-/* -*- mode: c; c-basic-offset: 8; indent-tabs-mode: nil; -*-
- * vim:expandtab:shiftwidth=8:tabstop=8:
- *
+/*
* GPL HEADER START
*
* DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
* GPL HEADER END
*/
/*
- * Copyright 2008 Sun Microsystems, Inc. All rights reserved.
+ * Copyright (c) 2008, 2010, Oracle and/or its affiliates. All rights reserved.
* Use is subject to license terms.
+ *
+ * Copyright (c) 2011, 2013, Intel Corporation.
*/
/*
* This file is part of Lustre, http://www.lustre.org/
* Implementation of cl_io for LOV layer.
*
* Author: Nikita Danilov <nikita.danilov@sun.com>
+ * Author: Jinshan Xiong <jinshan.xiong@whamcloud.com>
*/
#define DEBUG_SUBSYSTEM S_LOV
#include "lov_cl_internal.h"
-/** \addtogroup lov lov @{ */
+/** \addtogroup lov
+ * @{
+ */
-static void lov_sub_enter(struct lov_io_sub *sub)
+static inline void lov_sub_enter(struct lov_io_sub *sub)
{
- ENTRY;
- if (sub->sub_reenter++ == 0) {
- sub->sub_cookie = cl_env_reenter();
- cl_env_implant(sub->sub_env, &sub->sub_refcheck2);
- }
- EXIT;
+ sub->sub_reenter++;
}
-
-static void lov_sub_exit(struct lov_io_sub *sub)
+static inline void lov_sub_exit(struct lov_io_sub *sub)
{
- ENTRY;
- if (--sub->sub_reenter == 0) {
- cl_env_unplant(sub->sub_env, &sub->sub_refcheck2);
- cl_env_reexit(sub->sub_cookie);
- }
- EXIT;
+ sub->sub_reenter--;
}
static void lov_io_sub_fini(const struct lu_env *env, struct lov_io *lio,
static void lov_io_sub_inherit(struct cl_io *io, struct lov_io *lio,
int stripe, loff_t start, loff_t end)
{
- struct lov_stripe_md *lsm = lov_r0(lio->lis_object)->lo_lsm;
- struct cl_io *parent = lio->lis_cl.cis_io;
+ struct lov_stripe_md *lsm = lio->lis_object->lo_lsm;
+ struct cl_io *parent = lio->lis_cl.cis_io;
switch(io->ci_type) {
- case CIT_TRUNC: {
- size_t new_size = parent->u.ci_truncate.tr_size;
-
- new_size = lov_size_to_stripe(lsm, new_size, stripe);
- io->u.ci_truncate.tr_capa = parent->u.ci_truncate.tr_capa;
- io->u.ci_truncate.tr_size = new_size;
+ case CIT_SETATTR: {
+ io->u.ci_setattr.sa_attr = parent->u.ci_setattr.sa_attr;
+ io->u.ci_setattr.sa_valid = parent->u.ci_setattr.sa_valid;
+ io->u.ci_setattr.sa_capa = parent->u.ci_setattr.sa_capa;
+ if (cl_io_is_trunc(io)) {
+ loff_t new_size = parent->u.ci_setattr.sa_attr.lvb_size;
+
+ new_size = lov_size_to_stripe(lsm, new_size, stripe);
+ io->u.ci_setattr.sa_attr.lvb_size = new_size;
+ }
break;
}
case CIT_FAULT: {
io->u.ci_fault.ft_index = cl_index(obj, off);
break;
}
- case CIT_READ:
- case CIT_WRITE: {
- io->u.ci_rw.crw_pos = start;
- io->u.ci_rw.crw_count = end - start;
+ case CIT_FSYNC: {
+ io->u.ci_fsync.fi_start = start;
+ io->u.ci_fsync.fi_end = end;
+ io->u.ci_fsync.fi_capa = parent->u.ci_fsync.fi_capa;
+ io->u.ci_fsync.fi_fid = parent->u.ci_fsync.fi_fid;
+ io->u.ci_fsync.fi_mode = parent->u.ci_fsync.fi_mode;
+ break;
+ }
+ case CIT_READ:
+ case CIT_WRITE: {
+ io->u.ci_wr.wr_sync = cl_io_is_sync_write(parent);
+ if (cl_io_is_append(parent)) {
+ io->u.ci_wr.wr_append = 1;
+ } else {
+ io->u.ci_rw.crw_pos = start;
+ io->u.ci_rw.crw_count = end - start;
+ }
break;
}
default:
sub->sub_io_initialized = 0;
sub->sub_borrowed = 0;
- /*
- * First sub-io. Use ->lis_single_subio and current environment, to
- * avoid dynamic allocation.
- */
- if (lio->lis_active_subios == 0) {
- sub->sub_io = &lio->lis_single_subio;
- lio->lis_single_subio_index = stripe;
- sub->sub_env = cl_env_get(&sub->sub_refcheck);
- LASSERT(sub->sub_env == env);
- } else if (lio->lis_mem_frozen) {
- LASSERT(mutex_is_locked(&ld->ld_mutex));
+ if (lio->lis_mem_frozen) {
+ LASSERT(mutex_is_locked(&ld->ld_mutex));
sub->sub_io = &ld->ld_emrg[stripe]->emrg_subio;
sub->sub_env = ld->ld_emrg[stripe]->emrg_env;
sub->sub_borrowed = 1;
cookie = cl_env_reenter();
sub->sub_env = cl_env_get(&sub->sub_refcheck);
cl_env_reexit(cookie);
-
- OBD_ALLOC_PTR(sub->sub_io);
if (IS_ERR(sub->sub_env))
result = PTR_ERR(sub->sub_env);
- else if (sub->sub_io == NULL)
- result = -ENOMEM;
+
+ if (result == 0) {
+ /*
+ * First sub-io. Use ->lis_single_subio to
+ * avoid dynamic allocation.
+ */
+ if (lio->lis_active_subios == 0) {
+ sub->sub_io = &lio->lis_single_subio;
+ lio->lis_single_subio_index = stripe;
+ } else {
+ OBD_ALLOC_PTR(sub->sub_io);
+ if (sub->sub_io == NULL)
+ result = -ENOMEM;
+ }
+ }
}
if (result == 0) {
sub_io->ci_lockreq = io->ci_lockreq;
sub_io->ci_type = io->ci_type;
sub_io->ci_no_srvlock = io->ci_no_srvlock;
+ sub_io->ci_noatime = io->ci_noatime;
lov_sub_enter(sub);
result = cl_io_sub_init(sub->sub_env, sub_io,
*
*/
-static int lov_page_stripe(const struct cl_page *page)
+int lov_page_stripe(const struct cl_page *page)
{
- struct lovsub_object *subobj;
+ struct lovsub_object *subobj;
+ const struct cl_page_slice *slice;
+ ENTRY;
- ENTRY;
- subobj = lu2lovsub(
- lu_object_locate(page->cp_child->cp_obj->co_lu.lo_header,
- &lovsub_device_type));
- LASSERT(subobj != NULL);
- RETURN(subobj->lso_index);
+ slice = cl_page_at(page, &lovsub_device_type);
+ LASSERT(slice != NULL);
+ LASSERT(slice->cpl_obj != NULL);
+
+ subobj = cl2lovsub(slice->cpl_obj);
+ RETURN(subobj->lso_index);
}
struct lov_io_sub *lov_page_subio(const struct lu_env *env, struct lov_io *lio,
const struct cl_page_slice *slice)
{
- struct lov_stripe_md *lsm = lov_r0(lio->lis_object)->lo_lsm;
- struct cl_page *page = slice->cpl_page;
- int stripe;
+ struct lov_stripe_md *lsm = lio->lis_object->lo_lsm;
+ struct cl_page *page = slice->cpl_page;
+ int stripe;
LASSERT(lio->lis_cl.cis_io != NULL);
LASSERT(cl2lov(slice->cpl_obj) == lio->lis_object);
static int lov_io_subio_init(const struct lu_env *env, struct lov_io *lio,
struct cl_io *io)
{
- struct lov_object *lov = lio->lis_object;
- struct lov_stripe_md *lsm = lov_r0(lov)->lo_lsm;
- int result;
+ struct lov_stripe_md *lsm;
+ int result;
+ ENTRY;
- LASSERT(lio->lis_object != NULL);
- ENTRY;
+ LASSERT(lio->lis_object != NULL);
+ lsm = lio->lis_object->lo_lsm;
/*
* Need to be optimized, we can't afford to allocate a piece of memory
* when writing a page. -jay
*/
- OBD_ALLOC(lio->lis_subs,
- lsm->lsm_stripe_count * sizeof lio->lis_subs[0]);
+ OBD_ALLOC_LARGE(lio->lis_subs,
+ lsm->lsm_stripe_count * sizeof lio->lis_subs[0]);
if (lio->lis_subs != NULL) {
lio->lis_nr_subios = lio->lis_stripe_count;
lio->lis_single_subio_index = -1;
static void lov_io_slice_init(struct lov_io *lio,
struct lov_object *obj, struct cl_io *io)
{
- struct lov_stripe_md *lsm = lov_r0(obj)->lo_lsm;
+ ENTRY;
- LASSERT(lsm != NULL);
- ENTRY;
+ io->ci_result = 0;
+ lio->lis_object = obj;
- io->ci_result = 0;
- lio->lis_object = obj;
- lio->lis_stripe_count = lsm->lsm_stripe_count;
+ LASSERT(obj->lo_lsm != NULL);
+ lio->lis_stripe_count = obj->lo_lsm->lsm_stripe_count;
switch (io->ci_type) {
case CIT_READ:
}
break;
- case CIT_TRUNC:
- lio->lis_pos = io->u.ci_truncate.tr_size;
+ 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;
break;
}
+ case CIT_FSYNC: {
+ lio->lis_pos = io->u.ci_fsync.fi_start;
+ lio->lis_endpos = io->u.ci_fsync.fi_end;
+ break;
+ }
+
case CIT_MISC:
lio->lis_pos = 0;
lio->lis_endpos = OBD_OBJECT_EOF;
static void lov_io_fini(const struct lu_env *env, const struct cl_io_slice *ios)
{
- struct lov_io *lio = cl2lov_io(env, ios);
- int i;
+ struct lov_io *lio = cl2lov_io(env, ios);
+ struct lov_object *lov = cl2lov(ios->cis_obj);
+ int i;
ENTRY;
if (lio->lis_subs != NULL) {
for (i = 0; i < lio->lis_nr_subios; i++)
lov_io_sub_fini(env, lio, &lio->lis_subs[i]);
- OBD_FREE(lio->lis_subs,
+ OBD_FREE_LARGE(lio->lis_subs,
lio->lis_nr_subios * sizeof lio->lis_subs[0]);
lio->lis_nr_subios = 0;
}
- EXIT;
+
+ LASSERT(cfs_atomic_read(&lov->lo_active_ios) > 0);
+ if (cfs_atomic_dec_and_test(&lov->lo_active_ios))
+ wake_up_all(&lov->lo_waitq);
+ EXIT;
}
static obd_off lov_offset_mod(obd_off val, int delta)
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_stripe_md *lsm = lov_r0(lio->lis_object)->lo_lsm;
+ struct lov_io *lio = cl2lov_io(env, ios);
+ struct lov_stripe_md *lsm = lio->lis_object->lo_lsm;
struct lov_io_sub *sub;
obd_off endpos;
obd_off start;
start, end);
rc = cl_io_iter_init(sub->sub_env, sub->sub_io);
lov_sub_put(sub);
- CDEBUG(D_VFSTRACE, "shrink: %i [%llu, %llu)\n",
+ CDEBUG(D_VFSTRACE, "shrink: %d ["LPU64", "LPU64")\n",
stripe, start, end);
} else
rc = PTR_ERR(sub);
if (!rc)
- list_add_tail(&sub->sub_linkage, &lio->lis_active);
+ cfs_list_add_tail(&sub->sub_linkage, &lio->lis_active);
else
break;
}
static int lov_io_rw_iter_init(const struct lu_env *env,
const struct cl_io_slice *ios)
{
- struct lov_io *lio = cl2lov_io(env, ios);
- struct cl_io *io = ios->cis_io;
- struct lov_stripe_md *lsm = lov_r0(cl2lov(ios->cis_obj))->lo_lsm;
+ struct lov_io *lio = cl2lov_io(env, ios);
+ struct cl_io *io = ios->cis_io;
+ struct lov_stripe_md *lsm = lio->lis_object->lo_lsm;
loff_t start = io->u.ci_rw.crw_pos;
loff_t next;
- int ssize = lsm->lsm_stripe_size;
+ unsigned long ssize = lsm->lsm_stripe_size;
LASSERT(io->ci_type == CIT_READ || io->ci_type == CIT_WRITE);
ENTRY;
/* fast path for common case. */
if (lio->lis_nr_subios != 1 && !cl_io_is_append(io)) {
- do_div(start, ssize);
- next = (start + 1) * ssize;
- if (next <= start * ssize)
- next = ~0ull;
+ lov_do_div64(start, ssize);
+ next = (start + 1) * ssize;
+ if (next <= start * ssize)
+ next = ~0ull;
io->ci_continue = next < lio->lis_io_endpos;
io->u.ci_rw.crw_count = min_t(loff_t, lio->lis_io_endpos,
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: %llu chunk: [%llu, %llu) %llu\n",
- (__u64)start, lio->lis_pos, lio->lis_endpos,
- (__u64)lio->lis_io_endpos);
- }
- /*
- * XXX The following call should be optimized: we know, that
- * [lio->lis_pos, lio->lis_endpos) intersects with exactly one stripe.
- */
- RETURN(lov_io_iter_init(env, ios));
+ CDEBUG(D_VFSTRACE, "stripe: "LPU64" chunk: ["LPU64", "LPU64") "
+ LPU64"\n", (__u64)start, lio->lis_pos, lio->lis_endpos,
+ (__u64)lio->lis_io_endpos);
+ }
+ /*
+ * XXX The following call should be optimized: we know, that
+ * [lio->lis_pos, lio->lis_endpos) intersects with exactly one stripe.
+ */
+ RETURN(lov_io_iter_init(env, ios));
}
static int lov_io_call(const struct lu_env *env, struct lov_io *lio,
int (*iofunc)(const struct lu_env *, struct cl_io *))
{
+ struct cl_io *parent = lio->lis_cl.cis_io;
struct lov_io_sub *sub;
int rc = 0;
ENTRY;
- list_for_each_entry(sub, &lio->lis_active, sub_linkage) {
+ cfs_list_for_each_entry(sub, &lio->lis_active, sub_linkage) {
lov_sub_enter(sub);
rc = iofunc(sub->sub_env, sub->sub_io);
lov_sub_exit(sub);
if (rc)
break;
- }
- RETURN(rc);
+
+ if (parent->ci_result == 0)
+ parent->ci_result = sub->sub_io->ci_result;
+ }
+ RETURN(rc);
}
static int lov_io_lock(const struct lu_env *env, const struct cl_io_slice *ios)
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);
+ while (!cfs_list_empty(&lio->lis_active))
+ cfs_list_del_init(lio->lis_active.next);
EXIT;
}
EXIT;
}
-
-static struct cl_page_list *lov_io_submit_qin(struct lov_device *ld,
- struct cl_page_list *qin,
- int idx, int alloc)
-{
- return alloc ? &qin[idx] : &ld->ld_emrg[idx]->emrg_page_list;
-}
-
/**
* lov implementation of cl_operations::cio_submit() method. It takes a list
* of pages in \a queue, splits it into per-stripe sub-lists, invokes
* lov_device::ld_mutex mutex.
*/
static int lov_io_submit(const struct lu_env *env,
- const struct cl_io_slice *ios,
- enum cl_req_type crt, struct cl_2queue *queue,
- enum cl_req_priority priority)
-{
- struct lov_io *lio = cl2lov_io(env, ios);
- struct lov_object *obj = lio->lis_object;
- struct lov_device *ld = lu2lov_dev(lov2cl(obj)->co_lu.lo_dev);
- struct cl_page_list *qin = &queue->c2_qin;
- struct cl_2queue *cl2q = &lov_env_info(env)->lti_cl2q;
- struct cl_page_list *stripes_qin = NULL;
- struct cl_page *page;
- struct cl_page *tmp;
- int stripe;
+ const struct cl_io_slice *ios,
+ enum cl_req_type crt, struct cl_2queue *queue)
+{
+ struct cl_page_list *qin = &queue->c2_qin;
+ struct lov_io *lio = cl2lov_io(env, ios);
+ struct lov_io_sub *sub;
+ struct cl_page_list *plist = &lov_env_info(env)->lti_plist;
+ struct cl_page *page;
+ int stripe;
+ int rc = 0;
+ ENTRY;
-#define QIN(stripe) lov_io_submit_qin(ld, stripes_qin, stripe, alloc)
-
- int rc = 0;
- int alloc =
-#if defined(__KERNEL__) && defined(__linux__)
- !(current->flags & PF_MEMALLOC);
-#else
- 1;
-#endif
- ENTRY;
if (lio->lis_active_subios == 1) {
int idx = lio->lis_single_subio_index;
- struct lov_io_sub *sub;
LASSERT(idx < lio->lis_nr_subios);
sub = lov_sub_get(env, lio, idx);
LASSERT(!IS_ERR(sub));
LASSERT(sub->sub_io == &lio->lis_single_subio);
rc = cl_io_submit_rw(sub->sub_env, sub->sub_io,
- crt, queue, priority);
+ crt, queue);
lov_sub_put(sub);
RETURN(rc);
}
LASSERT(lio->lis_subs != NULL);
- if (alloc) {
- OBD_ALLOC(stripes_qin,
- sizeof(*stripes_qin) * lio->lis_nr_subios);
- if (stripes_qin == NULL)
- RETURN(-ENOMEM);
-
- for (stripe = 0; stripe < lio->lis_nr_subios; stripe++)
- cl_page_list_init(&stripes_qin[stripe]);
- } else {
- /*
- * If we get here, it means pageout & swap doesn't help.
- * In order to not make things worse, even don't try to
- * allocate the memory with __GFP_NOWARN. -jay
- */
- mutex_lock(&ld->ld_mutex);
- lio->lis_mem_frozen = 1;
- }
-
- cl_2queue_init(cl2q);
- cl_page_list_for_each_safe(page, tmp, qin) {
- stripe = lov_page_stripe(page);
- cl_page_list_move(QIN(stripe), qin, page);
- }
-
- for (stripe = 0; stripe < lio->lis_nr_subios; stripe++) {
- struct lov_io_sub *sub;
- struct cl_page_list *sub_qin = QIN(stripe);
-
- if (list_empty(&sub_qin->pl_pages))
- continue;
-
- cl_page_list_splice(sub_qin, &cl2q->c2_qin);
- sub = lov_sub_get(env, lio, stripe);
- if (!IS_ERR(sub)) {
- rc = cl_io_submit_rw(sub->sub_env, sub->sub_io,
- crt, cl2q, priority);
- lov_sub_put(sub);
- } else
- rc = PTR_ERR(sub);
- cl_page_list_splice(&cl2q->c2_qin, &queue->c2_qin);
- cl_page_list_splice(&cl2q->c2_qout, &queue->c2_qout);
- if (rc != 0)
- break;
- }
- for (stripe = 0; stripe < lio->lis_nr_subios; stripe++) {
- struct cl_page_list *sub_qin = QIN(stripe);
-
- if (list_empty(&sub_qin->pl_pages))
- continue;
-
- cl_page_list_splice(sub_qin, qin);
- }
-
- if (alloc) {
- OBD_FREE(stripes_qin,
- sizeof(*stripes_qin) * lio->lis_nr_subios);
- } else {
- int i;
-
- for (i = 0; i < lio->lis_nr_subios; i++) {
- struct cl_io *cio = lio->lis_subs[i].sub_io;
-
- if (cio && cio == &ld->ld_emrg[i]->emrg_subio)
- lov_io_sub_fini(env, lio, &lio->lis_subs[i]);
- }
- lio->lis_mem_frozen = 0;
- mutex_unlock(&ld->ld_mutex);
- }
+ cl_page_list_init(plist);
+ while (qin->pl_nr > 0) {
+ struct cl_2queue *cl2q = &lov_env_info(env)->lti_cl2q;
- RETURN(rc);
-#undef QIN
-}
+ cl_2queue_init(cl2q);
-static int lov_io_prepare_write(const struct lu_env *env,
- const struct cl_io_slice *ios,
- const struct cl_page_slice *slice,
- unsigned from, unsigned to)
-{
- struct lov_io *lio = cl2lov_io(env, ios);
- struct cl_page *sub_page = lov_sub_page(slice);
- struct lov_io_sub *sub;
- int result;
+ page = cl_page_list_first(qin);
+ cl_page_list_move(&cl2q->c2_qin, qin, page);
- ENTRY;
- sub = lov_page_subio(env, lio, slice);
- if (!IS_ERR(sub)) {
- result = cl_io_prepare_write(sub->sub_env, sub->sub_io,
- sub_page, from, to);
- lov_sub_put(sub);
- } else
- result = PTR_ERR(sub);
- RETURN(result);
-}
+ stripe = lov_page_stripe(page);
+ while (qin->pl_nr > 0) {
+ page = cl_page_list_first(qin);
+ if (stripe != lov_page_stripe(page))
+ break;
-static int lov_io_commit_write(const struct lu_env *env,
- const struct cl_io_slice *ios,
- const struct cl_page_slice *slice,
- unsigned from, unsigned to)
-{
- struct lov_io *lio = cl2lov_io(env, ios);
- struct cl_page *sub_page = lov_sub_page(slice);
- struct lov_io_sub *sub;
- int result;
+ cl_page_list_move(&cl2q->c2_qin, qin, page);
+ }
- ENTRY;
- sub = lov_page_subio(env, lio, slice);
- if (!IS_ERR(sub)) {
- result = cl_io_commit_write(sub->sub_env, sub->sub_io,
- sub_page, from, to);
- lov_sub_put(sub);
- } else
- result = PTR_ERR(sub);
- RETURN(result);
+ sub = lov_sub_get(env, lio, stripe);
+ if (!IS_ERR(sub)) {
+ rc = cl_io_submit_rw(sub->sub_env, sub->sub_io,
+ crt, cl2q);
+ lov_sub_put(sub);
+ } else {
+ rc = PTR_ERR(sub);
+ }
+
+ cl_page_list_splice(&cl2q->c2_qin, plist);
+ cl_page_list_splice(&cl2q->c2_qout, &queue->c2_qout);
+ cl_2queue_fini(env, cl2q);
+
+ if (rc != 0)
+ break;
+ }
+
+ cl_page_list_splice(plist, qin);
+ cl_page_list_fini(env, plist);
+
+ RETURN(rc);
+}
+
+static int lov_io_commit_async(const struct lu_env *env,
+ const struct cl_io_slice *ios,
+ struct cl_page_list *queue, int from, int to,
+ 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_sub *sub;
+ struct cl_page *page;
+ int rc = 0;
+ ENTRY;
+
+ if (lio->lis_active_subios == 1) {
+ int idx = lio->lis_single_subio_index;
+
+ LASSERT(idx < lio->lis_nr_subios);
+ sub = lov_sub_get(env, lio, idx);
+ LASSERT(!IS_ERR(sub));
+ LASSERT(sub->sub_io == &lio->lis_single_subio);
+ rc = cl_io_commit_async(sub->sub_env, sub->sub_io, queue,
+ from, to, cb);
+ lov_sub_put(sub);
+ RETURN(rc);
+ }
+
+ LASSERT(lio->lis_subs != NULL);
+
+ cl_page_list_init(plist);
+ while (queue->pl_nr > 0) {
+ int stripe_to = to;
+ int stripe;
+
+ LASSERT(plist->pl_nr == 0);
+ page = cl_page_list_first(queue);
+ cl_page_list_move(plist, queue, page);
+
+ stripe = lov_page_stripe(page);
+ while (queue->pl_nr > 0) {
+ page = cl_page_list_first(queue);
+ if (stripe != lov_page_stripe(page))
+ break;
+
+ cl_page_list_move(plist, queue, page);
+ }
+
+ if (queue->pl_nr > 0) /* still has more pages */
+ stripe_to = PAGE_SIZE;
+
+ sub = lov_sub_get(env, lio, stripe);
+ if (!IS_ERR(sub)) {
+ rc = cl_io_commit_async(sub->sub_env, sub->sub_io,
+ plist, from, stripe_to, cb);
+ lov_sub_put(sub);
+ } else {
+ rc = PTR_ERR(sub);
+ break;
+ }
+
+ if (plist->pl_nr > 0) /* short write */
+ break;
+
+ from = 0;
+ }
+
+ /* for error case, add the page back into the qin list */
+ LASSERT(ergo(rc == 0, plist->pl_nr == 0));
+ while (plist->pl_nr > 0) {
+ /* error occurred, add the uncommitted pages back into queue */
+ page = cl_page_list_last(plist);
+ cl_page_list_move_head(queue, plist, page);
+ }
+
+ RETURN(rc);
}
static int lov_io_fault_start(const struct lu_env *env,
RETURN(lov_io_start(env, ios));
}
+static void lov_io_fsync_end(const struct lu_env *env,
+ const struct cl_io_slice *ios)
+{
+ struct lov_io *lio = cl2lov_io(env, ios);
+ struct lov_io_sub *sub;
+ unsigned int *written = &ios->cis_io->u.ci_fsync.fi_nr_written;
+ ENTRY;
+
+ *written = 0;
+ cfs_list_for_each_entry(sub, &lio->lis_active, sub_linkage) {
+ struct cl_io *subio = sub->sub_io;
+
+ lov_sub_enter(sub);
+ lov_io_end_wrapper(sub->sub_env, subio);
+ lov_sub_exit(sub);
+
+ if (subio->ci_result == 0)
+ *written += subio->u.ci_fsync.fi_nr_written;
+ }
+ RETURN_EXIT;
+}
+
static const struct cl_io_operations lov_io_ops = {
.op = {
[CIT_READ] = {
.cio_start = lov_io_start,
.cio_end = lov_io_end
},
- [CIT_TRUNC] = {
+ [CIT_SETATTR] = {
.cio_fini = lov_io_fini,
.cio_iter_init = lov_io_iter_init,
.cio_iter_fini = lov_io_iter_fini,
.cio_start = lov_io_fault_start,
.cio_end = lov_io_end
},
- [CIT_MISC] = {
- .cio_fini = lov_io_fini
- }
- },
- .req_op = {
- [CRT_READ] = {
- .cio_submit = lov_io_submit
- },
- [CRT_WRITE] = {
- .cio_submit = lov_io_submit
- }
- },
- .cio_prepare_write = lov_io_prepare_write,
- .cio_commit_write = lov_io_commit_write
+ [CIT_FSYNC] = {
+ .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_fsync_end
+ },
+ [CIT_MISC] = {
+ .cio_fini = lov_io_fini
+ }
+ },
+ .cio_submit = lov_io_submit,
+ .cio_commit_async = lov_io_commit_async,
};
/*****************************************************************************
static void lov_empty_io_fini(const struct lu_env *env,
const struct cl_io_slice *ios)
{
- ENTRY;
- EXIT;
+ struct lov_object *lov = cl2lov(ios->cis_obj);
+ ENTRY;
+
+ if (cfs_atomic_dec_and_test(&lov->lo_active_ios))
+ wake_up_all(&lov->lo_waitq);
+ EXIT;
}
static void lov_empty_impossible(const struct lu_env *env,
static const struct cl_io_operations lov_empty_io_ops = {
.op = {
[CIT_READ] = {
-#if 0
.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_start = LOV_EMPTY_IMPOSSIBLE,
.cio_end = LOV_EMPTY_IMPOSSIBLE
},
- [CIT_TRUNC] = {
+ [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_MISC] = {
- .cio_fini = lov_empty_io_fini
- }
- },
- .req_op = {
- [CRT_READ] = {
- .cio_submit = LOV_EMPTY_IMPOSSIBLE
- },
- [CRT_WRITE] = {
- .cio_submit = LOV_EMPTY_IMPOSSIBLE
- }
- },
- .cio_commit_write = LOV_EMPTY_IMPOSSIBLE
+ [CIT_FSYNC] = {
+ .cio_fini = lov_empty_io_fini
+ },
+ [CIT_MISC] = {
+ .cio_fini = lov_empty_io_fini
+ }
+ },
+ .cio_submit = LOV_EMPTY_IMPOSSIBLE,
+ .cio_commit_async = LOV_EMPTY_IMPOSSIBLE
};
int lov_io_init_raid0(const struct lu_env *env, struct cl_object *obj,
CFS_INIT_LIST_HEAD(&lio->lis_active);
lov_io_slice_init(lio, lov, io);
if (io->ci_result == 0) {
- LASSERT(lov_r0(lov)->lo_lsm != NULL);
io->ci_result = lov_io_subio_init(env, lio, io);
- if (io->ci_result == 0)
+ if (io->ci_result == 0) {
cl_io_slice_add(io, &lio->lis_cl, obj, &lov_io_ops);
+ cfs_atomic_inc(&lov->lo_active_ios);
+ }
}
RETURN(io->ci_result);
}
int lov_io_init_empty(const struct lu_env *env, struct cl_object *obj,
struct cl_io *io)
{
- struct lov_io *lio = lov_env_io(env);
- int result;
-
- ENTRY;
- switch (io->ci_type) {
- default:
- LBUG();
- case CIT_MISC:
- case CIT_READ:
- result = 0;
- break;
+ struct lov_object *lov = cl2lov(obj);
+ struct lov_io *lio = lov_env_io(env);
+ int result;
+ ENTRY;
+
+ lio->lis_object = lov;
+ switch (io->ci_type) {
+ default:
+ LBUG();
+ case CIT_MISC:
+ case CIT_READ:
+ result = 0;
+ break;
+ case CIT_FSYNC:
+ case CIT_SETATTR:
+ result = +1;
+ break;
case CIT_WRITE:
- case CIT_TRUNC:
result = -EBADF;
break;
case CIT_FAULT:
PFID(lu_object_fid(&obj->co_lu)));
break;
}
- if (result == 0)
+ if (result == 0) {
cl_io_slice_add(io, &lio->lis_cl, obj, &lov_empty_io_ops);
- io->ci_result = result;
- RETURN(result != 0);
+ cfs_atomic_inc(&lov->lo_active_ios);
+ }
+
+ io->ci_result = result < 0 ? result : 0;
+ RETURN(result != 0);
+}
+
+int lov_io_init_released(const struct lu_env *env, struct cl_object *obj,
+ struct cl_io *io)
+{
+ struct lov_object *lov = cl2lov(obj);
+ struct lov_io *lio = lov_env_io(env);
+ int result;
+ ENTRY;
+
+ LASSERT(lov->lo_lsm != NULL);
+ lio->lis_object = lov;
+
+ switch (io->ci_type) {
+ default:
+ LASSERTF(0, "invalid type %d\n", io->ci_type);
+ case CIT_MISC:
+ case CIT_FSYNC:
+ result = 1;
+ break;
+ case CIT_SETATTR:
+ /* the truncate to 0 is managed by MDT:
+ * - in open, for open O_TRUNC
+ * - in setattr, for truncate
+ */
+ /* the truncate is for size > 0 so triggers a restore */
+ if (cl_io_is_trunc(io))
+ io->ci_restore_needed = 1;
+ result = -ENODATA;
+ break;
+ case CIT_READ:
+ case CIT_WRITE:
+ case CIT_FAULT:
+ io->ci_restore_needed = 1;
+ result = -ENODATA;
+ break;
+ }
+ if (result == 0) {
+ cl_io_slice_add(io, &lio->lis_cl, obj, &lov_empty_io_ops);
+ cfs_atomic_inc(&lov->lo_active_ios);
+ }
+
+ io->ci_result = result < 0 ? result : 0;
+ RETURN(result != 0);
}
-
/** @} lov */