ctx->dirty_regions);
638 rcu_read_unlock();
639 return 0;
640
641 undo:
642 BUG(); // TODO
643 }
644
--
0-DAY CI Kernel Test Service
https://01.org/lkp
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2022-04-23 13:56 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2022-04-23 13:56 [ammarfaizi2-block:dhowells/linux-fs/netfs-maple 39/44] fs/netfs/buffered_flush.c:545:56: sparse: sparse: Using plain integer as NULL pointer kernel test robot
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox