[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-block] [PATCH V3] block/iscsi: allow caching of the allocation
From: |
Paolo Bonzini |
Subject: |
Re: [Qemu-block] [PATCH V3] block/iscsi: allow caching of the allocation map |
Date: |
Mon, 13 Jun 2016 11:45:31 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.1.0 |
On 24/05/2016 10:40, Peter Lieven wrote:
> until now the allocation map was used only as a hint if a cluster
> is allocated or not. If a block was not allocated (or Qemu had
> no info about the allocation status) a get_block_status call was
> issued to check the allocation status and possibly avoid
> a subsequent read of unallocated sectors. If a block known to be
> allocated the get_block_status call was omitted. In the other case
> a get_block_status call was issued before every read to avoid
> the necessity for a consistent allocation map. To avoid the
> potential overhead of calling get_block_status for each and
> every read request this took only place for the bigger requests.
>
> This patch enhances this mechanism to cache the allocation
> status and avoid calling get_block_status for blocks where
> the allocation status has been queried before. This allows
> for bypassing the read request even for smaller requests and
> additionally omits calling get_block_status for known to be
> unallocated blocks.
>
> Signed-off-by: Peter Lieven <address@hidden>
> ---
> v2->v3: - fix wording errors [Fam]
> - reinit allocmap only if allocmap is present in
> bdrv_reopen_commit
> v1->v2: - add more comments [Fam]
> - free allocmap if allocation of allocmap_valid fails [Fam]
> - fix indent and whitespace errors [Fam]
> - account for cache mode changes on reopen
>
> block/iscsi.c | 231
> +++++++++++++++++++++++++++++++++++++++++++++-------------
> 1 file changed, 182 insertions(+), 49 deletions(-)
>
> diff --git a/block/iscsi.c b/block/iscsi.c
> index 2ca8e72..fb308f6 100644
> --- a/block/iscsi.c
> +++ b/block/iscsi.c
> @@ -2,7 +2,7 @@
> * QEMU Block driver for iSCSI images
> *
> * Copyright (c) 2010-2011 Ronnie Sahlberg <address@hidden>
> - * Copyright (c) 2012-2015 Peter Lieven <address@hidden>
> + * Copyright (c) 2012-2016 Peter Lieven <address@hidden>
> *
> * Permission is hereby granted, free of charge, to any person obtaining a
> copy
> * of this software and associated documentation files (the "Software"), to
> deal
> @@ -62,7 +62,23 @@ typedef struct IscsiLun {
> struct scsi_inquiry_logical_block_provisioning lbp;
> struct scsi_inquiry_block_limits bl;
> unsigned char *zeroblock;
> - unsigned long *allocationmap;
> + /* The allocmap tracks which clusters (pages) on the iSCSI target are
> + * allocated and which are not. In case a target returns zeros for
> + * unallocated pages (iscsilun->lprz) we can directly return zeros
> instead
> + * of reading zeros over the wire if a read request falls within an
> + * unallocated block. As there are 3 possible states we need 2 bitmaps to
> + * track. allocmap_valid keeps track if QEMU's information about a page
> is
> + * valid. allocmap tracks if a page is allocated or not. In case QEMU
> has no
> + * valid information about a page the corresponding allocmap entry
> should be
> + * switched to unallocated as well to force a new lookup of the
> allocation
> + * status as lookups are generally skipped if a page is suspect to be
> + * allocated. If a iSCSI target is opened with cache.direct = on the
> + * allocmap_valid does not exist turning all cached information invalid
> so
> + * that a fresh lookup is made for any page even if allocmap entry
> returns
> + * it's unallocated. */
> + unsigned long *allocmap;
> + unsigned long *allocmap_valid;
> + long allocmap_size;
> int cluster_sectors;
> bool use_16_for_rw;
> bool write_protected;
> @@ -415,37 +431,132 @@ static bool is_request_lun_aligned(int64_t sector_num,
> int nb_sectors,
> return 1;
> }
>
> -static unsigned long *iscsi_allocationmap_init(IscsiLun *iscsilun)
> +static void iscsi_allocmap_free(IscsiLun *iscsilun)
> {
> - return bitmap_try_new(DIV_ROUND_UP(sector_lun2qemu(iscsilun->num_blocks,
> - iscsilun),
> - iscsilun->cluster_sectors));
> + g_free(iscsilun->allocmap);
> + g_free(iscsilun->allocmap_valid);
> + iscsilun->allocmap = NULL;
> + iscsilun->allocmap_valid = NULL;
> }
>
> -static void iscsi_allocationmap_set(IscsiLun *iscsilun, int64_t sector_num,
> - int nb_sectors)
> +
> +static int iscsi_allocmap_init(IscsiLun *iscsilun, int open_flags)
> {
> - if (iscsilun->allocationmap == NULL) {
> - return;
> + iscsi_allocmap_free(iscsilun);
> +
> + iscsilun->allocmap_size =
> + DIV_ROUND_UP(sector_lun2qemu(iscsilun->num_blocks, iscsilun),
> + iscsilun->cluster_sectors);
> +
> + iscsilun->allocmap = bitmap_try_new(iscsilun->allocmap_size);
> + if (!iscsilun->allocmap) {
> + return -ENOMEM;
> + }
> +
> + if (open_flags & BDRV_O_NOCACHE) {
> + /* in case that cache.direct = on all allocmap entries are
> + * treated as invalid to force a relookup of the block
> + * status on every read request */
> + return 0;
> + }
> +
> + iscsilun->allocmap_valid = bitmap_try_new(iscsilun->allocmap_size);
> + if (!iscsilun->allocmap_valid) {
> + /* if we are under memory pressure free the allocmap as well */
> + iscsi_allocmap_free(iscsilun);
> + return -ENOMEM;
> }
> - bitmap_set(iscsilun->allocationmap,
> - sector_num / iscsilun->cluster_sectors,
> - DIV_ROUND_UP(nb_sectors, iscsilun->cluster_sectors));
> +
> + return 0;
> }
>
> -static void iscsi_allocationmap_clear(IscsiLun *iscsilun, int64_t sector_num,
> - int nb_sectors)
> +static void
> +iscsi_allocmap_update(IscsiLun *iscsilun, int64_t sector_num,
> + int nb_sectors, bool allocated, bool valid)
> {
> int64_t cluster_num, nb_clusters;
> - if (iscsilun->allocationmap == NULL) {
> +
> + if (iscsilun->allocmap == NULL) {
> return;
> }
> cluster_num = DIV_ROUND_UP(sector_num, iscsilun->cluster_sectors);
> nb_clusters = (sector_num + nb_sectors) / iscsilun->cluster_sectors
> - cluster_num;
> - if (nb_clusters > 0) {
> - bitmap_clear(iscsilun->allocationmap, cluster_num, nb_clusters);
> + if (allocated) {
> + bitmap_set(iscsilun->allocmap,
> + sector_num / iscsilun->cluster_sectors,
> + DIV_ROUND_UP(nb_sectors, iscsilun->cluster_sectors));
> + } else if (nb_clusters > 0) {
> + bitmap_clear(iscsilun->allocmap, cluster_num, nb_clusters);
> + }
> +
> + if (iscsilun->allocmap_valid == NULL) {
> + return;
> + }
> + if (valid) {
> + if (nb_clusters > 0) {
> + bitmap_set(iscsilun->allocmap_valid, cluster_num, nb_clusters);
> + }
> + } else {
> + bitmap_clear(iscsilun->allocmap_valid,
> + sector_num / iscsilun->cluster_sectors,
> + DIV_ROUND_UP(nb_sectors, iscsilun->cluster_sectors));
> + }
> +}
> +
> +static void
> +iscsi_allocmap_set_allocated(IscsiLun *iscsilun, int64_t sector_num,
> + int nb_sectors)
> +{
> + iscsi_allocmap_update(iscsilun, sector_num, nb_sectors, true, true);
> +}
> +
> +static void
> +iscsi_allocmap_set_unallocated(IscsiLun *iscsilun, int64_t sector_num,
> + int nb_sectors)
> +{
> + iscsi_allocmap_update(iscsilun, sector_num, nb_sectors, false, true);
> +}
> +
> +static void iscsi_allocmap_set_invalid(IscsiLun *iscsilun, int64_t
> sector_num,
> + int nb_sectors)
> +{
> + iscsi_allocmap_update(iscsilun, sector_num, nb_sectors, false, false);
> +}
> +
> +static void iscsi_allocmap_invalidate(IscsiLun *iscsilun)
> +{
> + if (iscsilun->allocmap) {
> + bitmap_zero(iscsilun->allocmap, iscsilun->allocmap_size);
> }
> + if (iscsilun->allocmap_valid) {
> + bitmap_zero(iscsilun->allocmap_valid, iscsilun->allocmap_size);
> + }
> +}
> +
> +static inline bool
> +iscsi_allocmap_is_allocated(IscsiLun *iscsilun, int64_t sector_num,
> + int nb_sectors)
> +{
> + unsigned long size;
> + if (iscsilun->allocmap == NULL) {
> + return true;
> + }
> + size = DIV_ROUND_UP(sector_num + nb_sectors, iscsilun->cluster_sectors);
> + return !(find_next_bit(iscsilun->allocmap, size,
> + sector_num / iscsilun->cluster_sectors) == size);
> +}
> +
> +static inline bool iscsi_allocmap_is_valid(IscsiLun *iscsilun,
> + int64_t sector_num, int
> nb_sectors)
> +{
> + unsigned long size;
> + if (iscsilun->allocmap_valid == NULL) {
> + return false;
> + }
> + size = DIV_ROUND_UP(sector_num + nb_sectors, iscsilun->cluster_sectors);
> + return (find_next_zero_bit(iscsilun->allocmap_valid, size,
> + sector_num / iscsilun->cluster_sectors) ==
> size);
> }
>
> static int coroutine_fn
> @@ -507,26 +618,16 @@ retry:
> }
>
> if (iTask.status != SCSI_STATUS_GOOD) {
> + iscsi_allocmap_set_invalid(iscsilun, sector_num, nb_sectors);
> return iTask.err_code;
> }
>
> - iscsi_allocationmap_set(iscsilun, sector_num, nb_sectors);
> + iscsi_allocmap_set_allocated(iscsilun, sector_num, nb_sectors);
>
> return 0;
> }
>
>
> -static bool iscsi_allocationmap_is_allocated(IscsiLun *iscsilun,
> - int64_t sector_num, int
> nb_sectors)
> -{
> - unsigned long size;
> - if (iscsilun->allocationmap == NULL) {
> - return true;
> - }
> - size = DIV_ROUND_UP(sector_num + nb_sectors, iscsilun->cluster_sectors);
> - return !(find_next_bit(iscsilun->allocationmap, size,
> - sector_num / iscsilun->cluster_sectors) == size);
> -}
>
> static int64_t coroutine_fn iscsi_co_get_block_status(BlockDriverState *bs,
> int64_t sector_num,
> @@ -611,9 +712,9 @@ retry:
> }
>
> if (ret & BDRV_BLOCK_ZERO) {
> - iscsi_allocationmap_clear(iscsilun, sector_num, *pnum);
> + iscsi_allocmap_set_unallocated(iscsilun, sector_num, *pnum);
> } else {
> - iscsi_allocationmap_set(iscsilun, sector_num, *pnum);
> + iscsi_allocmap_set_allocated(iscsilun, sector_num, *pnum);
> }
>
> if (*pnum > nb_sectors) {
> @@ -648,16 +749,32 @@ static int coroutine_fn iscsi_co_readv(BlockDriverState
> *bs,
> return -EINVAL;
> }
>
> - if (iscsilun->lbprz && nb_sectors >= ISCSI_CHECKALLOC_THRES &&
> - !iscsi_allocationmap_is_allocated(iscsilun, sector_num, nb_sectors))
> {
> - int64_t ret;
> + /* if cache.direct is off and we have a valid entry in our allocation map
> + * we can skip checking the block status and directly return zeroes if
> + * the request falls within an unallocated area */
> + if (iscsi_allocmap_is_valid(iscsilun, sector_num, nb_sectors) &&
> + !iscsi_allocmap_is_allocated(iscsilun, sector_num, nb_sectors)) {
> + qemu_iovec_memset(iov, 0, 0x00, iov->size);
> + return 0;
> + }
> +
> + if (nb_sectors >= ISCSI_CHECKALLOC_THRES &&
> + !iscsi_allocmap_is_valid(iscsilun, sector_num, nb_sectors) &&
> + !iscsi_allocmap_is_allocated(iscsilun, sector_num, nb_sectors)) {
> int pnum;
> BlockDriverState *file;
> - ret = iscsi_co_get_block_status(bs, sector_num, INT_MAX, &pnum,
> &file);
> + /* check the block status from the beginning of the cluster
> + * containing the start sector */
> + int64_t ret = iscsi_co_get_block_status(bs,
> + sector_num - sector_num %
> iscsilun->cluster_sectors,
> + INT_MAX, &pnum, &file);
> if (ret < 0) {
> return ret;
> }
> - if (ret & BDRV_BLOCK_ZERO && pnum >= nb_sectors) {
> + /* if the whole request falls into an unallocated area we can avoid
> + * to read and directly return zeroes instead */
> + if (ret & BDRV_BLOCK_ZERO &&
> + pnum >= nb_sectors + sector_num % iscsilun->cluster_sectors) {
> qemu_iovec_memset(iov, 0, 0x00, iov->size);
> return 0;
> }
> @@ -964,7 +1081,7 @@ retry:
> return iTask.err_code;
> }
>
> - iscsi_allocationmap_clear(iscsilun, sector_num, nb_sectors);
> + iscsi_allocmap_set_invalid(iscsilun, sector_num, nb_sectors);
>
> return 0;
> }
> @@ -1054,13 +1171,14 @@ retry:
> }
>
> if (iTask.status != SCSI_STATUS_GOOD) {
> + iscsi_allocmap_set_invalid(iscsilun, sector_num, nb_sectors);
> return iTask.err_code;
> }
>
> if (flags & BDRV_REQ_MAY_UNMAP) {
> - iscsi_allocationmap_clear(iscsilun, sector_num, nb_sectors);
> + iscsi_allocmap_set_invalid(iscsilun, sector_num, nb_sectors);
> } else {
> - iscsi_allocationmap_set(iscsilun, sector_num, nb_sectors);
> + iscsi_allocmap_set_allocated(iscsilun, sector_num, nb_sectors);
> }
>
> return 0;
> @@ -1634,10 +1752,7 @@ static int iscsi_open(BlockDriverState *bs, QDict
> *options, int flags,
> iscsilun->cluster_sectors = (iscsilun->bl.opt_unmap_gran *
> iscsilun->block_size) >>
> BDRV_SECTOR_BITS;
> if (iscsilun->lbprz) {
> - iscsilun->allocationmap = iscsi_allocationmap_init(iscsilun);
> - if (iscsilun->allocationmap == NULL) {
> - ret = -ENOMEM;
> - }
> + ret = iscsi_allocmap_init(iscsilun, bs->open_flags);
> }
> }
>
> @@ -1674,7 +1789,7 @@ static void iscsi_close(BlockDriverState *bs)
> }
> iscsi_destroy_context(iscsi);
> g_free(iscsilun->zeroblock);
> - g_free(iscsilun->allocationmap);
> + iscsi_allocmap_free(iscsilun);
> memset(iscsilun, 0, sizeof(IscsiLun));
> }
>
> @@ -1732,6 +1847,16 @@ static int iscsi_reopen_prepare(BDRVReopenState *state,
> return 0;
> }
>
> +static void iscsi_reopen_commit(BDRVReopenState *reopen_state)
> +{
> + IscsiLun *iscsilun = reopen_state->bs->opaque;
> +
> + /* the cache.direct status might have changed */
> + if (iscsilun->allocmap != NULL) {
> + iscsi_allocmap_init(iscsilun, reopen_state->flags);
> + }
> +}
> +
> static int iscsi_truncate(BlockDriverState *bs, int64_t offset)
> {
> IscsiLun *iscsilun = bs->opaque;
> @@ -1751,9 +1876,8 @@ static int iscsi_truncate(BlockDriverState *bs, int64_t
> offset)
> return -EINVAL;
> }
>
> - if (iscsilun->allocationmap != NULL) {
> - g_free(iscsilun->allocationmap);
> - iscsilun->allocationmap = iscsi_allocationmap_init(iscsilun);
> + if (iscsilun->allocmap != NULL) {
> + iscsi_allocmap_init(iscsilun, bs->open_flags);
> }
>
> return 0;
> @@ -1813,6 +1937,13 @@ static int iscsi_get_info(BlockDriverState *bs,
> BlockDriverInfo *bdi)
> return 0;
> }
>
> +static void iscsi_invalidate_cache(BlockDriverState *bs,
> + Error **errp)
> +{
> + IscsiLun *iscsilun = bs->opaque;
> + iscsi_allocmap_invalidate(iscsilun);
> +}
> +
> static QemuOptsList iscsi_create_opts = {
> .name = "iscsi-create-opts",
> .head = QTAILQ_HEAD_INITIALIZER(iscsi_create_opts.head),
> @@ -1836,7 +1967,9 @@ static BlockDriver bdrv_iscsi = {
> .bdrv_close = iscsi_close,
> .bdrv_create = iscsi_create,
> .create_opts = &iscsi_create_opts,
> - .bdrv_reopen_prepare = iscsi_reopen_prepare,
> + .bdrv_reopen_prepare = iscsi_reopen_prepare,
> + .bdrv_reopen_commit = iscsi_reopen_commit,
> + .bdrv_invalidate_cache = iscsi_invalidate_cache,
>
> .bdrv_getlength = iscsi_getlength,
> .bdrv_get_info = iscsi_get_info,
>
Queued, thanks.
Paolo