X-Git-Url: https://git.whamcloud.com/?a=blobdiff_plain;f=lustre%2Flov%2Flov_io.c;h=464c8f2c24ecc600f035f48fecb2fd1287f251b4;hb=8d6d2914cf85be0508805cd5195163e1959652f4;hp=c07540b5db05383f657e6a36a8778d27c623b671;hpb=ec7eb7a932b7d95b9b3d12800edb7416bfb4f04a;p=fs%2Flustre-release.git diff --git a/lustre/lov/lov_io.c b/lustre/lov/lov_io.c index c07540b..464c8f2 100644 --- a/lustre/lov/lov_io.c +++ b/lustre/lov/lov_io.c @@ -23,7 +23,7 @@ * Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved. * Use is subject to license terms. * - * Copyright (c) 2011, 2016, Intel Corporation. + * Copyright (c) 2011, 2017, Intel Corporation. */ /* * This file is part of Lustre, http://www.lustre.org/ @@ -56,7 +56,7 @@ static inline struct lov_io_sub *lov_sub_alloc(struct lov_io *lio, int index) OBD_ALLOC_PTR(sub); } - if (sub != NULL) { + if (sub) { INIT_LIST_HEAD(&sub->sub_list); INIT_LIST_HEAD(&sub->sub_linkage); sub->sub_subio_index = index; @@ -82,7 +82,7 @@ static void lov_io_sub_fini(const struct lu_env *env, struct lov_io *lio, cl_io_fini(sub->sub_env, &sub->sub_io); - if (sub->sub_env != NULL && !IS_ERR(sub->sub_env)) { + if (sub->sub_env && !IS_ERR(sub->sub_env)) { cl_env_put(sub->sub_env, &sub->sub_refcheck); sub->sub_env = NULL; } @@ -122,8 +122,10 @@ static int lov_io_sub_init(const struct lu_env *env, struct lov_io *lio, /* obtain new environment */ sub->sub_env = cl_env_get(&sub->sub_refcheck); - if (IS_ERR(sub->sub_env)) + if (IS_ERR(sub->sub_env)) { result = PTR_ERR(sub->sub_env); + RETURN(result); + } sub_obj = lovsub2cl(lov_r0(lov, index)->lo_sub[stripe]); sub_io = &sub->sub_io; @@ -136,7 +138,7 @@ static int lov_io_sub_init(const struct lu_env *env, struct lov_io *lio, sub_io->ci_type = io->ci_type; sub_io->ci_no_srvlock = io->ci_no_srvlock; sub_io->ci_noatime = io->ci_noatime; - sub_io->ci_pio = io->ci_pio; + sub_io->ci_async_readahead = io->ci_async_readahead; sub_io->ci_lock_no_expand = io->ci_lock_no_expand; sub_io->ci_ndelay = io->ci_ndelay; sub_io->ci_layout_version = io->ci_layout_version; @@ -166,7 +168,7 @@ struct lov_io_sub *lov_sub_get(const struct lu_env *env, if (rc == 0) { sub = lov_sub_alloc(lio, index); - if (sub == NULL) + if (!sub) GOTO(out, rc = -ENOMEM); rc = lov_io_sub_init(env, lio, sub); @@ -189,19 +191,6 @@ out: * Lov io operations. * */ - -int lov_page_index(const struct cl_page *page) -{ - const struct cl_page_slice *slice; - ENTRY; - - slice = cl_page_at(page, &lov_device_type); - LASSERT(slice != NULL); - LASSERT(slice->cpl_obj != NULL); - - RETURN(cl2lov_page(slice)->lps_index); -} - static int lov_io_subio_init(const struct lu_env *env, struct lov_io *lio, struct cl_io *io) { @@ -237,8 +226,18 @@ static int lov_io_mirror_write_intent(struct lov_io *lio, cl_io_is_mkwrite(io))) RETURN(0); + /* + * FLR: check if it needs to send a write intent RPC to server. + * Writing to sync_pending file needs write intent RPC to change + * the file state back to write_pending, so that the layout version + * can be increased when the state changes to sync_pending at a later + * time. Otherwise there exists a chance that an evicted client may + * dirty the file data while resync client is working on it. + * Designated I/O is allowed for resync workload. + */ if (lov_flr_state(obj) == LCM_FL_RDONLY || - lov_flr_state(obj) == LCM_FL_SYNC_PENDING) { + (lov_flr_state(obj) == LCM_FL_SYNC_PENDING && + io->ci_designated_mirror == 0)) { io->ci_need_write_intent = 1; RETURN(0); } @@ -246,8 +245,10 @@ static int lov_io_mirror_write_intent(struct lov_io *lio, LASSERT((lov_flr_state(obj) == LCM_FL_WRITE_PENDING)); LASSERT(comp->lo_preferred_mirror >= 0); - /* need to iterate all components to see if there are - * multiple components covering the writing component */ + /* + * need to iterate all components to see if there are + * multiple components covering the writing component + */ primary = &comp->lo_mirrors[comp->lo_preferred_mirror]; LASSERT(!primary->lre_stale); lov_foreach_mirror_layout_entry(obj, lle, primary) { @@ -308,6 +309,47 @@ static int lov_io_mirror_init(struct lov_io *lio, struct lov_object *obj, RETURN(0); } + /* transfer the layout version for verification */ + if (io->ci_layout_version == 0) + io->ci_layout_version = obj->lo_lsm->lsm_layout_gen; + + /* find the corresponding mirror for designated mirror IO */ + if (io->ci_designated_mirror > 0) { + struct lov_mirror_entry *entry; + + LASSERT(!io->ci_ndelay); + + CDEBUG(D_LAYOUT, "designated I/O mirror state: %d\n", + lov_flr_state(obj)); + + if ((cl_io_is_trunc(io) || io->ci_type == CIT_WRITE) && + (io->ci_layout_version != obj->lo_lsm->lsm_layout_gen)) { + /* + * For resync I/O, the ci_layout_version was the layout + * version when resync starts. If it doesn't match the + * current object layout version, it means the layout + * has been changed + */ + RETURN(-ESTALE); + } + + io->ci_layout_version |= LU_LAYOUT_RESYNC; + + index = 0; + lio->lis_mirror_index = -1; + lov_foreach_mirror_entry(obj, entry) { + if (entry->lre_mirror_id == + io->ci_designated_mirror) { + lio->lis_mirror_index = index; + break; + } + + index++; + } + + RETURN(lio->lis_mirror_index < 0 ? -EINVAL : 0); + } + result = lov_io_mirror_write_intent(lio, obj, io); if (result) RETURN(result); @@ -317,13 +359,20 @@ static int lov_io_mirror_init(struct lov_io *lio, struct lov_object *obj, PFID(lu_object_fid(lov2lu(obj))), lio->lis_pos, lio->lis_endpos); + if (cl_io_is_trunc(io)) { + /** + * for truncate, we uses [size, EOF) to judge whether + * a write intent needs to be send, but we need to + * restore the write extent to [0, size). + */ + io->ci_write_intent.e_start = 0; + io->ci_write_intent.e_end = + io->u.ci_setattr.sa_attr.lvb_size; + } /* stop cl_io_init() loop */ RETURN(1); } - /* transfer the layout version for verification */ - io->ci_layout_version = obj->lo_lsm->lsm_layout_gen; - if (io->ci_ndelay_tried == 0 || /* first time to try */ /* reset the mirror index if layout has changed */ lio->lis_mirror_layout_gen != obj->lo_lsm->lsm_layout_gen) { @@ -381,13 +430,15 @@ static int lov_io_mirror_init(struct lov_io *lio, struct lov_object *obj, lio->lis_mirror_index = index; - /* FLR: if all mirrors have been tried once, most likely the network + /* + * FLR: if all mirrors have been tried once, most likely the network * of this client has been partitioned. We should relinquish CPU for - * a while before trying again. */ + * a while before trying again. + */ ++io->ci_ndelay_tried; if (io->ci_ndelay && io->ci_ndelay_tried >= comp->lo_mirror_count) { set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(msecs_to_jiffies(MSEC_PER_SEC)); /* 10ms */ + schedule_timeout(cfs_time_seconds(1)); /* 10ms */ if (signal_pending(current)) RETURN(-EINTR); @@ -404,7 +455,6 @@ static int lov_io_mirror_init(struct lov_io *lio, struct lov_object *obj, static int lov_io_slice_init(struct lov_io *lio, struct lov_object *obj, struct cl_io *io) { - struct lu_extent ext; int index; int result = 0; ENTRY; @@ -417,14 +467,16 @@ static int lov_io_slice_init(struct lov_io *lio, switch (io->ci_type) { case CIT_READ: case CIT_WRITE: - lio->lis_pos = io->u.ci_rw.rw_range.cir_pos; - lio->lis_endpos = lio->lis_pos + io->u.ci_rw.rw_range.cir_count; + lio->lis_pos = io->u.ci_rw.crw_pos; + lio->lis_endpos = io->u.ci_rw.crw_pos + io->u.ci_rw.crw_count; lio->lis_io_endpos = lio->lis_endpos; if (cl_io_is_append(io)) { LASSERT(io->ci_type == CIT_WRITE); - /* If there is LOV EA hole, then we may cannot locate - * the current file-tail exactly. */ + /* + * If there is LOV EA hole, then we may cannot locate + * the current file-tail exactly. + */ if (unlikely(obj->lo_lsm->lsm_entries[0]->lsme_pattern & LOV_PATTERN_F_HOLE)) GOTO(out, result = -EIO); @@ -434,25 +486,26 @@ static int lov_io_slice_init(struct lov_io *lio, } break; - case CIT_SETATTR: - if (cl_io_is_trunc(io)) - lio->lis_pos = io->u.ci_setattr.sa_attr.lvb_size; - else - lio->lis_pos = 0; - lio->lis_endpos = OBD_OBJECT_EOF; - break; + case CIT_SETATTR: + if (cl_io_is_trunc(io)) + lio->lis_pos = io->u.ci_setattr.sa_attr.lvb_size; + else + lio->lis_pos = 0; + lio->lis_endpos = OBD_OBJECT_EOF; + break; case CIT_DATA_VERSION: lio->lis_pos = 0; lio->lis_endpos = OBD_OBJECT_EOF; break; - case CIT_FAULT: { - pgoff_t index = io->u.ci_fault.ft_index; - lio->lis_pos = cl_offset(io->ci_obj, index); - lio->lis_endpos = cl_offset(io->ci_obj, index + 1); - break; - } + case CIT_FAULT: { + pgoff_t index = io->u.ci_fault.ft_index; + + lio->lis_pos = cl_offset(io->ci_obj, index); + lio->lis_endpos = cl_offset(io->ci_obj, index + 1); + break; + } case CIT_FSYNC: { lio->lis_pos = io->u.ci_fsync.fi_start; @@ -476,14 +529,14 @@ static int lov_io_slice_init(struct lov_io *lio, GOTO(out, result = 1); break; - case CIT_MISC: - lio->lis_pos = 0; - lio->lis_endpos = OBD_OBJECT_EOF; - break; + case CIT_MISC: + lio->lis_pos = 0; + lio->lis_endpos = OBD_OBJECT_EOF; + break; - default: - LBUG(); - } + default: + LBUG(); + } result = lov_io_mirror_init(lio, obj, io); if (result) @@ -494,24 +547,52 @@ static int lov_io_slice_init(struct lov_io *lio, (cl_io_is_trunc(io) && io->u.ci_setattr.sa_attr.lvb_size > 0))) GOTO(out, result = 0); - ext.e_start = lio->lis_pos; - ext.e_end = lio->lis_endpos; - - /* for truncate, it only needs to instantiate the components - * before the truncated size. */ + /* + * for truncate, it only needs to instantiate the components + * before the truncated size. + */ if (cl_io_is_trunc(io)) { - ext.e_start = 0; - ext.e_end = io->u.ci_setattr.sa_attr.lvb_size; + io->ci_write_intent.e_start = 0; + /* for writes, e_end is endpos, the location of the file + * pointer after the write is completed, so it is not accessed. + * For truncate, 'end' is the size, and *is* acccessed. + * In other words, writes are [start, end), but truncate is + * [start, size], where both are included. So add 1 to the + * size when creating the write intent to account for this. + */ + io->ci_write_intent.e_end = + io->u.ci_setattr.sa_attr.lvb_size + 1; + } else { + io->ci_write_intent.e_start = lio->lis_pos; + io->ci_write_intent.e_end = lio->lis_endpos; } index = 0; - lov_foreach_io_layout(index, lio, &ext) { + lov_foreach_io_layout(index, lio, &io->ci_write_intent) { if (!lsm_entry_inited(obj->lo_lsm, index)) { io->ci_need_write_intent = 1; - io->ci_write_intent = ext; - GOTO(out, result = 1); + break; } } + + if (io->ci_need_write_intent && io->ci_designated_mirror > 0) { + /* + * REINT_SYNC RPC has already tried to instantiate all of the + * components involved, obviously it didn't succeed. Skip this + * mirror for now. The server won't be able to figure out + * which mirror it should instantiate components + */ + CERROR(DFID": trying to instantiate components for designated " + "I/O, file state: %d\n", + PFID(lu_object_fid(lov2lu(obj))), lov_flr_state(obj)); + + io->ci_need_write_intent = 0; + GOTO(out, result = -EIO); + } + + if (io->ci_need_write_intent) + GOTO(out, result = 1); + EXIT; out: @@ -555,13 +636,13 @@ static void lov_io_sub_inherit(struct lov_io_sub *sub, struct lov_io *lio, int index = lov_comp_entry(sub->sub_subio_index); int stripe = lov_comp_stripe(sub->sub_subio_index); - io->ci_pio = parent->ci_pio; switch (io->ci_type) { case CIT_SETATTR: { io->u.ci_setattr.sa_attr = parent->u.ci_setattr.sa_attr; io->u.ci_setattr.sa_attr_flags = parent->u.ci_setattr.sa_attr_flags; - io->u.ci_setattr.sa_valid = parent->u.ci_setattr.sa_valid; + io->u.ci_setattr.sa_avalid = parent->u.ci_setattr.sa_avalid; + io->u.ci_setattr.sa_xvalid = parent->u.ci_setattr.sa_xvalid; io->u.ci_setattr.sa_stripe_index = stripe; io->u.ci_setattr.sa_parent_fid = parent->u.ci_setattr.sa_parent_fid; @@ -600,16 +681,12 @@ static void lov_io_sub_inherit(struct lov_io_sub *sub, struct lov_io *lio, } case CIT_READ: case CIT_WRITE: { - io->u.ci_rw.rw_ptask = parent->u.ci_rw.rw_ptask; - io->u.ci_rw.rw_iter = parent->u.ci_rw.rw_iter; - io->u.ci_rw.rw_iocb = parent->u.ci_rw.rw_iocb; - io->u.ci_rw.rw_file = parent->u.ci_rw.rw_file; - io->u.ci_rw.rw_sync = parent->u.ci_rw.rw_sync; + io->u.ci_wr.wr_sync = cl_io_is_sync_write(parent); if (cl_io_is_append(parent)) { - io->u.ci_rw.rw_append = 1; + io->u.ci_wr.wr_append = 1; } else { - io->u.ci_rw.rw_range.cir_pos = start; - io->u.ci_rw.rw_range.cir_count = end - start; + io->u.ci_rw.crw_pos = start; + io->u.ci_rw.crw_count = end - start; } break; } @@ -630,28 +707,29 @@ static void lov_io_sub_inherit(struct lov_io_sub *sub, struct lov_io *lio, static loff_t lov_offset_mod(loff_t val, int delta) { - if (val != OBD_OBJECT_EOF) - val += delta; - return val; + if (val != OBD_OBJECT_EOF) + val += delta; + return val; } static int lov_io_iter_init(const struct lu_env *env, const struct cl_io_slice *ios) { - struct lov_io *lio = cl2lov_io(env, ios); + struct lov_io *lio = cl2lov_io(env, ios); struct lov_stripe_md *lsm = lio->lis_object->lo_lsm; - struct lov_io_sub *sub; + struct lov_io_sub *sub; struct lu_extent ext; int index; int rc = 0; - ENTRY; + ENTRY; ext.e_start = lio->lis_pos; ext.e_end = lio->lis_endpos; lov_foreach_io_layout(index, lio, &ext) { - struct lov_layout_raid0 *r0 = lov_r0(lio->lis_object, index); + struct lov_layout_entry *le = lov_entry(lio->lis_object, index); + struct lov_layout_raid0 *r0 = &le->lle_raid0; u64 start; u64 end; int stripe; @@ -659,17 +737,25 @@ static int lov_io_iter_init(const struct lu_env *env, CDEBUG(D_VFSTRACE, "component[%d] flags %#x\n", index, lsm->lsm_entries[index]->lsme_flags); if (!lsm_entry_inited(lsm, index)) { - /* Read from uninitialized components should return - * zero filled pages. */ + /* + * Read from uninitialized components should return + * zero filled pages. + */ continue; } + if (!le->lle_valid && !ios->cis_io->ci_designated_mirror) { + CERROR("I/O to invalid component: %d, mirror: %d\n", + index, lio->lis_mirror_index); + RETURN(-EIO); + } + for (stripe = 0; stripe < r0->lo_nr; stripe++) { if (!lov_stripe_intersects(lsm, index, stripe, &ext, &start, &end)) continue; - if (unlikely(r0->lo_sub[stripe] == NULL)) { + if (unlikely(!r0->lo_sub[stripe])) { if (ios->cis_io->ci_type == CIT_READ || ios->cis_io->ci_type == CIT_WRITE || ios->cis_io->ci_type == CIT_FAULT) @@ -693,9 +779,8 @@ static int lov_io_iter_init(const struct lu_env *env, if (rc != 0) break; - CDEBUG(D_VFSTRACE, - "shrink stripe: {%d, %d} range: [%llu, %llu)\n", - index, stripe, start, end); + CDEBUG(D_VFSTRACE, "shrink: %d [%llu, %llu)\n", + stripe, start, end); list_add_tail(&sub->sub_linkage, &lio->lis_active); } @@ -708,11 +793,10 @@ static int lov_io_iter_init(const struct lu_env *env, static int lov_io_rw_iter_init(const struct lu_env *env, const struct cl_io_slice *ios) { - struct cl_io *io = ios->cis_io; struct lov_io *lio = cl2lov_io(env, ios); + struct cl_io *io = ios->cis_io; struct lov_stripe_md_entry *lse; - struct cl_io_range *range = &io->u.ci_rw.rw_range; - loff_t start = range->cir_pos; + loff_t start = io->u.ci_rw.crw_pos; loff_t next; int index; @@ -722,14 +806,14 @@ static int lov_io_rw_iter_init(const struct lu_env *env, if (cl_io_is_append(io)) RETURN(lov_io_iter_init(env, ios)); - index = lov_io_layout_at(lio, range->cir_pos); + index = lov_io_layout_at(lio, io->u.ci_rw.crw_pos); if (index < 0) { /* non-existing layout component */ if (io->ci_type == CIT_READ) { - /* TODO: it needs to detect the next component and - * then set the next pos */ + /* + * TODO: it needs to detect the next component and + * then set the next pos + */ io->ci_continue = 0; - /* execute it in main thread */ - io->ci_pio = 0; RETURN(lov_io_iter_init(env, ios)); } @@ -737,6 +821,10 @@ static int lov_io_rw_iter_init(const struct lu_env *env, RETURN(-ENODATA); } + if (!lov_entry(lio->lis_object, index)->lle_valid && + !io->ci_designated_mirror) + RETURN(io->ci_type == CIT_READ ? -EAGAIN : -EIO); + lse = lov_lse(lio->lis_object, index); next = MAX_LFS_FILESIZE; @@ -749,28 +837,20 @@ static int lov_io_rw_iter_init(const struct lu_env *env, next = MAX_LFS_FILESIZE; } - LASSERTF(range->cir_pos >= lse->lsme_extent.e_start, - "pos %lld, [%lld, %lld)\n", range->cir_pos, + LASSERTF(io->u.ci_rw.crw_pos >= lse->lsme_extent.e_start, + "pos %lld, [%lld, %lld)\n", io->u.ci_rw.crw_pos, lse->lsme_extent.e_start, lse->lsme_extent.e_end); next = min_t(__u64, next, lse->lsme_extent.e_end); next = min_t(loff_t, next, lio->lis_io_endpos); - io->ci_continue = next < lio->lis_io_endpos; - range->cir_count = next - range->cir_pos; - lio->lis_pos = range->cir_pos; - lio->lis_endpos = range->cir_pos + range->cir_count; + io->ci_continue = next < lio->lis_io_endpos; + io->u.ci_rw.crw_count = next - io->u.ci_rw.crw_pos; + lio->lis_pos = io->u.ci_rw.crw_pos; + lio->lis_endpos = io->u.ci_rw.crw_pos + io->u.ci_rw.crw_count; CDEBUG(D_VFSTRACE, - "stripe: {%d, %llu} range: [%llu, %llu) end: %llu, count: %zd\n", - index, start, lio->lis_pos, lio->lis_endpos, - lio->lis_io_endpos, range->cir_count); - - if (!io->ci_continue) { - /* the last piece of IO, execute it in main thread */ - io->ci_pio = 0; - } - - if (io->ci_pio) - RETURN(0); + "stripe: %llu chunk: [%llu, %llu) %llu, %zd\n", + (__u64)start, lio->lis_pos, lio->lis_endpos, + (__u64)lio->lis_io_endpos, io->u.ci_rw.crw_count); /* * XXX The following call should be optimized: we know, that @@ -818,49 +898,49 @@ static int lov_io_call(const struct lu_env *env, struct lov_io *lio, static int lov_io_lock(const struct lu_env *env, const struct cl_io_slice *ios) { - ENTRY; - RETURN(lov_io_call(env, cl2lov_io(env, ios), cl_io_lock)); + ENTRY; + RETURN(lov_io_call(env, cl2lov_io(env, ios), cl_io_lock)); } static int lov_io_start(const struct lu_env *env, const struct cl_io_slice *ios) { - ENTRY; - RETURN(lov_io_call(env, cl2lov_io(env, ios), cl_io_start)); + ENTRY; + RETURN(lov_io_call(env, cl2lov_io(env, ios), cl_io_start)); } static int lov_io_end_wrapper(const struct lu_env *env, struct cl_io *io) { - ENTRY; - /* - * It's possible that lov_io_start() wasn't called against this - * sub-io, either because previous sub-io failed, or upper layer - * completed IO. - */ - if (io->ci_state == CIS_IO_GOING) - cl_io_end(env, io); - else - io->ci_state = CIS_IO_FINISHED; - RETURN(0); + ENTRY; + /* + * It's possible that lov_io_start() wasn't called against this + * sub-io, either because previous sub-io failed, or upper layer + * completed IO. + */ + if (io->ci_state == CIS_IO_GOING) + cl_io_end(env, io); + else + io->ci_state = CIS_IO_FINISHED; + RETURN(0); } static int lov_io_iter_fini_wrapper(const struct lu_env *env, struct cl_io *io) { - cl_io_iter_fini(env, io); - RETURN(0); + cl_io_iter_fini(env, io); + RETURN(0); } static int lov_io_unlock_wrapper(const struct lu_env *env, struct cl_io *io) { - cl_io_unlock(env, io); - RETURN(0); + cl_io_unlock(env, io); + RETURN(0); } static void lov_io_end(const struct lu_env *env, const struct cl_io_slice *ios) { - int rc; + int rc; - rc = lov_io_call(env, cl2lov_io(env, ios), lov_io_end_wrapper); - LASSERT(rc == 0); + rc = lov_io_call(env, cl2lov_io(env, ios), lov_io_end_wrapper); + LASSERT(rc == 0); } static void @@ -875,7 +955,7 @@ lov_io_data_version_end(const struct lu_env *env, const struct cl_io_slice *ios) list_for_each_entry(sub, &lio->lis_active, sub_linkage) { struct cl_data_version_io *sdv = &sub->sub_io.u.ci_data_version; - lov_io_end_wrapper(env, &sub->sub_io); + lov_io_end_wrapper(sub->sub_env, &sub->sub_io); pdv->dv_data_version += sdv->dv_data_version; if (pdv->dv_layout_version > sdv->dv_layout_version) @@ -891,26 +971,26 @@ lov_io_data_version_end(const struct lu_env *env, const struct cl_io_slice *ios) static void lov_io_iter_fini(const struct lu_env *env, const struct cl_io_slice *ios) { - struct lov_io *lio = cl2lov_io(env, ios); - int rc; + struct lov_io *lio = cl2lov_io(env, ios); + int rc; - ENTRY; - rc = lov_io_call(env, lio, lov_io_iter_fini_wrapper); - LASSERT(rc == 0); + ENTRY; + rc = lov_io_call(env, lio, lov_io_iter_fini_wrapper); + LASSERT(rc == 0); while (!list_empty(&lio->lis_active)) list_del_init(lio->lis_active.next); - EXIT; + EXIT; } static void lov_io_unlock(const struct lu_env *env, const struct cl_io_slice *ios) { - int rc; + int rc; - ENTRY; - rc = lov_io_call(env, cl2lov_io(env, ios), lov_io_unlock_wrapper); - LASSERT(rc == 0); - EXIT; + ENTRY; + rc = lov_io_call(env, cl2lov_io(env, ios), lov_io_unlock_wrapper); + LASSERT(rc == 0); + EXIT; } static int lov_io_read_ahead(const struct lu_env *env, @@ -943,7 +1023,7 @@ static int lov_io_read_ahead(const struct lu_env *env, stripe = lov_stripe_number(loo->lo_lsm, index, offset); r0 = lov_r0(loo, index); - if (unlikely(r0->lo_sub[stripe] == NULL)) + if (unlikely(!r0->lo_sub[stripe])) RETURN(-EIO); sub = lov_sub_get(env, lio, lov_comp_index(index, stripe)); @@ -1018,6 +1098,7 @@ static int lov_io_submit(const struct lu_env *env, struct lov_io_sub *sub; struct cl_page_list *plist = &lov_env_info(env)->lti_plist; struct cl_page *page; + struct cl_page *tmp; int index; int rc = 0; ENTRY; @@ -1030,7 +1111,12 @@ static int lov_io_submit(const struct lu_env *env, if (lov_page_is_empty(page)) { cl_page_list_move(&queue->c2_qout, qin, page); - cl_page_prep(env, ios->cis_io, page, crt); + /* + * it could only be mirror read to get here therefore + * the pages will be transient. We don't care about + * the return code of cl_page_prep() at all. + */ + (void) cl_page_prep(env, ios->cis_io, page, crt); cl_page_completion(env, page, crt, 0); continue; } @@ -1038,11 +1124,11 @@ static int lov_io_submit(const struct lu_env *env, cl_2queue_init(cl2q); cl_page_list_move(&cl2q->c2_qin, qin, page); - index = lov_page_index(page); - while (qin->pl_nr > 0) { - page = cl_page_list_first(qin); - if (index != lov_page_index(page)) - break; + index = page->cp_lov_index; + cl_page_list_for_each_safe(page, tmp, qin) { + /* this page is not on this stripe */ + if (index != page->cp_lov_index) + continue; cl_page_list_move(&cl2q->c2_qin, qin, page); } @@ -1075,7 +1161,7 @@ static int lov_io_commit_async(const struct lu_env *env, cl_commit_cbt cb) { struct cl_page_list *plist = &lov_env_info(env)->lti_plist; - struct lov_io *lio = cl2lov_io(env, ios); + struct lov_io *lio = cl2lov_io(env, ios); struct lov_io_sub *sub; struct cl_page *page; int rc = 0; @@ -1105,10 +1191,10 @@ static int lov_io_commit_async(const struct lu_env *env, cl_page_list_move(plist, queue, page); - index = lov_page_index(page); + index = page->cp_lov_index; while (queue->pl_nr > 0) { page = cl_page_list_first(queue); - if (index != lov_page_index(page)) + if (index != page->cp_lov_index) break; cl_page_list_move(plist, queue, page); @@ -1154,7 +1240,7 @@ static int lov_io_fault_start(const struct lu_env *env, fio = &ios->cis_io->u.ci_fault; lio = cl2lov_io(env, ios); - sub = lov_sub_get(env, lio, lov_page_index(fio->ft_page)); + sub = lov_sub_get(env, lio, fio->ft_page->cp_lov_index); sub->sub_io.u.ci_fault.ft_nob = fio->ft_nob; RETURN(lov_io_start(env, ios)); @@ -1181,25 +1267,25 @@ static void lov_io_fsync_end(const struct lu_env *env, } static const struct cl_io_operations lov_io_ops = { - .op = { - [CIT_READ] = { - .cio_fini = lov_io_fini, - .cio_iter_init = lov_io_rw_iter_init, - .cio_iter_fini = lov_io_iter_fini, - .cio_lock = lov_io_lock, - .cio_unlock = lov_io_unlock, - .cio_start = lov_io_start, - .cio_end = lov_io_end - }, - [CIT_WRITE] = { - .cio_fini = lov_io_fini, - .cio_iter_init = lov_io_rw_iter_init, - .cio_iter_fini = lov_io_iter_fini, - .cio_lock = lov_io_lock, - .cio_unlock = lov_io_unlock, - .cio_start = lov_io_start, - .cio_end = lov_io_end - }, + .op = { + [CIT_READ] = { + .cio_fini = lov_io_fini, + .cio_iter_init = lov_io_rw_iter_init, + .cio_iter_fini = lov_io_iter_fini, + .cio_lock = lov_io_lock, + .cio_unlock = lov_io_unlock, + .cio_start = lov_io_start, + .cio_end = lov_io_end + }, + [CIT_WRITE] = { + .cio_fini = lov_io_fini, + .cio_iter_init = lov_io_rw_iter_init, + .cio_iter_fini = lov_io_iter_fini, + .cio_lock = lov_io_lock, + .cio_unlock = lov_io_unlock, + .cio_start = lov_io_start, + .cio_end = lov_io_end + }, [CIT_SETATTR] = { .cio_fini = lov_io_fini, .cio_iter_init = lov_io_setattr_iter_init, @@ -1210,23 +1296,23 @@ static const struct cl_io_operations lov_io_ops = { .cio_end = lov_io_end }, [CIT_DATA_VERSION] = { - .cio_fini = lov_io_fini, - .cio_iter_init = lov_io_iter_init, - .cio_iter_fini = lov_io_iter_fini, - .cio_lock = lov_io_lock, - .cio_unlock = lov_io_unlock, - .cio_start = lov_io_start, - .cio_end = lov_io_data_version_end, + .cio_fini = lov_io_fini, + .cio_iter_init = lov_io_iter_init, + .cio_iter_fini = lov_io_iter_fini, + .cio_lock = lov_io_lock, + .cio_unlock = lov_io_unlock, + .cio_start = lov_io_start, + .cio_end = lov_io_data_version_end, + }, + [CIT_FAULT] = { + .cio_fini = lov_io_fini, + .cio_iter_init = lov_io_iter_init, + .cio_iter_fini = lov_io_iter_fini, + .cio_lock = lov_io_lock, + .cio_unlock = lov_io_unlock, + .cio_start = lov_io_fault_start, + .cio_end = lov_io_end }, - [CIT_FAULT] = { - .cio_fini = lov_io_fini, - .cio_iter_init = lov_io_iter_init, - .cio_iter_fini = lov_io_iter_fini, - .cio_lock = lov_io_lock, - .cio_unlock = lov_io_unlock, - .cio_start = lov_io_fault_start, - .cio_end = lov_io_end - }, [CIT_FSYNC] = { .cio_fini = lov_io_fini, .cio_iter_init = lov_io_iter_init, @@ -1252,7 +1338,7 @@ static const struct cl_io_operations lov_io_ops = { .cio_fini = lov_io_fini } }, - .cio_read_ahead = lov_io_read_ahead, + .cio_read_ahead = lov_io_read_ahead, .cio_submit = lov_io_submit, .cio_commit_async = lov_io_commit_async, }; @@ -1284,7 +1370,7 @@ static int lov_empty_io_submit(const struct lu_env *env, static void lov_empty_impossible(const struct lu_env *env, struct cl_io_slice *ios) { - LBUG(); + LBUG(); } #define LOV_EMPTY_IMPOSSIBLE ((void *)lov_empty_impossible) @@ -1293,37 +1379,37 @@ static void lov_empty_impossible(const struct lu_env *env, * An io operation vector for files without stripes. */ static const struct cl_io_operations lov_empty_io_ops = { - .op = { - [CIT_READ] = { - .cio_fini = lov_empty_io_fini, + .op = { + [CIT_READ] = { + .cio_fini = lov_empty_io_fini, #if 0 - .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, - .cio_lock = LOV_EMPTY_IMPOSSIBLE, - .cio_start = LOV_EMPTY_IMPOSSIBLE, - .cio_end = LOV_EMPTY_IMPOSSIBLE + .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, + .cio_lock = LOV_EMPTY_IMPOSSIBLE, + .cio_start = LOV_EMPTY_IMPOSSIBLE, + .cio_end = LOV_EMPTY_IMPOSSIBLE #endif - }, - [CIT_WRITE] = { - .cio_fini = lov_empty_io_fini, - .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, - .cio_lock = LOV_EMPTY_IMPOSSIBLE, - .cio_start = LOV_EMPTY_IMPOSSIBLE, - .cio_end = LOV_EMPTY_IMPOSSIBLE - }, - [CIT_SETATTR] = { - .cio_fini = lov_empty_io_fini, - .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, - .cio_lock = LOV_EMPTY_IMPOSSIBLE, - .cio_start = LOV_EMPTY_IMPOSSIBLE, - .cio_end = LOV_EMPTY_IMPOSSIBLE - }, - [CIT_FAULT] = { - .cio_fini = lov_empty_io_fini, - .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, - .cio_lock = LOV_EMPTY_IMPOSSIBLE, - .cio_start = LOV_EMPTY_IMPOSSIBLE, - .cio_end = LOV_EMPTY_IMPOSSIBLE - }, + }, + [CIT_WRITE] = { + .cio_fini = lov_empty_io_fini, + .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, + .cio_lock = LOV_EMPTY_IMPOSSIBLE, + .cio_start = LOV_EMPTY_IMPOSSIBLE, + .cio_end = LOV_EMPTY_IMPOSSIBLE + }, + [CIT_SETATTR] = { + .cio_fini = lov_empty_io_fini, + .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, + .cio_lock = LOV_EMPTY_IMPOSSIBLE, + .cio_start = LOV_EMPTY_IMPOSSIBLE, + .cio_end = LOV_EMPTY_IMPOSSIBLE + }, + [CIT_FAULT] = { + .cio_fini = lov_empty_io_fini, + .cio_iter_init = LOV_EMPTY_IMPOSSIBLE, + .cio_lock = LOV_EMPTY_IMPOSSIBLE, + .cio_start = LOV_EMPTY_IMPOSSIBLE, + .cio_end = LOV_EMPTY_IMPOSSIBLE + }, [CIT_FSYNC] = { .cio_fini = lov_empty_io_fini }, @@ -1344,9 +1430,10 @@ static const struct cl_io_operations lov_empty_io_ops = { int lov_io_init_composite(const struct lu_env *env, struct cl_object *obj, struct cl_io *io) { - struct lov_io *lio = lov_env_io(env); - struct lov_object *lov = cl2lov(obj); + struct lov_io *lio = lov_env_io(env); + struct lov_object *lov = cl2lov(obj); int result; + ENTRY; INIT_LIST_HEAD(&lio->lis_active); @@ -1430,7 +1517,8 @@ int lov_io_init_released(const struct lu_env *env, struct cl_object *obj, result = 1; break; case CIT_SETATTR: - /* the truncate to 0 is managed by MDT: + /* + * the truncate to 0 is managed by MDT: * - in open, for open O_TRUNC * - in setattr, for truncate */