diff options
Diffstat (limited to 'drivers/block')
42 files changed, 2690 insertions, 975 deletions
diff --git a/drivers/block/Kconfig b/drivers/block/Kconfig index 86b9f37d102e..014a1cfc41c5 100644 --- a/drivers/block/Kconfig +++ b/drivers/block/Kconfig @@ -108,6 +108,8 @@ source "drivers/block/paride/Kconfig" source "drivers/block/mtip32xx/Kconfig" +source "drivers/block/zram/Kconfig" + config BLK_CPQ_DA tristate "Compaq SMART2 support" depends on PCI && VIRT_TO_BUS && 0 @@ -368,7 +370,8 @@ config BLK_DEV_RAM For details, read <file:Documentation/blockdev/ramdisk.txt>. To compile this driver as a module, choose M here: the - module will be called rd. + module will be called brd. An alias "rd" has been defined + for historical reasons. Most normal users won't need the RAM disk functionality, and can thus say N here. diff --git a/drivers/block/Makefile b/drivers/block/Makefile index 8cc98cd0d4a8..02b688d1438d 100644 --- a/drivers/block/Makefile +++ b/drivers/block/Makefile @@ -31,7 +31,6 @@ obj-$(CONFIG_BLK_DEV_NBD) += nbd.o obj-$(CONFIG_BLK_DEV_CRYPTOLOOP) += cryptoloop.o obj-$(CONFIG_VIRTIO_BLK) += virtio_blk.o -obj-$(CONFIG_VIODASD) += viodasd.o obj-$(CONFIG_BLK_DEV_SX8) += sx8.o obj-$(CONFIG_BLK_DEV_HD) += hd.o @@ -43,6 +42,7 @@ obj-$(CONFIG_BLK_DEV_PCIESSD_MTIP32XX) += mtip32xx/ obj-$(CONFIG_BLK_DEV_RSXX) += rsxx/ obj-$(CONFIG_BLK_DEV_NULL_BLK) += null_blk.o +obj-$(CONFIG_ZRAM) += zram/ nvme-y := nvme-core.o nvme-scsi.o skd-y := skd_main.o diff --git a/drivers/block/aoe/aoe.h b/drivers/block/aoe/aoe.h index 14a9d1912318..9220f8e833d0 100644 --- a/drivers/block/aoe/aoe.h +++ b/drivers/block/aoe/aoe.h @@ -100,11 +100,8 @@ enum { struct buf { ulong nframesout; - ulong resid; - ulong bv_resid; - sector_t sector; struct bio *bio; - struct bio_vec *bv; + struct bvec_iter iter; struct request *rq; }; @@ -120,13 +117,10 @@ struct frame { ulong waited; ulong waited_total; struct aoetgt *t; /* parent target I belong to */ - sector_t lba; struct sk_buff *skb; /* command skb freed on module exit */ struct sk_buff *r_skb; /* response skb for async processing */ struct buf *buf; - struct bio_vec *bv; - ulong bcnt; - ulong bv_off; + struct bvec_iter iter; char flags; }; diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c index d2515435e23f..8184451b57c0 100644 --- a/drivers/block/aoe/aoecmd.c +++ b/drivers/block/aoe/aoecmd.c @@ -196,8 +196,7 @@ aoe_freetframe(struct frame *f) t = f->t; f->buf = NULL; - f->lba = 0; - f->bv = NULL; + memset(&f->iter, 0, sizeof(f->iter)); f->r_skb = NULL; f->flags = 0; list_add(&f->head, &t->ffree); @@ -295,21 +294,14 @@ newframe(struct aoedev *d) } static void -skb_fillup(struct sk_buff *skb, struct bio_vec *bv, ulong off, ulong cnt) +skb_fillup(struct sk_buff *skb, struct bio *bio, struct bvec_iter iter) { int frag = 0; - ulong fcnt; -loop: - fcnt = bv->bv_len - (off - bv->bv_offset); - if (fcnt > cnt) - fcnt = cnt; - skb_fill_page_desc(skb, frag++, bv->bv_page, off, fcnt); - cnt -= fcnt; - if (cnt <= 0) - return; - bv++; - off = bv->bv_offset; - goto loop; + struct bio_vec bv; + + __bio_for_each_segment(bv, bio, iter, iter) + skb_fill_page_desc(skb, frag++, bv.bv_page, + bv.bv_offset, bv.bv_len); } static void @@ -346,12 +338,10 @@ ata_rw_frameinit(struct frame *f) t->nout++; f->waited = 0; f->waited_total = 0; - if (f->buf) - f->lba = f->buf->sector; /* set up ata header */ - ah->scnt = f->bcnt >> 9; - put_lba(ah, f->lba); + ah->scnt = f->iter.bi_size >> 9; + put_lba(ah, f->iter.bi_sector); if (t->d->flags & DEVFL_EXT) { ah->aflags |= AOEAFL_EXT; } else { @@ -360,11 +350,11 @@ ata_rw_frameinit(struct frame *f) ah->lba3 |= 0xe0; /* LBA bit + obsolete 0xa0 */ } if (f->buf && bio_data_dir(f->buf->bio) == WRITE) { - skb_fillup(skb, f->bv, f->bv_off, f->bcnt); + skb_fillup(skb, f->buf->bio, f->iter); ah->aflags |= AOEAFL_WRITE; - skb->len += f->bcnt; - skb->data_len = f->bcnt; - skb->truesize += f->bcnt; + skb->len += f->iter.bi_size; + skb->data_len = f->iter.bi_size; + skb->truesize += f->iter.bi_size; t->wpkts++; } else { t->rpkts++; @@ -382,7 +372,6 @@ aoecmd_ata_rw(struct aoedev *d) struct buf *buf; struct sk_buff *skb; struct sk_buff_head queue; - ulong bcnt, fbcnt; buf = nextbuf(d); if (buf == NULL) @@ -390,39 +379,22 @@ aoecmd_ata_rw(struct aoedev *d) f = newframe(d); if (f == NULL) return 0; - bcnt = d->maxbcnt; - if (bcnt == 0) - bcnt = DEFAULTBCNT; - if (bcnt > buf->resid) - bcnt = buf->resid; - fbcnt = bcnt; - f->bv = buf->bv; - f->bv_off = f->bv->bv_offset + (f->bv->bv_len - buf->bv_resid); - do { - if (fbcnt < buf->bv_resid) { - buf->bv_resid -= fbcnt; - buf->resid -= fbcnt; - break; - } - fbcnt -= buf->bv_resid; - buf->resid -= buf->bv_resid; - if (buf->resid == 0) { - d->ip.buf = NULL; - break; - } - buf->bv++; - buf->bv_resid = buf->bv->bv_len; - WARN_ON(buf->bv_resid == 0); - } while (fbcnt); /* initialize the headers & frame */ f->buf = buf; - f->bcnt = bcnt; - ata_rw_frameinit(f); + f->iter = buf->iter; + f->iter.bi_size = min_t(unsigned long, + d->maxbcnt ?: DEFAULTBCNT, + f->iter.bi_size); + bio_advance_iter(buf->bio, &buf->iter, f->iter.bi_size); + + if (!buf->iter.bi_size) + d->ip.buf = NULL; /* mark all tracking fields and load out */ buf->nframesout += 1; - buf->sector += bcnt >> 9; + + ata_rw_frameinit(f); skb = skb_clone(f->skb, GFP_ATOMIC); if (skb) { @@ -613,10 +585,7 @@ reassign_frame(struct frame *f) skb = nf->skb; nf->skb = f->skb; nf->buf = f->buf; - nf->bcnt = f->bcnt; - nf->lba = f->lba; - nf->bv = f->bv; - nf->bv_off = f->bv_off; + nf->iter = f->iter; nf->waited = 0; nf->waited_total = f->waited_total; nf->sent = f->sent; @@ -648,19 +617,19 @@ probe(struct aoetgt *t) } f->flags |= FFL_PROBE; ifrotate(t); - f->bcnt = t->d->maxbcnt ? t->d->maxbcnt : DEFAULTBCNT; + f->iter.bi_size = t->d->maxbcnt ? t->d->maxbcnt : DEFAULTBCNT; ata_rw_frameinit(f); skb = f->skb; - for (frag = 0, n = f->bcnt; n > 0; ++frag, n -= m) { + for (frag = 0, n = f->iter.bi_size; n > 0; ++frag, n -= m) { if (n < PAGE_SIZE) m = n; else m = PAGE_SIZE; skb_fill_page_desc(skb, frag, empty_page, 0, m); } - skb->len += f->bcnt; - skb->data_len = f->bcnt; - skb->truesize += f->bcnt; + skb->len += f->iter.bi_size; + skb->data_len = f->iter.bi_size; + skb->truesize += f->iter.bi_size; skb = skb_clone(f->skb, GFP_ATOMIC); if (skb) { @@ -897,15 +866,15 @@ rqbiocnt(struct request *r) static void bio_pageinc(struct bio *bio) { - struct bio_vec *bv; + struct bio_vec bv; struct page *page; - int i; + struct bvec_iter iter; - bio_for_each_segment(bv, bio, i) { + bio_for_each_segment(bv, bio, iter) { /* Non-zero page count for non-head members of * compound pages is no longer allowed by the kernel. */ - page = compound_trans_head(bv->bv_page); + page = compound_trans_head(bv.bv_page); atomic_inc(&page->_count); } } @@ -913,12 +882,12 @@ bio_pageinc(struct bio *bio) static void bio_pagedec(struct bio *bio) { - struct bio_vec *bv; struct page *page; - int i; + struct bio_vec bv; + struct bvec_iter iter; - bio_for_each_segment(bv, bio, i) { - page = compound_trans_head(bv->bv_page); + bio_for_each_segment(bv, bio, iter) { + page = compound_trans_head(bv.bv_page); atomic_dec(&page->_count); } } @@ -929,12 +898,8 @@ bufinit(struct buf *buf, struct request *rq, struct bio *bio) memset(buf, 0, sizeof(*buf)); buf->rq = rq; buf->bio = bio; - buf->resid = bio->bi_size; - buf->sector = bio->bi_sector; + buf->iter = bio->bi_iter; bio_pageinc(bio); - buf->bv = bio_iovec(bio); - buf->bv_resid = buf->bv->bv_len; - WARN_ON(buf->bv_resid == 0); } static struct buf * @@ -1119,24 +1084,18 @@ gettgt(struct aoedev *d, char *addr) } static void -bvcpy(struct bio_vec *bv, ulong off, struct sk_buff *skb, long cnt) +bvcpy(struct sk_buff *skb, struct bio *bio, struct bvec_iter iter, long cnt) { - ulong fcnt; - char *p; int soff = 0; -loop: - fcnt = bv->bv_len - (off - bv->bv_offset); - if (fcnt > cnt) - fcnt = cnt; - p = page_address(bv->bv_page) + off; - skb_copy_bits(skb, soff, p, fcnt); - soff += fcnt; - cnt -= fcnt; - if (cnt <= 0) - return; - bv++; - off = bv->bv_offset; - goto loop; + struct bio_vec bv; + + iter.bi_size = cnt; + + __bio_for_each_segment(bv, bio, iter, iter) { + char *p = page_address(bv.bv_page) + bv.bv_offset; + skb_copy_bits(skb, soff, p, bv.bv_len); + soff += bv.bv_len; + } } void @@ -1152,7 +1111,7 @@ aoe_end_request(struct aoedev *d, struct request *rq, int fastfail) do { bio = rq->bio; bok = !fastfail && test_bit(BIO_UPTODATE, &bio->bi_flags); - } while (__blk_end_request(rq, bok ? 0 : -EIO, bio->bi_size)); + } while (__blk_end_request(rq, bok ? 0 : -EIO, bio->bi_iter.bi_size)); /* cf. http://lkml.org/lkml/2006/10/31/28 */ if (!fastfail) @@ -1229,7 +1188,15 @@ noskb: if (buf) clear_bit(BIO_UPTODATE, &buf->bio->bi_flags); break; } - bvcpy(f->bv, f->bv_off, skb, n); + if (n > f->iter.bi_size) { + pr_err_ratelimited("%s e%ld.%d. bytes=%ld need=%u\n", + "aoe: too-large data size in read from", + (long) d->aoemajor, d->aoeminor, + n, f->iter.bi_size); + clear_bit(BIO_UPTODATE, &buf->bio->bi_flags); + break; + } + bvcpy(skb, f->buf->bio, f->iter, n); case ATA_CMD_PIO_WRITE: case ATA_CMD_PIO_WRITE_EXT: spin_lock_irq(&d->lock); @@ -1272,7 +1239,7 @@ out: aoe_freetframe(f); - if (buf && --buf->nframesout == 0 && buf->resid == 0) + if (buf && --buf->nframesout == 0 && buf->iter.bi_size == 0) aoe_end_buf(d, buf); spin_unlock_irq(&d->lock); @@ -1727,7 +1694,7 @@ aoe_failbuf(struct aoedev *d, struct buf *buf) { if (buf == NULL) return; - buf->resid = 0; + buf->iter.bi_size = 0; clear_bit(BIO_UPTODATE, &buf->bio->bi_flags); if (buf->nframesout == 0) aoe_end_buf(d, buf); diff --git a/drivers/block/brd.c b/drivers/block/brd.c index d91f1a56e861..e73b85cf0756 100644 --- a/drivers/block/brd.c +++ b/drivers/block/brd.c @@ -328,18 +328,18 @@ static void brd_make_request(struct request_queue *q, struct bio *bio) struct block_device *bdev = bio->bi_bdev; struct brd_device *brd = bdev->bd_disk->private_data; int rw; - struct bio_vec *bvec; + struct bio_vec bvec; sector_t sector; - int i; + struct bvec_iter iter; int err = -EIO; - sector = bio->bi_sector; + sector = bio->bi_iter.bi_sector; if (bio_end_sector(bio) > get_capacity(bdev->bd_disk)) goto out; if (unlikely(bio->bi_rw & REQ_DISCARD)) { err = 0; - discard_from_brd(brd, sector, bio->bi_size); + discard_from_brd(brd, sector, bio->bi_iter.bi_size); goto out; } @@ -347,10 +347,10 @@ static void brd_make_request(struct request_queue *q, struct bio *bio) if (rw == READA) rw = READ; - bio_for_each_segment(bvec, bio, i) { - unsigned int len = bvec->bv_len; - err = brd_do_bvec(brd, bvec->bv_page, len, - bvec->bv_offset, rw, sector); + bio_for_each_segment(bvec, bio, iter) { + unsigned int len = bvec.bv_len; + err = brd_do_bvec(brd, bvec.bv_page, len, + bvec.bv_offset, rw, sector); if (err) break; sector += len >> SECTOR_SHIFT; diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index b35fc4f5237c..036e8ab86c71 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -5004,7 +5004,7 @@ reinit_after_soft_reset: i = alloc_cciss_hba(pdev); if (i < 0) - return -1; + return -ENOMEM; h = hba[i]; h->pdev = pdev; @@ -5205,7 +5205,7 @@ clean_no_release_regions: */ pci_set_drvdata(pdev, NULL); free_hba(h); - return -1; + return -ENODEV; } static void cciss_shutdown(struct pci_dev *pdev) diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c index 28c73ca320a8..a9b13f2cc420 100644 --- a/drivers/block/drbd/drbd_actlog.c +++ b/drivers/block/drbd/drbd_actlog.c @@ -159,7 +159,7 @@ static int _drbd_md_sync_page_io(struct drbd_conf *mdev, bio = bio_alloc_drbd(GFP_NOIO); bio->bi_bdev = bdev->md_bdev; - bio->bi_sector = sector; + bio->bi_iter.bi_sector = sector; err = -EIO; if (bio_add_page(bio, page, size, 0) != size) goto out; diff --git a/drivers/block/drbd/drbd_bitmap.c b/drivers/block/drbd/drbd_bitmap.c index b12c11ec4bd2..597f111df67b 100644 --- a/drivers/block/drbd/drbd_bitmap.c +++ b/drivers/block/drbd/drbd_bitmap.c @@ -1028,7 +1028,7 @@ static void bm_page_io_async(struct bm_aio_ctx *ctx, int page_nr, int rw) __must } else page = b->bm_pages[page_nr]; bio->bi_bdev = mdev->ldev->md_bdev; - bio->bi_sector = on_disk_sector; + bio->bi_iter.bi_sector = on_disk_sector; /* bio_add_page of a single page to an empty bio will always succeed, * according to api. Do we want to assert that? */ bio_add_page(bio, page, len, 0); diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c index 9e3818b1bc83..929468e1512a 100644 --- a/drivers/block/drbd/drbd_main.c +++ b/drivers/block/drbd/drbd_main.c @@ -1537,15 +1537,17 @@ static int _drbd_send_page(struct drbd_conf *mdev, struct page *page, static int _drbd_send_bio(struct drbd_conf *mdev, struct bio *bio) { - struct bio_vec *bvec; - int i; + struct bio_vec bvec; + struct bvec_iter iter; + /* hint all but last page with MSG_MORE */ - bio_for_each_segment(bvec, bio, i) { + bio_for_each_segment(bvec, bio, iter) { int err; - err = _drbd_no_send_page(mdev, bvec->bv_page, - bvec->bv_offset, bvec->bv_len, - i == bio->bi_vcnt - 1 ? 0 : MSG_MORE); + err = _drbd_no_send_page(mdev, bvec.bv_page, + bvec.bv_offset, bvec.bv_len, + bio_iter_last(bvec, iter) + ? 0 : MSG_MORE); if (err) return err; } @@ -1554,15 +1556,16 @@ static int _drbd_send_bio(struct drbd_conf *mdev, struct bio *bio) static int _drbd_send_zc_bio(struct drbd_conf *mdev, struct bio *bio) { - struct bio_vec *bvec; - int i; + struct bio_vec bvec; + struct bvec_iter iter; + /* hint all but last page with MSG_MORE */ - bio_for_each_segment(bvec, bio, i) { + bio_for_each_segment(bvec, bio, iter) { int err; - err = _drbd_send_page(mdev, bvec->bv_page, - bvec->bv_offset, bvec->bv_len, - i == bio->bi_vcnt - 1 ? 0 : MSG_MORE); + err = _drbd_send_page(mdev, bvec.bv_page, + bvec.bv_offset, bvec.bv_len, + bio_iter_last(bvec, iter) ? 0 : MSG_MORE); if (err) return err; } diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index 6fa6673b36b3..d073305ffd5e 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -1333,7 +1333,7 @@ next_bio: goto fail; } /* > peer_req->i.sector, unless this is the first bio */ - bio->bi_sector = sector; + bio->bi_iter.bi_sector = sector; bio->bi_bdev = mdev->ldev->backing_bdev; bio->bi_rw = rw; bio->bi_private = peer_req; @@ -1353,7 +1353,7 @@ next_bio: dev_err(DEV, "bio_add_page failed for len=%u, " "bi_vcnt=0 (bi_sector=%llu)\n", - len, (unsigned long long)bio->bi_sector); + len, (uint64_t)bio->bi_iter.bi_sector); err = -ENOSPC; goto fail; } @@ -1595,9 +1595,10 @@ static int drbd_drain_block(struct drbd_conf *mdev, int data_size) static int recv_dless_read(struct drbd_conf *mdev, struct drbd_request *req, sector_t sector, int data_size) { - struct bio_vec *bvec; + struct bio_vec bvec; + struct bvec_iter iter; struct bio *bio; - int dgs, err, i, expect; + int dgs, err, expect; void *dig_in = mdev->tconn->int_dig_in; void *dig_vv = mdev->tconn->int_dig_vv; @@ -1615,13 +1616,13 @@ static int recv_dless_read(struct drbd_conf *mdev, struct drbd_request *req, mdev->recv_cnt += data_size>>9; bio = req->master_bio; - D_ASSERT(sector == bio->bi_sector); + D_ASSERT(sector == bio->bi_iter.bi_sector); - bio_for_each_segment(bvec, bio, i) { - void *mapped = kmap(bvec->bv_page) + bvec->bv_offset; - expect = min_t(int, data_size, bvec->bv_len); + bio_for_each_segment(bvec, bio, iter) { + void *mapped = kmap(bvec.bv_page) + bvec.bv_offset; + expect = min_t(int, data_size, bvec.bv_len); err = drbd_recv_all_warn(mdev->tconn, mapped, expect); - kunmap(bvec->bv_page); + kunmap(bvec.bv_page); if (err) return err; data_size -= expect; diff --git a/drivers/block/drbd/drbd_req.c b/drivers/block/drbd/drbd_req.c index fec7bef44994..104a040f24de 100644 --- a/drivers/block/drbd/drbd_req.c +++ b/drivers/block/drbd/drbd_req.c @@ -77,8 +77,8 @@ static struct drbd_request *drbd_req_new(struct drbd_conf *mdev, req->epoch = 0; drbd_clear_interval(&req->i); - req->i.sector = bio_src->bi_sector; - req->i.size = bio_src->bi_size; + req->i.sector = bio_src->bi_iter.bi_sector; + req->i.size = bio_src->bi_iter.bi_size; req->i.local = true; req->i.waiting = false; @@ -1280,7 +1280,7 @@ void drbd_make_request(struct request_queue *q, struct bio *bio) /* * what we "blindly" assume: */ - D_ASSERT(IS_ALIGNED(bio->bi_size, 512)); + D_ASSERT(IS_ALIGNED(bio->bi_iter.bi_size, 512)); inc_ap_bio(mdev); __drbd_make_request(mdev, bio, start_time); diff --git a/drivers/block/drbd/drbd_req.h b/drivers/block/drbd/drbd_req.h index 978cb1addc98..28e15d91197a 100644 --- a/drivers/block/drbd/drbd_req.h +++ b/drivers/block/drbd/drbd_req.h @@ -269,7 +269,7 @@ static inline void drbd_req_make_private_bio(struct drbd_request *req, struct bi /* Short lived temporary struct on the stack. * We could squirrel the error to be returned into - * bio->bi_size, or similar. But that would be too ugly. */ + * bio->bi_iter.bi_size, or similar. But that would be too ugly. */ struct bio_and_error { struct bio *bio; int error; diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c index 891c0ecaa292..84d3175d493a 100644 --- a/drivers/block/drbd/drbd_worker.c +++ b/drivers/block/drbd/drbd_worker.c @@ -313,8 +313,8 @@ void drbd_csum_bio(struct drbd_conf *mdev, struct crypto_hash *tfm, struct bio * { struct hash_desc desc; struct scatterlist sg; - struct bio_vec *bvec; - int i; + struct bio_vec bvec; + struct bvec_iter iter; desc.tfm = tfm; desc.flags = 0; @@ -322,8 +322,8 @@ void drbd_csum_bio(struct drbd_conf *mdev, struct crypto_hash *tfm, struct bio * sg_init_table(&sg, 1); crypto_hash_init(&desc); - bio_for_each_segment(bvec, bio, i) { - sg_set_page(&sg, bvec->bv_page, bvec->bv_len, bvec->bv_offset); + bio_for_each_segment(bvec, bio, iter) { + sg_set_page(&sg, bvec.bv_page, bvec.bv_len, bvec.bv_offset); crypto_hash_update(&desc, &sg, sg.length); } crypto_hash_final(&desc, digest); diff --git a/drivers/block/floppy.c b/drivers/block/floppy.c index 000abe2f105c..2023043ce7c0 100644 --- a/drivers/block/floppy.c +++ b/drivers/block/floppy.c @@ -2351,7 +2351,7 @@ static void rw_interrupt(void) /* Compute maximal contiguous buffer size. */ static int buffer_chain_size(void) { - struct bio_vec *bv; + struct bio_vec bv; int size; struct req_iterator iter; char *base; @@ -2360,10 +2360,10 @@ static int buffer_chain_size(void) size = 0; rq_for_each_segment(bv, current_req, iter) { - if (page_address(bv->bv_page) + bv->bv_offset != base + size) + if (page_address(bv.bv_page) + bv.bv_offset != base + size) break; - size += bv->bv_len; + size += bv.bv_len; } return size >> 9; @@ -2389,7 +2389,7 @@ static int transfer_size(int ssize, int max_sector, int max_size) static void copy_buffer(int ssize, int max_sector, int max_sector_2) { int remaining; /* number of transferred 512-byte sectors */ - struct bio_vec *bv; + struct bio_vec bv; char *buffer; char *dma_buffer; int size; @@ -2427,10 +2427,10 @@ static void copy_buffer(int ssize, int max_sector, int max_sector_2) if (!remaining) break; - size = bv->bv_len; + size = bv.bv_len; SUPBOUND(size, remaining); - buffer = page_address(bv->bv_page) + bv->bv_offset; + buffer = page_address(bv.bv_page) + bv.bv_offset; if (dma_buffer + size > floppy_track_buffer + (max_buffer_sectors << 10) || dma_buffer < floppy_track_buffer) { @@ -3691,9 +3691,12 @@ static int floppy_open(struct block_device *bdev, fmode_t mode) if (!(mode & FMODE_NDELAY)) { if (mode & (FMODE_READ|FMODE_WRITE)) { UDRS->last_checked = 0; + clear_bit(FD_OPEN_SHOULD_FAIL_BIT, &UDRS->flags); check_disk_change(bdev); if (test_bit(FD_DISK_CHANGED_BIT, &UDRS->flags)) goto out; + if (test_bit(FD_OPEN_SHOULD_FAIL_BIT, &UDRS->flags)) + goto out; } res = -EROFS; if ((mode & FMODE_WRITE) && @@ -3746,17 +3749,29 @@ static unsigned int floppy_check_events(struct gendisk *disk, * a disk in the drive, and whether that disk is writable. */ -static void floppy_rb0_complete(struct bio *bio, int err) +struct rb0_cbdata { + int drive; + struct completion complete; +}; + +static void floppy_rb0_cb(struct bio *bio, int err) { - complete((struct completion *)bio->bi_private); + struct rb0_cbdata *cbdata = (struct rb0_cbdata *)bio->bi_private; + int drive = cbdata->drive; + + if (err) { + pr_info("floppy: error %d while reading block 0", err); + set_bit(FD_OPEN_SHOULD_FAIL_BIT, &UDRS->flags); + } + complete(&cbdata->complete); } -static int __floppy_read_block_0(struct block_device *bdev) +static int __floppy_read_block_0(struct block_device *bdev, int drive) { struct bio bio; struct bio_vec bio_vec; - struct completion complete; struct page *page; + struct rb0_cbdata cbdata; size_t size; page = alloc_page(GFP_NOIO); @@ -3769,23 +3784,26 @@ static int __floppy_read_block_0(struct block_device *bdev) if (!size) size = 1024; + cbdata.drive = drive; + bio_init(&bio); bio.bi_io_vec = &bio_vec; bio_vec.bv_page = page; bio_vec.bv_len = size; bio_vec.bv_offset = 0; bio.bi_vcnt = 1; - bio.bi_size = size; + bio.bi_iter.bi_size = size; bio.bi_bdev = bdev; - bio.bi_sector = 0; + bio.bi_iter.bi_sector = 0; bio.bi_flags = (1 << BIO_QUIET); - init_completion(&complete); - bio.bi_private = &complete; - bio.bi_end_io = floppy_rb0_complete; + bio.bi_private = &cbdata; + bio.bi_end_io = floppy_rb0_cb; submit_bio(READ, &bio); process_fd_request(); - wait_for_completion(&complete); + + init_completion(&cbdata.complete); + wait_for_completion(&cbdata.complete); __free_page(page); @@ -3827,7 +3845,7 @@ static int floppy_revalidate(struct gendisk *disk) UDRS->generation++; if (drive_no_geom(drive)) { /* auto-sensing */ - res = __floppy_read_block_0(opened_bdev[drive]); + res = __floppy_read_block_0(opened_bdev[drive], drive); } else { if (cf) poll_drive(false, FD_RAW_NEED_DISK); diff --git a/drivers/block/loop.c b/drivers/block/loop.c index c8dac7305244..66e8c3b94ef3 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -288,9 +288,10 @@ static int lo_send(struct loop_device *lo, struct bio *bio, loff_t pos) { int (*do_lo_send)(struct loop_device *, struct bio_vec *, loff_t, struct page *page); - struct bio_vec *bvec; + struct bio_vec bvec; + struct bvec_iter iter; struct page *page = NULL; - int i, ret = 0; + int ret = 0; if (lo->transfer != transfer_none) { page = alloc_page(GFP_NOIO | __GFP_HIGHMEM); @@ -302,11 +303,11 @@ static int lo_send(struct loop_device *lo, struct bio *bio, loff_t pos) do_lo_send = do_lo_send_direct_write; } - bio_for_each_segment(bvec, bio, i) { - ret = do_lo_send(lo, bvec, pos, page); + bio_for_each_segment(bvec, bio, iter) { + ret = do_lo_send(lo, &bvec, pos, page); if (ret < 0) break; - pos += bvec->bv_len; + pos += bvec.bv_len; } if (page) { kunmap(page); @@ -392,20 +393,20 @@ do_lo_receive(struct loop_device *lo, static int lo_receive(struct loop_device *lo, struct bio *bio, int bsize, loff_t pos) { - struct bio_vec *bvec; + struct bio_vec bvec; + struct bvec_iter iter; ssize_t s; - int i; - bio_for_each_segment(bvec, bio, i) { - s = do_lo_receive(lo, bvec, bsize, pos); + bio_for_each_segment(bvec, bio, iter) { + s = do_lo_receive(lo, &bvec, bsize, pos); if (s < 0) return s; - if (s != bvec->bv_len) { + if (s != bvec.bv_len) { zero_fill_bio(bio); break; } - pos += bvec->bv_len; + pos += bvec.bv_len; } return 0; } @@ -415,7 +416,7 @@ static int do_bio_filebacked(struct loop_device *lo, struct bio *bio) loff_t pos; int ret; - pos = ((loff_t) bio->bi_sector << 9) + lo->lo_offset; + pos = ((loff_t) bio->bi_iter.bi_sector << 9) + lo->lo_offset; if (bio_rw(bio) == WRITE) { struct file *file = lo->lo_backing_file; @@ -444,7 +445,7 @@ static int do_bio_filebacked(struct loop_device *lo, struct bio *bio) goto out; } ret = file->f_op->fallocate(file, mode, pos, - bio->bi_size); + bio->bi_iter.bi_size); if (unlikely(ret && ret != -EINVAL && ret != -EOPNOTSUPP)) ret = -EIO; @@ -798,7 +799,7 @@ static void loop_config_discard(struct loop_device *lo) /* * We use punch hole to reclaim the free space used by the - * image a.k.a. discard. However we do support discard if + * image a.k.a. discard. However we do not support discard if * encryption is enabled, because it may give an attacker * useful information. */ diff --git a/drivers/block/mg_disk.c b/drivers/block/mg_disk.c index 7bc363f1ee82..eb59b1241366 100644 --- a/drivers/block/mg_disk.c +++ b/drivers/block/mg_disk.c @@ -915,7 +915,7 @@ static int mg_probe(struct platform_device *plat_dev) /* disk reset */ if (prv_data->dev_attr == MG_STORAGE_DEV) { - /* If POR seq. not yet finised, wait */ + /* If POR seq. not yet finished, wait */ err = mg_wait_rstout(host->rstout, MG_TMAX_RSTOUT); if (err) goto probe_err_3b; diff --git a/drivers/block/mtip32xx/mtip32xx.c b/drivers/block/mtip32xx/mtip32xx.c index 050c71267f14..516026954be6 100644 --- a/drivers/block/mtip32xx/mtip32xx.c +++ b/drivers/block/mtip32xx/mtip32xx.c @@ -41,10 +41,31 @@ #include "mtip32xx.h" #define HW_CMD_SLOT_SZ (MTIP_MAX_COMMAND_SLOTS * 32) -#define HW_CMD_TBL_SZ (AHCI_CMD_TBL_HDR_SZ + (MTIP_MAX_SG * 16)) -#define HW_CMD_TBL_AR_SZ (HW_CMD_TBL_SZ * MTIP_MAX_COMMAND_SLOTS) -#define HW_PORT_PRIV_DMA_SZ \ - (HW_CMD_SLOT_SZ + HW_CMD_TBL_AR_SZ + AHCI_RX_FIS_SZ) + +/* DMA region containing RX Fis, Identify, RLE10, and SMART buffers */ +#define AHCI_RX_FIS_SZ 0x100 +#define AHCI_RX_FIS_OFFSET 0x0 +#define AHCI_IDFY_SZ ATA_SECT_SIZE +#define AHCI_IDFY_OFFSET 0x400 +#define AHCI_SECTBUF_SZ ATA_SECT_SIZE +#define AHCI_SECTBUF_OFFSET 0x800 +#define AHCI_SMARTBUF_SZ ATA_SECT_SIZE +#define AHCI_SMARTBUF_OFFSET 0xC00 +/* 0x100 + 0x200 + 0x200 + 0x200 is smaller than 4k but we pad it out */ +#define BLOCK_DMA_ALLOC_SZ 4096 + +/* DMA region containing command table (should be 8192 bytes) */ +#define AHCI_CMD_SLOT_SZ sizeof(struct mtip_cmd_hdr) +#define AHCI_CMD_TBL_SZ (MTIP_MAX_COMMAND_SLOTS * AHCI_CMD_SLOT_SZ) +#define AHCI_CMD_TBL_OFFSET 0x0 + +/* DMA region per command (contains header and SGL) */ +#define AHCI_CMD_TBL_HDR_SZ 0x80 +#define AHCI_CMD_TBL_HDR_OFFSET 0x0 +#define AHCI_CMD_TBL_SGL_SZ (MTIP_MAX_SG * sizeof(struct mtip_cmd_sg)) +#define AHCI_CMD_TBL_SGL_OFFSET AHCI_CMD_TBL_HDR_SZ +#define CMD_DMA_ALLOC_SZ (AHCI_CMD_TBL_SGL_SZ + AHCI_CMD_TBL_HDR_SZ) + #define HOST_CAP_NZDMA (1 << 19) #define HOST_HSORG 0xFC @@ -899,8 +920,9 @@ static void mtip_handle_tfe(struct driver_data *dd) fail_reason = "thermal shutdown"; } if (buf[288] == 0xBF) { + set_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag); dev_info(&dd->pdev->dev, - "Drive indicates rebuild has failed.\n"); + "Drive indicates rebuild has failed. Secure erase required.\n"); fail_all_ncq_cmds = 1; fail_reason = "rebuild failed"; } @@ -1566,6 +1588,12 @@ static int mtip_get_identify(struct mtip_port *port, void __user *user_buffer) } #endif + /* Check security locked state */ + if (port->identify[128] & 0x4) + set_bit(MTIP_DDF_SEC_LOCK_BIT, &port->dd->dd_flag); + else + clear_bit(MTIP_DDF_SEC_LOCK_BIT, &port->dd->dd_flag); + #ifdef MTIP_TRIM /* Disabling TRIM support temporarily */ /* Demux ID.DRAT & ID.RZAT to determine trim support */ if (port->identify[69] & (1 << 14) && port->identify[69] & (1 << 5)) @@ -1887,6 +1915,10 @@ static void mtip_dump_identify(struct mtip_port *port) strlcpy(cbuf, (char *)(port->identify+27), 41); dev_info(&port->dd->pdev->dev, "Model: %s\n", cbuf); + dev_info(&port->dd->pdev->dev, "Security: %04x %s\n", + port->identify[128], + port->identify[128] & 0x4 ? "(LOCKED)" : ""); + if (mtip_hw_get_capacity(port->dd, §ors)) dev_info(&port->dd->pdev->dev, "Capacity: %llu sectors (%llu MB)\n", @@ -3313,6 +3345,118 @@ st_out: } /* + * DMA region teardown + * + * @dd Pointer to driver_data structure + * + * return value + * None + */ +static void mtip_dma_free(struct driver_data *dd) +{ + int i; + struct mtip_port *port = dd->port; + + if (port->block1) + dmam_free_coherent(&dd->pdev->dev, BLOCK_DMA_ALLOC_SZ, + port->block1, port->block1_dma); + + if (port->command_list) { + dmam_free_coherent(&dd->pdev->dev, AHCI_CMD_TBL_SZ, + port->command_list, port->command_list_dma); + } + + for (i = 0; i < MTIP_MAX_COMMAND_SLOTS; i++) { + if (port->commands[i].command) + dmam_free_coherent(&dd->pdev->dev, CMD_DMA_ALLOC_SZ, + port->commands[i].command, + port->commands[i].command_dma); + } +} + +/* + * DMA region setup + * + * @dd Pointer to driver_data structure + * + * return value + * -ENOMEM Not enough free DMA region space to initialize driver + */ +static int mtip_dma_alloc(struct driver_data *dd) +{ + struct mtip_port *port = dd->port; + int i, rv = 0; + u32 host_cap_64 = readl(dd->mmio + HOST_CAP) & HOST_CAP_64; + + /* Allocate dma memory for RX Fis, Identify, and Sector Bufffer */ + port->block1 = + dmam_alloc_coherent(&dd->pdev->dev, BLOCK_DMA_ALLOC_SZ, + &port->block1_dma, GFP_KERNEL); + if (!port->block1) + return -ENOMEM; + memset(port->block1, 0, BLOCK_DMA_ALLOC_SZ); + + /* Allocate dma memory for command list */ + port->command_list = + dmam_alloc_coherent(&dd->pdev->dev, AHCI_CMD_TBL_SZ, + &port->command_list_dma, GFP_KERNEL); + if (!port->command_list) { + dmam_free_coherent(&dd->pdev->dev, BLOCK_DMA_ALLOC_SZ, + port->block1, port->block1_dma); + port->block1 = NULL; + port->block1_dma = 0; + return -ENOMEM; + } + memset(port->command_list, 0, AHCI_CMD_TBL_SZ); + + /* Setup all pointers into first DMA region */ + port->rxfis = port->block1 + AHCI_RX_FIS_OFFSET; + port->rxfis_dma = port->block1_dma + AHCI_RX_FIS_OFFSET; + port->identify = port->block1 + AHCI_IDFY_OFFSET; + port->identify_dma = port->block1_dma + AHCI_IDFY_OFFSET; + port->log_buf = port->block1 + AHCI_SECTBUF_OFFSET; + port->log_buf_dma = port->block1_dma + AHCI_SECTBUF_OFFSET; + port->smart_buf = port->block1 + AHCI_SMARTBUF_OFFSET; + port->smart_buf_dma = port->block1_dma + AHCI_SMARTBUF_OFFSET; + + /* Setup per command SGL DMA region */ + + /* Point the command headers at the command tables */ + for (i = 0; i < MTIP_MAX_COMMAND_SLOTS; i++) { + port->commands[i].command = + dmam_alloc_coherent(&dd->pdev->dev, CMD_DMA_ALLOC_SZ, + &port->commands[i].command_dma, GFP_KERNEL); + if (!port->commands[i].command) { + rv = -ENOMEM; + mtip_dma_free(dd); + return rv; + } + memset(port->commands[i].command, 0, CMD_DMA_ALLOC_SZ); + + port->commands[i].command_header = port->command_list + + (sizeof(struct mtip_cmd_hdr) * i); + port->commands[i].command_header_dma = + dd->port->command_list_dma + + (sizeof(struct mtip_cmd_hdr) * i); + + if (host_cap_64) + port->commands[i].command_header->ctbau = + __force_bit2int cpu_to_le32( + (port->commands[i].command_dma >> 16) >> 16); + + port->commands[i].command_header->ctba = + __force_bit2int cpu_to_le32( + port->commands[i].command_dma & 0xFFFFFFFF); + + sg_init_table(port->commands[i].sg, MTIP_MAX_SG); + + /* Mark command as currently inactive */ + atomic_set(&dd->port->commands[i].active, 0); + } + return 0; +} + +/* * Called once for each card. * * @dd Pointer to the driver data structure. @@ -3370,83 +3514,10 @@ static int mtip_hw_init(struct driver_data *dd) dd->port->mmio = dd->mmio + PORT_OFFSET; dd->port->dd = dd; - /* Allocate memory for the command list. */ - dd->port->command_list = - dmam_alloc_coherent(&dd->pdev->dev, - HW_PORT_PRIV_DMA_SZ + (ATA_SECT_SIZE * 4), - &dd->port->command_list_dma, - GFP_KERNEL); - if (!dd->port->command_list) { - dev_err(&dd->pdev->dev, - "Memory allocation: command list\n"); - rv = -ENOMEM; + /* DMA allocations */ + rv = mtip_dma_alloc(dd); + if (rv < 0) goto out1; - } - - /* Clear the memory we have allocated. */ - memset(dd->port->command_list, - 0, - HW_PORT_PRIV_DMA_SZ + (ATA_SECT_SIZE * 4)); - - /* Setup the addresse of the RX FIS. */ - dd->port->rxfis = dd->port->command_list + HW_CMD_SLOT_SZ; - dd->port->rxfis_dma = dd->port->command_list_dma + HW_CMD_SLOT_SZ; - - /* Setup the address of the command tables. */ - dd->port->command_table = dd->port->rxfis + AHCI_RX_FIS_SZ; - dd->port->command_tbl_dma = dd->port->rxfis_dma + AHCI_RX_FIS_SZ; - - /* Setup the address of the identify data. */ - dd->port->identify = dd->port->command_table + - HW_CMD_TBL_AR_SZ; - dd->port->identify_dma = dd->port->command_tbl_dma + - HW_CMD_TBL_AR_SZ; - - /* Setup the address of the sector buffer - for some non-ncq cmds */ - dd->port->sector_buffer = (void *) dd->port->identify + ATA_SECT_SIZE; - dd->port->sector_buffer_dma = dd->port->identify_dma + ATA_SECT_SIZE; - - /* Setup the address of the log buf - for read log command */ - dd->port->log_buf = (void *)dd->port->sector_buffer + ATA_SECT_SIZE; - dd->port->log_buf_dma = dd->port->sector_buffer_dma + ATA_SECT_SIZE; - - /* Setup the address of the smart buf - for smart read data command */ - dd->port->smart_buf = (void *)dd->port->log_buf + ATA_SECT_SIZE; - dd->port->smart_buf_dma = dd->port->log_buf_dma + ATA_SECT_SIZE; - - - /* Point the command headers at the command tables. */ - for (i = 0; i < num_command_slots; i++) { - dd->port->commands[i].command_header = - dd->port->command_list + - (sizeof(struct mtip_cmd_hdr) * i); - dd->port->commands[i].command_header_dma = - dd->port->command_list_dma + - (sizeof(struct mtip_cmd_hdr) * i); - - dd->port->commands[i].command = - dd->port->command_table + (HW_CMD_TBL_SZ * i); - dd->port->commands[i].command_dma = - dd->port->command_tbl_dma + (HW_CMD_TBL_SZ * i); - - if (readl(dd->mmio + HOST_CAP) & HOST_CAP_64) - dd->port->commands[i].command_header->ctbau = - __force_bit2int cpu_to_le32( - (dd->port->commands[i].command_dma >> 16) >> 16); - dd->port->commands[i].command_header->ctba = - __force_bit2int cpu_to_le32( - dd->port->commands[i].command_dma & 0xFFFFFFFF); - - /* - * If this is not done, a bug is reported by the stock - * FC11 i386. Due to the fact that it has lots of kernel - * debugging enabled. - */ - sg_init_table(dd->port->commands[i].sg, MTIP_MAX_SG); - - /* Mark all commands as currently inactive.*/ - atomic_set(&dd->port->commands[i].active, 0); - } /* Setup the pointers to the extended s_active and CI registers. */ for (i = 0; i < dd->slot_groups; i++) { @@ -3594,12 +3665,8 @@ out3: out2: mtip_deinit_port(dd->port); + mtip_dma_free(dd); - /* Free the command/command header memory. */ - dmam_free_coherent(&dd->pdev->dev, - HW_PORT_PRIV_DMA_SZ + (ATA_SECT_SIZE * 4), - dd->port->command_list, - dd->port->command_list_dma); out1: /* Free the memory allocated for the for structure. */ kfree(dd->port); @@ -3622,7 +3689,8 @@ static int mtip_hw_exit(struct driver_data *dd) * saves its state. */ if (!dd->sr) { - if (!test_bit(MTIP_DDF_REBUILD_FAILED_BIT, &dd->dd_flag)) + if (!test_bit(MTIP_PF_REBUILD_BIT, &dd->port->flags) && + !test_bit(MTIP_DDF_SEC_LOCK_BIT, &dd->dd_flag)) if (mtip_standby_immediate(dd->port)) dev_warn(&dd->pdev->dev, "STANDBY IMMEDIATE failed\n"); @@ -3641,11 +3709,9 @@ static int mtip_hw_exit(struct driver_data *dd) irq_set_affinity_hint(dd->pdev->irq, NULL); devm_free_irq(&dd->pdev->dev, dd->pdev->irq, dd); - /* Free the command/command header memory. */ - dmam_free_coherent(&dd->pdev->dev, - HW_PORT_PRIV_DMA_SZ + (ATA_SECT_SIZE * 4), - dd->port->command_list, - dd->port->command_list_dma); + /* Free dma regions */ + mtip_dma_free(dd); + /* Free the memory allocated for the for structure. */ kfree(dd->port); dd->port = NULL; @@ -3962,8 +4028,9 @@ static void mtip_make_request(struct request_queue *queue, struct bio *bio) { struct driver_data *dd = queue->queuedata; struct scatterlist *sg; - struct bio_vec *bvec; - int i, nents = 0; + struct bio_vec bvec; + struct bvec_iter iter; + int nents = 0; int tag = 0, unaligned = 0; if (unlikely(dd->dd_flag & MTIP_DDF_STOP_IO)) { @@ -3993,7 +4060,7 @@ static void mtip_make_request(struct request_queue *queue, struct bio *bio) } if (unlikely(bio->bi_rw & REQ_DISCARD)) { - bio_endio(bio, mtip_send_trim(dd, bio->bi_sector, + bio_endio(bio, mtip_send_trim(dd, bio->bi_iter.bi_sector, bio_sectors(bio))); return; } @@ -4006,7 +4073,8 @@ static void mtip_make_request(struct request_queue *queue, struct bio *bio) if (bio_data_dir(bio) == WRITE && bio_sectors(bio) <= 64 && dd->unal_qdepth) { - if (bio->bi_sector % 8 != 0) /* Unaligned on 4k boundaries */ + if (bio->bi_iter.bi_sector % 8 != 0) + /* Unaligned on 4k boundaries */ unaligned = 1; else if (bio_sectors(bio) % 8 != 0) /* Aligned but not 4k/8k */ unaligned = 1; @@ -4025,17 +4093,17 @@ static void mtip_make_request(struct request_queue *queue, struct bio *bio) } /* Create the scatter list for this bio. */ - bio_for_each_segment(bvec, bio, i) { + bio_for_each_segment(bvec, bio, iter) { sg_set_page(&sg[nents], - bvec->bv_page, - bvec->bv_len, - bvec->bv_offset); + bvec.bv_page, + bvec.bv_len, + bvec.bv_offset); nents++; } /* Issue the read/write. */ mtip_hw_submit_io(dd, - bio->bi_sector, + bio->bi_iter.bi_sector, bio_sectors(bio), nents, tag, diff --git a/drivers/block/mtip32xx/mtip32xx.h b/drivers/block/mtip32xx/mtip32xx.h index 9be7a1582ad3..b52e9a6d6aad 100644 --- a/drivers/block/mtip32xx/mtip32xx.h +++ b/drivers/block/mtip32xx/mtip32xx.h @@ -69,7 +69,7 @@ * Maximum number of scatter gather entries * a single command may have. */ -#define MTIP_MAX_SG 128 +#define MTIP_MAX_SG 504 /* * Maximum number of slot groups (Command Issue & s_active registers) @@ -92,7 +92,7 @@ /* Driver name and version strings */ #define MTIP_DRV_NAME "mtip32xx" -#define MTIP_DRV_VERSION "1.2.6os3" +#define MTIP_DRV_VERSION "1.3.0" /* Maximum number of minor device numbers per device. */ #define MTIP_MAX_MINORS 16 @@ -391,15 +391,13 @@ struct mtip_port { */ dma_addr_t rxfis_dma; /* - * Pointer to the beginning of the command table memory as used - * by the driver. + * Pointer to the DMA region for RX Fis, Identify, RLE10, and SMART */ - void *command_table; + void *block1; /* - * Pointer to the beginning of the command table memory as used - * by the DMA. + * DMA address of region for RX Fis, Identify, RLE10, and SMART */ - dma_addr_t command_tbl_dma; + dma_addr_t block1_dma; /* * Pointer to the beginning of the identify data memory as used * by the driver. diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index 2dc3b5153f0d..55298db36b2d 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -271,18 +271,18 @@ static int nbd_send_req(struct nbd_device *nbd, struct request *req) if (nbd_cmd(req) == NBD_CMD_WRITE) { struct req_iterator iter; - struct bio_vec *bvec; + struct bio_vec bvec; /* * we are really probing at internals to determine * whether to set MSG_MORE or not... */ rq_for_each_segment(bvec, req, iter) { flags = 0; - if (!rq_iter_last(req, iter)) + if (!rq_iter_last(bvec, iter)) flags = MSG_MORE; dprintk(DBG_TX, "%s: request %p: sending %d bytes data\n", - nbd->disk->disk_name, req, bvec->bv_len); - result = sock_send_bvec(nbd, bvec, flags); + nbd->disk->disk_name, req, bvec.bv_len); + result = sock_send_bvec(nbd, &bvec, flags); if (result <= 0) { dev_err(disk_to_dev(nbd->disk), "Send data failed (result %d)\n", @@ -378,10 +378,10 @@ static struct request *nbd_read_stat(struct nbd_device *nbd) nbd->disk->disk_name, req); if (nbd_cmd(req) == NBD_CMD_READ) { struct req_iterator iter; - struct bio_vec *bvec; + struct bio_vec bvec; rq_for_each_segment(bvec, req, iter) { - result = sock_recv_bvec(nbd, bvec); + result = sock_recv_bvec(nbd, &bvec); if (result <= 0) { dev_err(disk_to_dev(nbd->disk), "Receive data failed (result %d)\n", result); @@ -389,7 +389,7 @@ static struct request *nbd_read_stat(struct nbd_device *nbd) return req; } dprintk(DBG_RX, "%s: request %p: got %d bytes data\n", - nbd->disk->disk_name, req, bvec->bv_len); + nbd->disk->disk_name, req, bvec.bv_len); } } return req; diff --git a/drivers/block/null_blk.c b/drivers/block/null_blk.c index f370fc13aea5..091b9ea14feb 100644 --- a/drivers/block/null_blk.c +++ b/drivers/block/null_blk.c @@ -1,4 +1,5 @@ #include <linux/module.h> + #include <linux/moduleparam.h> #include <linux/sched.h> #include <linux/fs.h> @@ -59,13 +60,15 @@ enum { NULL_IRQ_NONE = 0, NULL_IRQ_SOFTIRQ = 1, NULL_IRQ_TIMER = 2, +}; +enum { NULL_Q_BIO = 0, NULL_Q_RQ = 1, NULL_Q_MQ = 2, }; -static int submit_queues = 1; +static int submit_queues; module_param(submit_queues, int, S_IRUGO); MODULE_PARM_DESC(submit_queues, "Number of submission queues"); @@ -101,9 +104,9 @@ static int hw_queue_depth = 64; module_param(hw_queue_depth, int, S_IRUGO); MODULE_PARM_DESC(hw_queue_depth, "Queue depth for each hardware queue. Default: 64"); -static bool use_per_node_hctx = true; +static bool use_per_node_hctx = false; module_param(use_per_node_hctx, bool, S_IRUGO); -MODULE_PARM_DESC(use_per_node_hctx, "Use per-node allocation for hardware context queues. Default: true"); +MODULE_PARM_DESC(use_per_node_hctx, "Use per-node allocation for hardware context queues. Default: false"); static void put_tag(struct nullb_queue *nq, unsigned int tag) { @@ -171,18 +174,20 @@ static struct nullb_cmd *alloc_cmd(struct nullb_queue *nq, int can_wait) static void end_cmd(struct nullb_cmd *cmd) { - if (cmd->rq) { - if (queue_mode == NULL_Q_MQ) - blk_mq_end_io(cmd->rq, 0); - else { - INIT_LIST_HEAD(&cmd->rq->queuelist); - blk_end_request_all(cmd->rq, 0); - } - } else if (cmd->bio) + switch (queue_mode) { + case NULL_Q_MQ: + blk_mq_end_io(cmd->rq, 0); + return; + case NULL_Q_RQ: + INIT_LIST_HEAD(&cmd->rq->queuelist); + blk_end_request_all(cmd->rq, 0); + break; + case NULL_Q_BIO: bio_endio(cmd->bio, 0); + break; + } - if (queue_mode != NULL_Q_MQ) - free_cmd(cmd); + free_cmd(cmd); } static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer) @@ -194,6 +199,7 @@ static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer) cq = &per_cpu(completion_queues, smp_processor_id()); while ((entry = llist_del_all(&cq->list)) != NULL) { + entry = llist_reverse_order(entry); do { cmd = container_of(entry, struct nullb_cmd, ll_list); end_cmd(cmd); @@ -220,61 +226,31 @@ static void null_cmd_end_timer(struct nullb_cmd *cmd) static void null_softirq_done_fn(struct request *rq) { - blk_end_request_all(rq, 0); -} - -#ifdef CONFIG_SMP - -static void null_ipi_cmd_end_io(void *data) -{ - struct completion_queue *cq; - struct llist_node *entry, *next; - struct nullb_cmd *cmd; - - cq = &per_cpu(completion_queues, smp_processor_id()); - - entry = llist_del_all(&cq->list); - - while (entry) { - next = entry->next; - cmd = llist_entry(entry, struct nullb_cmd, ll_list); - end_cmd(cmd); - entry = next; - } -} - -static void null_cmd_end_ipi(struct nullb_cmd *cmd) -{ - struct call_single_data *data = &cmd->csd; - int cpu = get_cpu(); - struct completion_queue *cq = &per_cpu(completion_queues, cpu); - - cmd->ll_list.next = NULL; - - if (llist_add(&cmd->ll_list, &cq->list)) { - data->func = null_ipi_cmd_end_io; - data->flags = 0; - __smp_call_function_single(cpu, data, 0); - } - - put_cpu(); + end_cmd(rq->special); } -#endif /* CONFIG_SMP */ - static inline void null_handle_cmd(struct nullb_cmd *cmd) { /* Complete IO by inline, softirq or timer */ switch (irqmode) { - case NULL_IRQ_NONE: - end_cmd(cmd); - break; case NULL_IRQ_SOFTIRQ: -#ifdef CONFIG_SMP - null_cmd_end_ipi(cmd); -#else + switch (queue_mode) { + case NULL_Q_MQ: + blk_mq_complete_request(cmd->rq); + break; + case NULL_Q_RQ: + blk_complete_request(cmd->rq); + break; + case NULL_Q_BIO: + /* + * XXX: no proper submitting cpu information available. + */ + end_cmd(cmd); + break; + } + break; + case NULL_IRQ_NONE: end_cmd(cmd); -#endif break; case NULL_IRQ_TIMER: null_cmd_end_timer(cmd); @@ -346,8 +322,37 @@ static int null_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *rq) static struct blk_mq_hw_ctx *null_alloc_hctx(struct blk_mq_reg *reg, unsigned int hctx_index) { - return kzalloc_node(sizeof(struct blk_mq_hw_ctx), GFP_KERNEL, - hctx_index); + int b_size = DIV_ROUND_UP(reg->nr_hw_queues, nr_online_nodes); + int tip = (reg->nr_hw_queues % nr_online_nodes); + int node = 0, i, n; + + /* + * Split submit queues evenly wrt to the number of nodes. If uneven, + * fill the first buckets with one extra, until the rest is filled with + * no extra. + */ + for (i = 0, n = 1; i < hctx_index; i++, n++) { + if (n % b_size == 0) { + n = 0; + node++; + + tip--; + if (!tip) + b_size = reg->nr_hw_queues / nr_online_nodes; + } + } + + /* + * A node might not be online, therefore map the relative node id to the + * real node id. + */ + for_each_online_node(n) { + if (!node) + break; + node--; + } + + return kzalloc_node(sizeof(struct blk_mq_hw_ctx), GFP_KERNEL, n); } static void null_free_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_index) @@ -355,16 +360,24 @@ static void null_free_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_index) kfree(hctx); } +static void null_init_queue(struct nullb *nullb, struct nullb_queue *nq) +{ + BUG_ON(!nullb); + BUG_ON(!nq); + + init_waitqueue_head(&nq->wait); + nq->queue_depth = nullb->queue_depth; +} + static int null_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, unsigned int index) { struct nullb *nullb = data; struct nullb_queue *nq = &nullb->queues[index]; - init_waitqueue_head(&nq->wait); - nq->queue_depth = nullb->queue_depth; - nullb->nr_queues++; hctx->driver_data = nq; + null_init_queue(nullb, nq); + nullb->nr_queues++; return 0; } @@ -373,6 +386,7 @@ static struct blk_mq_ops null_mq_ops = { .queue_rq = null_queue_rq, .map_queue = blk_mq_map_queue, .init_hctx = null_init_hctx, + .complete = null_softirq_done_fn, }; static struct blk_mq_reg null_mq_reg = { @@ -387,10 +401,7 @@ static void null_del_dev(struct nullb *nullb) list_del_init(&nullb->list); del_gendisk(nullb->disk); - if (queue_mode == NULL_Q_MQ) - blk_mq_free_queue(nullb->q); - else - blk_cleanup_queue(nullb->q); + blk_cleanup_queue(nullb->q); put_disk(nullb->disk); kfree(nullb); } @@ -417,13 +428,13 @@ static int setup_commands(struct nullb_queue *nq) nq->cmds = kzalloc(nq->queue_depth * sizeof(*cmd), GFP_KERNEL); if (!nq->cmds) - return 1; + return -ENOMEM; tag_size = ALIGN(nq->queue_depth, BITS_PER_LONG) / BITS_PER_LONG; nq->tag_map = kzalloc(tag_size * sizeof(unsigned long), GFP_KERNEL); if (!nq->tag_map) { kfree(nq->cmds); - return 1; + return -ENOMEM; } for (i = 0; i < nq->queue_depth; i++) { @@ -454,33 +465,37 @@ static void cleanup_queues(struct nullb *nullb) static int setup_queues(struct nullb *nullb) { - struct nullb_queue *nq; - int i; - - nullb->queues = kzalloc(submit_queues * sizeof(*nq), GFP_KERNEL); + nullb->queues = kzalloc(submit_queues * sizeof(struct nullb_queue), + GFP_KERNEL); if (!nullb->queues) - return 1; + return -ENOMEM; nullb->nr_queues = 0; nullb->queue_depth = hw_queue_depth; - if (queue_mode == NULL_Q_MQ) - return 0; + return 0; +} + +static int init_driver_queues(struct nullb *nullb) +{ + struct nullb_queue *nq; + int i, ret = 0; for (i = 0; i < submit_queues; i++) { nq = &nullb->queues[i]; - init_waitqueue_head(&nq->wait); - nq->queue_depth = hw_queue_depth; - if (setup_commands(nq)) - break; + + null_init_queue(nullb, nq); + + ret = setup_commands(nq); + if (ret) + goto err_queue; nullb->nr_queues++; } - if (i == submit_queues) - return 0; - + return 0; +err_queue: cleanup_queues(nullb); - return 1; + return ret; } static int null_add_dev(void) @@ -518,11 +533,13 @@ static int null_add_dev(void) } else if (queue_mode == NULL_Q_BIO) { nullb->q = blk_alloc_queue_node(GFP_KERNEL, home_node); blk_queue_make_request(nullb->q, null_queue_bio); + init_driver_queues(nullb); } else { nullb->q = blk_init_queue_node(null_request_fn, &nullb->lock, home_node); blk_queue_prep_rq(nullb->q, null_rq_prep_fn); if (nullb->q) blk_queue_softirq_done(nullb->q, null_softirq_done_fn); + init_driver_queues(nullb); } if (!nullb->q) @@ -534,10 +551,7 @@ static int null_add_dev(void) disk = nullb->disk = alloc_disk_node(1, home_node); if (!disk) { queue_fail: - if (queue_mode == NULL_Q_MQ) - blk_mq_free_queue(nullb->q); - else - blk_cleanup_queue(nullb->q); + blk_cleanup_queue(nullb->q); cleanup_queues(nullb); err: kfree(nullb); @@ -571,15 +585,19 @@ static int __init null_init(void) { unsigned int i; -#if !defined(CONFIG_SMP) - if (irqmode == NULL_IRQ_SOFTIRQ) { - pr_warn("null_blk: softirq completions not available.\n"); - pr_warn("null_blk: using direct completions.\n"); - irqmode = NULL_IRQ_NONE; + if (bs > PAGE_SIZE) { + pr_warn("null_blk: invalid block size\n"); + pr_warn("null_blk: defaults block size to %lu\n", PAGE_SIZE); + bs = PAGE_SIZE; } -#endif - if (submit_queues > nr_cpu_ids) + if (queue_mode == NULL_Q_MQ && use_per_node_hctx) { + if (submit_queues < nr_online_nodes) { + pr_warn("null_blk: submit_queues param is set to %u.", + nr_online_nodes); + submit_queues = nr_online_nodes; + } + } else if (submit_queues > nr_cpu_ids) submit_queues = nr_cpu_ids; else if (!submit_queues) submit_queues = 1; diff --git a/drivers/block/nvme-core.c b/drivers/block/nvme-core.c index 26d03fa0bf26..51824d1f23ea 100644 --- a/drivers/block/nvme-core.c +++ b/drivers/block/nvme-core.c @@ -46,7 +46,6 @@ #define NVME_Q_DEPTH 1024 #define SQ_SIZE(depth) (depth * sizeof(struct nvme_command)) #define CQ_SIZE(depth) (depth * sizeof(struct nvme_completion)) -#define NVME_MINORS 64 #define ADMIN_TIMEOUT (60 * HZ) static int nvme_major; @@ -58,6 +57,17 @@ module_param(use_threaded_interrupts, int, 0); static DEFINE_SPINLOCK(dev_list_lock); static LIST_HEAD(dev_list); static struct task_struct *nvme_thread; +static struct workqueue_struct *nvme_workq; + +static void nvme_reset_failed_dev(struct work_struct *ws); + +struct async_cmd_info { + struct kthread_work work; + struct kthread_worker *worker; + u32 result; + int status; + void *ctx; +}; /* * An NVM Express queue. Each device has at least two (one for admin @@ -66,6 +76,7 @@ static struct task_struct *nvme_thread; struct nvme_queue { struct device *q_dmadev; struct nvme_dev *dev; + char irqname[24]; /* nvme4294967295-65535\0 */ spinlock_t q_lock; struct nvme_command *sq_cmds; volatile struct nvme_completion *cqes; @@ -80,9 +91,11 @@ struct nvme_queue { u16 sq_head; u16 sq_tail; u16 cq_head; + u16 qid; u8 cq_phase; u8 cqe_seen; u8 q_suspended; + struct async_cmd_info cmdinfo; unsigned long cmdid_data[]; }; @@ -97,6 +110,7 @@ static inline void _nvme_check_size(void) BUILD_BUG_ON(sizeof(struct nvme_delete_queue) != 64); BUILD_BUG_ON(sizeof(struct nvme_features) != 64); BUILD_BUG_ON(sizeof(struct nvme_format_cmd) != 64); + BUILD_BUG_ON(sizeof(struct nvme_abort_cmd) != 64); BUILD_BUG_ON(sizeof(struct nvme_command) != 64); BUILD_BUG_ON(sizeof(struct nvme_id_ctrl) != 4096); BUILD_BUG_ON(sizeof(struct nvme_id_ns) != 4096); @@ -111,6 +125,7 @@ struct nvme_cmd_info { nvme_completion_fn fn; void *ctx; unsigned long timeout; + int aborted; }; static struct nvme_cmd_info *nvme_cmd_info(struct nvme_queue *nvmeq) @@ -154,6 +169,7 @@ static int alloc_cmdid(struct nvme_queue *nvmeq, void *ctx, info[cmdid].fn = handler; info[cmdid].ctx = ctx; info[cmdid].timeout = jiffies + timeout; + info[cmdid].aborted = 0; return cmdid; } @@ -172,6 +188,7 @@ static int alloc_cmdid_killable(struct nvme_queue *nvmeq, void *ctx, #define CMD_CTX_COMPLETED (0x310 + CMD_CTX_BASE) #define CMD_CTX_INVALID (0x314 + CMD_CTX_BASE) #define CMD_CTX_FLUSH (0x318 + CMD_CTX_BASE) +#define CMD_CTX_ABORT (0x31C + CMD_CTX_BASE) static void special_completion(struct nvme_dev *dev, void *ctx, struct nvme_completion *cqe) @@ -180,6 +197,10 @@ static void special_completion(struct nvme_dev *dev, void *ctx, return; if (ctx == CMD_CTX_FLUSH) return; + if (ctx == CMD_CTX_ABORT) { + ++dev->abort_limit; + return; + } if (ctx == CMD_CTX_COMPLETED) { dev_warn(&dev->pci_dev->dev, "completed id %d twice on queue %d\n", @@ -196,6 +217,15 @@ static void special_completion(struct nvme_dev *dev, void *ctx, dev_warn(&dev->pci_dev->dev, "Unknown special completion %p\n", ctx); } +static void async_completion(struct nvme_dev *dev, void *ctx, + struct nvme_completion *cqe) +{ + struct async_cmd_info *cmdinfo = ctx; + cmdinfo->result = le32_to_cpup(&cqe->result); + cmdinfo->status = le16_to_cpup(&cqe->status) >> 1; + queue_kthread_work(cmdinfo->worker, &cmdinfo->work); +} + /* * Called with local interrupts disabled and the q_lock held. May not sleep. */ @@ -441,104 +471,19 @@ int nvme_setup_prps(struct nvme_dev *dev, struct nvme_common_command *cmd, return total_len; } -struct nvme_bio_pair { - struct bio b1, b2, *parent; - struct bio_vec *bv1, *bv2; - int err; - atomic_t cnt; -}; - -static void nvme_bio_pair_endio(struct bio *bio, int err) -{ - struct nvme_bio_pair *bp = bio->bi_private; - - if (err) - bp->err = err; - - if (atomic_dec_and_test(&bp->cnt)) { - bio_endio(bp->parent, bp->err); - kfree(bp->bv1); - kfree(bp->bv2); - kfree(bp); - } -} - -static struct nvme_bio_pair *nvme_bio_split(struct bio *bio, int idx, - int len, int offset) -{ - struct nvme_bio_pair *bp; - - BUG_ON(len > bio->bi_size); - BUG_ON(idx > bio->bi_vcnt); - - bp = kmalloc(sizeof(*bp), GFP_ATOMIC); - if (!bp) - return NULL; - bp->err = 0; - - bp->b1 = *bio; - bp->b2 = *bio; - - bp->b1.bi_size = len; - bp->b2.bi_size -= len; - bp->b1.bi_vcnt = idx; - bp->b2.bi_idx = idx; - bp->b2.bi_sector += len >> 9; - - if (offset) { - bp->bv1 = kmalloc(bio->bi_max_vecs * sizeof(struct bio_vec), - GFP_ATOMIC); - if (!bp->bv1) - goto split_fail_1; - - bp->bv2 = kmalloc(bio->bi_max_vecs * sizeof(struct bio_vec), - GFP_ATOMIC); - if (!bp->bv2) - goto split_fail_2; - - memcpy(bp->bv1, bio->bi_io_vec, - bio->bi_max_vecs * sizeof(struct bio_vec)); - memcpy(bp->bv2, bio->bi_io_vec, - bio->bi_max_vecs * sizeof(struct bio_vec)); - - bp->b1.bi_io_vec = bp->bv1; - bp->b2.bi_io_vec = bp->bv2; - bp->b2.bi_io_vec[idx].bv_offset += offset; - bp->b2.bi_io_vec[idx].bv_len -= offset; - bp->b1.bi_io_vec[idx].bv_len = offset; - bp->b1.bi_vcnt++; - } else - bp->bv1 = bp->bv2 = NULL; - - bp->b1.bi_private = bp; - bp->b2.bi_private = bp; - - bp->b1.bi_end_io = nvme_bio_pair_endio; - bp->b2.bi_end_io = nvme_bio_pair_endio; - - bp->parent = bio; - atomic_set(&bp->cnt, 2); - - return bp; - - split_fail_2: - kfree(bp->bv1); - split_fail_1: - kfree(bp); - return NULL; -} - static int nvme_split_and_submit(struct bio *bio, struct nvme_queue *nvmeq, - int idx, int len, int offset) + int len) { - struct nvme_bio_pair *bp = nvme_bio_split(bio, idx, len, offset); - if (!bp) + struct bio *split = bio_split(bio, len >> 9, GFP_ATOMIC, NULL); + if (!split) return -ENOMEM; + bio_chain(split, bio); + if (bio_list_empty(&nvmeq->sq_cong)) add_wait_queue(&nvmeq->sq_full, &nvmeq->sq_cong_wait); - bio_list_add(&nvmeq->sq_cong, &bp->b1); - bio_list_add(&nvmeq->sq_cong, &bp->b2); + bio_list_add(&nvmeq->sq_cong, split); + bio_list_add(&nvmeq->sq_cong, bio); return 0; } @@ -550,41 +495,44 @@ static int nvme_split_and_submit(struct bio *bio, struct nvme_queue *nvmeq, static int nvme_map_bio(struct nvme_queue *nvmeq, struct nvme_iod *iod, struct bio *bio, enum dma_data_direction dma_dir, int psegs) { - struct bio_vec *bvec, *bvprv = NULL; + struct bio_vec bvec, bvprv; + struct bvec_iter iter; struct scatterlist *sg = NULL; - int i, length = 0, nsegs = 0, split_len = bio->bi_size; + int length = 0, nsegs = 0, split_len = bio->bi_iter.bi_size; + int first = 1; if (nvmeq->dev->stripe_size) split_len = nvmeq->dev->stripe_size - - ((bio->bi_sector << 9) & (nvmeq->dev->stripe_size - 1)); + ((bio->bi_iter.bi_sector << 9) & + (nvmeq->dev->stripe_size - 1)); sg_init_table(iod->sg, psegs); - bio_for_each_segment(bvec, bio, i) { - if (bvprv && BIOVEC_PHYS_MERGEABLE(bvprv, bvec)) { - sg->length += bvec->bv_len; + bio_for_each_segment(bvec, bio, iter) { + if (!first && BIOVEC_PHYS_MERGEABLE(&bvprv, &bvec)) { + sg->length += bvec.bv_len; } else { - if (bvprv && BIOVEC_NOT_VIRT_MERGEABLE(bvprv, bvec)) - return nvme_split_and_submit(bio, nvmeq, i, - length, 0); + if (!first && BIOVEC_NOT_VIRT_MERGEABLE(&bvprv, &bvec)) + return nvme_split_and_submit(bio, nvmeq, + length); sg = sg ? sg + 1 : iod->sg; - sg_set_page(sg, bvec->bv_page, bvec->bv_len, - bvec->bv_offset); + sg_set_page(sg, bvec.bv_page, + bvec.bv_len, bvec.bv_offset); nsegs++; } - if (split_len - length < bvec->bv_len) - return nvme_split_and_submit(bio, nvmeq, i, split_len, - split_len - length); - length += bvec->bv_len; + if (split_len - length < bvec.bv_len) + return nvme_split_and_submit(bio, nvmeq, split_len); + length += bvec.bv_len; bvprv = bvec; + first = 0; } iod->nents = nsegs; sg_mark_end(sg); if (dma_map_sg(nvmeq->q_dmadev, iod->sg, iod->nents, dma_dir) == 0) return -ENOMEM; - BUG_ON(length != bio->bi_size); + BUG_ON(length != bio->bi_iter.bi_size); return length; } @@ -608,8 +556,8 @@ static int nvme_submit_discard(struct nvme_queue *nvmeq, struct nvme_ns *ns, iod->npages = 0; range->cattr = cpu_to_le32(0); - range->nlb = cpu_to_le32(bio->bi_size >> ns->lba_shift); - range->slba = cpu_to_le64(nvme_block_nr(ns, bio->bi_sector)); + range->nlb = cpu_to_le32(bio->bi_iter.bi_size >> ns->lba_shift); + range->slba = cpu_to_le64(nvme_block_nr(ns, bio->bi_iter.bi_sector)); memset(cmnd, 0, sizeof(*cmnd)); cmnd->dsm.opcode = nvme_cmd_dsm; @@ -674,7 +622,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, } result = -ENOMEM; - iod = nvme_alloc_iod(psegs, bio->bi_size, GFP_ATOMIC); + iod = nvme_alloc_iod(psegs, bio->bi_iter.bi_size, GFP_ATOMIC); if (!iod) goto nomem; iod->private = bio; @@ -723,7 +671,7 @@ static int nvme_submit_bio_queue(struct nvme_queue *nvmeq, struct nvme_ns *ns, cmnd->rw.nsid = cpu_to_le32(ns->ns_id); length = nvme_setup_prps(nvmeq->dev, &cmnd->common, iod, length, GFP_ATOMIC); - cmnd->rw.slba = cpu_to_le64(nvme_block_nr(ns, bio->bi_sector)); + cmnd->rw.slba = cpu_to_le64(nvme_block_nr(ns, bio->bi_iter.bi_sector)); cmnd->rw.length = cpu_to_le16((length >> ns->lba_shift) - 1); cmnd->rw.control = cpu_to_le16(control); cmnd->rw.dsmgmt = cpu_to_le32(dsmgmt); @@ -775,7 +723,7 @@ static int nvme_process_cq(struct nvme_queue *nvmeq) if (head == nvmeq->cq_head && phase == nvmeq->cq_phase) return 0; - writel(head, nvmeq->q_db + (1 << nvmeq->dev->db_stride)); + writel(head, nvmeq->q_db + nvmeq->dev->db_stride); nvmeq->cq_head = head; nvmeq->cq_phase = phase; @@ -886,12 +834,34 @@ int nvme_submit_sync_cmd(struct nvme_queue *nvmeq, struct nvme_command *cmd, return cmdinfo.status; } +static int nvme_submit_async_cmd(struct nvme_queue *nvmeq, + struct nvme_command *cmd, + struct async_cmd_info *cmdinfo, unsigned timeout) +{ + int cmdid; + + cmdid = alloc_cmdid_killable(nvmeq, cmdinfo, async_completion, timeout); + if (cmdid < 0) + return cmdid; + cmdinfo->status = -EINTR; + cmd->common.command_id = cmdid; + nvme_submit_cmd(nvmeq, cmd); + return 0; +} + int nvme_submit_admin_cmd(struct nvme_dev *dev, struct nvme_command *cmd, u32 *result) { return nvme_submit_sync_cmd(dev->queues[0], cmd, result, ADMIN_TIMEOUT); } +static int nvme_submit_admin_cmd_async(struct nvme_dev *dev, + struct nvme_command *cmd, struct async_cmd_info *cmdinfo) +{ + return nvme_submit_async_cmd(dev->queues[0], cmd, cmdinfo, + ADMIN_TIMEOUT); +} + static int adapter_delete_queue(struct nvme_dev *dev, u8 opcode, u16 id) { int status; @@ -1002,6 +972,56 @@ int nvme_set_features(struct nvme_dev *dev, unsigned fid, unsigned dword11, } /** + * nvme_abort_cmd - Attempt aborting a command + * @cmdid: Command id of a timed out IO + * @queue: The queue with timed out IO + * + * Schedule controller reset if the command was already aborted once before and + * still hasn't been returned to the driver, or if this is the admin queue. + */ +static void nvme_abort_cmd(int cmdid, struct nvme_queue *nvmeq) +{ + int a_cmdid; + struct nvme_command cmd; + struct nvme_dev *dev = nvmeq->dev; + struct nvme_cmd_info *info = nvme_cmd_info(nvmeq); + + if (!nvmeq->qid || info[cmdid].aborted) { + if (work_busy(&dev->reset_work)) + return; + list_del_init(&dev->node); + dev_warn(&dev->pci_dev->dev, + "I/O %d QID %d timeout, reset controller\n", cmdid, + nvmeq->qid); + PREPARE_WORK(&dev->reset_work, nvme_reset_failed_dev); + queue_work(nvme_workq, &dev->reset_work); + return; + } + + if (!dev->abort_limit) + return; + + a_cmdid = alloc_cmdid(dev->queues[0], CMD_CTX_ABORT, special_completion, + ADMIN_TIMEOUT); + if (a_cmdid < 0) + return; + + memset(&cmd, 0, sizeof(cmd)); + cmd.abort.opcode = nvme_admin_abort_cmd; + cmd.abort.cid = cmdid; + cmd.abort.sqid = cpu_to_le16(nvmeq->qid); + cmd.abort.command_id = a_cmdid; + + --dev->abort_limit; + info[cmdid].aborted = 1; + info[cmdid].timeout = jiffies + ADMIN_TIMEOUT; + + dev_warn(nvmeq->q_dmadev, "Aborting I/O %d QID %d\n", cmdid, + nvmeq->qid); + nvme_submit_cmd(dev->queues[0], &cmd); +} + +/** * nvme_cancel_ios - Cancel outstanding I/Os * @queue: The queue to cancel I/Os on * @timeout: True to only cancel I/Os which have timed out @@ -1024,7 +1044,12 @@ static void nvme_cancel_ios(struct nvme_queue *nvmeq, bool timeout) continue; if (info[cmdid].ctx == CMD_CTX_CANCELLED) continue; - dev_warn(nvmeq->q_dmadev, "Cancelling I/O %d\n", cmdid); + if (timeout && nvmeq->dev->initialized) { + nvme_abort_cmd(cmdid, nvmeq); + continue; + } + dev_warn(nvmeq->q_dmadev, "Cancelling I/O %d QID %d\n", cmdid, + nvmeq->qid); ctx = cancel_cmdid(nvmeq, cmdid, &fn); fn(nvmeq->dev, ctx, &cqe); } @@ -1046,26 +1071,31 @@ static void nvme_free_queue(struct nvme_queue *nvmeq) kfree(nvmeq); } -static void nvme_free_queues(struct nvme_dev *dev) +static void nvme_free_queues(struct nvme_dev *dev, int lowest) { int i; - for (i = dev->queue_count - 1; i >= 0; i--) { + for (i = dev->queue_count - 1; i >= lowest; i--) { nvme_free_queue(dev->queues[i]); dev->queue_count--; dev->queues[i] = NULL; } } -static void nvme_disable_queue(struct nvme_dev *dev, int qid) +/** + * nvme_suspend_queue - put queue into suspended state + * @nvmeq - queue to suspend + * + * Returns 1 if already suspended, 0 otherwise. + */ +static int nvme_suspend_queue(struct nvme_queue *nvmeq) { - struct nvme_queue *nvmeq = dev->queues[qid]; - int vector = dev->entry[nvmeq->cq_vector].vector; + int vector = nvmeq->dev->entry[nvmeq->cq_vector].vector; spin_lock_irq(&nvmeq->q_lock); if (nvmeq->q_suspended) { spin_unlock_irq(&nvmeq->q_lock); - return; + return 1; } nvmeq->q_suspended = 1; spin_unlock_irq(&nvmeq->q_lock); @@ -1073,18 +1103,35 @@ static void nvme_disable_queue(struct nvme_dev *dev, int qid) irq_set_affinity_hint(vector, NULL); free_irq(vector, nvmeq); - /* Don't tell the adapter to delete the admin queue */ - if (qid) { - adapter_delete_sq(dev, qid); - adapter_delete_cq(dev, qid); - } + return 0; +} +static void nvme_clear_queue(struct nvme_queue *nvmeq) +{ spin_lock_irq(&nvmeq->q_lock); nvme_process_cq(nvmeq); nvme_cancel_ios(nvmeq, false); spin_unlock_irq(&nvmeq->q_lock); } +static void nvme_disable_queue(struct nvme_dev *dev, int qid) +{ + struct nvme_queue *nvmeq = dev->queues[qid]; + + if (!nvmeq) + return; + if (nvme_suspend_queue(nvmeq)) + return; + + /* Don't tell the adapter to delete the admin queue. + * Don't tell a removed adapter to delete IO queues. */ + if (qid && readl(&dev->bar->csts) != -1) { + adapter_delete_sq(dev, qid); + adapter_delete_cq(dev, qid); + } + nvme_clear_queue(nvmeq); +} + static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth, int vector) { @@ -1107,15 +1154,18 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, nvmeq->q_dmadev = dmadev; nvmeq->dev = dev; + snprintf(nvmeq->irqname, sizeof(nvmeq->irqname), "nvme%dq%d", + dev->instance, qid); spin_lock_init(&nvmeq->q_lock); nvmeq->cq_head = 0; nvmeq->cq_phase = 1; init_waitqueue_head(&nvmeq->sq_full); init_waitqueue_entry(&nvmeq->sq_cong_wait, nvme_thread); bio_list_init(&nvmeq->sq_cong); - nvmeq->q_db = &dev->dbs[qid << (dev->db_stride + 1)]; + nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride]; nvmeq->q_depth = depth; nvmeq->cq_vector = vector; + nvmeq->qid = qid; nvmeq->q_suspended = 1; dev->queue_count++; @@ -1134,11 +1184,10 @@ static int queue_request_irq(struct nvme_dev *dev, struct nvme_queue *nvmeq, { if (use_threaded_interrupts) return request_threaded_irq(dev->entry[nvmeq->cq_vector].vector, - nvme_irq_check, nvme_irq, - IRQF_DISABLED | IRQF_SHARED, + nvme_irq_check, nvme_irq, IRQF_SHARED, name, nvmeq); return request_irq(dev->entry[nvmeq->cq_vector].vector, nvme_irq, - IRQF_DISABLED | IRQF_SHARED, name, nvmeq); + IRQF_SHARED, name, nvmeq); } static void nvme_init_queue(struct nvme_queue *nvmeq, u16 qid) @@ -1149,7 +1198,7 @@ static void nvme_init_queue(struct nvme_queue *nvmeq, u16 qid) nvmeq->sq_tail = 0; nvmeq->cq_head = 0; nvmeq->cq_phase = 1; - nvmeq->q_db = &dev->dbs[qid << (dev->db_stride + 1)]; + nvmeq->q_db = &dev->dbs[qid * 2 * dev->db_stride]; memset(nvmeq->cmdid_data, 0, extra); memset((void *)nvmeq->cqes, 0, CQ_SIZE(nvmeq->q_depth)); nvme_cancel_ios(nvmeq, false); @@ -1169,13 +1218,13 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid) if (result < 0) goto release_cq; - result = queue_request_irq(dev, nvmeq, "nvme"); + result = queue_request_irq(dev, nvmeq, nvmeq->irqname); if (result < 0) goto release_sq; - spin_lock(&nvmeq->q_lock); + spin_lock_irq(&nvmeq->q_lock); nvme_init_queue(nvmeq, qid); - spin_unlock(&nvmeq->q_lock); + spin_unlock_irq(&nvmeq->q_lock); return result; @@ -1287,13 +1336,13 @@ static int nvme_configure_admin_queue(struct nvme_dev *dev) if (result) return result; - result = queue_request_irq(dev, nvmeq, "nvme admin"); + result = queue_request_irq(dev, nvmeq, nvmeq->irqname); if (result) return result; - spin_lock(&nvmeq->q_lock); + spin_lock_irq(&nvmeq->q_lock); nvme_init_queue(nvmeq, 0); - spin_unlock(&nvmeq->q_lock); + spin_unlock_irq(&nvmeq->q_lock); return result; } @@ -1569,10 +1618,47 @@ static int nvme_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd, } } +#ifdef CONFIG_COMPAT +static int nvme_compat_ioctl(struct block_device *bdev, fmode_t mode, + unsigned int cmd, unsigned long arg) +{ + struct nvme_ns *ns = bdev->bd_disk->private_data; + + switch (cmd) { + case SG_IO: + return nvme_sg_io32(ns, arg); + } + return nvme_ioctl(bdev, mode, cmd, arg); +} +#else +#define nvme_compat_ioctl NULL +#endif + +static int nvme_open(struct block_device *bdev, fmode_t mode) +{ + struct nvme_ns *ns = bdev->bd_disk->private_data; + struct nvme_dev *dev = ns->dev; + + kref_get(&dev->kref); + return 0; +} + +static void nvme_free_dev(struct kref *kref); + +static void nvme_release(struct gendisk *disk, fmode_t mode) +{ + struct nvme_ns *ns = disk->private_data; + struct nvme_dev *dev = ns->dev; + + kref_put(&dev->kref, nvme_free_dev); +} + static const struct block_device_operations nvme_fops = { .owner = THIS_MODULE, .ioctl = nvme_ioctl, - .compat_ioctl = nvme_ioctl, + .compat_ioctl = nvme_compat_ioctl, + .open = nvme_open, + .release = nvme_release, }; static void nvme_resubmit_bios(struct nvme_queue *nvmeq) @@ -1596,13 +1682,25 @@ static void nvme_resubmit_bios(struct nvme_queue *nvmeq) static int nvme_kthread(void *data) { - struct nvme_dev *dev; + struct nvme_dev *dev, *next; while (!kthread_should_stop()) { set_current_state(TASK_INTERRUPTIBLE); spin_lock(&dev_list_lock); - list_for_each_entry(dev, &dev_list, node) { + list_for_each_entry_safe(dev, next, &dev_list, node) { int i; + if (readl(&dev->bar->csts) & NVME_CSTS_CFS && + dev->initialized) { + if (work_busy(&dev->reset_work)) + continue; + list_del_init(&dev->node); + dev_warn(&dev->pci_dev->dev, + "Failed status, reset controller\n"); + PREPARE_WORK(&dev->reset_work, + nvme_reset_failed_dev); + queue_work(nvme_workq, &dev->reset_work); + continue; + } for (i = 0; i < dev->queue_count; i++) { struct nvme_queue *nvmeq = dev->queues[i]; if (!nvmeq) @@ -1623,33 +1721,6 @@ static int nvme_kthread(void *data) return 0; } -static DEFINE_IDA(nvme_index_ida); - -static int nvme_get_ns_idx(void) -{ - int index, error; - - do { - if (!ida_pre_get(&nvme_index_ida, GFP_KERNEL)) - return -1; - - spin_lock(&dev_list_lock); - error = ida_get_new(&nvme_index_ida, &index); - spin_unlock(&dev_list_lock); - } while (error == -EAGAIN); - - if (error) - index = -1; - return index; -} - -static void nvme_put_ns_idx(int index) -{ - spin_lock(&dev_list_lock); - ida_remove(&nvme_index_ida, index); - spin_unlock(&dev_list_lock); -} - static void nvme_config_discard(struct nvme_ns *ns) { u32 logical_block_size = queue_logical_block_size(ns->queue); @@ -1683,7 +1754,7 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, unsigned nsid, ns->dev = dev; ns->queue->queuedata = ns; - disk = alloc_disk(NVME_MINORS); + disk = alloc_disk(0); if (!disk) goto out_free_queue; ns->ns_id = nsid; @@ -1696,12 +1767,12 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, unsigned nsid, blk_queue_max_hw_sectors(ns->queue, dev->max_hw_sectors); disk->major = nvme_major; - disk->minors = NVME_MINORS; - disk->first_minor = NVME_MINORS * nvme_get_ns_idx(); + disk->first_minor = 0; disk->fops = &nvme_fops; disk->private_data = ns; disk->queue = ns->queue; disk->driverfs_dev = &dev->pci_dev->dev; + disk->flags = GENHD_FL_EXT_DEVT; sprintf(disk->disk_name, "nvme%dn%d", dev->instance, nsid); set_capacity(disk, le64_to_cpup(&id->nsze) << (ns->lba_shift - 9)); @@ -1717,15 +1788,6 @@ static struct nvme_ns *nvme_alloc_ns(struct nvme_dev *dev, unsigned nsid, return NULL; } -static void nvme_ns_free(struct nvme_ns *ns) -{ - int index = ns->disk->first_minor / NVME_MINORS; - put_disk(ns->disk); - nvme_put_ns_idx(index); - blk_cleanup_queue(ns->queue); - kfree(ns); -} - static int set_queue_count(struct nvme_dev *dev, int count) { int status; @@ -1741,11 +1803,12 @@ static int set_queue_count(struct nvme_dev *dev, int count) static size_t db_bar_size(struct nvme_dev *dev, unsigned nr_io_queues) { - return 4096 + ((nr_io_queues + 1) << (dev->db_stride + 3)); + return 4096 + ((nr_io_queues + 1) * 8 * dev->db_stride); } static int nvme_setup_io_queues(struct nvme_dev *dev) { + struct nvme_queue *adminq = dev->queues[0]; struct pci_dev *pdev = dev->pci_dev; int result, cpu, i, vecs, nr_io_queues, size, q_depth; @@ -1772,7 +1835,7 @@ static int nvme_setup_io_queues(struct nvme_dev *dev) } /* Deregister the admin queue's interrupt */ - free_irq(dev->entry[0].vector, dev->queues[0]); + free_irq(dev->entry[0].vector, adminq); vecs = nr_io_queues; for (i = 0; i < vecs; i++) @@ -1810,9 +1873,9 @@ static int nvme_setup_io_queues(struct nvme_dev *dev) */ nr_io_queues = vecs; - result = queue_request_irq(dev, dev->queues[0], "nvme admin"); + result = queue_request_irq(dev, adminq, adminq->irqname); if (result) { - dev->queues[0]->q_suspended = 1; + adminq->q_suspended = 1; goto free_queues; } @@ -1821,9 +1884,9 @@ static int nvme_setup_io_queues(struct nvme_dev *dev) for (i = dev->queue_count - 1; i > nr_io_queues; i--) { struct nvme_queue *nvmeq = dev->queues[i]; - spin_lock(&nvmeq->q_lock); + spin_lock_irq(&nvmeq->q_lock); nvme_cancel_ios(nvmeq, false); - spin_unlock(&nvmeq->q_lock); + spin_unlock_irq(&nvmeq->q_lock); nvme_free_queue(nvmeq); dev->queue_count--; @@ -1864,7 +1927,7 @@ static int nvme_setup_io_queues(struct nvme_dev *dev) return 0; free_queues: - nvme_free_queues(dev); + nvme_free_queues(dev, 1); return result; } @@ -1876,6 +1939,7 @@ static int nvme_setup_io_queues(struct nvme_dev *dev) */ static int nvme_dev_add(struct nvme_dev *dev) { + struct pci_dev *pdev = dev->pci_dev; int res; unsigned nn, i; struct nvme_ns *ns; @@ -1885,8 +1949,7 @@ static int nvme_dev_add(struct nvme_dev *dev) dma_addr_t dma_addr; int shift = NVME_CAP_MPSMIN(readq(&dev->bar->cap)) + 12; - mem = dma_alloc_coherent(&dev->pci_dev->dev, 8192, &dma_addr, - GFP_KERNEL); + mem = dma_alloc_coherent(&pdev->dev, 8192, &dma_addr, GFP_KERNEL); if (!mem) return -ENOMEM; @@ -1899,13 +1962,14 @@ static int nvme_dev_add(struct nvme_dev *dev) ctrl = mem; nn = le32_to_cpup(&ctrl->nn); dev->oncs = le16_to_cpup(&ctrl->oncs); + dev->abort_limit = ctrl->acl + 1; memcpy(dev->serial, ctrl->sn, sizeof(ctrl->sn)); memcpy(dev->model, ctrl->mn, sizeof(ctrl->mn)); memcpy(dev->firmware_rev, ctrl->fr, sizeof(ctrl->fr)); if (ctrl->mdts) dev->max_hw_sectors = 1 << (ctrl->mdts + shift - 9); - if ((dev->pci_dev->vendor == PCI_VENDOR_ID_INTEL) && - (dev->pci_dev->device == 0x0953) && ctrl->vs[3]) + if ((pdev->vendor == PCI_VENDOR_ID_INTEL) && + (pdev->device == 0x0953) && ctrl->vs[3]) dev->stripe_size = 1 << (ctrl->vs[3] + shift); id_ns = mem; @@ -1953,16 +2017,21 @@ static int nvme_dev_map(struct nvme_dev *dev) dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) goto disable; - pci_set_drvdata(pdev, dev); dev->bar = ioremap(pci_resource_start(pdev, 0), 8192); if (!dev->bar) goto disable; - - dev->db_stride = NVME_CAP_STRIDE(readq(&dev->bar->cap)); + if (readl(&dev->bar->csts) == -1) { + result = -ENODEV; + goto unmap; + } + dev->db_stride = 1 << NVME_CAP_STRIDE(readq(&dev->bar->cap)); dev->dbs = ((void __iomem *)dev->bar) + 4096; return 0; + unmap: + iounmap(dev->bar); + dev->bar = NULL; disable: pci_release_regions(pdev); disable_pci: @@ -1980,37 +2049,183 @@ static void nvme_dev_unmap(struct nvme_dev *dev) if (dev->bar) { iounmap(dev->bar); dev->bar = NULL; + pci_release_regions(dev->pci_dev); } - pci_release_regions(dev->pci_dev); if (pci_is_enabled(dev->pci_dev)) pci_disable_device(dev->pci_dev); } +struct nvme_delq_ctx { + struct task_struct *waiter; + struct kthread_worker *worker; + atomic_t refcount; +}; + +static void nvme_wait_dq(struct nvme_delq_ctx *dq, struct nvme_dev *dev) +{ + dq->waiter = current; + mb(); + + for (;;) { + set_current_state(TASK_KILLABLE); + if (!atomic_read(&dq->refcount)) + break; + if (!schedule_timeout(ADMIN_TIMEOUT) || + fatal_signal_pending(current)) { + set_current_state(TASK_RUNNING); + + nvme_disable_ctrl(dev, readq(&dev->bar->cap)); + nvme_disable_queue(dev, 0); + + send_sig(SIGKILL, dq->worker->task, 1); + flush_kthread_worker(dq->worker); + return; + } + } + set_current_state(TASK_RUNNING); +} + +static void nvme_put_dq(struct nvme_delq_ctx *dq) +{ + atomic_dec(&dq->refcount); + if (dq->waiter) + wake_up_process(dq->waiter); +} + +static struct nvme_delq_ctx *nvme_get_dq(struct nvme_delq_ctx *dq) +{ + atomic_inc(&dq->refcount); + return dq; +} + +static void nvme_del_queue_end(struct nvme_queue *nvmeq) +{ + struct nvme_delq_ctx *dq = nvmeq->cmdinfo.ctx; + + nvme_clear_queue(nvmeq); + nvme_put_dq(dq); +} + +static int adapter_async_del_queue(struct nvme_queue *nvmeq, u8 opcode, + kthread_work_func_t fn) +{ + struct nvme_command c; + + memset(&c, 0, sizeof(c)); + c.delete_queue.opcode = opcode; + c.delete_queue.qid = cpu_to_le16(nvmeq->qid); + + init_kthread_work(&nvmeq->cmdinfo.work, fn); + return nvme_submit_admin_cmd_async(nvmeq->dev, &c, &nvmeq->cmdinfo); +} + +static void nvme_del_cq_work_handler(struct kthread_work *work) +{ + struct nvme_queue *nvmeq = container_of(work, struct nvme_queue, + cmdinfo.work); + nvme_del_queue_end(nvmeq); +} + +static int nvme_delete_cq(struct nvme_queue *nvmeq) +{ + return adapter_async_del_queue(nvmeq, nvme_admin_delete_cq, + nvme_del_cq_work_handler); +} + +static void nvme_del_sq_work_handler(struct kthread_work *work) +{ + struct nvme_queue *nvmeq = container_of(work, struct nvme_queue, + cmdinfo.work); + int status = nvmeq->cmdinfo.status; + + if (!status) + status = nvme_delete_cq(nvmeq); + if (status) + nvme_del_queue_end(nvmeq); +} + +static int nvme_delete_sq(struct nvme_queue *nvmeq) +{ + return adapter_async_del_queue(nvmeq, nvme_admin_delete_sq, + nvme_del_sq_work_handler); +} + +static void nvme_del_queue_start(struct kthread_work *work) +{ + struct nvme_queue *nvmeq = container_of(work, struct nvme_queue, + cmdinfo.work); + allow_signal(SIGKILL); + if (nvme_delete_sq(nvmeq)) + nvme_del_queue_end(nvmeq); +} + +static void nvme_disable_io_queues(struct nvme_dev *dev) +{ + int i; + DEFINE_KTHREAD_WORKER_ONSTACK(worker); + struct nvme_delq_ctx dq; + struct task_struct *kworker_task = kthread_run(kthread_worker_fn, + &worker, "nvme%d", dev->instance); + + if (IS_ERR(kworker_task)) { + dev_err(&dev->pci_dev->dev, + "Failed to create queue del task\n"); + for (i = dev->queue_count - 1; i > 0; i--) + nvme_disable_queue(dev, i); + return; + } + + dq.waiter = NULL; + atomic_set(&dq.refcount, 0); + dq.worker = &worker; + for (i = dev->queue_count - 1; i > 0; i--) { + struct nvme_queue *nvmeq = dev->queues[i]; + + if (nvme_suspend_queue(nvmeq)) + continue; + nvmeq->cmdinfo.ctx = nvme_get_dq(&dq); + nvmeq->cmdinfo.worker = dq.worker; + init_kthread_work(&nvmeq->cmdinfo.work, nvme_del_queue_start); + queue_kthread_work(dq.worker, &nvmeq->cmdinfo.work); + } + nvme_wait_dq(&dq, dev); + kthread_stop(kworker_task); +} + static void nvme_dev_shutdown(struct nvme_dev *dev) { int i; - for (i = dev->queue_count - 1; i >= 0; i--) - nvme_disable_queue(dev, i); + dev->initialized = 0; spin_lock(&dev_list_lock); list_del_init(&dev->node); spin_unlock(&dev_list_lock); - if (dev->bar) + if (!dev->bar || (dev->bar && readl(&dev->bar->csts) == -1)) { + for (i = dev->queue_count - 1; i >= 0; i--) { + struct nvme_queue *nvmeq = dev->queues[i]; + nvme_suspend_queue(nvmeq); + nvme_clear_queue(nvmeq); + } + } else { + nvme_disable_io_queues(dev); nvme_shutdown_ctrl(dev); + nvme_disable_queue(dev, 0); + } nvme_dev_unmap(dev); } static void nvme_dev_remove(struct nvme_dev *dev) { - struct nvme_ns *ns, *next; + struct nvme_ns *ns; - list_for_each_entry_safe(ns, next, &dev->namespaces, list) { - list_del(&ns->list); - del_gendisk(ns->disk); - nvme_ns_free(ns); + list_for_each_entry(ns, &dev->namespaces, list) { + if (ns->disk->flags & GENHD_FL_UP) + del_gendisk(ns->disk); + if (!blk_queue_dying(ns->queue)) + blk_cleanup_queue(ns->queue); } } @@ -2067,14 +2282,22 @@ static void nvme_release_instance(struct nvme_dev *dev) spin_unlock(&dev_list_lock); } +static void nvme_free_namespaces(struct nvme_dev *dev) +{ + struct nvme_ns *ns, *next; + + list_for_each_entry_safe(ns, next, &dev->namespaces, list) { + list_del(&ns->list); + put_disk(ns->disk); + kfree(ns); + } +} + static void nvme_free_dev(struct kref *kref) { struct nvme_dev *dev = container_of(kref, struct nvme_dev, kref); - nvme_dev_remove(dev); - nvme_dev_shutdown(dev); - nvme_free_queues(dev); - nvme_release_instance(dev); - nvme_release_prp_pools(dev); + + nvme_free_namespaces(dev); kfree(dev->queues); kfree(dev->entry); kfree(dev); @@ -2138,6 +2361,7 @@ static int nvme_dev_start(struct nvme_dev *dev) return result; disable: + nvme_disable_queue(dev, 0); spin_lock(&dev_list_lock); list_del_init(&dev->node); spin_unlock(&dev_list_lock); @@ -2146,6 +2370,71 @@ static int nvme_dev_start(struct nvme_dev *dev) return result; } +static int nvme_remove_dead_ctrl(void *arg) +{ + struct nvme_dev *dev = (struct nvme_dev *)arg; + struct pci_dev *pdev = dev->pci_dev; + + if (pci_get_drvdata(pdev)) + pci_stop_and_remove_bus_device(pdev); + kref_put(&dev->kref, nvme_free_dev); + return 0; +} + +static void nvme_remove_disks(struct work_struct *ws) +{ + int i; + struct nvme_dev *dev = container_of(ws, struct nvme_dev, reset_work); + + nvme_dev_remove(dev); + spin_lock(&dev_list_lock); + for (i = dev->queue_count - 1; i > 0; i--) { + BUG_ON(!dev->queues[i] || !dev->queues[i]->q_suspended); + nvme_free_queue(dev->queues[i]); + dev->queue_count--; + dev->queues[i] = NULL; + } + spin_unlock(&dev_list_lock); +} + +static int nvme_dev_resume(struct nvme_dev *dev) +{ + int ret; + + ret = nvme_dev_start(dev); + if (ret && ret != -EBUSY) + return ret; + if (ret == -EBUSY) { + spin_lock(&dev_list_lock); + PREPARE_WORK(&dev->reset_work, nvme_remove_disks); + queue_work(nvme_workq, &dev->reset_work); + spin_unlock(&dev_list_lock); + } + dev->initialized = 1; + return 0; +} + +static void nvme_dev_reset(struct nvme_dev *dev) +{ + nvme_dev_shutdown(dev); + if (nvme_dev_resume(dev)) { + dev_err(&dev->pci_dev->dev, "Device failed to resume\n"); + kref_get(&dev->kref); + if (IS_ERR(kthread_run(nvme_remove_dead_ctrl, dev, "nvme%d", + dev->instance))) { + dev_err(&dev->pci_dev->dev, + "Failed to start controller remove task\n"); + kref_put(&dev->kref, nvme_free_dev); + } + } +} + +static void nvme_reset_failed_dev(struct work_struct *ws) +{ + struct nvme_dev *dev = container_of(ws, struct nvme_dev, reset_work); + nvme_dev_reset(dev); +} + static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) { int result = -ENOMEM; @@ -2164,8 +2453,9 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto free; INIT_LIST_HEAD(&dev->namespaces); + INIT_WORK(&dev->reset_work, nvme_reset_failed_dev); dev->pci_dev = pdev; - + pci_set_drvdata(pdev, dev); result = nvme_set_instance(dev); if (result) goto free; @@ -2181,6 +2471,7 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto release_pools; } + kref_init(&dev->kref); result = nvme_dev_add(dev); if (result) goto shutdown; @@ -2195,15 +2486,16 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (result) goto remove; - kref_init(&dev->kref); + dev->initialized = 1; return 0; remove: nvme_dev_remove(dev); + nvme_free_namespaces(dev); shutdown: nvme_dev_shutdown(dev); release_pools: - nvme_free_queues(dev); + nvme_free_queues(dev, 0); nvme_release_prp_pools(dev); release: nvme_release_instance(dev); @@ -2214,10 +2506,28 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) return result; } +static void nvme_shutdown(struct pci_dev *pdev) +{ + struct nvme_dev *dev = pci_get_drvdata(pdev); + nvme_dev_shutdown(dev); +} + static void nvme_remove(struct pci_dev *pdev) { struct nvme_dev *dev = pci_get_drvdata(pdev); + + spin_lock(&dev_list_lock); + list_del_init(&dev->node); + spin_unlock(&dev_list_lock); + + pci_set_drvdata(pdev, NULL); + flush_work(&dev->reset_work); misc_deregister(&dev->miscdev); + nvme_dev_remove(dev); + nvme_dev_shutdown(dev); + nvme_free_queues(dev, 0); + nvme_release_instance(dev); + nvme_release_prp_pools(dev); kref_put(&dev->kref, nvme_free_dev); } @@ -2241,13 +2551,12 @@ static int nvme_resume(struct device *dev) { struct pci_dev *pdev = to_pci_dev(dev); struct nvme_dev *ndev = pci_get_drvdata(pdev); - int ret; - ret = nvme_dev_start(ndev); - /* XXX: should remove gendisks if resume fails */ - if (ret) - nvme_free_queues(ndev); - return ret; + if (nvme_dev_resume(ndev) && !work_busy(&ndev->reset_work)) { + PREPARE_WORK(&ndev->reset_work, nvme_reset_failed_dev); + queue_work(nvme_workq, &ndev->reset_work); + } + return 0; } static SIMPLE_DEV_PM_OPS(nvme_dev_pm_ops, nvme_suspend, nvme_resume); @@ -2274,6 +2583,7 @@ static struct pci_driver nvme_driver = { .id_table = nvme_id_table, .probe = nvme_probe, .remove = nvme_remove, + .shutdown = nvme_shutdown, .driver = { .pm = &nvme_dev_pm_ops, }, @@ -2288,9 +2598,14 @@ static int __init nvme_init(void) if (IS_ERR(nvme_thread)) return PTR_ERR(nvme_thread); + result = -ENOMEM; + nvme_workq = create_singlethread_workqueue("nvme"); + if (!nvme_workq) + goto kill_kthread; + result = register_blkdev(nvme_major, "nvme"); if (result < 0) - goto kill_kthread; + goto kill_workq; else if (result > 0) nvme_major = result; @@ -2301,6 +2616,8 @@ static int __init nvme_init(void) unregister_blkdev: unregister_blkdev(nvme_major, "nvme"); + kill_workq: + destroy_workqueue(nvme_workq); kill_kthread: kthread_stop(nvme_thread); return result; @@ -2310,6 +2627,7 @@ static void __exit nvme_exit(void) { pci_unregister_driver(&nvme_driver); unregister_blkdev(nvme_major, "nvme"); + destroy_workqueue(nvme_workq); kthread_stop(nvme_thread); } diff --git a/drivers/block/nvme-scsi.c b/drivers/block/nvme-scsi.c index 4a4ff4eb8e23..4a0ceb64e269 100644 --- a/drivers/block/nvme-scsi.c +++ b/drivers/block/nvme-scsi.c @@ -25,6 +25,7 @@ #include <linux/bio.h> #include <linux/bitops.h> #include <linux/blkdev.h> +#include <linux/compat.h> #include <linux/delay.h> #include <linux/errno.h> #include <linux/fs.h> @@ -3038,6 +3039,152 @@ int nvme_sg_io(struct nvme_ns *ns, struct sg_io_hdr __user *u_hdr) return retcode; } +#ifdef CONFIG_COMPAT +typedef struct sg_io_hdr32 { + compat_int_t interface_id; /* [i] 'S' for SCSI generic (required) */ + compat_int_t dxfer_direction; /* [i] data transfer direction */ + unsigned char cmd_len; /* [i] SCSI command length ( <= 16 bytes) */ + unsigned char mx_sb_len; /* [i] max length to write to sbp */ + unsigned short iovec_count; /* [i] 0 implies no scatter gather */ + compat_uint_t dxfer_len; /* [i] byte count of data transfer */ + compat_uint_t dxferp; /* [i], [*io] points to data transfer memory + or scatter gather list */ + compat_uptr_t cmdp; /* [i], [*i] points to command to perform */ + compat_uptr_t sbp; /* [i], [*o] points to sense_buffer memory */ + compat_uint_t timeout; /* [i] MAX_UINT->no timeout (unit: millisec) */ + compat_uint_t flags; /* [i] 0 -> default, see SG_FLAG... */ + compat_int_t pack_id; /* [i->o] unused internally (normally) */ + compat_uptr_t usr_ptr; /* [i->o] unused internally */ + unsigned char status; /* [o] scsi status */ + unsigned char masked_status; /* [o] shifted, masked scsi status */ + unsigned char msg_status; /* [o] messaging level data (optional) */ + unsigned char sb_len_wr; /* [o] byte count actually written to sbp */ + unsigned short host_status; /* [o] errors from host adapter */ + unsigned short driver_status; /* [o] errors from software driver */ + compat_int_t resid; /* [o] dxfer_len - actual_transferred */ + compat_uint_t duration; /* [o] time taken by cmd (unit: millisec) */ + compat_uint_t info; /* [o] auxiliary information */ +} sg_io_hdr32_t; /* 64 bytes long (on sparc32) */ + +typedef struct sg_iovec32 { + compat_uint_t iov_base; + compat_uint_t iov_len; +} sg_iovec32_t; + +static int sg_build_iovec(sg_io_hdr_t __user *sgio, void __user *dxferp, u16 iovec_count) +{ + sg_iovec_t __user *iov = (sg_iovec_t __user *) (sgio + 1); + sg_iovec32_t __user *iov32 = dxferp; + int i; + + for (i = 0; i < iovec_count; i++) { + u32 base, len; + + if (get_user(base, &iov32[i].iov_base) || + get_user(len, &iov32[i].iov_len) || + put_user(compat_ptr(base), &iov[i].iov_base) || + put_user(len, &iov[i].iov_len)) + return -EFAULT; + } + + if (put_user(iov, &sgio->dxferp)) + return -EFAULT; + return 0; +} + +int nvme_sg_io32(struct nvme_ns *ns, unsigned long arg) +{ + sg_io_hdr32_t __user *sgio32 = (sg_io_hdr32_t __user *)arg; + sg_io_hdr_t __user *sgio; + u16 iovec_count; + u32 data; + void __user *dxferp; + int err; + int interface_id; + + if (get_user(interface_id, &sgio32->interface_id)) + return -EFAULT; + if (interface_id != 'S') + return -EINVAL; + + if (get_user(iovec_count, &sgio32->iovec_count)) + return -EFAULT; + + { + void __user *top = compat_alloc_user_space(0); + void __user *new = compat_alloc_user_space(sizeof(sg_io_hdr_t) + + (iovec_count * sizeof(sg_iovec_t))); + if (new > top) + return -EINVAL; + + sgio = new; + } + + /* Ok, now construct. */ + if (copy_in_user(&sgio->interface_id, &sgio32->interface_id, + (2 * sizeof(int)) + + (2 * sizeof(unsigned char)) + + (1 * sizeof(unsigned short)) + + (1 * sizeof(unsigned int)))) + return -EFAULT; + + if (get_user(data, &sgio32->dxferp)) + return -EFAULT; + dxferp = compat_ptr(data); + if (iovec_count) { + if (sg_build_iovec(sgio, dxferp, iovec_count)) + return -EFAULT; + } else { + if (put_user(dxferp, &sgio->dxferp)) + return -EFAULT; + } + + { + unsigned char __user *cmdp; + unsigned char __user *sbp; + + if (get_user(data, &sgio32->cmdp)) + return -EFAULT; + cmdp = compat_ptr(data); + + if (get_user(data, &sgio32->sbp)) + return -EFAULT; + sbp = compat_ptr(data); + + if (put_user(cmdp, &sgio->cmdp) || + put_user(sbp, &sgio->sbp)) + return -EFAULT; + } + + if (copy_in_user(&sgio->timeout, &sgio32->timeout, + 3 * sizeof(int))) + return -EFAULT; + + if (get_user(data, &sgio32->usr_ptr)) + return -EFAULT; + if (put_user(compat_ptr(data), &sgio->usr_ptr)) + return -EFAULT; + + err = nvme_sg_io(ns, sgio); + if (err >= 0) { + void __user *datap; + + if (copy_in_user(&sgio32->pack_id, &sgio->pack_id, + sizeof(int)) || + get_user(datap, &sgio->usr_ptr) || + put_user((u32)(unsigned long)datap, + &sgio32->usr_ptr) || + copy_in_user(&sgio32->status, &sgio->status, + (4 * sizeof(unsigned char)) + + (2 * sizeof(unsigned short)) + + (3 * sizeof(int)))) + err = -EFAULT; + } + + return err; +} +#endif + int nvme_sg_get_version_num(int __user *ip) { return put_user(sg_version_num, ip); diff --git a/drivers/block/paride/pg.c b/drivers/block/paride/pg.c index 4a27b1de5fcb..2ce3dfd7e6b9 100644 --- a/drivers/block/paride/pg.c +++ b/drivers/block/paride/pg.c @@ -581,7 +581,7 @@ static ssize_t pg_write(struct file *filp, const char __user *buf, size_t count, if (hdr.magic != PG_MAGIC) return -EINVAL; - if (hdr.dlen > PG_MAX_DATA) + if (hdr.dlen < 0 || hdr.dlen > PG_MAX_DATA) return -EINVAL; if ((count - hs) > PG_MAX_DATA) return -EINVAL; diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index ff8668c5efb1..a2af73db187b 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -651,7 +651,7 @@ static struct pkt_rb_node *pkt_rbtree_find(struct pktcdvd_device *pd, sector_t s for (;;) { tmp = rb_entry(n, struct pkt_rb_node, rb_node); - if (s <= tmp->bio->bi_sector) + if (s <= tmp->bio->bi_iter.bi_sector) next = n->rb_left; else next = n->rb_right; @@ -660,12 +660,12 @@ static struct pkt_rb_node *pkt_rbtree_find(struct pktcdvd_device *pd, sector_t s n = next; } - if (s > tmp->bio->bi_sector) { + if (s > tmp->bio->bi_iter.bi_sector) { tmp = pkt_rbtree_next(tmp); if (!tmp) return NULL; } - BUG_ON(s > tmp->bio->bi_sector); + BUG_ON(s > tmp->bio->bi_iter.bi_sector); return tmp; } @@ -676,13 +676,13 @@ static void pkt_rbtree_insert(struct pktcdvd_device *pd, struct pkt_rb_node *nod { struct rb_node **p = &pd->bio_queue.rb_node; struct rb_node *parent = NULL; - sector_t s = node->bio->bi_sector; + sector_t s = node->bio->bi_iter.bi_sector; struct pkt_rb_node *tmp; while (*p) { parent = *p; tmp = rb_entry(parent, struct pkt_rb_node, rb_node); - if (s < tmp->bio->bi_sector) + if (s < tmp->bio->bi_iter.bi_sector) p = &(*p)->rb_left; else p = &(*p)->rb_right; @@ -706,7 +706,9 @@ static int pkt_generic_packet(struct pktcdvd_device *pd, struct packet_command * WRITE : READ, __GFP_WAIT); if (cgc->buflen) { - if (blk_rq_map_kern(q, rq, cgc->buffer, cgc->buflen, __GFP_WAIT)) + ret = blk_rq_map_kern(q, rq, cgc->buffer, cgc->buflen, + __GFP_WAIT); + if (ret) goto out; } @@ -857,7 +859,8 @@ static void pkt_iosched_process_queue(struct pktcdvd_device *pd) spin_lock(&pd->iosched.lock); bio = bio_list_peek(&pd->iosched.write_queue); spin_unlock(&pd->iosched.lock); - if (bio && (bio->bi_sector == pd->iosched.last_write)) + if (bio && (bio->bi_iter.bi_sector == + pd->iosched.last_write)) need_write_seek = 0; if (need_write_seek && reads_queued) { if (atomic_read(&pd->cdrw.pending_bios) > 0) { @@ -888,7 +891,8 @@ static void pkt_iosched_process_queue(struct pktcdvd_device *pd) continue; if (bio_data_dir(bio) == READ) - pd->iosched.successive_reads += bio->bi_size >> 10; + pd->iosched.successive_reads += + bio->bi_iter.bi_size >> 10; else { pd->iosched.successive_reads = 0; pd->iosched.last_write = bio_end_sector(bio); @@ -978,7 +982,7 @@ static void pkt_end_io_read(struct bio *bio, int err) pkt_dbg(2, pd, "bio=%p sec0=%llx sec=%llx err=%d\n", bio, (unsigned long long)pkt->sector, - (unsigned long long)bio->bi_sector, err); + (unsigned long long)bio->bi_iter.bi_sector, err); if (err) atomic_inc(&pkt->io_errors); @@ -1026,8 +1030,9 @@ static void pkt_gather_data(struct pktcdvd_device *pd, struct packet_data *pkt) memset(written, 0, sizeof(written)); spin_lock(&pkt->lock); bio_list_for_each(bio, &pkt->orig_bios) { - int first_frame = (bio->bi_sector - pkt->sector) / (CD_FRAMESIZE >> 9); - int num_frames = bio->bi_size / CD_FRAMESIZE; + int first_frame = (bio->bi_iter.bi_sector - pkt->sector) / + (CD_FRAMESIZE >> 9); + int num_frames = bio->bi_iter.bi_size / CD_FRAMESIZE; pd->stats.secs_w += num_frames * (CD_FRAMESIZE >> 9); BUG_ON(first_frame < 0); BUG_ON(first_frame + num_frames > pkt->frames); @@ -1053,7 +1058,7 @@ static void pkt_gather_data(struct pktcdvd_device *pd, struct packet_data *pkt) bio = pkt->r_bios[f]; bio_reset(bio); - bio->bi_sector = pkt->sector + f * (CD_FRAMESIZE >> 9); + bio->bi_iter.bi_sector = pkt->sector + f * (CD_FRAMESIZE >> 9); bio->bi_bdev = pd->bdev; bio->bi_end_io = pkt_end_io_read; bio->bi_private = pkt; @@ -1150,8 +1155,8 @@ static int pkt_start_recovery(struct packet_data *pkt) bio_reset(pkt->bio); pkt->bio->bi_bdev = pd->bdev; pkt->bio->bi_rw = REQ_WRITE; - pkt->bio->bi_sector = new_sector; - pkt->bio->bi_size = pkt->frames * CD_FRAMESIZE; + pkt->bio->bi_iter.bi_sector = new_sector; + pkt->bio->bi_iter.bi_size = pkt->frames * CD_FRAMESIZE; pkt->bio->bi_vcnt = pkt->frames; pkt->bio->bi_end_io = pkt_end_io_packet_write; @@ -1213,7 +1218,7 @@ static int pkt_handle_queue(struct pktcdvd_device *pd) node = first_node; while (node) { bio = node->bio; - zone = get_zone(bio->bi_sector, pd); + zone = get_zone(bio->bi_iter.bi_sector, pd); list_for_each_entry(p, &pd->cdrw.pkt_active_list, list) { if (p->sector == zone) { bio = NULL; @@ -1252,14 +1257,14 @@ try_next_bio: pkt_dbg(2, pd, "looking for zone %llx\n", (unsigned long long)zone); while ((node = pkt_rbtree_find(pd, zone)) != NULL) { bio = node->bio; - pkt_dbg(2, pd, "found zone=%llx\n", - (unsigned long long)get_zone(bio->bi_sector, pd)); - if (get_zone(bio->bi_sector, pd) != zone) + pkt_dbg(2, pd, "found zone=%llx\n", (unsigned long long) + get_zone(bio->bi_iter.bi_sector, pd)); + if (get_zone(bio->bi_iter.bi_sector, pd) != zone) break; pkt_rbtree_erase(pd, node); spin_lock(&pkt->lock); bio_list_add(&pkt->orig_bios, bio); - pkt->write_size += bio->bi_size / CD_FRAMESIZE; + pkt->write_size += bio->bi_iter.bi_size / CD_FRAMESIZE; spin_unlock(&pkt->lock); } /* check write congestion marks, and if bio_queue_size is @@ -1293,7 +1298,7 @@ static void pkt_start_write(struct pktcdvd_device *pd, struct packet_data *pkt) struct bio_vec *bvec = pkt->w_bio->bi_io_vec; bio_reset(pkt->w_bio); - pkt->w_bio->bi_sector = pkt->sector; + pkt->w_bio->bi_iter.bi_sector = pkt->sector; pkt->w_bio->bi_bdev = pd->bdev; pkt->w_bio->bi_end_io = pkt_end_io_packet_write; pkt->w_bio->bi_private = pkt; @@ -2335,75 +2340,29 @@ static void pkt_end_io_read_cloned(struct bio *bio, int err) pkt_bio_finished(pd); } -static void pkt_make_request(struct request_queue *q, struct bio *bio) +static void pkt_make_request_read(struct pktcdvd_device *pd, struct bio *bio) { - struct pktcdvd_device *pd; - char b[BDEVNAME_SIZE]; + struct bio *cloned_bio = bio_clone(bio, GFP_NOIO); + struct packet_stacked_data *psd = mempool_alloc(psd_pool, GFP_NOIO); + + psd->pd = pd; + psd->bio = bio; + cloned_bio->bi_bdev = pd->bdev; + cloned_bio->bi_private = psd; + cloned_bio->bi_end_io = pkt_end_io_read_cloned; + pd->stats.secs_r += bio_sectors(bio); + pkt_queue_bio(pd, cloned_bio); +} + +static void pkt_make_request_write(struct request_queue *q, struct bio *bio) +{ + struct pktcdvd_device *pd = q->queuedata; sector_t zone; struct packet_data *pkt; int was_empty, blocked_bio; struct pkt_rb_node *node; - pd = q->queuedata; - if (!pd) { - pr_err("%s incorrect request queue\n", - bdevname(bio->bi_bdev, b)); - goto end_io; - } - - /* - * Clone READ bios so we can have our own bi_end_io callback. - */ - if (bio_data_dir(bio) == READ) { - struct bio *cloned_bio = bio_clone(bio, GFP_NOIO); - struct packet_stacked_data *psd = mempool_alloc(psd_pool, GFP_NOIO); - - psd->pd = pd; - psd->bio = bio; - cloned_bio->bi_bdev = pd->bdev; - cloned_bio->bi_private = psd; - cloned_bio->bi_end_io = pkt_end_io_read_cloned; - pd->stats.secs_r += bio_sectors(bio); - pkt_queue_bio(pd, cloned_bio); - return; - } - - if (!test_bit(PACKET_WRITABLE, &pd->flags)) { - pkt_notice(pd, "WRITE for ro device (%llu)\n", - (unsigned long long)bio->bi_sector); - goto end_io; - } - - if (!bio->bi_size || (bio->bi_size % CD_FRAMESIZE)) { - pkt_err(pd, "wrong bio size\n"); - goto end_io; - } - - blk_queue_bounce(q, &bio); - - zone = get_zone(bio->bi_sector, pd); - pkt_dbg(2, pd, "start = %6llx stop = %6llx\n", - (unsigned long long)bio->bi_sector, - (unsigned long long)bio_end_sector(bio)); - - /* Check if we have to split the bio */ - { - struct bio_pair *bp; - sector_t last_zone; - int first_sectors; - - last_zone = get_zone(bio_end_sector(bio) - 1, pd); - if (last_zone != zone) { - BUG_ON(last_zone != zone + pd->settings.size); - first_sectors = last_zone - bio->bi_sector; - bp = bio_split(bio, first_sectors); - BUG_ON(!bp); - pkt_make_request(q, &bp->bio1); - pkt_make_request(q, &bp->bio2); - bio_pair_release(bp); - return; - } - } + zone = get_zone(bio->bi_iter.bi_sector, pd); /* * If we find a matching packet in state WAITING or READ_WAIT, we can @@ -2417,7 +2376,8 @@ static void pkt_make_request(struct request_queue *q, struct bio *bio) if ((pkt->state == PACKET_WAITING_STATE) || (pkt->state == PACKET_READ_WAIT_STATE)) { bio_list_add(&pkt->orig_bios, bio); - pkt->write_size += bio->bi_size / CD_FRAMESIZE; + pkt->write_size += + bio->bi_iter.bi_size / CD_FRAMESIZE; if ((pkt->write_size >= pkt->frames) && (pkt->state == PACKET_WAITING_STATE)) { atomic_inc(&pkt->run_sm); @@ -2476,6 +2436,64 @@ static void pkt_make_request(struct request_queue *q, struct bio *bio) */ wake_up(&pd->wqueue); } +} + +static void pkt_make_request(struct request_queue *q, struct bio *bio) +{ + struct pktcdvd_device *pd; + char b[BDEVNAME_SIZE]; + struct bio *split; + + pd = q->queuedata; + if (!pd) { + pr_err("%s incorrect request queue\n", + bdevname(bio->bi_bdev, b)); + goto end_io; + } + + pkt_dbg(2, pd, "start = %6llx stop = %6llx\n", + (unsigned long long)bio->bi_iter.bi_sector, + (unsigned long long)bio_end_sector(bio)); + + /* + * Clone READ bios so we can have our own bi_end_io callback. + */ + if (bio_data_dir(bio) == READ) { + pkt_make_request_read(pd, bio); + return; + } + + if (!test_bit(PACKET_WRITABLE, &pd->flags)) { + pkt_notice(pd, "WRITE for ro device (%llu)\n", + (unsigned long long)bio->bi_iter.bi_sector); + goto end_io; + } + + if (!bio->bi_iter.bi_size || (bio->bi_iter.bi_size % CD_FRAMESIZE)) { + pkt_err(pd, "wrong bio size\n"); + goto end_io; + } + + blk_queue_bounce(q, &bio); + + do { + sector_t zone = get_zone(bio->bi_iter.bi_sector, pd); + sector_t last_zone = get_zone(bio_end_sector(bio) - 1, pd); + + if (last_zone != zone) { + BUG_ON(last_zone != zone + pd->settings.size); + + split = bio_split(bio, last_zone - + bio->bi_iter.bi_sector, + GFP_NOIO, fs_bio_set); + bio_chain(split, bio); + } else { + split = bio; + } + + pkt_make_request_write(q, split); + } while (split != bio); + return; end_io: bio_io_error(bio); diff --git a/drivers/block/ps3disk.c b/drivers/block/ps3disk.c index d754a88d7585..c120d70d3fb3 100644 --- a/drivers/block/ps3disk.c +++ b/drivers/block/ps3disk.c @@ -94,26 +94,25 @@ static void ps3disk_scatter_gather(struct ps3_storage_device *dev, { unsigned int offset = 0; struct req_iterator iter; - struct bio_vec *bvec; + struct bio_vec bvec; unsigned int i = 0; size_t size; void *buf; rq_for_each_segment(bvec, req, iter) { unsigned long flags; - dev_dbg(&dev->sbd.core, - "%s:%u: bio %u: %u segs %u sectors from %lu\n", - __func__, __LINE__, i, bio_segments(iter.bio), - bio_sectors(iter.bio), iter.bio->bi_sector); + dev_dbg(&dev->sbd.core, "%s:%u: bio %u: %u sectors from %lu\n", + __func__, __LINE__, i, bio_sectors(iter.bio), + iter.bio->bi_iter.bi_sector); - size = bvec->bv_len; - buf = bvec_kmap_irq(bvec, &flags); + size = bvec.bv_len; + buf = bvec_kmap_irq(&bvec, &flags); if (gather) memcpy(dev->bounce_buf+offset, buf, size); else memcpy(buf, dev->bounce_buf+offset, size); offset += size; - flush_kernel_dcache_page(bvec->bv_page); + flush_kernel_dcache_page(bvec.bv_page); bvec_kunmap_irq(buf, &flags); i++; } @@ -130,7 +129,7 @@ static int ps3disk_submit_request_sg(struct ps3_storage_device *dev, #ifdef DEBUG unsigned int n = 0; - struct bio_vec *bv; + struct bio_vec bv; struct req_iterator iter; rq_for_each_segment(bv, req, iter) diff --git a/drivers/block/ps3vram.c b/drivers/block/ps3vram.c index 06a2e53e5f37..ef45cfb98fd2 100644 --- a/drivers/block/ps3vram.c +++ b/drivers/block/ps3vram.c @@ -553,16 +553,16 @@ static struct bio *ps3vram_do_bio(struct ps3_system_bus_device *dev, struct ps3vram_priv *priv = ps3_system_bus_get_drvdata(dev); int write = bio_data_dir(bio) == WRITE; const char *op = write ? "write" : "read"; - loff_t offset = bio->bi_sector << 9; + loff_t offset = bio->bi_iter.bi_sector << 9; int error = 0; - struct bio_vec *bvec; - unsigned int i; + struct bio_vec bvec; + struct bvec_iter iter; struct bio *next; - bio_for_each_segment(bvec, bio, i) { + bio_for_each_segment(bvec, bio, iter) { /* PS3 is ppc64, so we don't handle highmem */ - char *ptr = page_address(bvec->bv_page) + bvec->bv_offset; - size_t len = bvec->bv_len, retlen; + char *ptr = page_address(bvec.bv_page) + bvec.bv_offset; + size_t len = bvec.bv_len, retlen; dev_dbg(&dev->core, " %s %zu bytes at offset %llu\n", op, len, offset); diff --git a/drivers/block/rbd.c b/drivers/block/rbd.c index cb1db2979d3d..b365e0dfccb6 100644 --- a/drivers/block/rbd.c +++ b/drivers/block/rbd.c @@ -41,6 +41,7 @@ #include <linux/fs.h> #include <linux/blkdev.h> #include <linux/slab.h> +#include <linux/idr.h> #include "rbd_types.h" @@ -89,9 +90,9 @@ static int atomic_dec_return_safe(atomic_t *v) } #define RBD_DRV_NAME "rbd" -#define RBD_DRV_NAME_LONG "rbd (rados block device)" -#define RBD_MINORS_PER_MAJOR 256 /* max minors per blkdev */ +#define RBD_MINORS_PER_MAJOR 256 +#define RBD_SINGLE_MAJOR_PART_SHIFT 4 #define RBD_SNAP_DEV_NAME_PREFIX "snap_" #define RBD_MAX_SNAP_NAME_LEN \ @@ -323,6 +324,7 @@ struct rbd_device { int dev_id; /* blkdev unique id */ int major; /* blkdev assigned major */ + int minor; struct gendisk *disk; /* blkdev's gendisk and rq */ u32 image_format; /* Either 1 or 2 */ @@ -386,6 +388,17 @@ static struct kmem_cache *rbd_img_request_cache; static struct kmem_cache *rbd_obj_request_cache; static struct kmem_cache *rbd_segment_name_cache; +static int rbd_major; +static DEFINE_IDA(rbd_dev_id_ida); + +/* + * Default to false for now, as single-major requires >= 0.75 version of + * userspace rbd utility. + */ +static bool single_major = false; +module_param(single_major, bool, S_IRUGO); +MODULE_PARM_DESC(single_major, "Use a single major number for all rbd devices (default: false)"); + static int rbd_img_request_submit(struct rbd_img_request *img_request); static void rbd_dev_device_release(struct device *dev); @@ -394,18 +407,52 @@ static ssize_t rbd_add(struct bus_type *bus, const char *buf, size_t count); static ssize_t rbd_remove(struct bus_type *bus, const char *buf, size_t count); +static ssize_t rbd_add_single_major(struct bus_type *bus, const char *buf, + size_t count); +static ssize_t rbd_remove_single_major(struct bus_type *bus, const char *buf, + size_t count); static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping); static void rbd_spec_put(struct rbd_spec *spec); +static int rbd_dev_id_to_minor(int dev_id) +{ + return dev_id << RBD_SINGLE_MAJOR_PART_SHIFT; +} + +static int minor_to_rbd_dev_id(int minor) +{ + return minor >> RBD_SINGLE_MAJOR_PART_SHIFT; +} + static BUS_ATTR(add, S_IWUSR, NULL, rbd_add); static BUS_ATTR(remove, S_IWUSR, NULL, rbd_remove); +static BUS_ATTR(add_single_major, S_IWUSR, NULL, rbd_add_single_major); +static BUS_ATTR(remove_single_major, S_IWUSR, NULL, rbd_remove_single_major); static struct attribute *rbd_bus_attrs[] = { &bus_attr_add.attr, &bus_attr_remove.attr, + &bus_attr_add_single_major.attr, + &bus_attr_remove_single_major.attr, NULL, }; -ATTRIBUTE_GROUPS(rbd_bus); + +static umode_t rbd_bus_is_visible(struct kobject *kobj, + struct attribute *attr, int index) +{ + if (!single_major && + (attr == &bus_attr_add_single_major.attr || + attr == &bus_attr_remove_single_major.attr)) + return 0; + + return attr->mode; +} + +static const struct attribute_group rbd_bus_group = { + .attrs = rbd_bus_attrs, + .is_visible = rbd_bus_is_visible, +}; +__ATTRIBUTE_GROUPS(rbd_bus); static struct bus_type rbd_bus_type = { .name = "rbd", @@ -1041,9 +1088,9 @@ static const char *rbd_segment_name(struct rbd_device *rbd_dev, u64 offset) name_format = "%s.%012llx"; if (rbd_dev->image_format == 2) name_format = "%s.%016llx"; - ret = snprintf(name, MAX_OBJ_NAME_SIZE + 1, name_format, + ret = snprintf(name, CEPH_MAX_OID_NAME_LEN + 1, name_format, rbd_dev->header.object_prefix, segment); - if (ret < 0 || ret > MAX_OBJ_NAME_SIZE) { + if (ret < 0 || ret > CEPH_MAX_OID_NAME_LEN) { pr_err("error formatting segment name for #%llu (%d)\n", segment, ret); kfree(name); @@ -1109,23 +1156,23 @@ static void bio_chain_put(struct bio *chain) */ static void zero_bio_chain(struct bio *chain, int start_ofs) { - struct bio_vec *bv; + struct bio_vec bv; + struct bvec_iter iter; unsigned long flags; void *buf; - int i; int pos = 0; while (chain) { - bio_for_each_segment(bv, chain, i) { - if (pos + bv->bv_len > start_ofs) { + bio_for_each_segment(bv, chain, iter) { + if (pos + bv.bv_len > start_ofs) { int remainder = max(start_ofs - pos, 0); - buf = bvec_kmap_irq(bv, &flags); + buf = bvec_kmap_irq(&bv, &flags); memset(buf + remainder, 0, - bv->bv_len - remainder); - flush_dcache_page(bv->bv_page); + bv.bv_len - remainder); + flush_dcache_page(bv.bv_page); bvec_kunmap_irq(buf, &flags); } - pos += bv->bv_len; + pos += bv.bv_len; } chain = chain->bi_next; @@ -1173,74 +1220,14 @@ static struct bio *bio_clone_range(struct bio *bio_src, unsigned int len, gfp_t gfpmask) { - struct bio_vec *bv; - unsigned int resid; - unsigned short idx; - unsigned int voff; - unsigned short end_idx; - unsigned short vcnt; struct bio *bio; - /* Handle the easy case for the caller */ - - if (!offset && len == bio_src->bi_size) - return bio_clone(bio_src, gfpmask); - - if (WARN_ON_ONCE(!len)) - return NULL; - if (WARN_ON_ONCE(len > bio_src->bi_size)) - return NULL; - if (WARN_ON_ONCE(offset > bio_src->bi_size - len)) - return NULL; - - /* Find first affected segment... */ - - resid = offset; - bio_for_each_segment(bv, bio_src, idx) { - if (resid < bv->bv_len) - break; - resid -= bv->bv_len; - } - voff = resid; - - /* ...and the last affected segment */ - - resid += len; - __bio_for_each_segment(bv, bio_src, end_idx, idx) { - if (resid <= bv->bv_len) - break; - resid -= bv->bv_len; - } - vcnt = end_idx - idx + 1; - - /* Build the clone */ - - bio = bio_alloc(gfpmask, (unsigned int) vcnt); + bio = bio_clone(bio_src, gfpmask); if (!bio) return NULL; /* ENOMEM */ - bio->bi_bdev = bio_src->bi_bdev; - bio->bi_sector = bio_src->bi_sector + (offset >> SECTOR_SHIFT); - bio->bi_rw = bio_src->bi_rw; - bio->bi_flags |= 1 << BIO_CLONED; - - /* - * Copy over our part of the bio_vec, then update the first - * and last (or only) entries. - */ - memcpy(&bio->bi_io_vec[0], &bio_src->bi_io_vec[idx], - vcnt * sizeof (struct bio_vec)); - bio->bi_io_vec[0].bv_offset += voff; - if (vcnt > 1) { - bio->bi_io_vec[0].bv_len -= voff; - bio->bi_io_vec[vcnt - 1].bv_len = resid; - } else { - bio->bi_io_vec[0].bv_len = len; - } - - bio->bi_vcnt = vcnt; - bio->bi_size = len; - bio->bi_idx = 0; + bio_advance(bio, offset); + bio->bi_iter.bi_size = len; return bio; } @@ -1271,7 +1258,7 @@ static struct bio *bio_chain_clone_range(struct bio **bio_src, /* Build up a chain of clone bios up to the limit */ - if (!bi || off >= bi->bi_size || !len) + if (!bi || off >= bi->bi_iter.bi_size || !len) return NULL; /* Nothing to clone */ end = &chain; @@ -1283,7 +1270,7 @@ static struct bio *bio_chain_clone_range(struct bio **bio_src, rbd_warn(NULL, "bio_chain exhausted with %u left", len); goto out_err; /* EINVAL; ran out of bio's */ } - bi_size = min_t(unsigned int, bi->bi_size - off, len); + bi_size = min_t(unsigned int, bi->bi_iter.bi_size - off, len); bio = bio_clone_range(bi, off, bi_size, gfpmask); if (!bio) goto out_err; /* ENOMEM */ @@ -1292,7 +1279,7 @@ static struct bio *bio_chain_clone_range(struct bio **bio_src, end = &bio->bi_next; off += bi_size; - if (off == bi->bi_size) { + if (off == bi->bi_iter.bi_size) { bi = bi->bi_next; off = 0; } @@ -1761,11 +1748,8 @@ static struct ceph_osd_request *rbd_osd_req_create( osd_req->r_callback = rbd_osd_req_callback; osd_req->r_priv = obj_request; - osd_req->r_oid_len = strlen(obj_request->object_name); - rbd_assert(osd_req->r_oid_len < sizeof (osd_req->r_oid)); - memcpy(osd_req->r_oid, obj_request->object_name, osd_req->r_oid_len); - - osd_req->r_file_layout = rbd_dev->layout; /* struct */ + osd_req->r_base_oloc.pool = ceph_file_layout_pg_pool(rbd_dev->layout); + ceph_oid_set_name(&osd_req->r_base_oid, obj_request->object_name); return osd_req; } @@ -1802,11 +1786,8 @@ rbd_osd_req_create_copyup(struct rbd_obj_request *obj_request) osd_req->r_callback = rbd_osd_req_callback; osd_req->r_priv = obj_request; - osd_req->r_oid_len = strlen(obj_request->object_name); - rbd_assert(osd_req->r_oid_len < sizeof (osd_req->r_oid)); - memcpy(osd_req->r_oid, obj_request->object_name, osd_req->r_oid_len); - - osd_req->r_file_layout = rbd_dev->layout; /* struct */ + osd_req->r_base_oloc.pool = ceph_file_layout_pg_pool(rbd_dev->layout); + ceph_oid_set_name(&osd_req->r_base_oid, obj_request->object_name); return osd_req; } @@ -2186,7 +2167,8 @@ static int rbd_img_request_fill(struct rbd_img_request *img_request, if (type == OBJ_REQUEST_BIO) { bio_list = data_desc; - rbd_assert(img_offset == bio_list->bi_sector << SECTOR_SHIFT); + rbd_assert(img_offset == + bio_list->bi_iter.bi_sector << SECTOR_SHIFT); } else { rbd_assert(type == OBJ_REQUEST_PAGES); pages = data_desc; @@ -2866,7 +2848,7 @@ static void rbd_watch_cb(u64 ver, u64 notify_id, u8 opcode, void *data) * Request sync osd watch/unwatch. The value of "start" determines * whether a watch request is being initiated or torn down. */ -static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, bool start) +static int __rbd_dev_header_watch_sync(struct rbd_device *rbd_dev, bool start) { struct ceph_osd_client *osdc = &rbd_dev->rbd_client->client->osdc; struct rbd_obj_request *obj_request; @@ -2941,6 +2923,22 @@ out_cancel: return ret; } +static int rbd_dev_header_watch_sync(struct rbd_device *rbd_dev) +{ + return __rbd_dev_header_watch_sync(rbd_dev, true); +} + +static void rbd_dev_header_unwatch_sync(struct rbd_device *rbd_dev) +{ + int ret; + + ret = __rbd_dev_header_watch_sync(rbd_dev, false); + if (ret) { + rbd_warn(rbd_dev, "unable to tear down watch request: %d\n", + ret); + } +} + /* * Synchronous osd object method call. Returns the number of bytes * returned in the outbound buffer, or a negative error code. @@ -3388,14 +3386,18 @@ static int rbd_init_disk(struct rbd_device *rbd_dev) u64 segment_size; /* create gendisk info */ - disk = alloc_disk(RBD_MINORS_PER_MAJOR); + disk = alloc_disk(single_major ? + (1 << RBD_SINGLE_MAJOR_PART_SHIFT) : + RBD_MINORS_PER_MAJOR); if (!disk) return -ENOMEM; snprintf(disk->disk_name, sizeof(disk->disk_name), RBD_DRV_NAME "%d", rbd_dev->dev_id); disk->major = rbd_dev->major; - disk->first_minor = 0; + disk->first_minor = rbd_dev->minor; + if (single_major) + disk->flags |= GENHD_FL_EXT_DEVT; disk->fops = &rbd_bd_ops; disk->private_data = rbd_dev; @@ -3467,7 +3469,14 @@ static ssize_t rbd_major_show(struct device *dev, return sprintf(buf, "%d\n", rbd_dev->major); return sprintf(buf, "(none)\n"); +} + +static ssize_t rbd_minor_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); + return sprintf(buf, "%d\n", rbd_dev->minor); } static ssize_t rbd_client_id_show(struct device *dev, @@ -3589,6 +3598,7 @@ static ssize_t rbd_image_refresh(struct device *dev, static DEVICE_ATTR(size, S_IRUGO, rbd_size_show, NULL); static DEVICE_ATTR(features, S_IRUGO, rbd_features_show, NULL); static DEVICE_ATTR(major, S_IRUGO, rbd_major_show, NULL); +static DEVICE_ATTR(minor, S_IRUGO, rbd_minor_show, NULL); static DEVICE_ATTR(client_id, S_IRUGO, rbd_client_id_show, NULL); static DEVICE_ATTR(pool, S_IRUGO, rbd_pool_show, NULL); static DEVICE_ATTR(pool_id, S_IRUGO, rbd_pool_id_show, NULL); @@ -3602,6 +3612,7 @@ static struct attribute *rbd_attrs[] = { &dev_attr_size.attr, &dev_attr_features.attr, &dev_attr_major.attr, + &dev_attr_minor.attr, &dev_attr_client_id.attr, &dev_attr_pool.attr, &dev_attr_pool_id.attr, @@ -4372,21 +4383,29 @@ static void rbd_bus_del_dev(struct rbd_device *rbd_dev) device_unregister(&rbd_dev->dev); } -static atomic64_t rbd_dev_id_max = ATOMIC64_INIT(0); - /* * Get a unique rbd identifier for the given new rbd_dev, and add - * the rbd_dev to the global list. The minimum rbd id is 1. + * the rbd_dev to the global list. */ -static void rbd_dev_id_get(struct rbd_device *rbd_dev) +static int rbd_dev_id_get(struct rbd_device *rbd_dev) { - rbd_dev->dev_id = atomic64_inc_return(&rbd_dev_id_max); + int new_dev_id; + + new_dev_id = ida_simple_get(&rbd_dev_id_ida, + 0, minor_to_rbd_dev_id(1 << MINORBITS), + GFP_KERNEL); + if (new_dev_id < 0) + return new_dev_id; + + rbd_dev->dev_id = new_dev_id; spin_lock(&rbd_dev_list_lock); list_add_tail(&rbd_dev->node, &rbd_dev_list); spin_unlock(&rbd_dev_list_lock); - dout("rbd_dev %p given dev id %llu\n", rbd_dev, - (unsigned long long) rbd_dev->dev_id); + + dout("rbd_dev %p given dev id %d\n", rbd_dev, rbd_dev->dev_id); + + return 0; } /* @@ -4395,49 +4414,13 @@ static void rbd_dev_id_get(struct rbd_device *rbd_dev) */ static void rbd_dev_id_put(struct rbd_device *rbd_dev) { - struct list_head *tmp; - int rbd_id = rbd_dev->dev_id; - int max_id; - - rbd_assert(rbd_id > 0); - - dout("rbd_dev %p released dev id %llu\n", rbd_dev, - (unsigned long long) rbd_dev->dev_id); spin_lock(&rbd_dev_list_lock); list_del_init(&rbd_dev->node); - - /* - * If the id being "put" is not the current maximum, there - * is nothing special we need to do. - */ - if (rbd_id != atomic64_read(&rbd_dev_id_max)) { - spin_unlock(&rbd_dev_list_lock); - return; - } - - /* - * We need to update the current maximum id. Search the - * list to find out what it is. We're more likely to find - * the maximum at the end, so search the list backward. - */ - max_id = 0; - list_for_each_prev(tmp, &rbd_dev_list) { - struct rbd_device *rbd_dev; - - rbd_dev = list_entry(tmp, struct rbd_device, node); - if (rbd_dev->dev_id > max_id) - max_id = rbd_dev->dev_id; - } spin_unlock(&rbd_dev_list_lock); - /* - * The max id could have been updated by rbd_dev_id_get(), in - * which case it now accurately reflects the new maximum. - * Be careful not to overwrite the maximum value in that - * case. - */ - atomic64_cmpxchg(&rbd_dev_id_max, rbd_id, max_id); - dout(" max dev id has been reset\n"); + ida_simple_remove(&rbd_dev_id_ida, rbd_dev->dev_id); + + dout("rbd_dev %p released dev id %d\n", rbd_dev, rbd_dev->dev_id); } /* @@ -4860,20 +4843,29 @@ static int rbd_dev_device_setup(struct rbd_device *rbd_dev) { int ret; - /* generate unique id: find highest unique id, add one */ - rbd_dev_id_get(rbd_dev); + /* Get an id and fill in device name. */ + + ret = rbd_dev_id_get(rbd_dev); + if (ret) + return ret; - /* Fill in the device name, now that we have its id. */ BUILD_BUG_ON(DEV_NAME_LEN < sizeof (RBD_DRV_NAME) + MAX_INT_FORMAT_WIDTH); sprintf(rbd_dev->name, "%s%d", RBD_DRV_NAME, rbd_dev->dev_id); - /* Get our block major device number. */ + /* Record our major and minor device numbers. */ - ret = register_blkdev(0, rbd_dev->name); - if (ret < 0) - goto err_out_id; - rbd_dev->major = ret; + if (!single_major) { + ret = register_blkdev(0, rbd_dev->name); + if (ret < 0) + goto err_out_id; + + rbd_dev->major = ret; + rbd_dev->minor = 0; + } else { + rbd_dev->major = rbd_major; + rbd_dev->minor = rbd_dev_id_to_minor(rbd_dev->dev_id); + } /* Set up the blkdev mapping. */ @@ -4905,7 +4897,8 @@ err_out_mapping: err_out_disk: rbd_free_disk(rbd_dev); err_out_blkdev: - unregister_blkdev(rbd_dev->major, rbd_dev->name); + if (!single_major) + unregister_blkdev(rbd_dev->major, rbd_dev->name); err_out_id: rbd_dev_id_put(rbd_dev); rbd_dev_mapping_clear(rbd_dev); @@ -4961,7 +4954,6 @@ static void rbd_dev_image_release(struct rbd_device *rbd_dev) static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping) { int ret; - int tmp; /* * Get the id from the image id object. Unless there's an @@ -4980,7 +4972,7 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping) goto err_out_format; if (mapping) { - ret = rbd_dev_header_watch_sync(rbd_dev, true); + ret = rbd_dev_header_watch_sync(rbd_dev); if (ret) goto out_header_name; } @@ -5007,12 +4999,8 @@ static int rbd_dev_image_probe(struct rbd_device *rbd_dev, bool mapping) err_out_probe: rbd_dev_unprobe(rbd_dev); err_out_watch: - if (mapping) { - tmp = rbd_dev_header_watch_sync(rbd_dev, false); - if (tmp) - rbd_warn(rbd_dev, "unable to tear down " - "watch request (%d)\n", tmp); - } + if (mapping) + rbd_dev_header_unwatch_sync(rbd_dev); out_header_name: kfree(rbd_dev->header_name); rbd_dev->header_name = NULL; @@ -5026,9 +5014,9 @@ err_out_format: return ret; } -static ssize_t rbd_add(struct bus_type *bus, - const char *buf, - size_t count) +static ssize_t do_rbd_add(struct bus_type *bus, + const char *buf, + size_t count) { struct rbd_device *rbd_dev = NULL; struct ceph_options *ceph_opts = NULL; @@ -5090,6 +5078,12 @@ static ssize_t rbd_add(struct bus_type *bus, rc = rbd_dev_device_setup(rbd_dev); if (rc) { + /* + * rbd_dev_header_unwatch_sync() can't be moved into + * rbd_dev_image_release() without refactoring, see + * commit 1f3ef78861ac. + */ + rbd_dev_header_unwatch_sync(rbd_dev); rbd_dev_image_release(rbd_dev); goto err_out_module; } @@ -5110,6 +5104,23 @@ err_out_module: return (ssize_t)rc; } +static ssize_t rbd_add(struct bus_type *bus, + const char *buf, + size_t count) +{ + if (single_major) + return -EINVAL; + + return do_rbd_add(bus, buf, count); +} + +static ssize_t rbd_add_single_major(struct bus_type *bus, + const char *buf, + size_t count) +{ + return do_rbd_add(bus, buf, count); +} + static void rbd_dev_device_release(struct device *dev) { struct rbd_device *rbd_dev = dev_to_rbd_dev(dev); @@ -5117,8 +5128,8 @@ static void rbd_dev_device_release(struct device *dev) rbd_free_disk(rbd_dev); clear_bit(RBD_DEV_FLAG_EXISTS, &rbd_dev->flags); rbd_dev_mapping_clear(rbd_dev); - unregister_blkdev(rbd_dev->major, rbd_dev->name); - rbd_dev->major = 0; + if (!single_major) + unregister_blkdev(rbd_dev->major, rbd_dev->name); rbd_dev_id_put(rbd_dev); rbd_dev_mapping_clear(rbd_dev); } @@ -5149,9 +5160,9 @@ static void rbd_dev_remove_parent(struct rbd_device *rbd_dev) } } -static ssize_t rbd_remove(struct bus_type *bus, - const char *buf, - size_t count) +static ssize_t do_rbd_remove(struct bus_type *bus, + const char *buf, + size_t count) { struct rbd_device *rbd_dev = NULL; struct list_head *tmp; @@ -5191,16 +5202,14 @@ static ssize_t rbd_remove(struct bus_type *bus, if (ret < 0 || already) return ret; - ret = rbd_dev_header_watch_sync(rbd_dev, false); - if (ret) - rbd_warn(rbd_dev, "failed to cancel watch event (%d)\n", ret); - + rbd_dev_header_unwatch_sync(rbd_dev); /* * flush remaining watch callbacks - these must be complete * before the osd_client is shutdown */ dout("%s: flushing notifies", __func__); ceph_osdc_flush_notifies(&rbd_dev->rbd_client->client->osdc); + /* * Don't free anything from rbd_dev->disk until after all * notifies are completely processed. Otherwise @@ -5214,6 +5223,23 @@ static ssize_t rbd_remove(struct bus_type *bus, return count; } +static ssize_t rbd_remove(struct bus_type *bus, + const char *buf, + size_t count) +{ + if (single_major) + return -EINVAL; + + return do_rbd_remove(bus, buf, count); +} + +static ssize_t rbd_remove_single_major(struct bus_type *bus, + const char *buf, + size_t count) +{ + return do_rbd_remove(bus, buf, count); +} + /* * create control files in sysfs * /sys/bus/rbd/... @@ -5259,7 +5285,7 @@ static int rbd_slab_init(void) rbd_assert(!rbd_segment_name_cache); rbd_segment_name_cache = kmem_cache_create("rbd_segment_name", - MAX_OBJ_NAME_SIZE + 1, 1, 0, NULL); + CEPH_MAX_OID_NAME_LEN + 1, 1, 0, NULL); if (rbd_segment_name_cache) return 0; out_err: @@ -5295,24 +5321,45 @@ static int __init rbd_init(void) if (!libceph_compatible(NULL)) { rbd_warn(NULL, "libceph incompatibility (quitting)"); - return -EINVAL; } + rc = rbd_slab_init(); if (rc) return rc; + + if (single_major) { + rbd_major = register_blkdev(0, RBD_DRV_NAME); + if (rbd_major < 0) { + rc = rbd_major; + goto err_out_slab; + } + } + rc = rbd_sysfs_init(); if (rc) - rbd_slab_exit(); + goto err_out_blkdev; + + if (single_major) + pr_info("loaded (major %d)\n", rbd_major); else - pr_info("loaded " RBD_DRV_NAME_LONG "\n"); + pr_info("loaded\n"); + + return 0; +err_out_blkdev: + if (single_major) + unregister_blkdev(rbd_major, RBD_DRV_NAME); +err_out_slab: + rbd_slab_exit(); return rc; } static void __exit rbd_exit(void) { rbd_sysfs_cleanup(); + if (single_major) + unregister_blkdev(rbd_major, RBD_DRV_NAME); rbd_slab_exit(); } @@ -5322,9 +5369,8 @@ module_exit(rbd_exit); MODULE_AUTHOR("Alex Elder <elder@inktank.com>"); MODULE_AUTHOR("Sage Weil <sage@newdream.net>"); MODULE_AUTHOR("Yehuda Sadeh <yehuda@hq.newdream.net>"); -MODULE_DESCRIPTION("rados block device"); - /* following authorship retained from original osdblk.c */ MODULE_AUTHOR("Jeff Garzik <jeff@garzik.org>"); +MODULE_DESCRIPTION("RADOS Block Device (RBD) driver"); MODULE_LICENSE("GPL"); diff --git a/drivers/block/rsxx/dev.c b/drivers/block/rsxx/dev.c index 2284f5d3a54a..2839d37e5af7 100644 --- a/drivers/block/rsxx/dev.c +++ b/drivers/block/rsxx/dev.c @@ -174,7 +174,7 @@ static void rsxx_make_request(struct request_queue *q, struct bio *bio) if (!card) goto req_err; - if (bio->bi_sector + (bio->bi_size >> 9) > get_capacity(card->gendisk)) + if (bio_end_sector(bio) > get_capacity(card->gendisk)) goto req_err; if (unlikely(card->halt)) { @@ -187,7 +187,7 @@ static void rsxx_make_request(struct request_queue *q, struct bio *bio) goto req_err; } - if (bio->bi_size == 0) { + if (bio->bi_iter.bi_size == 0) { dev_err(CARD_TO_DEV(card), "size zero BIO!\n"); goto req_err; } @@ -208,7 +208,7 @@ static void rsxx_make_request(struct request_queue *q, struct bio *bio) dev_dbg(CARD_TO_DEV(card), "BIO[%c]: meta: %p addr8: x%llx size: %d\n", bio_data_dir(bio) ? 'W' : 'R', bio_meta, - (u64)bio->bi_sector << 9, bio->bi_size); + (u64)bio->bi_iter.bi_sector << 9, bio->bi_iter.bi_size); st = rsxx_dma_queue_bio(card, bio, &bio_meta->pending_dmas, bio_dma_done_cb, bio_meta); diff --git a/drivers/block/rsxx/dma.c b/drivers/block/rsxx/dma.c index fc88ba3e1bd2..cf8cd293abb5 100644 --- a/drivers/block/rsxx/dma.c +++ b/drivers/block/rsxx/dma.c @@ -684,7 +684,8 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card, void *cb_data) { struct list_head dma_list[RSXX_MAX_TARGETS]; - struct bio_vec *bvec; + struct bio_vec bvec; + struct bvec_iter iter; unsigned long long addr8; unsigned int laddr; unsigned int bv_len; @@ -696,7 +697,7 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card, int st; int i; - addr8 = bio->bi_sector << 9; /* sectors are 512 bytes */ + addr8 = bio->bi_iter.bi_sector << 9; /* sectors are 512 bytes */ atomic_set(n_dmas, 0); for (i = 0; i < card->n_targets; i++) { @@ -705,7 +706,7 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card, } if (bio->bi_rw & REQ_DISCARD) { - bv_len = bio->bi_size; + bv_len = bio->bi_iter.bi_size; while (bv_len > 0) { tgt = rsxx_get_dma_tgt(card, addr8); @@ -722,9 +723,9 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card, bv_len -= RSXX_HW_BLK_SIZE; } } else { - bio_for_each_segment(bvec, bio, i) { - bv_len = bvec->bv_len; - bv_off = bvec->bv_offset; + bio_for_each_segment(bvec, bio, iter) { + bv_len = bvec.bv_len; + bv_off = bvec.bv_offset; while (bv_len > 0) { tgt = rsxx_get_dma_tgt(card, addr8); @@ -736,7 +737,7 @@ int rsxx_dma_queue_bio(struct rsxx_cardinfo *card, st = rsxx_queue_dma(card, &dma_list[tgt], bio_data_dir(bio), dma_off, dma_len, - laddr, bvec->bv_page, + laddr, bvec.bv_page, bv_off, cb, cb_data); if (st) goto bvec_err; diff --git a/drivers/block/skd_main.c b/drivers/block/skd_main.c index 9199c93be926..eb6e1e0e8db2 100644 --- a/drivers/block/skd_main.c +++ b/drivers/block/skd_main.c @@ -5269,7 +5269,7 @@ const char *skd_skdev_state_to_str(enum skd_drvr_state state) } } -const char *skd_skmsg_state_to_str(enum skd_fit_msg_state state) +static const char *skd_skmsg_state_to_str(enum skd_fit_msg_state state) { switch (state) { case SKD_MSG_STATE_IDLE: @@ -5281,7 +5281,7 @@ const char *skd_skmsg_state_to_str(enum skd_fit_msg_state state) } } -const char *skd_skreq_state_to_str(enum skd_req_state state) +static const char *skd_skreq_state_to_str(enum skd_req_state state) { switch (state) { case SKD_REQ_STATE_IDLE: diff --git a/drivers/block/sx8.c b/drivers/block/sx8.c index 3fb6ab4c8b4e..d5e2d12b9d9e 100644 --- a/drivers/block/sx8.c +++ b/drivers/block/sx8.c @@ -1744,20 +1744,6 @@ static void carm_remove_one (struct pci_dev *pdev) kfree(host); pci_release_regions(pdev); pci_disable_device(pdev); - pci_set_drvdata(pdev, NULL); } -static int __init carm_init(void) -{ - return pci_register_driver(&carm_driver); -} - -static void __exit carm_exit(void) -{ - pci_unregister_driver(&carm_driver); -} - -module_init(carm_init); -module_exit(carm_exit); - - +module_pci_driver(carm_driver); diff --git a/drivers/block/umem.c b/drivers/block/umem.c index ad70868f8a96..4cf81b5bf0f7 100644 --- a/drivers/block/umem.c +++ b/drivers/block/umem.c @@ -108,8 +108,7 @@ struct cardinfo { * have been written */ struct bio *bio, *currentbio, **biotail; - int current_idx; - sector_t current_sector; + struct bvec_iter current_iter; struct request_queue *queue; @@ -118,7 +117,7 @@ struct cardinfo { struct mm_dma_desc *desc; int cnt, headcnt; struct bio *bio, **biotail; - int idx; + struct bvec_iter iter; } mm_pages[2]; #define DESC_PER_PAGE ((PAGE_SIZE*2)/sizeof(struct mm_dma_desc)) @@ -344,16 +343,13 @@ static int add_bio(struct cardinfo *card) dma_addr_t dma_handle; int offset; struct bio *bio; - struct bio_vec *vec; - int idx; + struct bio_vec vec; int rw; - int len; bio = card->currentbio; if (!bio && card->bio) { card->currentbio = card->bio; - card->current_idx = card->bio->bi_idx; - card->current_sector = card->bio->bi_sector; + card->current_iter = card->bio->bi_iter; card->bio = card->bio->bi_next; if (card->bio == NULL) card->biotail = &card->bio; @@ -362,18 +358,17 @@ static int add_bio(struct cardinfo *card) } if (!bio) return 0; - idx = card->current_idx; rw = bio_rw(bio); if (card->mm_pages[card->Ready].cnt >= DESC_PER_PAGE) return 0; - vec = bio_iovec_idx(bio, idx); - len = vec->bv_len; + vec = bio_iter_iovec(bio, card->current_iter); + dma_handle = pci_map_page(card->dev, - vec->bv_page, - vec->bv_offset, - len, + vec.bv_page, + vec.bv_offset, + vec.bv_len, (rw == READ) ? PCI_DMA_FROMDEVICE : PCI_DMA_TODEVICE); @@ -381,7 +376,7 @@ static int add_bio(struct cardinfo *card) desc = &p->desc[p->cnt]; p->cnt++; if (p->bio == NULL) - p->idx = idx; + p->iter = card->current_iter; if ((p->biotail) != &bio->bi_next) { *(p->biotail) = bio; p->biotail = &(bio->bi_next); @@ -391,8 +386,8 @@ static int add_bio(struct cardinfo *card) desc->data_dma_handle = dma_handle; desc->pci_addr = cpu_to_le64((u64)desc->data_dma_handle); - desc->local_addr = cpu_to_le64(card->current_sector << 9); - desc->transfer_size = cpu_to_le32(len); + desc->local_addr = cpu_to_le64(card->current_iter.bi_sector << 9); + desc->transfer_size = cpu_to_le32(vec.bv_len); offset = (((char *)&desc->sem_control_bits) - ((char *)p->desc)); desc->sem_addr = cpu_to_le64((u64)(p->page_dma+offset)); desc->zero1 = desc->zero2 = 0; @@ -407,10 +402,9 @@ static int add_bio(struct cardinfo *card) desc->control_bits |= cpu_to_le32(DMASCR_TRANSFER_READ); desc->sem_control_bits = desc->control_bits; - card->current_sector += (len >> 9); - idx++; - card->current_idx = idx; - if (idx >= bio->bi_vcnt) + + bio_advance_iter(bio, &card->current_iter, vec.bv_len); + if (!card->current_iter.bi_size) card->currentbio = NULL; return 1; @@ -439,23 +433,25 @@ static void process_page(unsigned long data) struct mm_dma_desc *desc = &page->desc[page->headcnt]; int control = le32_to_cpu(desc->sem_control_bits); int last = 0; - int idx; + struct bio_vec vec; if (!(control & DMASCR_DMA_COMPLETE)) { control = dma_status; last = 1; } + page->headcnt++; - idx = page->idx; - page->idx++; - if (page->idx >= bio->bi_vcnt) { + vec = bio_iter_iovec(bio, page->iter); + bio_advance_iter(bio, &page->iter, vec.bv_len); + + if (!page->iter.bi_size) { page->bio = bio->bi_next; if (page->bio) - page->idx = page->bio->bi_idx; + page->iter = page->bio->bi_iter; } pci_unmap_page(card->dev, desc->data_dma_handle, - bio_iovec_idx(bio, idx)->bv_len, + vec.bv_len, (control & DMASCR_TRANSFER_READ) ? PCI_DMA_TODEVICE : PCI_DMA_FROMDEVICE); if (control & DMASCR_HARD_ERROR) { @@ -532,7 +528,8 @@ static void mm_make_request(struct request_queue *q, struct bio *bio) { struct cardinfo *card = q->queuedata; pr_debug("mm_make_request %llu %u\n", - (unsigned long long)bio->bi_sector, bio->bi_size); + (unsigned long long)bio->bi_iter.bi_sector, + bio->bi_iter.bi_size); spin_lock_irq(&card->lock); *card->biotail = bio; diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 6a680d4de7f1..b1cb3f4c4db4 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -110,9 +110,9 @@ static int __virtblk_add_req(struct virtqueue *vq, return virtqueue_add_sgs(vq, sgs, num_out, num_in, vbr, GFP_ATOMIC); } -static inline void virtblk_request_done(struct virtblk_req *vbr) +static inline void virtblk_request_done(struct request *req) { - struct request *req = vbr->req; + struct virtblk_req *vbr = req->special; int error = virtblk_result(vbr); if (req->cmd_type == REQ_TYPE_BLOCK_PC) { @@ -138,7 +138,7 @@ static void virtblk_done(struct virtqueue *vq) do { virtqueue_disable_cb(vq); while ((vbr = virtqueue_get_buf(vblk->vq, &len)) != NULL) { - virtblk_request_done(vbr); + blk_mq_complete_request(vbr->req); req_done = true; } if (unlikely(virtqueue_is_broken(vq))) @@ -479,6 +479,7 @@ static struct blk_mq_ops virtio_mq_ops = { .map_queue = blk_mq_map_queue, .alloc_hctx = blk_mq_alloc_single_hw_queue, .free_hctx = blk_mq_free_single_hw_queue, + .complete = virtblk_request_done, }; static struct blk_mq_reg virtio_mq_reg = { diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c index 6620b73d0490..64c60edcdfbc 100644 --- a/drivers/block/xen-blkback/blkback.c +++ b/drivers/block/xen-blkback/blkback.c @@ -299,7 +299,7 @@ static void free_persistent_gnts(struct xen_blkif *blkif, struct rb_root *root, BUG_ON(num != 0); } -static void unmap_purged_grants(struct work_struct *work) +void xen_blkbk_unmap_purged_grants(struct work_struct *work) { struct gnttab_unmap_grant_ref unmap[BLKIF_MAX_SEGMENTS_PER_REQUEST]; struct page *pages[BLKIF_MAX_SEGMENTS_PER_REQUEST]; @@ -375,7 +375,7 @@ static void purge_persistent_gnt(struct xen_blkif *blkif) pr_debug(DRV_PFX "Going to purge %u persistent grants\n", num_clean); - INIT_LIST_HEAD(&blkif->persistent_purge_list); + BUG_ON(!list_empty(&blkif->persistent_purge_list)); root = &blkif->persistent_gnts; purge_list: foreach_grant_safe(persistent_gnt, n, root, node) { @@ -420,7 +420,6 @@ finished: blkif->vbd.overflow_max_grants = 0; /* We can defer this work */ - INIT_WORK(&blkif->persistent_purge_work, unmap_purged_grants); schedule_work(&blkif->persistent_purge_work); pr_debug(DRV_PFX "Purged %u/%u\n", (total - num_clean), total); return; @@ -625,9 +624,23 @@ purge_gnt_list: print_stats(blkif); } - /* Since we are shutting down remove all pages from the buffer */ - shrink_free_pagepool(blkif, 0 /* All */); + /* Drain pending purge work */ + flush_work(&blkif->persistent_purge_work); + if (log_stats) + print_stats(blkif); + + blkif->xenblkd = NULL; + xen_blkif_put(blkif); + + return 0; +} + +/* + * Remove persistent grants and empty the pool of free pages + */ +void xen_blkbk_free_caches(struct xen_blkif *blkif) +{ /* Free all persistent grant pages */ if (!RB_EMPTY_ROOT(&blkif->persistent_gnts)) free_persistent_gnts(blkif, &blkif->persistent_gnts, @@ -636,13 +649,8 @@ purge_gnt_list: BUG_ON(!RB_EMPTY_ROOT(&blkif->persistent_gnts)); blkif->persistent_gnt_c = 0; - if (log_stats) - print_stats(blkif); - - blkif->xenblkd = NULL; - xen_blkif_put(blkif); - - return 0; + /* Since we are shutting down remove all pages from the buffer */ + shrink_free_pagepool(blkif, 0 /* All */); } /* @@ -838,7 +846,7 @@ static int xen_blkbk_parse_indirect(struct blkif_request *req, struct grant_page **pages = pending_req->indirect_pages; struct xen_blkif *blkif = pending_req->blkif; int indirect_grefs, rc, n, nseg, i; - struct blkif_request_segment_aligned *segments = NULL; + struct blkif_request_segment *segments = NULL; nseg = pending_req->nr_pages; indirect_grefs = INDIRECT_PAGES(nseg); @@ -934,9 +942,7 @@ static void xen_blk_drain_io(struct xen_blkif *blkif) { atomic_set(&blkif->drain, 1); do { - /* The initial value is one, and one refcnt taken at the - * start of the xen_blkif_schedule thread. */ - if (atomic_read(&blkif->refcnt) <= 2) + if (atomic_read(&blkif->inflight) == 0) break; wait_for_completion_interruptible_timeout( &blkif->drain_complete, HZ); @@ -976,17 +982,30 @@ static void __end_block_io_op(struct pending_req *pending_req, int error) * the proper response on the ring. */ if (atomic_dec_and_test(&pending_req->pendcnt)) { - xen_blkbk_unmap(pending_req->blkif, + struct xen_blkif *blkif = pending_req->blkif; + + xen_blkbk_unmap(blkif, pending_req->segments, pending_req->nr_pages); - make_response(pending_req->blkif, pending_req->id, + make_response(blkif, pending_req->id, pending_req->operation, pending_req->status); - xen_blkif_put(pending_req->blkif); - if (atomic_read(&pending_req->blkif->refcnt) <= 2) { - if (atomic_read(&pending_req->blkif->drain)) - complete(&pending_req->blkif->drain_complete); + free_req(blkif, pending_req); + /* + * Make sure the request is freed before releasing blkif, + * or there could be a race between free_req and the + * cleanup done in xen_blkif_free during shutdown. + * + * NB: The fact that we might try to wake up pending_free_wq + * before drain_complete (in case there's a drain going on) + * it's not a problem with our current implementation + * because we can assure there's no thread waiting on + * pending_free_wq if there's a drain going on, but it has + * to be taken into account if the current model is changed. + */ + if (atomic_dec_and_test(&blkif->inflight) && atomic_read(&blkif->drain)) { + complete(&blkif->drain_complete); } - free_req(pending_req->blkif, pending_req); + xen_blkif_put(blkif); } } @@ -1240,6 +1259,7 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif, * below (in "!bio") if we are handling a BLKIF_OP_DISCARD. */ xen_blkif_get(blkif); + atomic_inc(&blkif->inflight); for (i = 0; i < nseg; i++) { while ((bio == NULL) || @@ -1257,7 +1277,7 @@ static int dispatch_rw_block_io(struct xen_blkif *blkif, bio->bi_bdev = preq.bdev; bio->bi_private = pending_req; bio->bi_end_io = end_block_io_op; - bio->bi_sector = preq.sector_number; + bio->bi_iter.bi_sector = preq.sector_number; } preq.sector_number += seg[i].nsec; diff --git a/drivers/block/xen-blkback/common.h b/drivers/block/xen-blkback/common.h index 8d8807563d99..be052773ad03 100644 --- a/drivers/block/xen-blkback/common.h +++ b/drivers/block/xen-blkback/common.h @@ -57,7 +57,7 @@ #define MAX_INDIRECT_SEGMENTS 256 #define SEGS_PER_INDIRECT_FRAME \ - (PAGE_SIZE/sizeof(struct blkif_request_segment_aligned)) + (PAGE_SIZE/sizeof(struct blkif_request_segment)) #define MAX_INDIRECT_PAGES \ ((MAX_INDIRECT_SEGMENTS + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME) #define INDIRECT_PAGES(_segs) \ @@ -278,6 +278,7 @@ struct xen_blkif { /* for barrier (drain) requests */ struct completion drain_complete; atomic_t drain; + atomic_t inflight; /* One thread per one blkif. */ struct task_struct *xenblkd; unsigned int waiting_reqs; @@ -376,6 +377,7 @@ int xen_blkif_xenbus_init(void); irqreturn_t xen_blkif_be_int(int irq, void *dev_id); int xen_blkif_schedule(void *arg); int xen_blkif_purge_persistent(void *arg); +void xen_blkbk_free_caches(struct xen_blkif *blkif); int xen_blkbk_flush_diskcache(struct xenbus_transaction xbt, struct backend_info *be, int state); @@ -383,6 +385,7 @@ int xen_blkbk_flush_diskcache(struct xenbus_transaction xbt, int xen_blkbk_barrier(struct xenbus_transaction xbt, struct backend_info *be, int state); struct xenbus_device *xen_blkbk_xenbus(struct backend_info *be); +void xen_blkbk_unmap_purged_grants(struct work_struct *work); static inline void blkif_get_x86_32_req(struct blkif_request *dst, struct blkif_x86_32_request *src) diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c index c2014a0aa206..9a547e6b6ebf 100644 --- a/drivers/block/xen-blkback/xenbus.c +++ b/drivers/block/xen-blkback/xenbus.c @@ -125,8 +125,11 @@ static struct xen_blkif *xen_blkif_alloc(domid_t domid) blkif->persistent_gnts.rb_node = NULL; spin_lock_init(&blkif->free_pages_lock); INIT_LIST_HEAD(&blkif->free_pages); + INIT_LIST_HEAD(&blkif->persistent_purge_list); blkif->free_pages_num = 0; atomic_set(&blkif->persistent_gnt_in_use, 0); + atomic_set(&blkif->inflight, 0); + INIT_WORK(&blkif->persistent_purge_work, xen_blkbk_unmap_purged_grants); INIT_LIST_HEAD(&blkif->pending_free); @@ -259,6 +262,17 @@ static void xen_blkif_free(struct xen_blkif *blkif) if (!atomic_dec_and_test(&blkif->refcnt)) BUG(); + /* Remove all persistent grants and the cache of ballooned pages. */ + xen_blkbk_free_caches(blkif); + + /* Make sure everything is drained before shutting down */ + BUG_ON(blkif->persistent_gnt_c != 0); + BUG_ON(atomic_read(&blkif->persistent_gnt_in_use) != 0); + BUG_ON(blkif->free_pages_num != 0); + BUG_ON(!list_empty(&blkif->persistent_purge_list)); + BUG_ON(!list_empty(&blkif->free_pages)); + BUG_ON(!RB_EMPTY_ROOT(&blkif->persistent_gnts)); + /* Check that there is no request in use */ list_for_each_entry_safe(req, n, &blkif->pending_free, free_list) { list_del(&req->free_list); diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c index c4a4c9006288..efe1b4761735 100644 --- a/drivers/block/xen-blkfront.c +++ b/drivers/block/xen-blkfront.c @@ -162,7 +162,7 @@ static DEFINE_SPINLOCK(minor_lock); #define DEV_NAME "xvd" /* name in /dev */ #define SEGS_PER_INDIRECT_FRAME \ - (PAGE_SIZE/sizeof(struct blkif_request_segment_aligned)) + (PAGE_SIZE/sizeof(struct blkif_request_segment)) #define INDIRECT_GREFS(_segs) \ ((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME) @@ -393,7 +393,7 @@ static int blkif_queue_request(struct request *req) unsigned long id; unsigned int fsect, lsect; int i, ref, n; - struct blkif_request_segment_aligned *segments = NULL; + struct blkif_request_segment *segments = NULL; /* * Used to store if we are able to queue the request by just using @@ -550,7 +550,7 @@ static int blkif_queue_request(struct request *req) } else { n = i % SEGS_PER_INDIRECT_FRAME; segments[n] = - (struct blkif_request_segment_aligned) { + (struct blkif_request_segment) { .gref = ref, .first_sect = fsect, .last_sect = lsect }; @@ -1356,7 +1356,7 @@ static int blkfront_probe(struct xenbus_device *dev, char *type; int len; /* no unplug has been done: do not hook devices != xen vbds */ - if (xen_platform_pci_unplug & XEN_UNPLUG_UNNECESSARY) { + if (xen_has_pv_and_legacy_disk_devices()) { int major; if (!VDEV_IS_EXTENDED(vdevice)) @@ -1547,7 +1547,7 @@ static int blkif_recover(struct blkfront_info *info) for (i = 0; i < pending; i++) { offset = (i * segs * PAGE_SIZE) >> 9; size = min((unsigned int)(segs * PAGE_SIZE) >> 9, - (unsigned int)(bio->bi_size >> 9) - offset); + (unsigned int)bio_sectors(bio) - offset); cloned_bio = bio_clone(bio, GFP_NOIO); BUG_ON(cloned_bio == NULL); bio_trim(cloned_bio, offset, size); @@ -1904,13 +1904,16 @@ static void blkback_changed(struct xenbus_device *dev, case XenbusStateReconfiguring: case XenbusStateReconfigured: case XenbusStateUnknown: - case XenbusStateClosed: break; case XenbusStateConnected: blkfront_connect(info); break; + case XenbusStateClosed: + if (dev->state == XenbusStateClosed) + break; + /* Missed the backend's Closing state -- fallthrough */ case XenbusStateClosing: blkfront_closing(info); break; @@ -2079,7 +2082,7 @@ static int __init xlblk_init(void) if (!xen_domain()) return -ENODEV; - if (xen_hvm_domain() && !xen_platform_pci_unplug) + if (!xen_has_pv_disk_devices()) return -ENODEV; if (register_blkdev(XENVBD_MAJOR, DEV_NAME)) { diff --git a/drivers/block/z2ram.c b/drivers/block/z2ram.c index 5a95baf4b104..27de5046708a 100644 --- a/drivers/block/z2ram.c +++ b/drivers/block/z2ram.c @@ -43,9 +43,6 @@ #include <linux/zorro.h> -extern int m68k_realnum_memory; -extern struct mem_info m68k_memory[NUM_MEMINFO]; - #define Z2MINOR_COMBINED (0) #define Z2MINOR_Z2ONLY (1) #define Z2MINOR_CHIPONLY (2) @@ -116,8 +113,8 @@ get_z2ram( void ) if ( test_bit( i, zorro_unused_z2ram ) ) { z2_count++; - z2ram_map[ z2ram_size++ ] = - ZTWO_VADDR( Z2RAM_START ) + ( i << Z2RAM_CHUNKSHIFT ); + z2ram_map[z2ram_size++] = (unsigned long)ZTWO_VADDR(Z2RAM_START) + + (i << Z2RAM_CHUNKSHIFT); clear_bit( i, zorro_unused_z2ram ); } } diff --git a/drivers/block/zram/Kconfig b/drivers/block/zram/Kconfig new file mode 100644 index 000000000000..3450be850399 --- /dev/null +++ b/drivers/block/zram/Kconfig @@ -0,0 +1,24 @@ +config ZRAM + tristate "Compressed RAM block device support" + depends on BLOCK && SYSFS && ZSMALLOC + select LZO_COMPRESS + select LZO_DECOMPRESS + default n + help + Creates virtual block devices called /dev/zramX (X = 0, 1, ...). + Pages written to these disks are compressed and stored in memory + itself. These disks allow very fast I/O and compression provides + good amounts of memory savings. + + It has several use cases, for example: /tmp storage, use as swap + disks and maybe many more. + + See zram.txt for more information. + +config ZRAM_DEBUG + bool "Compressed RAM block device debug support" + depends on ZRAM + default n + help + This option adds additional debugging code to the compressed + RAM block device driver. diff --git a/drivers/block/zram/Makefile b/drivers/block/zram/Makefile new file mode 100644 index 000000000000..cb0f9ced6a93 --- /dev/null +++ b/drivers/block/zram/Makefile @@ -0,0 +1,3 @@ +zram-y := zram_drv.o + +obj-$(CONFIG_ZRAM) += zram.o diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c new file mode 100644 index 000000000000..011e55d820b1 --- /dev/null +++ b/drivers/block/zram/zram_drv.c @@ -0,0 +1,958 @@ +/* + * Compressed RAM block device + * + * Copyright (C) 2008, 2009, 2010 Nitin Gupta + * 2012, 2013 Minchan Kim + * + * This code is released using a dual license strategy: BSD/GPL + * You can choose the licence that better fits your requirements. + * + * Released under the terms of 3-clause BSD License + * Released under the terms of GNU General Public License Version 2.0 + * + */ + +#define KMSG_COMPONENT "zram" +#define pr_fmt(fmt) KMSG_COMPONENT ": " fmt + +#ifdef CONFIG_ZRAM_DEBUG +#define DEBUG +#endif + +#include <linux/module.h> +#include <linux/kernel.h> +#include <linux/bio.h> +#include <linux/bitops.h> +#include <linux/blkdev.h> +#include <linux/buffer_head.h> +#include <linux/device.h> +#include <linux/genhd.h> +#include <linux/highmem.h> +#include <linux/slab.h> +#include <linux/lzo.h> +#include <linux/string.h> +#include <linux/vmalloc.h> + +#include "zram_drv.h" + +/* Globals */ +static int zram_major; +static struct zram *zram_devices; + +/* Module params (documentation at end) */ +static unsigned int num_devices = 1; + +static inline struct zram *dev_to_zram(struct device *dev) +{ + return (struct zram *)dev_to_disk(dev)->private_data; +} + +static ssize_t disksize_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%llu\n", zram->disksize); +} + +static ssize_t initstate_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%u\n", zram->init_done); +} + +static ssize_t num_reads_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%llu\n", + (u64)atomic64_read(&zram->stats.num_reads)); +} + +static ssize_t num_writes_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%llu\n", + (u64)atomic64_read(&zram->stats.num_writes)); +} + +static ssize_t invalid_io_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%llu\n", + (u64)atomic64_read(&zram->stats.invalid_io)); +} + +static ssize_t notify_free_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%llu\n", + (u64)atomic64_read(&zram->stats.notify_free)); +} + +static ssize_t zero_pages_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%u\n", atomic_read(&zram->stats.pages_zero)); +} + +static ssize_t orig_data_size_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%llu\n", + (u64)(atomic_read(&zram->stats.pages_stored)) << PAGE_SHIFT); +} + +static ssize_t compr_data_size_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct zram *zram = dev_to_zram(dev); + + return sprintf(buf, "%llu\n", + (u64)atomic64_read(&zram->stats.compr_size)); +} + +static ssize_t mem_used_total_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + u64 val = 0; + struct zram *zram = dev_to_zram(dev); + struct zram_meta *meta = zram->meta; + + down_read(&zram->init_lock); + if (zram->init_done) + val = zs_get_total_size_bytes(meta->mem_pool); + up_read(&zram->init_lock); + + return sprintf(buf, "%llu\n", val); +} + +/* flag operations needs meta->tb_lock */ +static int zram_test_flag(struct zram_meta *meta, u32 index, + enum zram_pageflags flag) +{ + return meta->table[index].flags & BIT(flag); +} + +static void zram_set_flag(struct zram_meta *meta, u32 index, + enum zram_pageflags flag) +{ + meta->table[index].flags |= BIT(flag); +} + +static void zram_clear_flag(struct zram_meta *meta, u32 index, + enum zram_pageflags flag) +{ + meta->table[index].flags &= ~BIT(flag); +} + +static inline int is_partial_io(struct bio_vec *bvec) +{ + return bvec->bv_len != PAGE_SIZE; +} + +/* + * Check if request is within bounds and aligned on zram logical blocks. + */ +static inline int valid_io_request(struct zram *zram, struct bio *bio) +{ + u64 start, end, bound; + + /* unaligned request */ + if (unlikely(bio->bi_iter.bi_sector & + (ZRAM_SECTOR_PER_LOGICAL_BLOCK - 1))) + return 0; + if (unlikely(bio->bi_iter.bi_size & (ZRAM_LOGICAL_BLOCK_SIZE - 1))) + return 0; + + start = bio->bi_iter.bi_sector; + end = start + (bio->bi_iter.bi_size >> SECTOR_SHIFT); + bound = zram->disksize >> SECTOR_SHIFT; + /* out of range range */ + if (unlikely(start >= bound || end > bound || start > end)) + return 0; + + /* I/O request is valid */ + return 1; +} + +static void zram_meta_free(struct zram_meta *meta) +{ + zs_destroy_pool(meta->mem_pool); + kfree(meta->compress_workmem); + free_pages((unsigned long)meta->compress_buffer, 1); + vfree(meta->table); + kfree(meta); +} + +static struct zram_meta *zram_meta_alloc(u64 disksize) +{ + size_t num_pages; + struct zram_meta *meta = kmalloc(sizeof(*meta), GFP_KERNEL); + if (!meta) + goto out; + + meta->compress_workmem = kzalloc(LZO1X_MEM_COMPRESS, GFP_KERNEL); + if (!meta->compress_workmem) + goto free_meta; + + meta->compress_buffer = + (void *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, 1); + if (!meta->compress_buffer) { + pr_err("Error allocating compressor buffer space\n"); + goto free_workmem; + } + + num_pages = disksize >> PAGE_SHIFT; + meta->table = vzalloc(num_pages * sizeof(*meta->table)); + if (!meta->table) { + pr_err("Error allocating zram address table\n"); + goto free_buffer; + } + + meta->mem_pool = zs_create_pool(GFP_NOIO | __GFP_HIGHMEM); + if (!meta->mem_pool) { + pr_err("Error creating memory pool\n"); + goto free_table; + } + + rwlock_init(&meta->tb_lock); + mutex_init(&meta->buffer_lock); + return meta; + +free_table: + vfree(meta->table); +free_buffer: + free_pages((unsigned long)meta->compress_buffer, 1); +free_workmem: + kfree(meta->compress_workmem); +free_meta: + kfree(meta); + meta = NULL; +out: + return meta; +} + +static void update_position(u32 *index, int *offset, struct bio_vec *bvec) +{ + if (*offset + bvec->bv_len >= PAGE_SIZE) + (*index)++; + *offset = (*offset + bvec->bv_len) % PAGE_SIZE; +} + +static int page_zero_filled(void *ptr) +{ + unsigned int pos; + unsigned long *page; + + page = (unsigned long *)ptr; + + for (pos = 0; pos != PAGE_SIZE / sizeof(*page); pos++) { + if (page[pos]) + return 0; + } + + return 1; +} + +static void handle_zero_page(struct bio_vec *bvec) +{ + struct page *page = bvec->bv_page; + void *user_mem; + + user_mem = kmap_atomic(page); + if (is_partial_io(bvec)) + memset(user_mem + bvec->bv_offset, 0, bvec->bv_len); + else + clear_page(user_mem); + kunmap_atomic(user_mem); + + flush_dcache_page(page); +} + +/* NOTE: caller should hold meta->tb_lock with write-side */ +static void zram_free_page(struct zram *zram, size_t index) +{ + struct zram_meta *meta = zram->meta; + unsigned long handle = meta->table[index].handle; + u16 size = meta->table[index].size; + + if (unlikely(!handle)) { + /* + * No memory is allocated for zero filled pages. + * Simply clear zero page flag. + */ + if (zram_test_flag(meta, index, ZRAM_ZERO)) { + zram_clear_flag(meta, index, ZRAM_ZERO); + atomic_dec(&zram->stats.pages_zero); + } + return; + } + + if (unlikely(size > max_zpage_size)) + atomic_dec(&zram->stats.bad_compress); + + zs_free(meta->mem_pool, handle); + + if (size <= PAGE_SIZE / 2) + atomic_dec(&zram->stats.good_compress); + + atomic64_sub(meta->table[index].size, &zram->stats.compr_size); + atomic_dec(&zram->stats.pages_stored); + + meta->table[index].handle = 0; + meta->table[index].size = 0; +} + +static int zram_decompress_page(struct zram *zram, char *mem, u32 index) +{ + int ret = LZO_E_OK; + size_t clen = PAGE_SIZE; + unsigned char *cmem; + struct zram_meta *meta = zram->meta; + unsigned long handle; + u16 size; + + read_lock(&meta->tb_lock); + handle = meta->table[index].handle; + size = meta->table[index].size; + + if (!handle || zram_test_flag(meta, index, ZRAM_ZERO)) { + read_unlock(&meta->tb_lock); + clear_page(mem); + return 0; + } + + cmem = zs_map_object(meta->mem_pool, handle, ZS_MM_RO); + if (size == PAGE_SIZE) + copy_page(mem, cmem); + else + ret = lzo1x_decompress_safe(cmem, size, mem, &clen); + zs_unmap_object(meta->mem_pool, handle); + read_unlock(&meta->tb_lock); + + /* Should NEVER happen. Return bio error if it does. */ + if (unlikely(ret != LZO_E_OK)) { + pr_err("Decompression failed! err=%d, page=%u\n", ret, index); + atomic64_inc(&zram->stats.failed_reads); + return ret; + } + + return 0; +} + +static int zram_bvec_read(struct zram *zram, struct bio_vec *bvec, + u32 index, int offset, struct bio *bio) +{ + int ret; + struct page *page; + unsigned char *user_mem, *uncmem = NULL; + struct zram_meta *meta = zram->meta; + page = bvec->bv_page; + + read_lock(&meta->tb_lock); + if (unlikely(!meta->table[index].handle) || + zram_test_flag(meta, index, ZRAM_ZERO)) { + read_unlock(&meta->tb_lock); + handle_zero_page(bvec); + return 0; + } + read_unlock(&meta->tb_lock); + + if (is_partial_io(bvec)) + /* Use a temporary buffer to decompress the page */ + uncmem = kmalloc(PAGE_SIZE, GFP_NOIO); + + user_mem = kmap_atomic(page); + if (!is_partial_io(bvec)) + uncmem = user_mem; + + if (!uncmem) { + pr_info("Unable to allocate temp memory\n"); + ret = -ENOMEM; + goto out_cleanup; + } + + ret = zram_decompress_page(zram, uncmem, index); + /* Should NEVER happen. Return bio error if it does. */ + if (unlikely(ret != LZO_E_OK)) + goto out_cleanup; + + if (is_partial_io(bvec)) + memcpy(user_mem + bvec->bv_offset, uncmem + offset, + bvec->bv_len); + + flush_dcache_page(page); + ret = 0; +out_cleanup: + kunmap_atomic(user_mem); + if (is_partial_io(bvec)) + kfree(uncmem); + return ret; +} + +static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, u32 index, + int offset) +{ + int ret = 0; + size_t clen; + unsigned long handle; + struct page *page; + unsigned char *user_mem, *cmem, *src, *uncmem = NULL; + struct zram_meta *meta = zram->meta; + bool locked = false; + + page = bvec->bv_page; + src = meta->compress_buffer; + + if (is_partial_io(bvec)) { + /* + * This is a partial IO. We need to read the full page + * before to write the changes. + */ + uncmem = kmalloc(PAGE_SIZE, GFP_NOIO); + if (!uncmem) { + ret = -ENOMEM; + goto out; + } + ret = zram_decompress_page(zram, uncmem, index); + if (ret) + goto out; + } + + mutex_lock(&meta->buffer_lock); + locked = true; + user_mem = kmap_atomic(page); + + if (is_partial_io(bvec)) { + memcpy(uncmem + offset, user_mem + bvec->bv_offset, + bvec->bv_len); + kunmap_atomic(user_mem); + user_mem = NULL; + } else { + uncmem = user_mem; + } + + if (page_zero_filled(uncmem)) { + kunmap_atomic(user_mem); + /* Free memory associated with this sector now. */ + write_lock(&zram->meta->tb_lock); + zram_free_page(zram, index); + zram_set_flag(meta, index, ZRAM_ZERO); + write_unlock(&zram->meta->tb_lock); + + atomic_inc(&zram->stats.pages_zero); + ret = 0; + goto out; + } + + ret = lzo1x_1_compress(uncmem, PAGE_SIZE, src, &clen, + meta->compress_workmem); + if (!is_partial_io(bvec)) { + kunmap_atomic(user_mem); + user_mem = NULL; + uncmem = NULL; + } + + if (unlikely(ret != LZO_E_OK)) { + pr_err("Compression failed! err=%d\n", ret); + goto out; + } + + if (unlikely(clen > max_zpage_size)) { + atomic_inc(&zram->stats.bad_compress); + clen = PAGE_SIZE; + src = NULL; + if (is_partial_io(bvec)) + src = uncmem; + } + + handle = zs_malloc(meta->mem_pool, clen); + if (!handle) { + pr_info("Error allocating memory for compressed page: %u, size=%zu\n", + index, clen); + ret = -ENOMEM; + goto out; + } + cmem = zs_map_object(meta->mem_pool, handle, ZS_MM_WO); + + if ((clen == PAGE_SIZE) && !is_partial_io(bvec)) { + src = kmap_atomic(page); + copy_page(cmem, src); + kunmap_atomic(src); + } else { + memcpy(cmem, src, clen); + } + + zs_unmap_object(meta->mem_pool, handle); + + /* + * Free memory associated with this sector + * before overwriting unused sectors. + */ + write_lock(&zram->meta->tb_lock); + zram_free_page(zram, index); + + meta->table[index].handle = handle; + meta->table[index].size = clen; + write_unlock(&zram->meta->tb_lock); + + /* Update stats */ + atomic64_add(clen, &zram->stats.compr_size); + atomic_inc(&zram->stats.pages_stored); + if (clen <= PAGE_SIZE / 2) + atomic_inc(&zram->stats.good_compress); + +out: + if (locked) + mutex_unlock(&meta->buffer_lock); + if (is_partial_io(bvec)) + kfree(uncmem); + + if (ret) + atomic64_inc(&zram->stats.failed_writes); + return ret; +} + +static int zram_bvec_rw(struct zram *zram, struct bio_vec *bvec, u32 index, + int offset, struct bio *bio, int rw) +{ + int ret; + + if (rw == READ) + ret = zram_bvec_read(zram, bvec, index, offset, bio); + else + ret = zram_bvec_write(zram, bvec, index, offset); + + return ret; +} + +static void zram_reset_device(struct zram *zram, bool reset_capacity) +{ + size_t index; + struct zram_meta *meta; + + down_write(&zram->init_lock); + if (!zram->init_done) { + up_write(&zram->init_lock); + return; + } + + meta = zram->meta; + zram->init_done = 0; + + /* Free all pages that are still in this zram device */ + for (index = 0; index < zram->disksize >> PAGE_SHIFT; index++) { + unsigned long handle = meta->table[index].handle; + if (!handle) + continue; + + zs_free(meta->mem_pool, handle); + } + + zram_meta_free(zram->meta); + zram->meta = NULL; + /* Reset stats */ + memset(&zram->stats, 0, sizeof(zram->stats)); + + zram->disksize = 0; + if (reset_capacity) + set_capacity(zram->disk, 0); + up_write(&zram->init_lock); +} + +static void zram_init_device(struct zram *zram, struct zram_meta *meta) +{ + if (zram->disksize > 2 * (totalram_pages << PAGE_SHIFT)) { + pr_info( + "There is little point creating a zram of greater than " + "twice the size of memory since we expect a 2:1 compression " + "ratio. Note that zram uses about 0.1%% of the size of " + "the disk when not in use so a huge zram is " + "wasteful.\n" + "\tMemory Size: %lu kB\n" + "\tSize you selected: %llu kB\n" + "Continuing anyway ...\n", + (totalram_pages << PAGE_SHIFT) >> 10, zram->disksize >> 10 + ); + } + + /* zram devices sort of resembles non-rotational disks */ + queue_flag_set_unlocked(QUEUE_FLAG_NONROT, zram->disk->queue); + + zram->meta = meta; + zram->init_done = 1; + + pr_debug("Initialization done!\n"); +} + +static ssize_t disksize_store(struct device *dev, + struct device_attribute *attr, const char *buf, size_t len) +{ + u64 disksize; + struct zram_meta *meta; + struct zram *zram = dev_to_zram(dev); + + disksize = memparse(buf, NULL); + if (!disksize) + return -EINVAL; + + disksize = PAGE_ALIGN(disksize); + meta = zram_meta_alloc(disksize); + down_write(&zram->init_lock); + if (zram->init_done) { + up_write(&zram->init_lock); + zram_meta_free(meta); + pr_info("Cannot change disksize for initialized device\n"); + return -EBUSY; + } + + zram->disksize = disksize; + set_capacity(zram->disk, zram->disksize >> SECTOR_SHIFT); + zram_init_device(zram, meta); + up_write(&zram->init_lock); + + return len; +} + +static ssize_t reset_store(struct device *dev, + struct device_attribute *attr, const char *buf, size_t len) +{ + int ret; + unsigned short do_reset; + struct zram *zram; + struct block_device *bdev; + + zram = dev_to_zram(dev); + bdev = bdget_disk(zram->disk, 0); + + if (!bdev) + return -ENOMEM; + + /* Do not reset an active device! */ + if (bdev->bd_holders) { + ret = -EBUSY; + goto out; + } + + ret = kstrtou16(buf, 10, &do_reset); + if (ret) + goto out; + + if (!do_reset) { + ret = -EINVAL; + goto out; + } + + /* Make sure all pending I/O is finished */ + fsync_bdev(bdev); + bdput(bdev); + + zram_reset_device(zram, true); + return len; + +out: + bdput(bdev); + return ret; +} + +static void __zram_make_request(struct zram *zram, struct bio *bio, int rw) +{ + int offset; + u32 index; + struct bio_vec bvec; + struct bvec_iter iter; + + switch (rw) { + case READ: + atomic64_inc(&zram->stats.num_reads); + break; + case WRITE: + atomic64_inc(&zram->stats.num_writes); + break; + } + + index = bio->bi_iter.bi_sector >> SECTORS_PER_PAGE_SHIFT; + offset = (bio->bi_iter.bi_sector & + (SECTORS_PER_PAGE - 1)) << SECTOR_SHIFT; + + 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; + + if (zram_bvec_rw(zram, &bv, index, offset, bio, 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, bio, rw) < 0) + goto out; + } else + if (zram_bvec_rw(zram, &bvec, index, offset, bio, rw) + < 0) + goto out; + + update_position(&index, &offset, &bvec); + } + + set_bit(BIO_UPTODATE, &bio->bi_flags); + bio_endio(bio, 0); + return; + +out: + bio_io_error(bio); +} + +/* + * Handler function for all zram I/O requests. + */ +static void zram_make_request(struct request_queue *queue, struct bio *bio) +{ + struct zram *zram = queue->queuedata; + + down_read(&zram->init_lock); + if (unlikely(!zram->init_done)) + goto error; + + if (!valid_io_request(zram, bio)) { + atomic64_inc(&zram->stats.invalid_io); + goto error; + } + + __zram_make_request(zram, bio, bio_data_dir(bio)); + up_read(&zram->init_lock); + + return; + +error: + up_read(&zram->init_lock); + bio_io_error(bio); +} + +static void zram_slot_free_notify(struct block_device *bdev, + unsigned long index) +{ + struct zram *zram; + struct zram_meta *meta; + + zram = bdev->bd_disk->private_data; + meta = zram->meta; + + write_lock(&meta->tb_lock); + zram_free_page(zram, index); + write_unlock(&meta->tb_lock); + atomic64_inc(&zram->stats.notify_free); +} + +static const struct block_device_operations zram_devops = { + .swap_slot_free_notify = zram_slot_free_notify, + .owner = THIS_MODULE +}; + +static DEVICE_ATTR(disksize, S_IRUGO | S_IWUSR, + disksize_show, disksize_store); +static DEVICE_ATTR(initstate, S_IRUGO, initstate_show, NULL); +static DEVICE_ATTR(reset, S_IWUSR, NULL, reset_store); +static DEVICE_ATTR(num_reads, S_IRUGO, num_reads_show, NULL); +static DEVICE_ATTR(num_writes, S_IRUGO, num_writes_show, NULL); +static DEVICE_ATTR(invalid_io, S_IRUGO, invalid_io_show, NULL); +static DEVICE_ATTR(notify_free, S_IRUGO, notify_free_show, NULL); +static DEVICE_ATTR(zero_pages, S_IRUGO, zero_pages_show, NULL); +static DEVICE_ATTR(orig_data_size, S_IRUGO, orig_data_size_show, NULL); +static DEVICE_ATTR(compr_data_size, S_IRUGO, compr_data_size_show, NULL); +static DEVICE_ATTR(mem_used_total, S_IRUGO, mem_used_total_show, NULL); + +static struct attribute *zram_disk_attrs[] = { + &dev_attr_disksize.attr, + &dev_attr_initstate.attr, + &dev_attr_reset.attr, + &dev_attr_num_reads.attr, + &dev_attr_num_writes.attr, + &dev_attr_invalid_io.attr, + &dev_attr_notify_free.attr, + &dev_attr_zero_pages.attr, + &dev_attr_orig_data_size.attr, + &dev_attr_compr_data_size.attr, + &dev_attr_mem_used_total.attr, + NULL, +}; + +static struct attribute_group zram_disk_attr_group = { + .attrs = zram_disk_attrs, +}; + +static int create_device(struct zram *zram, int device_id) +{ + int ret = -ENOMEM; + + init_rwsem(&zram->init_lock); + + zram->queue = blk_alloc_queue(GFP_KERNEL); + if (!zram->queue) { + pr_err("Error allocating disk queue for device %d\n", + device_id); + goto out; + } + + blk_queue_make_request(zram->queue, zram_make_request); + zram->queue->queuedata = zram; + + /* gendisk structure */ + zram->disk = alloc_disk(1); + if (!zram->disk) { + pr_warn("Error allocating disk structure for device %d\n", + device_id); + goto out_free_queue; + } + + zram->disk->major = zram_major; + zram->disk->first_minor = device_id; + zram->disk->fops = &zram_devops; + zram->disk->queue = zram->queue; + zram->disk->private_data = zram; + snprintf(zram->disk->disk_name, 16, "zram%d", device_id); + + /* Actual capacity set using syfs (/sys/block/zram<id>/disksize */ + set_capacity(zram->disk, 0); + + /* + * To ensure that we always get PAGE_SIZE aligned + * and n*PAGE_SIZED sized I/O requests. + */ + blk_queue_physical_block_size(zram->disk->queue, PAGE_SIZE); + blk_queue_logical_block_size(zram->disk->queue, + ZRAM_LOGICAL_BLOCK_SIZE); + blk_queue_io_min(zram->disk->queue, PAGE_SIZE); + blk_queue_io_opt(zram->disk->queue, PAGE_SIZE); + + add_disk(zram->disk); + + ret = sysfs_create_group(&disk_to_dev(zram->disk)->kobj, + &zram_disk_attr_group); + if (ret < 0) { + pr_warn("Error creating sysfs group"); + goto out_free_disk; + } + + zram->init_done = 0; + return 0; + +out_free_disk: + del_gendisk(zram->disk); + put_disk(zram->disk); +out_free_queue: + blk_cleanup_queue(zram->queue); +out: + return ret; +} + +static void destroy_device(struct zram *zram) +{ + sysfs_remove_group(&disk_to_dev(zram->disk)->kobj, + &zram_disk_attr_group); + + del_gendisk(zram->disk); + put_disk(zram->disk); + + blk_cleanup_queue(zram->queue); +} + +static int __init zram_init(void) +{ + int ret, dev_id; + + if (num_devices > max_num_devices) { + pr_warn("Invalid value for num_devices: %u\n", + num_devices); + ret = -EINVAL; + goto out; + } + + zram_major = register_blkdev(0, "zram"); + if (zram_major <= 0) { + pr_warn("Unable to get major number\n"); + ret = -EBUSY; + goto out; + } + + /* Allocate the device array and initialize each one */ + zram_devices = kzalloc(num_devices * sizeof(struct zram), GFP_KERNEL); + if (!zram_devices) { + ret = -ENOMEM; + goto unregister; + } + + for (dev_id = 0; dev_id < num_devices; dev_id++) { + ret = create_device(&zram_devices[dev_id], dev_id); + if (ret) + goto free_devices; + } + + pr_info("Created %u device(s) ...\n", num_devices); + + return 0; + +free_devices: + while (dev_id) + destroy_device(&zram_devices[--dev_id]); + kfree(zram_devices); +unregister: + unregister_blkdev(zram_major, "zram"); +out: + return ret; +} + +static void __exit zram_exit(void) +{ + int i; + struct zram *zram; + + for (i = 0; i < num_devices; i++) { + zram = &zram_devices[i]; + + destroy_device(zram); + /* + * Shouldn't access zram->disk after destroy_device + * because destroy_device already released zram->disk. + */ + zram_reset_device(zram, false); + } + + unregister_blkdev(zram_major, "zram"); + + kfree(zram_devices); + pr_debug("Cleanup done!\n"); +} + +module_init(zram_init); +module_exit(zram_exit); + +module_param(num_devices, uint, 0); +MODULE_PARM_DESC(num_devices, "Number of zram devices"); + +MODULE_LICENSE("Dual BSD/GPL"); +MODULE_AUTHOR("Nitin Gupta <ngupta@vflare.org>"); +MODULE_DESCRIPTION("Compressed RAM Block Device"); diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h new file mode 100644 index 000000000000..ad8aa35bae00 --- /dev/null +++ b/drivers/block/zram/zram_drv.h @@ -0,0 +1,109 @@ +/* + * Compressed RAM block device + * + * Copyright (C) 2008, 2009, 2010 Nitin Gupta + * 2012, 2013 Minchan Kim + * + * This code is released using a dual license strategy: BSD/GPL + * You can choose the licence that better fits your requirements. + * + * Released under the terms of 3-clause BSD License + * Released under the terms of GNU General Public License Version 2.0 + * + */ + +#ifndef _ZRAM_DRV_H_ +#define _ZRAM_DRV_H_ + +#include <linux/spinlock.h> +#include <linux/mutex.h> +#include <linux/zsmalloc.h> + +/* + * Some arbitrary value. This is just to catch + * invalid value for num_devices module parameter. + */ +static const unsigned max_num_devices = 32; + +/*-- Configurable parameters */ + +/* + * Pages that compress to size greater than this are stored + * uncompressed in memory. + */ +static const size_t max_zpage_size = PAGE_SIZE / 4 * 3; + +/* + * NOTE: max_zpage_size must be less than or equal to: + * ZS_MAX_ALLOC_SIZE. Otherwise, zs_malloc() would + * always return failure. + */ + +/*-- End of configurable params */ + +#define SECTOR_SHIFT 9 +#define SECTOR_SIZE (1 << SECTOR_SHIFT) +#define SECTORS_PER_PAGE_SHIFT (PAGE_SHIFT - SECTOR_SHIFT) +#define SECTORS_PER_PAGE (1 << SECTORS_PER_PAGE_SHIFT) +#define ZRAM_LOGICAL_BLOCK_SHIFT 12 +#define ZRAM_LOGICAL_BLOCK_SIZE (1 << ZRAM_LOGICAL_BLOCK_SHIFT) +#define ZRAM_SECTOR_PER_LOGICAL_BLOCK \ + (1 << (ZRAM_LOGICAL_BLOCK_SHIFT - SECTOR_SHIFT)) + +/* Flags for zram pages (table[page_no].flags) */ +enum zram_pageflags { + /* Page consists entirely of zeros */ + ZRAM_ZERO, + + __NR_ZRAM_PAGEFLAGS, +}; + +/*-- Data structures */ + +/* Allocated for each disk page */ +struct table { + unsigned long handle; + u16 size; /* object size (excluding header) */ + u8 count; /* object ref count (not yet used) */ + u8 flags; +} __aligned(4); + +struct zram_stats { + atomic64_t compr_size; /* compressed size of pages stored */ + atomic64_t num_reads; /* failed + successful */ + atomic64_t num_writes; /* --do-- */ + atomic64_t failed_reads; /* should NEVER! happen */ + atomic64_t failed_writes; /* can happen when memory is too low */ + atomic64_t invalid_io; /* non-page-aligned I/O requests */ + atomic64_t notify_free; /* no. of swap slot free notifications */ + atomic_t pages_zero; /* no. of zero filled pages */ + atomic_t pages_stored; /* no. of pages currently stored */ + atomic_t good_compress; /* % of pages with compression ratio<=50% */ + atomic_t bad_compress; /* % of pages with compression ratio>=75% */ +}; + +struct zram_meta { + rwlock_t tb_lock; /* protect table */ + void *compress_workmem; + void *compress_buffer; + struct table *table; + struct zs_pool *mem_pool; + struct mutex buffer_lock; /* protect compress buffers */ +}; + +struct zram { + struct zram_meta *meta; + struct request_queue *queue; + struct gendisk *disk; + int init_done; + /* Prevent concurrent execution of device init, reset and R/W request */ + struct rw_semaphore init_lock; + /* + * This is the limit on amount of *uncompressed* worth of data + * we can store in a disk. + */ + u64 disksize; /* bytes */ + + struct zram_stats stats; +}; +#endif |