Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1569134
| From | James Simmons <jsimmons@infradead.org> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | [PATCH 52/60] staging: lustre: linkea: linkEA size limitation |
| Date | 2017-01-29 01:30 +0100 |
| Message-ID | <t4Lxw-8ps-23@gated-at.bofh.it> (permalink) |
| References | <t4Lea-8ik-3@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
From: Fan Yong <fan.yong@intel.com>
Under DNE mode, if we do not restrict the linkEA size, and if there
are too many cross-MDTs hard links to the same object, then it will
casue the llog overflow. On the other hand, too many linkEA entries
in the linkEA will serious affect the linkEA performance because we
only support to locate linkEA entry consecutively.
So we need to restrict the linkEA size. Currently, it is 4096 bytes,
that is independent from the backend. If too many hard links caused
the linkEA overflowed, we will add overflow timestamp in the linkEA
header. Such overflow timestamp has some functionalities:
1. It will prevent the object being migrated to other MDT, because
some name entries may be not in the linkEA, so we cannot update
these name entries for the migration.
2. It will tell the namespace LFSCK that the 'nlink' attribute may
be more trustable than the linkEA, then avoid misguiding the
namespace LFSCK to repair 'nlink' attribute based on linkEA.
There will be subsequent patch(es) for namespace LFSCK to handle the
linkEA size limitation and overflow cases.
Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8569
Reviewed-on: https://review.whamcloud.com/23500
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: wangdi <di.wang@intel.com>
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
.../lustre/lustre/include/lustre/lustre_idl.h | 5 +-
.../staging/lustre/lustre/include/lustre_linkea.h | 15 ++++-
drivers/staging/lustre/lustre/llite/llite_lib.c | 2 +-
drivers/staging/lustre/lustre/obdclass/linkea.c | 70 +++++++++++++++++-----
drivers/staging/lustre/lustre/ptlrpc/wiretest.c | 16 ++---
5 files changed, 81 insertions(+), 27 deletions(-)
diff --git a/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h b/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
index b0eb80d..fc960da 100644
--- a/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
+++ b/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
@@ -3217,9 +3217,8 @@ struct link_ea_header {
__u32 leh_magic;
__u32 leh_reccount;
__u64 leh_len; /* total size */
- /* future use */
- __u32 padding1;
- __u32 padding2;
+ __u32 leh_overflow_time;
+ __u32 leh_padding;
};
/** Hardlink data is name and parent fid.
diff --git a/drivers/staging/lustre/lustre/include/lustre_linkea.h b/drivers/staging/lustre/lustre/include/lustre_linkea.h
index 249e8bf..3ff008f 100644
--- a/drivers/staging/lustre/lustre/include/lustre_linkea.h
+++ b/drivers/staging/lustre/lustre/include/lustre_linkea.h
@@ -26,7 +26,19 @@
* Author: di wang <di.wang@intel.com>
*/
-#define DEFAULT_LINKEA_SIZE 4096
+/* There are several reasons to restrict the linkEA size:
+ *
+ * 1. Under DNE mode, if we do not restrict the linkEA size, and if there
+ * are too many cross-MDTs hard links to the same object, then it will
+ * casue the llog overflow.
+ *
+ * 2. Some backend has limited size for EA. For example, if without large
+ * EA enabled, the ldiskfs will make all EAs to share one (4K) EA block.
+ *
+ * 3. Too many entries in linkEA will seriously affect linkEA performance
+ * because we only support to locate linkEA entry consecutively.
+ */
+#define MAX_LINKEA_SIZE 4096
struct linkea_data {
/**
@@ -43,6 +55,7 @@ struct linkea_data {
int linkea_data_new(struct linkea_data *ldata, struct lu_buf *buf);
int linkea_init(struct linkea_data *ldata);
+int linkea_init_with_rec(struct linkea_data *ldata);
void linkea_entry_unpack(const struct link_ea_entry *lee, int *reclen,
struct lu_name *lname, struct lu_fid *pfid);
int linkea_entry_pack(struct link_ea_entry *lee, const struct lu_name *lname,
diff --git a/drivers/staging/lustre/lustre/llite/llite_lib.c b/drivers/staging/lustre/lustre/llite/llite_lib.c
index b229cbc..9a9cdb0 100644
--- a/drivers/staging/lustre/lustre/llite/llite_lib.c
+++ b/drivers/staging/lustre/lustre/llite/llite_lib.c
@@ -2553,7 +2553,7 @@ static int ll_linkea_decode(struct linkea_data *ldata, unsigned int linkno,
unsigned int idx;
int rc;
- rc = linkea_init(ldata);
+ rc = linkea_init_with_rec(ldata);
if (rc < 0)
return rc;
diff --git a/drivers/staging/lustre/lustre/obdclass/linkea.c b/drivers/staging/lustre/lustre/obdclass/linkea.c
index 0b1d2f0..dddd0c4 100644
--- a/drivers/staging/lustre/lustre/obdclass/linkea.c
+++ b/drivers/staging/lustre/lustre/obdclass/linkea.c
@@ -39,6 +39,8 @@ int linkea_data_new(struct linkea_data *ldata, struct lu_buf *buf)
ldata->ld_leh->leh_magic = LINK_EA_MAGIC;
ldata->ld_leh->leh_len = sizeof(struct link_ea_header);
ldata->ld_leh->leh_reccount = 0;
+ ldata->ld_leh->leh_overflow_time = 0;
+ ldata->ld_leh->leh_padding = 0;
return 0;
}
EXPORT_SYMBOL(linkea_data_new);
@@ -53,11 +55,15 @@ int linkea_init(struct linkea_data *ldata)
leh->leh_magic = LINK_EA_MAGIC;
leh->leh_reccount = __swab32(leh->leh_reccount);
leh->leh_len = __swab64(leh->leh_len);
- /* entries are swabbed by linkea_entry_unpack */
+ leh->leh_overflow_time = __swab32(leh->leh_overflow_time);
+ leh->leh_padding = __swab32(leh->leh_padding);
+ /* individual entries are swabbed by linkea_entry_unpack() */
}
+
if (leh->leh_magic != LINK_EA_MAGIC)
return -EINVAL;
- if (leh->leh_reccount == 0)
+
+ if (leh->leh_reccount == 0 && leh->leh_overflow_time == 0)
return -ENODATA;
ldata->ld_leh = leh;
@@ -65,6 +71,18 @@ int linkea_init(struct linkea_data *ldata)
}
EXPORT_SYMBOL(linkea_init);
+int linkea_init_with_rec(struct linkea_data *ldata)
+{
+ int rc;
+
+ rc = linkea_init(ldata);
+ if (!rc && ldata->ld_leh->leh_reccount == 0)
+ rc = -ENODATA;
+
+ return rc;
+}
+EXPORT_SYMBOL(linkea_init_with_rec);
+
/**
* Pack a link_ea_entry.
* All elements are stored as chars to avoid alignment issues.
@@ -94,6 +112,8 @@ int linkea_entry_pack(struct link_ea_entry *lee, const struct lu_name *lname,
void linkea_entry_unpack(const struct link_ea_entry *lee, int *reclen,
struct lu_name *lname, struct lu_fid *pfid)
{
+ LASSERT(lee);
+
*reclen = (lee->lee_reclen[0] << 8) | lee->lee_reclen[1];
memcpy(pfid, &lee->lee_parent_fid, sizeof(*pfid));
fid_be_to_cpu(pfid, pfid);
@@ -110,25 +130,45 @@ void linkea_entry_unpack(const struct link_ea_entry *lee, int *reclen,
int linkea_add_buf(struct linkea_data *ldata, const struct lu_name *lname,
const struct lu_fid *pfid)
{
- LASSERT(ldata->ld_leh);
+ struct link_ea_header *leh = ldata->ld_leh;
+ int reclen;
+
+ LASSERT(leh);
if (!lname || !pfid)
return -EINVAL;
- ldata->ld_reclen = lname->ln_namelen + sizeof(struct link_ea_entry);
- if (ldata->ld_leh->leh_len + ldata->ld_reclen >
- ldata->ld_buf->lb_len) {
+ reclen = lname->ln_namelen + sizeof(struct link_ea_entry);
+ if (unlikely(leh->leh_len + reclen > MAX_LINKEA_SIZE)) {
+ /*
+ * Use 32-bits to save the overflow time, although it will
+ * shrink the ktime_get_real_seconds() returned 64-bits value
+ * to 32-bits value, it is still quite large and can be used
+ * for about 140 years. That is enough.
+ */
+ leh->leh_overflow_time = ktime_get_real_seconds();
+ if (unlikely(leh->leh_overflow_time == 0))
+ leh->leh_overflow_time++;
+
+ CDEBUG(D_INODE, "No enough space to hold linkea entry '" DFID ": %.*s' at %u\n",
+ PFID(pfid), lname->ln_namelen,
+ lname->ln_name, leh->leh_overflow_time);
+ return 0;
+ }
+
+ if (leh->leh_len + reclen > ldata->ld_buf->lb_len) {
if (lu_buf_check_and_grow(ldata->ld_buf,
- ldata->ld_leh->leh_len +
- ldata->ld_reclen) < 0)
+ leh->leh_len + reclen) < 0)
return -ENOMEM;
+
+ ldata->ld_leh = ldata->ld_buf->lb_buf;
+ leh = ldata->ld_leh;
}
- ldata->ld_leh = ldata->ld_buf->lb_buf;
- ldata->ld_lee = ldata->ld_buf->lb_buf + ldata->ld_leh->leh_len;
+ ldata->ld_lee = ldata->ld_buf->lb_buf + leh->leh_len;
ldata->ld_reclen = linkea_entry_pack(ldata->ld_lee, lname, pfid);
- ldata->ld_leh->leh_len += ldata->ld_reclen;
- ldata->ld_leh->leh_reccount++;
+ leh->leh_len += ldata->ld_reclen;
+ leh->leh_reccount++;
CDEBUG(D_INODE, "New link_ea name '" DFID ":%.*s' is added\n",
PFID(pfid), lname->ln_namelen, lname->ln_name);
return 0;
@@ -139,6 +179,7 @@ int linkea_add_buf(struct linkea_data *ldata, const struct lu_name *lname,
void linkea_del_buf(struct linkea_data *ldata, const struct lu_name *lname)
{
LASSERT(ldata->ld_leh && ldata->ld_lee);
+ LASSERT(ldata->ld_leh->leh_reccount > 0);
ldata->ld_leh->leh_reccount--;
ldata->ld_leh->leh_len -= ldata->ld_reclen;
@@ -174,8 +215,9 @@ int linkea_links_find(struct linkea_data *ldata, const struct lu_name *lname,
LASSERT(ldata->ld_leh);
- /* link #0 */
- ldata->ld_lee = (struct link_ea_entry *)(ldata->ld_leh + 1);
+ /* link #0, if leh_reccount == 0 we skip the loop and return -ENOENT */
+ if (likely(ldata->ld_leh->leh_reccount > 0))
+ ldata->ld_lee = (struct link_ea_entry *)(ldata->ld_leh + 1);
for (count = 0; count < ldata->ld_leh->leh_reccount; count++) {
linkea_entry_unpack(ldata->ld_lee, &ldata->ld_reclen,
diff --git a/drivers/staging/lustre/lustre/ptlrpc/wiretest.c b/drivers/staging/lustre/lustre/ptlrpc/wiretest.c
index a04e36c..f166518 100644
--- a/drivers/staging/lustre/lustre/ptlrpc/wiretest.c
+++ b/drivers/staging/lustre/lustre/ptlrpc/wiretest.c
@@ -3820,14 +3820,14 @@ void lustre_assert_wire_constants(void)
(long long)(int)offsetof(struct link_ea_header, leh_len));
LASSERTF((int)sizeof(((struct link_ea_header *)0)->leh_len) == 8, "found %lld\n",
(long long)(int)sizeof(((struct link_ea_header *)0)->leh_len));
- LASSERTF((int)offsetof(struct link_ea_header, padding1) == 16, "found %lld\n",
- (long long)(int)offsetof(struct link_ea_header, padding1));
- LASSERTF((int)sizeof(((struct link_ea_header *)0)->padding1) == 4, "found %lld\n",
- (long long)(int)sizeof(((struct link_ea_header *)0)->padding1));
- LASSERTF((int)offsetof(struct link_ea_header, padding2) == 20, "found %lld\n",
- (long long)(int)offsetof(struct link_ea_header, padding2));
- LASSERTF((int)sizeof(((struct link_ea_header *)0)->padding2) == 4, "found %lld\n",
- (long long)(int)sizeof(((struct link_ea_header *)0)->padding2));
+ LASSERTF((int)offsetof(struct link_ea_header, leh_overflow_time) == 16, "found %lld\n",
+ (long long)(int)offsetof(struct link_ea_header, leh_overflow_time));
+ LASSERTF((int)sizeof(((struct link_ea_header *)0)->leh_overflow_time) == 4, "found %lld\n",
+ (long long)(int)sizeof(((struct link_ea_header *)0)->leh_overflow_time));
+ LASSERTF((int)offsetof(struct link_ea_header, leh_padding) == 20, "found %lld\n",
+ (long long)(int)offsetof(struct link_ea_header, leh_padding));
+ LASSERTF((int)sizeof(((struct link_ea_header *)0)->leh_padding) == 4, "found %lld\n",
+ (long long)(int)sizeof(((struct link_ea_header *)0)->leh_padding));
CLASSERT(LINK_EA_MAGIC == 0x11EAF1DFUL);
/* Checks for struct link_ea_entry */
--
1.8.3.1
Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
[PATCH 00/60] staging: lustre: batches of fixes for lustre client James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 21/60] staging: lustre: ptlrpc: correct use of list_add_tail() James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 11/60] staging: lustre: obd: RCU stalls in lu_cache_shrink_count() James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 12/60] staging: lustre: lmv: Error not handled for lmv_find_target James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 28/60] staging: lustre: ldlm: ASSERTION(flock->blocking_export!=0) failed James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 34/60] staging: lustre: libcfs: default CPT matches NUMA topology James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 25/60] staging: lustre: llite: normal user can't set FS default stripe James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 07/60] staging: lustre: ptlrpc: set proper mbits for EINPROGRESS resend James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 19/60] staging: lustre: llite: handle inactive OSTs better in statfs James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 05/60] staging: lustre: llite: check request != NULL in ll_migrate James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
Re: [PATCH 05/60] staging: lustre: llite: check request != NULL in ll_migrate Dan Carpenter <dan.carpenter@oracle.com> - 2017-01-30 12:40 +0100
[PATCH 03/60] staging: lustre: clio: add cl_page LRU shrinker James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 09/60] staging: lustre: lmv: Correctly generate target_obd James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 04/60] staging: lustre: mdc: quiet console message for known -EINTR James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 33/60] staging: lustre: libcfs: avoid stomping on module param cpu_pattern James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 31/60] staging: lustre: clio: sync write should update mtime James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 35/60] staging: lustre: lov: ld_target could be NULL James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 18/60] staging: lustre: ptlrpc: skip lock if export failed James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 27/60] staging: lustre: mgc: handle config_llog_data::cld_refcount properly James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 16/60] staging: lustre: llite: don't ignore layout for group lock request James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 36/60] staging: lustre: header: remove assert from interval_set() James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 13/60] staging: lustre: obdclass: health_check to report unhealthy upon LBUG James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
Re: [PATCH 13/60] staging: lustre: obdclass: health_check to report unhealthy upon LBUG Dan Carpenter <dan.carpenter@oracle.com> - 2017-01-30 13:50 +0100
[PATCH 15/60] staging: lustre: hsm: stack overrun in hai_dump_data_field James Simmons <jsimmons@infradead.org> - 2017-01-29 01:20 +0100
[PATCH 55/60] staging: lustre: obd: move s3 in lmd_parse to inner loop James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 42/60] staging: lustre: ptlrpc: allow blocking asts to be delayed James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 53/60] staging: lustre: ptlrpc: update replay cursor when close during replay James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 54/60] staging: lustre: fid: Change positional struct initializers to C99 James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 43/60] staging: lustre: obd: remove OBD_NOTIFY_CREATE James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 41/60] staging: lustre: osc: osc_match_base prototype differs from declaration James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 51/60] staging: lustre: ptlrpc: update MODULE_PARAM_DESC in ptlrpcd.c James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 50/60] staging: lustre: ptlrpc: remove unused pc->pc_env James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 49/60] staging: lustre: socklnd: remove socklnd_init_msg James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 52/60] staging: lustre: linkea: linkEA size limitation James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 45/60] staging: lustre: libcfs: Change positional struct initializers to C99 James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 39/60] staging: libcfs: remove integer types abstraction from libcfs James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 48/60] staging: lustre: ksocklnd: ignore timedout TX on closing connection James Simmons <jsimmons@infradead.org> - 2017-01-29 01:30 +0100
[PATCH 58/60] staging: lustre: osc: avoid 64 divide in osc_cache_too_much James Simmons <jsimmons@infradead.org> - 2017-01-29 01:40 +0100
[PATCH 57/60] staging: lustre: lmv: remove nlink check in lmv_revalidate_slaves James Simmons <jsimmons@infradead.org> - 2017-01-29 01:40 +0100
[PATCH 60/60] staging: lustre: libcfs: fix minimum size check for libcfs ioctl James Simmons <jsimmons@infradead.org> - 2017-01-29 01:40 +0100
Re: [PATCH 60/60] staging: lustre: libcfs: fix minimum size check for libcfs ioctl Dan Carpenter <dan.carpenter@oracle.com> - 2017-01-30 12:00 +0100
Re: [PATCH 60/60] staging: lustre: libcfs: fix minimum size check for libcfs ioctl Dan Carpenter <dan.carpenter@oracle.com> - 2017-01-30 12:00 +0100
[PATCH 56/60] staging: lustre: llite: don't invoke direct_IO for the EOF case James Simmons <jsimmons@infradead.org> - 2017-01-29 01:40 +0100
[PATCH 01/60] staging: lustre: llite: Remove access of stripe in ll_setattr_raw James Simmons <jsimmons@infradead.org> - 2017-01-29 01:40 +0100
[PATCH 59/60] staging: lustre: ptlrpc : remove userland usage from ptlrpc James Simmons <jsimmons@infradead.org> - 2017-01-29 01:40 +0100
[PATCH 44/60] staging: lustre: libcfs: fix error messages James Simmons <jsimmons@infradead.org> - 2017-01-29 01:50 +0100
[PATCH 38/60] staging: lustre: llite: Adding timed wait in ll_umount_begin James Simmons <jsimmons@infradead.org> - 2017-01-29 02:20 +0100
[PATCH 46/60] staging: lustre: mdc: Make IT_OPEN take lookup bits lock James Simmons <jsimmons@infradead.org> - 2017-01-29 02:50 +0100
[PATCH 47/60] staging: lustre: mdc: avoid returning freed request James Simmons <jsimmons@infradead.org> - 2017-01-29 02:50 +0100
[PATCH 40/60] staging: ptlrpc: leaked rs on difficult reply James Simmons <jsimmons@infradead.org> - 2017-01-29 06:00 +0100
csiph-web