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;
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;
}
!lov_r0(lov, index)->lo_sub[stripe]))
RETURN(-EIO);
- LASSERTF(is_index_within_mirror(lov, index, lio->lis_mirror_index),
+ LASSERTF(ergo(lov_is_flr(lov),
+ is_index_within_mirror(lov, index,
+ lio->lis_mirror_index)),
DFID "iot = %d, index = %d, mirror = %d\n",
PFID(lu_object_fid(lov2lu(lov))), io->ci_type, index,
lio->lis_mirror_index);
/* 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;
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;
+ sub_io->ci_tried_all_mirrors = io->ci_tried_all_mirrors;
result = cl_io_sub_init(sub->sub_env, sub_io, io->ci_type, sub_obj);
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);
out:
if (rc < 0)
sub = ERR_PTR(rc);
+ else
+ sub->sub_io.ci_noquota = lio->lis_cl.cis_io->ci_noquota;
RETURN(sub);
}
* 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)
{
cl_io_is_mkwrite(io)))
RETURN(0);
- /* FLR: check if it needs to send a write intent RPC to server.
+ /*
+ * 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
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) {
if (!lu_extent_is_overlapped(ext, lle->lle_extent))
continue;
- ext->e_start = MIN(ext->e_start, lle->lle_extent->e_start);
- ext->e_end = MAX(ext->e_end, lle->lle_extent->e_end);
+ ext->e_start = min(ext->e_start, lle->lle_extent->e_start);
+ ext->e_end = max(ext->e_end, lle->lle_extent->e_end);
++count;
}
if (count == 0) {
ENTRY;
if (!lov_is_flr(obj)) {
- LASSERT(comp->lo_preferred_mirror == 0);
- lio->lis_mirror_index = comp->lo_preferred_mirror;
+ /* only locks/pages are manipulated for CIT_MISC op, no
+ * cl_io_loop() will be called, don't check/set mirror info.
+ */
+ if (io->ci_type != CIT_MISC) {
+ LASSERT(comp->lo_preferred_mirror == 0);
+ lio->lis_mirror_index = comp->lo_preferred_mirror;
+ }
io->ci_ndelay = 0;
RETURN(0);
}
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
+ /*
+ * 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 */
+ * has been changed
+ */
RETURN(-ESTALE);
}
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], in truncate,
+ * the byte in the size position is accessed.
+ */
+ io->ci_write_intent.e_start = 0;
+ io->ci_write_intent.e_end =
+ io->u.ci_setattr.sa_attr.lvb_size + 1;
+ }
/* stop cl_io_init() loop */
RETURN(1);
}
found = true;
break;
}
- }
-
+ } /* each component of the mirror */
if (found) {
index = (index + i) % comp->lo_mirror_count;
break;
}
- }
+ } /* each mirror */
+
if (i == comp->lo_mirror_count) {
CERROR(DFID": failed to find a component covering "
"I/O region at %llu\n",
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. */
- ++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 */
+ * a while before trying again.
+ */
+ if (io->ci_ndelay && io->ci_ndelay_tried > 0 &&
+ (io->ci_ndelay_tried % comp->lo_mirror_count == 0)) {
+ schedule_timeout_interruptible(cfs_time_seconds(1) / 100);
if (signal_pending(current))
RETURN(-EINTR);
- /* reset retry counter */
- io->ci_ndelay_tried = 1;
+ /**
+ * we'd set ci_tried_all_mirrors to turn off fast mirror
+ * switching for read after we've tried all mirrors several
+ * rounds.
+ */
+ io->ci_tried_all_mirrors = io->ci_ndelay_tried %
+ (comp->lo_mirror_count * 4) == 0;
}
+ ++io->ci_ndelay_tried;
CDEBUG(D_VFSTRACE, "use %sdelayed RPC state for this IO\n",
io->ci_ndelay ? "non-" : "");
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;
io->ci_result = 0;
lio->lis_object = obj;
- LASSERT(obj->lo_lsm != NULL);
-
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);
}
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_fallocate(io)) {
+ lio->lis_pos = io->u.ci_setattr.sa_falloc_offset;
+ lio->lis_endpos = io->u.ci_setattr.sa_falloc_end;
+ } else if (cl_io_is_trunc(io)) {
+ lio->lis_pos = io->u.ci_setattr.sa_attr.lvb_size;
+ lio->lis_endpos = OBD_OBJECT_EOF;
+ } 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;
break;
}
+ case CIT_LSEEK: {
+ lio->lis_pos = io->u.ci_lseek.ls_start;
+ lio->lis_endpos = OBD_OBJECT_EOF;
+ break;
+ }
+
case CIT_GLIMPSE:
lio->lis_pos = 0;
lio->lis_endpos = OBD_OBJECT_EOF;
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();
+ }
+
+ /*
+ * CIT_MISC + ci_ignore_layout can identify the I/O from the OSC layer,
+ * it won't care/access lov layout related info.
+ */
+ if (io->ci_ignore_layout && io->ci_type == CIT_MISC)
+ GOTO(out, result = 0);
- default:
- LBUG();
- }
+ LASSERT(obj->lo_lsm != NULL);
result = lov_io_mirror_init(lio, obj, io);
if (result)
/* check if it needs to instantiate layout */
if (!(io->ci_type == CIT_WRITE || cl_io_is_mkwrite(io) ||
+ cl_io_is_fallocate(io) ||
(cl_io_is_trunc(io) && io->u.ci_setattr.sa_attr.lvb_size > 0)))
GOTO(out, result = 0);
- io->ci_write_intent.e_start = lio->lis_pos;
- io->ci_write_intent.e_end = lio->lis_endpos;
-
- ext = io->ci_write_intent;
- /* 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;
break;
}
if (io->ci_need_write_intent && io->ci_designated_mirror > 0) {
- /* REINT_SYNC RPC has already tried to instantiate all of the
+ /*
+ * 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 */
+ * 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));
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_falloc_mode =
+ parent->u.ci_setattr.sa_falloc_mode;
io->u.ci_setattr.sa_stripe_index = stripe;
io->u.ci_setattr.sa_parent_fid =
parent->u.ci_setattr.sa_parent_fid;
+ /* For SETATTR(fallocate) pass the subtype to lower IO */
+ io->u.ci_setattr.sa_subtype = parent->u.ci_setattr.sa_subtype;
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, index, new_size,
stripe);
io->u.ci_setattr.sa_attr.lvb_size = new_size;
+ } else if (cl_io_is_fallocate(io)) {
+ io->u.ci_setattr.sa_falloc_offset = start;
+ io->u.ci_setattr.sa_falloc_end = end;
+ io->u.ci_setattr.sa_attr.lvb_size =
+ parent->u.ci_setattr.sa_attr.lvb_size;
}
lov_lsm2layout(lsm, lsm->lsm_entries[index],
&io->u.ci_setattr.sa_layout);
}
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);
+ io->ci_tried_all_mirrors = parent->ci_tried_all_mirrors;
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;
}
io->u.ci_ladvise.li_flags = parent->u.ci_ladvise.li_flags;
break;
}
+ case CIT_LSEEK: {
+ io->u.ci_lseek.ls_start = start;
+ io->u.ci_lseek.ls_whence = parent->u.ci_lseek.ls_whence;
+ io->u.ci_lseek.ls_result = parent->u.ci_lseek.ls_result;
+ break;
+ }
case CIT_GLIMPSE:
case CIT_MISC:
default:
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_add_sub(const struct lu_env *env, struct lov_io *lio,
+ struct lov_io_sub *sub, u64 start, u64 end)
+{
+ int rc;
+
+ end = lov_offset_mod(end, 1);
+ lov_io_sub_inherit(sub, lio, start, end);
+ rc = cl_io_iter_init(sub->sub_env, &sub->sub_io);
+ if (rc != 0) {
+ cl_io_iter_fini(sub->sub_env, &sub->sub_io);
+ return rc;
+ }
+
+ list_add_tail(&sub->sub_linkage, &lio->lis_active);
+
+ return rc;
+}
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;
u64 start;
u64 end;
int stripe;
+ bool tested_trunc_stripe = false;
+
+ r0->lo_trunc_stripeno = -1;
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;
}
&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)
continue;
}
- end = lov_offset_mod(end, 1);
+ if (cl_io_is_trunc(ios->cis_io) &&
+ !tested_trunc_stripe) {
+ int prev;
+ u64 tr_start;
+
+ prev = (stripe == 0) ? r0->lo_nr - 1 :
+ stripe - 1;
+ /**
+ * Only involving previous stripe if the
+ * truncate in this component is at the
+ * beginning of this stripe.
+ */
+ tested_trunc_stripe = true;
+ if (ext.e_start < lsm->lsm_entries[index]->
+ lsme_extent.e_start) {
+ /* need previous stripe involvement */
+ r0->lo_trunc_stripeno = prev;
+ } else {
+ tr_start = ext.e_start;
+ tr_start = lov_do_div64(tr_start,
+ stripe_width(lsm, index));
+ /* tr_start %= stripe_swidth */
+ if (tr_start == stripe * lsm->
+ lsm_entries[index]->
+ lsme_stripe_size)
+ r0->lo_trunc_stripeno = prev;
+ }
+ }
+
+ /* if the last stripe is the trunc stripeno */
+ if (r0->lo_trunc_stripeno == stripe)
+ r0->lo_trunc_stripeno = -1;
+
sub = lov_sub_get(env, lio,
lov_comp_index(index, stripe));
- if (IS_ERR(sub)) {
- rc = PTR_ERR(sub);
+ if (IS_ERR(sub))
+ return PTR_ERR(sub);
+
+ rc = lov_io_add_sub(env, lio, sub, start, end);
+ if (rc != 0)
break;
+ }
+ if (rc != 0)
+ break;
+
+ if (r0->lo_trunc_stripeno != -1) {
+ stripe = r0->lo_trunc_stripeno;
+ if (unlikely(!r0->lo_sub[stripe])) {
+ r0->lo_trunc_stripeno = -1;
+ continue;
+ }
+ sub = lov_sub_get(env, lio,
+ lov_comp_index(index, stripe));
+ if (IS_ERR(sub))
+ return PTR_ERR(sub);
+
+ /**
+ * the prev sub could be used by another truncate, we'd
+ * skip it. LU-14128 happends when expand truncate +
+ * read get wrong kms.
+ */
+ if (!list_empty(&sub->sub_linkage)) {
+ r0->lo_trunc_stripeno = -1;
+ continue;
}
- lov_io_sub_inherit(sub, lio, start, end);
- rc = cl_io_iter_init(sub->sub_env, &sub->sub_io);
- if (rc != 0)
- cl_io_iter_fini(sub->sub_env, &sub->sub_io);
+ (void)lov_stripe_intersects(lsm, index, stripe, &ext,
+ &start, &end);
+ rc = lov_io_add_sub(env, lio, sub, start, end);
if (rc != 0)
break;
- CDEBUG(D_VFSTRACE,
- "shrink stripe: {%d, %d} range: [%llu, %llu)\n",
- index, stripe, start, end);
-
- list_add_tail(&sub->sub_linkage, &lio->lis_active);
}
- if (rc != 0)
- break;
}
RETURN(rc);
}
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;
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));
}
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
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
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)
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,
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));
ra);
CDEBUG(D_READA, DFID " cra_end = %lu, stripes = %d, rc = %d\n",
- PFID(lu_object_fid(lov2lu(loo))), ra->cra_end, r0->lo_nr, rc);
+ PFID(lu_object_fid(lov2lu(loo))), ra->cra_end_idx,
+ r0->lo_nr, rc);
if (rc != 0)
RETURN(rc);
*/
/* cra_end is stripe level, convert it into file level */
- ra_end = ra->cra_end;
+ ra_end = ra->cra_end_idx;
if (ra_end != CL_PAGE_EOF)
- ra->cra_end = lov_stripe_pgoff(loo->lo_lsm, index,
- ra_end, stripe);
+ ra->cra_end_idx = lov_stripe_pgoff(loo->lo_lsm, index,
+ ra_end, stripe);
/* boundary of current component */
ra_end = cl_index(obj, (loff_t)lov_io_extent(lio, index)->e_end);
- if (ra_end != CL_PAGE_EOF && ra->cra_end >= ra_end)
- ra->cra_end = ra_end - 1;
+ if (ra_end != CL_PAGE_EOF && ra->cra_end_idx >= ra_end)
+ ra->cra_end_idx = ra_end - 1;
if (r0->lo_nr == 1) /* single stripe file */
RETURN(0);
pps = lov_lse(loo, index)->lsme_stripe_size >> PAGE_SHIFT;
- CDEBUG(D_READA, DFID " max_index = %lu, pps = %u, index = %u, "
+ CDEBUG(D_READA, DFID " max_index = %lu, pps = %u, index = %d, "
"stripe_size = %u, stripe no = %u, start index = %lu\n",
- PFID(lu_object_fid(lov2lu(loo))), ra->cra_end, pps, index,
+ PFID(lu_object_fid(lov2lu(loo))), ra->cra_end_idx, pps, index,
lov_lse(loo, index)->lsme_stripe_size, stripe, start);
/* never exceed the end of the stripe */
- ra->cra_end = min_t(pgoff_t,
- ra->cra_end, start + pps - start % pps - 1);
+ ra->cra_end_idx = min_t(pgoff_t, ra->cra_end_idx,
+ start + pps - start % pps - 1);
RETURN(0);
}
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;
if (lov_page_is_empty(page)) {
cl_page_list_move(&queue->c2_qout, qin, page);
- /* it could only be mirror read to get here therefore
+ /*
+ * 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. */
+ * 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;
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);
}
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;
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);
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));
RETURN_EXIT;
}
+static void lov_io_lseek_end(const struct lu_env *env,
+ const struct cl_io_slice *ios)
+{
+ struct lov_io *lio = cl2lov_io(env, ios);
+ struct cl_io *io = lio->lis_cl.cis_io;
+ struct lov_stripe_md *lsm = lio->lis_object->lo_lsm;
+ struct lov_io_sub *sub;
+ loff_t offset = -ENXIO;
+ __u64 hole_off = 0;
+ bool seek_hole = io->u.ci_lseek.ls_whence == SEEK_HOLE;
+
+ ENTRY;
+
+ list_for_each_entry(sub, &lio->lis_active, sub_linkage) {
+ struct cl_io *subio = &sub->sub_io;
+ int index = lov_comp_entry(sub->sub_subio_index);
+ int stripe = lov_comp_stripe(sub->sub_subio_index);
+ loff_t sub_off, lov_off;
+ __u64 comp_end = lsm->lsm_entries[index]->lsme_extent.e_end;
+
+ lov_io_end_wrapper(sub->sub_env, subio);
+
+ if (io->ci_result == 0)
+ io->ci_result = sub->sub_io.ci_result;
+
+ if (io->ci_result)
+ continue;
+
+ CDEBUG(D_INFO, DFID": entry %x stripe %u: SEEK_%s from %lld\n",
+ PFID(lu_object_fid(lov2lu(lio->lis_object))),
+ index, stripe, seek_hole ? "HOLE" : "DATA",
+ subio->u.ci_lseek.ls_start);
+
+ /* first subio with positive result is what we need */
+ sub_off = subio->u.ci_lseek.ls_result;
+ /* Expected error, offset is out of stripe file size */
+ if (sub_off == -ENXIO)
+ continue;
+ /* Any other errors are not expected with ci_result == 0 */
+ if (sub_off < 0) {
+ CDEBUG(D_INFO, "unexpected error: rc = %lld\n",
+ sub_off);
+ io->ci_result = sub_off;
+ continue;
+ }
+ lov_off = lov_stripe_size(lsm, index, sub_off + 1, stripe) - 1;
+ if (lov_off < 0) {
+ /* the only way to get negatove lov_off here is too big
+ * result. Return -EOVERFLOW then.
+ */
+ io->ci_result = -EOVERFLOW;
+ CDEBUG(D_INFO, "offset %llu is too big: rc = %d\n",
+ (u64)lov_off, io->ci_result);
+ continue;
+ }
+ if (lov_off < io->u.ci_lseek.ls_start) {
+ io->ci_result = -EINVAL;
+ CDEBUG(D_INFO, "offset %lld < start %lld: rc = %d\n",
+ sub_off, io->u.ci_lseek.ls_start, io->ci_result);
+ continue;
+ }
+ /* resulting offset can be out of component range if stripe
+ * object is full and its file size was returned as virtual
+ * hole start. Skip this result, the next component will give
+ * us correct lseek result but keep possible hole offset in
+ * case there is no more components ahead
+ */
+ if (lov_off >= comp_end) {
+ /* must be SEEK_HOLE case */
+ if (likely(seek_hole)) {
+ /* save comp end as potential hole offset */
+ hole_off = max_t(__u64, comp_end, hole_off);
+ } else {
+ io->ci_result = -EINVAL;
+ CDEBUG(D_INFO,
+ "off %lld >= comp_end %llu: rc = %d\n",
+ lov_off, comp_end, io->ci_result);
+ }
+ continue;
+ }
+
+ CDEBUG(D_INFO, "SEEK_%s: %lld->%lld/%lld: rc = %d\n",
+ seek_hole ? "HOLE" : "DATA",
+ subio->u.ci_lseek.ls_start, sub_off, lov_off,
+ sub->sub_io.ci_result);
+ offset = min_t(__u64, offset, lov_off);
+ }
+ /* no result but some component returns hole as component end */
+ if (seek_hole && offset == -ENXIO && hole_off > 0)
+ offset = hole_off;
+
+ io->u.ci_lseek.ls_result = offset;
+ RETURN_EXIT;
+}
+
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,
.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,
.cio_start = lov_io_start,
.cio_end = lov_io_end
},
+ [CIT_LSEEK] = {
+ .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_lseek_end
+ },
[CIT_GLIMPSE] = {
.cio_fini = lov_io_fini,
},
.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,
};
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)
* 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
},
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);
break;
case CIT_FSYNC:
case CIT_LADVISE:
+ case CIT_LSEEK:
case CIT_SETATTR:
case CIT_DATA_VERSION:
result = +1;
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
*/
- /* the truncate is for size > 0 so triggers a restore */
- if (cl_io_is_trunc(io)) {
+ /*
+ * the truncate is for size > 0 so triggers a restore,
+ * also trigger a restore for prealloc/punch
+ */
+ if (cl_io_is_trunc(io) || cl_io_is_fallocate(io)) {
io->ci_restore_needed = 1;
result = -ENODATA;
} else
case CIT_READ:
case CIT_WRITE:
case CIT_FAULT:
+ case CIT_LSEEK:
io->ci_restore_needed = 1;
result = -ENODATA;
break;
git://git.whamcloud.com / fs / lustre-release.git / blobdiff