From 3623c8ecaa7ae8c0e0cb88dce9835ad33671e2a3 Mon Sep 17 00:00:00 2001 From: Minchan Kim Date: Wed, 3 May 2017 14:55:38 -0700 Subject: [PATCH] BACKPORT: zram: handle multiple pages attached bio's bvec Patch series "zram clean up", v2. This patchset aims to clean up zram . [1] clean up multiple pages's bvec handling. [2] clean up partial IO handling [3-6] clean up zram via using accessor and removing pointless structure. With [2-6] applied, we can get a few hundred bytes as well as huge readibility enhance. x86: 708 byte save add/remove: 1/1 grow/shrink: 0/11 up/down: 478/-1186 (-708) function old new delta zram_special_page_read - 478 +478 zram_reset_device 317 314 -3 mem_used_max_store 131 128 -3 compact_store 96 93 -3 mm_stat_show 203 197 -6 zram_add 719 712 -7 zram_slot_free_notify 229 214 -15 zram_make_request 819 803 -16 zram_meta_free 128 111 -17 zram_free_page 180 151 -29 disksize_store 432 361 -71 zram_decompress_page.isra 504 - -504 zram_bvec_rw 2592 2080 -512 Total: Before=25350773, After=25350065, chg -0.00% ppc64: 231 byte save add/remove: 2/0 grow/shrink: 1/9 up/down: 681/-912 (-231) function old new delta zram_special_page_read - 480 +480 zram_slot_lock - 200 +200 vermagic 39 40 +1 mm_stat_show 256 248 -8 zram_meta_free 200 184 -16 zram_add 944 912 -32 zram_free_page 348 308 -40 disksize_store 572 492 -80 zram_decompress_page 664 564 -100 zram_slot_free_notify 292 160 -132 zram_make_request 1132 1000 -132 zram_bvec_rw 2768 2396 -372 Total: Before=17565825, After=17565594, chg -0.00% This patch (of 6): Johannes Thumshirn reported system goes the panic when using NVMe over Fabrics loopback target with zram. The reason is zram expects each bvec in bio contains a single page but nvme can attach a huge bulk of pages attached to the bio's bvec so that zram's index arithmetic could be wrong so that out-of-bound access makes system panic. [1] in mainline solved solved the problem by limiting max_sectors with SECTORS_PER_PAGE but it makes zram slow because bio should split with each pages so this patch makes zram aware of multiple pages in a bvec so it could solve without any regression(ie, bio split). [1] 0bc315381fe9, zram: set physical queue limits to avoid array out of bounds accesses Link: http://lkml.kernel.org/r/20170413134057.GA27499@bbox Signed-off-by: Minchan Kim Reported-by: Johannes Thumshirn Tested-by: Johannes Thumshirn Reviewed-by: Johannes Thumshirn Reviewed-by: Sergey Senozhatsky Cc: Hannes Reinecke Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds (cherry picked from commit e86942c7b6c1e1dd5e539f3bf3cfb63799163048) Signed-off-by: Peter Kalauskas Bug: 112488418 Change-Id: Ibedc9e8163fc16a0c2569e8c3e33dd81bb325ee5 --- drivers/block/zram/zram_drv.c | 36 ++++++++++------------------------- 1 file changed, 10 insertions(+), 26 deletions(-) diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c index b064c0b3c4c5..90d9c29ebb74 100644 --- a/drivers/block/zram/zram_drv.c +++ b/drivers/block/zram/zram_drv.c @@ -136,8 +136,7 @@ static inline bool valid_io_request(struct zram *zram, static void update_position(u32 *index, int *offset, struct bio_vec *bvec) { - if (*offset + bvec->bv_len >= PAGE_SIZE) - (*index)++; + *index += (*offset + bvec->bv_len) / PAGE_SIZE; *offset = (*offset + bvec->bv_len) % PAGE_SIZE; } @@ -835,31 +834,20 @@ static void __zram_make_request(struct zram *zram, struct bio *bio) rw = bio_data_dir(bio); bio_for_each_segment(bvec, bio, iter) { - int max_transfer_size = PAGE_SIZE - offset; - - if (bvec.bv_len > max_transfer_size) { - /* - * zram_bvec_rw() can only make operation on a single - * zram page. Split the bio vector. - */ - struct bio_vec bv; - - bv.bv_page = bvec.bv_page; - bv.bv_len = max_transfer_size; - bv.bv_offset = bvec.bv_offset; + struct bio_vec bv = bvec; + unsigned int unwritten = bvec.bv_len; + do { + bv.bv_len = min_t(unsigned int, PAGE_SIZE - offset, + unwritten); if (zram_bvec_rw(zram, &bv, index, offset, rw) < 0) goto out; - bv.bv_len = bvec.bv_len - max_transfer_size; - bv.bv_offset += max_transfer_size; - if (zram_bvec_rw(zram, &bv, index + 1, 0, rw) < 0) - goto out; - } else - if (zram_bvec_rw(zram, &bvec, index, offset, rw) < 0) - goto out; + bv.bv_offset += bv.bv_len; + unwritten -= bv.bv_len; - update_position(&index, &offset, &bvec); + update_position(&index, &offset, &bv); + } while (unwritten); } bio_endio(bio); @@ -876,8 +864,6 @@ static blk_qc_t zram_make_request(struct request_queue *queue, struct bio *bio) { struct zram *zram = queue->queuedata; - blk_queue_split(queue, &bio, queue->bio_split); - if (!valid_io_request(zram, bio->bi_iter.bi_sector, bio->bi_iter.bi_size)) { atomic64_inc(&zram->stats.invalid_io); @@ -1185,8 +1171,6 @@ static int zram_add(void) blk_queue_io_min(zram->disk->queue, PAGE_SIZE); blk_queue_io_opt(zram->disk->queue, PAGE_SIZE); zram->disk->queue->limits.discard_granularity = PAGE_SIZE; - zram->disk->queue->limits.max_sectors = SECTORS_PER_PAGE; - zram->disk->queue->limits.chunk_sectors = 0; blk_queue_max_discard_sectors(zram->disk->queue, UINT_MAX); /* * zram_bio_discard() will clear all logical blocks if logical block -- 2.20.1