* [PATCH 01/22] staging: lustre: llite: clear LLIF_DATA_MODIFIED in atomic
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 02/22] staging: lustre: osc: fix debug log message formatting James Simmons
` (21 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Jinshan Xiong, James Simmons, James Simmons
From: Jinshan Xiong <jinshan.xiong@intel.com>
This flag should be cleared atomically after the op_data flag
MDS_DATA_MODIFIED is packed. Otherwise, if there exists an
operation to dirty the file again, the state may be missed on
the MDT.
Stop using spin lock lli_lock to protect operations of changing
file flags; using bit operations instead.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: James Simmons <uja.ornl@yahoo.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6377
Reviewed-on: http://review.whamcloud.com/14100
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/llite/file.c | 61 +++++++------------
.../staging/lustre/lustre/llite/llite_internal.h | 12 ++--
drivers/staging/lustre/lustre/llite/llite_lib.c | 18 +------
drivers/staging/lustre/lustre/llite/llite_mmap.c | 7 +--
drivers/staging/lustre/lustre/llite/vvp_io.c | 10 +--
drivers/staging/lustre/lustre/llite/xattr_cache.c | 7 +-
6 files changed, 38 insertions(+), 77 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/file.c b/drivers/staging/lustre/lustre/llite/file.c
index 6c2abb3..ea21e19 100644
--- a/drivers/staging/lustre/lustre/llite/file.c
+++ b/drivers/staging/lustre/lustre/llite/file.c
@@ -75,45 +75,39 @@ static void ll_file_data_put(struct ll_file_data *fd)
kmem_cache_free(ll_file_data_slab, fd);
}
-void ll_pack_inode2opdata(struct inode *inode, struct md_op_data *op_data,
- struct lustre_handle *fh)
+/**
+ * Packs all the attributes into @op_data for the CLOSE rpc.
+ */
+static void ll_prepare_close(struct inode *inode, struct md_op_data *op_data,
+ struct obd_client_handle *och)
{
- op_data->op_fid1 = ll_i2info(inode)->lli_fid;
+ struct ll_inode_info *lli = ll_i2info(inode);
+
+ ll_prep_md_op_data(op_data, inode, NULL, NULL,
+ 0, 0, LUSTRE_OPC_ANY, NULL);
+
op_data->op_attr.ia_mode = inode->i_mode;
op_data->op_attr.ia_atime = inode->i_atime;
op_data->op_attr.ia_mtime = inode->i_mtime;
op_data->op_attr.ia_ctime = inode->i_ctime;
op_data->op_attr.ia_size = i_size_read(inode);
+ op_data->op_attr.ia_valid |= ATTR_MODE | ATTR_ATIME | ATTR_ATIME_SET |
+ ATTR_MTIME | ATTR_MTIME_SET |
+ ATTR_CTIME | ATTR_CTIME_SET;
op_data->op_attr_blocks = inode->i_blocks;
op_data->op_attr_flags = ll_inode_to_ext_flags(inode->i_flags);
- if (fh)
- op_data->op_handle = *fh;
+ op_data->op_handle = och->och_fh;
- if (ll_i2info(inode)->lli_flags & LLIF_DATA_MODIFIED)
+ /*
+ * For HSM: if inode data has been modified, pack it so that
+ * MDT can set data dirty flag in the archive.
+ */
+ if (och->och_flags & FMODE_WRITE &&
+ test_and_clear_bit(LLIF_DATA_MODIFIED, &lli->lli_flags))
op_data->op_bias |= MDS_DATA_MODIFIED;
}
/**
- * Packs all the attributes into @op_data for the CLOSE rpc.
- */
-static void ll_prepare_close(struct inode *inode, struct md_op_data *op_data,
- struct obd_client_handle *och)
-{
- op_data->op_attr.ia_valid = ATTR_MODE | ATTR_ATIME | ATTR_ATIME_SET |
- ATTR_MTIME | ATTR_MTIME_SET |
- ATTR_CTIME | ATTR_CTIME_SET;
-
- if (!(och->och_flags & FMODE_WRITE))
- goto out;
-
- op_data->op_attr.ia_valid |= ATTR_SIZE | ATTR_BLOCKS;
-out:
- ll_pack_inode2opdata(inode, op_data, &och->och_fh);
- ll_prep_md_op_data(op_data, inode, NULL, NULL,
- 0, 0, LUSTRE_OPC_ANY, NULL);
-}
-
-/**
* Perform a close, possibly with a bias.
* The meaning of "data" depends on the value of "bias".
*
@@ -181,17 +175,6 @@ static int ll_close_inode_openhandle(struct obd_export *md_exp,
PFID(ll_inode2fid(inode)), rc);
}
- /* DATA_MODIFIED flag was successfully sent on close, cancel data
- * modification flag.
- */
- if (rc == 0 && (op_data->op_bias & MDS_DATA_MODIFIED)) {
- struct ll_inode_info *lli = ll_i2info(inode);
-
- spin_lock(&lli->lli_lock);
- lli->lli_flags &= ~LLIF_DATA_MODIFIED;
- spin_unlock(&lli->lli_lock);
- }
-
if (op_data->op_bias & (MDS_HSM_RELEASE | MDS_CLOSE_LAYOUT_SWAP) &&
!rc) {
struct mdt_body *body;
@@ -2888,6 +2871,8 @@ static int ll_inode_revalidate(struct dentry *dentry, __u64 ibits)
LTIME_S(inode->i_mtime) = ll_i2info(inode)->lli_mtime;
LTIME_S(inode->i_ctime) = ll_i2info(inode)->lli_ctime;
} else {
+ struct ll_inode_info *lli = ll_i2info(inode);
+
/* In case of restore, the MDT has the right size and has
* already send it back without granting the layout lock,
* inode is up-to-date so glimpse is useless.
@@ -2895,7 +2880,7 @@ static int ll_inode_revalidate(struct dentry *dentry, __u64 ibits)
* restore the MDT holds the layout lock so the glimpse will
* block up to the end of restore (getattr will block)
*/
- if (!(ll_i2info(inode)->lli_flags & LLIF_FILE_RESTORING))
+ if (!test_bit(LLIF_FILE_RESTORING, &lli->lli_flags))
rc = ll_glimpse_size(inode);
}
return rc;
diff --git a/drivers/staging/lustre/lustre/llite/llite_internal.h b/drivers/staging/lustre/lustre/llite/llite_internal.h
index 50bb328..ac4ce05 100644
--- a/drivers/staging/lustre/lustre/llite/llite_internal.h
+++ b/drivers/staging/lustre/lustre/llite/llite_internal.h
@@ -97,20 +97,20 @@ struct ll_grouplock {
unsigned long lg_gid;
};
-enum lli_flags {
+enum ll_file_flags {
/* File data is modified. */
- LLIF_DATA_MODIFIED = BIT(0),
+ LLIF_DATA_MODIFIED = 0,
/* File is being restored */
- LLIF_FILE_RESTORING = BIT(1),
+ LLIF_FILE_RESTORING = 1,
/* Xattr cache is attached to the file */
- LLIF_XATTR_CACHE = BIT(2),
+ LLIF_XATTR_CACHE = 2,
};
struct ll_inode_info {
__u32 lli_inode_magic;
- __u32 lli_flags;
spinlock_t lli_lock;
+ unsigned long lli_flags;
struct posix_acl *lli_posix_acl;
/* identifying fields for both metadata and data stacks. */
@@ -740,8 +740,6 @@ enum ldlm_mode ll_take_md_lock(struct inode *inode, __u64 bits,
int ll_file_release(struct inode *inode, struct file *file);
int ll_release_openhandle(struct inode *, struct lookup_intent *);
int ll_md_real_close(struct inode *inode, fmode_t fmode);
-void ll_pack_inode2opdata(struct inode *inode, struct md_op_data *op_data,
- struct lustre_handle *fh);
int ll_getattr(struct vfsmount *mnt, struct dentry *de, struct kstat *stat);
struct posix_acl *ll_get_acl(struct inode *inode, int type);
int ll_migrate(struct inode *parent, struct file *file, int mdtidx,
diff --git a/drivers/staging/lustre/lustre/llite/llite_lib.c b/drivers/staging/lustre/lustre/llite/llite_lib.c
index dcd9240..2a51efa 100644
--- a/drivers/staging/lustre/lustre/llite/llite_lib.c
+++ b/drivers/staging/lustre/lustre/llite/llite_lib.c
@@ -1532,9 +1532,6 @@ int ll_setattr_raw(struct dentry *dentry, struct iattr *attr, bool hsm_import)
* modified, flag it.
*/
attr->ia_valid |= MDS_OPEN_OWNEROVERRIDE;
- spin_lock(&lli->lli_lock);
- lli->lli_flags |= LLIF_DATA_MODIFIED;
- spin_unlock(&lli->lli_lock);
op_data->op_bias |= MDS_DATA_MODIFIED;
}
}
@@ -1545,13 +1542,6 @@ int ll_setattr_raw(struct dentry *dentry, struct iattr *attr, bool hsm_import)
if (rc)
goto out;
- /* RPC to MDT is sent, cancel data modification flag */
- if (op_data->op_bias & MDS_DATA_MODIFIED) {
- spin_lock(&lli->lli_lock);
- lli->lli_flags &= ~LLIF_DATA_MODIFIED;
- spin_unlock(&lli->lli_lock);
- }
-
if (!S_ISREG(inode->i_mode) || file_is_released) {
rc = 0;
goto out;
@@ -1822,7 +1812,7 @@ int ll_update_inode(struct inode *inode, struct lustre_md *md)
if (body->mbo_valid & OBD_MD_TSTATE) {
if (body->mbo_t_state & MS_RESTORE)
- lli->lli_flags |= LLIF_FILE_RESTORING;
+ set_bit(LLIF_FILE_RESTORING, &lli->lli_flags);
}
return 0;
@@ -2331,8 +2321,6 @@ struct md_op_data *ll_prep_md_op_data(struct md_op_data *op_data,
op_data->op_fsuid = from_kuid(&init_user_ns, current_fsuid());
op_data->op_fsgid = from_kgid(&init_user_ns, current_fsgid());
op_data->op_cap = cfs_curproc_cap_pack();
- op_data->op_bias = 0;
- op_data->op_cli_flags = 0;
if ((opc == LUSTRE_OPC_CREATE) && name &&
filename_is_volatile(name, namelen, &op_data->op_mds))
op_data->op_bias |= MDS_CREATE_VOLATILE;
@@ -2340,10 +2328,6 @@ struct md_op_data *ll_prep_md_op_data(struct md_op_data *op_data,
op_data->op_mds = 0;
op_data->op_data = data;
- /* When called by ll_setattr_raw, file is i1. */
- if (ll_i2info(i1)->lli_flags & LLIF_DATA_MODIFIED)
- op_data->op_bias |= MDS_DATA_MODIFIED;
-
return op_data;
}
diff --git a/drivers/staging/lustre/lustre/llite/llite_mmap.c b/drivers/staging/lustre/lustre/llite/llite_mmap.c
index b56203b..f50b6c1 100644
--- a/drivers/staging/lustre/lustre/llite/llite_mmap.c
+++ b/drivers/staging/lustre/lustre/llite/llite_mmap.c
@@ -215,11 +215,8 @@ static int ll_page_mkwrite0(struct vm_area_struct *vma, struct page *vmpage,
result = -EAGAIN;
}
- if (result == 0) {
- spin_lock(&lli->lli_lock);
- lli->lli_flags |= LLIF_DATA_MODIFIED;
- spin_unlock(&lli->lli_lock);
- }
+ if (!result)
+ set_bit(LLIF_DATA_MODIFIED, &lli->lli_flags);
}
out_io:
diff --git a/drivers/staging/lustre/lustre/llite/vvp_io.c b/drivers/staging/lustre/lustre/llite/vvp_io.c
index 2bf3e47..114906d 100644
--- a/drivers/staging/lustre/lustre/llite/vvp_io.c
+++ b/drivers/staging/lustre/lustre/llite/vvp_io.c
@@ -329,8 +329,8 @@ static void vvp_io_fini(const struct lu_env *env, const struct cl_io_slice *ios)
vio->vui_layout_gen, gen);
/* today successful restore is the only possible case */
/* restore was done, clear restoring state */
- ll_i2info(vvp_object_inode(obj))->lli_flags &=
- ~LLIF_FILE_RESTORING;
+ clear_bit(LLIF_FILE_RESTORING,
+ &ll_i2info(inode)->lli_flags);
}
}
}
@@ -982,11 +982,7 @@ static int vvp_io_write_start(const struct lu_env *env,
}
}
if (result > 0) {
- struct ll_inode_info *lli = ll_i2info(inode);
-
- spin_lock(&lli->lli_lock);
- lli->lli_flags |= LLIF_DATA_MODIFIED;
- spin_unlock(&lli->lli_lock);
+ set_bit(LLIF_DATA_MODIFIED, &(ll_i2info(inode))->lli_flags);
if (result < cnt)
io->ci_continue = 0;
diff --git a/drivers/staging/lustre/lustre/llite/xattr_cache.c b/drivers/staging/lustre/lustre/llite/xattr_cache.c
index 47c1d11..198b051 100644
--- a/drivers/staging/lustre/lustre/llite/xattr_cache.c
+++ b/drivers/staging/lustre/lustre/llite/xattr_cache.c
@@ -60,7 +60,7 @@ void ll_xattr_fini(void)
static void ll_xattr_cache_init(struct ll_inode_info *lli)
{
INIT_LIST_HEAD(&lli->lli_xattrs);
- lli->lli_flags |= LLIF_XATTR_CACHE;
+ set_bit(LLIF_XATTR_CACHE, &lli->lli_flags);
}
/**
@@ -216,7 +216,7 @@ static int ll_xattr_cache_list(struct list_head *cache,
*/
static int ll_xattr_cache_valid(struct ll_inode_info *lli)
{
- return !!(lli->lli_flags & LLIF_XATTR_CACHE);
+ return test_bit(LLIF_XATTR_CACHE, &lli->lli_flags);
}
/**
@@ -233,7 +233,8 @@ static int ll_xattr_cache_destroy_locked(struct ll_inode_info *lli)
while (ll_xattr_cache_del(&lli->lli_xattrs, NULL) == 0)
; /* empty loop */
- lli->lli_flags &= ~LLIF_XATTR_CACHE;
+
+ clear_bit(LLIF_XATTR_CACHE, &lli->lli_flags);
return 0;
}
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 02/22] staging: lustre: osc: fix debug log message formatting
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
2016-12-03 0:53 ` [PATCH 01/22] staging: lustre: llite: clear LLIF_DATA_MODIFIED in atomic James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 03/22] staging: lustre: mdt: race between open and migrate James Simmons
` (20 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Ashish Purkar, James Simmons
From: Ashish Purkar <ashish.purkar@seagate.com>
Corrected newline specifier in debug log message.
Signed-off-by: Ashish Purkar <ashish.purkar@seagate.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7029
Reviewed-on: http://review.whamcloud.com/16046
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/osc/osc_page.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/staging/lustre/lustre/osc/osc_page.c b/drivers/staging/lustre/lustre/osc/osc_page.c
index a80a847..c5129d1 100644
--- a/drivers/staging/lustre/lustre/osc/osc_page.c
+++ b/drivers/staging/lustre/lustre/osc/osc_page.c
@@ -462,7 +462,7 @@ static void osc_lru_del(struct client_obd *cli, struct osc_page *opg)
* stealing one of them.
*/
if (osc_cache_too_much(cli)) {
- CDEBUG(D_CACHE, "%s: queue LRU workn", cli_name(cli));
+ CDEBUG(D_CACHE, "%s: queue LRU work\n", cli_name(cli));
(void)ptlrpcd_queue_work(cli->cl_lru_work);
}
wake_up(&osc_lru_waitq);
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 03/22] staging: lustre: mdt: race between open and migrate
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
2016-12-03 0:53 ` [PATCH 01/22] staging: lustre: llite: clear LLIF_DATA_MODIFIED in atomic James Simmons
2016-12-03 0:53 ` [PATCH 02/22] staging: lustre: osc: fix debug log message formatting James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 04/22] staging: lustre: osc: handle osc eviction correctly James Simmons
` (19 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, wang di,
James Simmons
From: wang di <di.wang@intel.com>
During intent open, it was found that if the parent has
been migrated to another MDT, it should retry the open
request with the new object, so it needs to keep the
old object in the orphan list, which will be cleanup
during next recovery. Note: if the client still using
the old FID after next recovery, it will return -ENOENT
for the application. Also enqueue the lease lock of
the migrating file, then compare the lease before
migration to make sure no other clients open the file
at the same time.
Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6475
Reviewed-on: http://review.whamcloud.com/14497
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
.../lustre/lustre/include/lustre_req_layout.h | 1 +
drivers/staging/lustre/lustre/llite/dir.c | 2 +-
drivers/staging/lustre/lustre/llite/file.c | 76 ++++++++++++++++---
.../staging/lustre/lustre/llite/llite_internal.h | 2 +-
drivers/staging/lustre/lustre/mdc/mdc_lib.c | 59 +++++++++-------
drivers/staging/lustre/lustre/mdc/mdc_reint.c | 20 +++++-
drivers/staging/lustre/lustre/ptlrpc/layout.c | 18 +++++
7 files changed, 138 insertions(+), 40 deletions(-)
diff --git a/drivers/staging/lustre/lustre/include/lustre_req_layout.h b/drivers/staging/lustre/lustre/include/lustre_req_layout.h
index 7657132..fbcd395 100644
--- a/drivers/staging/lustre/lustre/include/lustre_req_layout.h
+++ b/drivers/staging/lustre/lustre/include/lustre_req_layout.h
@@ -167,6 +167,7 @@ void req_capsule_shrink(struct req_capsule *pill,
extern struct req_format RQF_MDS_REINT_SETXATTR;
extern struct req_format RQF_MDS_QUOTACTL;
extern struct req_format RQF_MDS_SWAP_LAYOUTS;
+extern struct req_format RQF_MDS_REINT_MIGRATE;
/* MDS hsm formats */
extern struct req_format RQF_MDS_HSM_STATE_GET;
extern struct req_format RQF_MDS_HSM_STATE_SET;
diff --git a/drivers/staging/lustre/lustre/llite/dir.c b/drivers/staging/lustre/lustre/llite/dir.c
index ce05493..351e900 100644
--- a/drivers/staging/lustre/lustre/llite/dir.c
+++ b/drivers/staging/lustre/lustre/llite/dir.c
@@ -1083,7 +1083,7 @@ static long ll_dir_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
goto out_free;
}
- rc = ll_get_fid_by_name(inode, filename, namelen, NULL);
+ rc = ll_get_fid_by_name(inode, filename, namelen, NULL, NULL);
if (rc < 0) {
CERROR("%s: lookup %.*s failed: rc = %d\n",
ll_get_fsname(inode->i_sb, NULL, 0), namelen,
diff --git a/drivers/staging/lustre/lustre/llite/file.c b/drivers/staging/lustre/lustre/llite/file.c
index ea21e19..aa29583 100644
--- a/drivers/staging/lustre/lustre/llite/file.c
+++ b/drivers/staging/lustre/lustre/llite/file.c
@@ -2531,7 +2531,8 @@ int ll_fsync(struct file *file, loff_t start, loff_t end, int datasync)
}
int ll_get_fid_by_name(struct inode *parent, const char *name,
- int namelen, struct lu_fid *fid)
+ int namelen, struct lu_fid *fid,
+ struct inode **inode)
{
struct md_op_data *op_data = NULL;
struct ptlrpc_request *req;
@@ -2543,7 +2544,7 @@ int ll_get_fid_by_name(struct inode *parent, const char *name,
if (IS_ERR(op_data))
return PTR_ERR(op_data);
- op_data->op_valid = OBD_MD_FLID;
+ op_data->op_valid = OBD_MD_FLID | OBD_MD_FLTYPE;
rc = md_getattr_name(ll_i2sbi(parent)->ll_md_exp, op_data, &req);
ll_finish_md_op_data(op_data);
if (rc < 0)
@@ -2556,6 +2557,9 @@ int ll_get_fid_by_name(struct inode *parent, const char *name,
}
if (fid)
*fid = body->mbo_fid1;
+
+ if (inode)
+ rc = ll_prep_inode(inode, req, parent->i_sb, NULL);
out_req:
ptlrpc_req_finished(req);
return rc;
@@ -2565,9 +2569,12 @@ int ll_migrate(struct inode *parent, struct file *file, int mdtidx,
const char *name, int namelen)
{
struct ptlrpc_request *request = NULL;
+ struct obd_client_handle *och = NULL;
struct inode *child_inode = NULL;
struct dentry *dchild = NULL;
struct md_op_data *op_data;
+ struct mdt_body *body;
+ u64 data_version = 0;
struct qstr qstr;
int rc;
@@ -2586,22 +2593,25 @@ int ll_migrate(struct inode *parent, struct file *file, int mdtidx,
dchild = d_lookup(file_dentry(file), &qstr);
if (dchild) {
op_data->op_fid3 = *ll_inode2fid(dchild->d_inode);
- if (dchild->d_inode) {
+ if (dchild->d_inode)
child_inode = igrab(dchild->d_inode);
- if (child_inode) {
- inode_lock(child_inode);
- op_data->op_fid3 = *ll_inode2fid(child_inode);
- ll_invalidate_aliases(child_inode);
- }
- }
dput(dchild);
- } else {
+ }
+
+ if (!child_inode) {
rc = ll_get_fid_by_name(parent, name, namelen,
- &op_data->op_fid3);
+ &op_data->op_fid3, &child_inode);
if (rc)
goto out_free;
}
+ if (!child_inode) {
+ rc = -EINVAL;
+ goto out_free;
+ }
+
+ inode_lock(child_inode);
+ op_data->op_fid3 = *ll_inode2fid(child_inode);
if (!fid_is_sane(&op_data->op_fid3)) {
CERROR("%s: migrate %s, but fid "DFID" is insane\n",
ll_get_fsname(parent->i_sb, NULL, 0), name,
@@ -2620,6 +2630,26 @@ int ll_migrate(struct inode *parent, struct file *file, int mdtidx,
rc = 0;
goto out_free;
}
+again:
+ if (S_ISREG(child_inode->i_mode)) {
+ och = ll_lease_open(child_inode, NULL, FMODE_WRITE, 0);
+ if (IS_ERR(och)) {
+ rc = PTR_ERR(och);
+ och = NULL;
+ goto out_free;
+ }
+
+ rc = ll_data_version(child_inode, &data_version,
+ LL_DV_WR_FLUSH);
+ if (rc)
+ goto out_free;
+
+ op_data->op_handle = och->och_fh;
+ op_data->op_data = och->och_mod;
+ op_data->op_data_version = data_version;
+ op_data->op_lease_handle = och->och_lease_handle;
+ op_data->op_bias |= MDS_RENAME_MIGRATE;
+ }
op_data->op_mds = mdtidx;
op_data->op_cli_flags = CLI_MIGRATE;
@@ -2628,10 +2658,32 @@ int ll_migrate(struct inode *parent, struct file *file, int mdtidx,
if (!rc)
ll_update_times(request, parent);
- ptlrpc_req_finished(request);
+ body = req_capsule_server_get(&request->rq_pill, &RMF_MDT_BODY);
+ if (!body) {
+ rc = -EPROTO;
+ goto out_free;
+ }
+
+ /*
+ * If the server does release layout lock, then we cleanup
+ * the client och here, otherwise release it in out_free:
+ */
+ if (och && body->mbo_valid & OBD_MD_CLOSE_INTENT_EXECED) {
+ obd_mod_put(och->och_mod);
+ md_clear_open_replay_data(ll_i2sbi(parent)->ll_md_exp, och);
+ och->och_fh.cookie = DEAD_HANDLE_MAGIC;
+ kfree(och);
+ och = NULL;
+ }
+ ptlrpc_req_finished(request);
+ /* Try again if the file layout has changed. */
+ if (rc == -EAGAIN && S_ISREG(child_inode->i_mode))
+ goto again;
out_free:
if (child_inode) {
+ if (och) /* close the file */
+ ll_lease_close(och, child_inode, NULL);
clear_nlink(child_inode);
inode_unlock(child_inode);
iput(child_inode);
diff --git a/drivers/staging/lustre/lustre/llite/llite_internal.h b/drivers/staging/lustre/lustre/llite/llite_internal.h
index ac4ce05..ae0bb09 100644
--- a/drivers/staging/lustre/lustre/llite/llite_internal.h
+++ b/drivers/staging/lustre/lustre/llite/llite_internal.h
@@ -745,7 +745,7 @@ enum ldlm_mode ll_take_md_lock(struct inode *inode, __u64 bits,
int ll_migrate(struct inode *parent, struct file *file, int mdtidx,
const char *name, int namelen);
int ll_get_fid_by_name(struct inode *parent, const char *name,
- int namelen, struct lu_fid *fid);
+ int namelen, struct lu_fid *fid, struct inode **inode);
int ll_inode_permission(struct inode *inode, int mask);
int ll_lov_setstripe_ea_info(struct inode *inode, struct dentry *dentry,
diff --git a/drivers/staging/lustre/lustre/mdc/mdc_lib.c b/drivers/staging/lustre/lustre/mdc/mdc_lib.c
index c1990f0..f35e1f9 100644
--- a/drivers/staging/lustre/lustre/mdc/mdc_lib.c
+++ b/drivers/staging/lustre/lustre/mdc/mdc_lib.c
@@ -376,6 +376,31 @@ void mdc_link_pack(struct ptlrpc_request *req, struct md_op_data *op_data)
mdc_pack_name(req, &RMF_NAME, op_data->op_name, op_data->op_namelen);
}
+static void mdc_intent_close_pack(struct ptlrpc_request *req,
+ struct md_op_data *op_data)
+{
+ enum mds_op_bias bias = op_data->op_bias;
+ struct close_data *data;
+ struct ldlm_lock *lock;
+
+ if (!(bias & (MDS_HSM_RELEASE | MDS_CLOSE_LAYOUT_SWAP |
+ MDS_RENAME_MIGRATE)))
+ return;
+
+ data = req_capsule_client_get(&req->rq_pill, &RMF_CLOSE_DATA);
+ LASSERT(data);
+
+ lock = ldlm_handle2lock(&op_data->op_lease_handle);
+ if (lock) {
+ data->cd_handle = lock->l_remote_handle;
+ LDLM_LOCK_PUT(lock);
+ }
+ ldlm_cli_cancel(&op_data->op_lease_handle, LCF_LOCAL);
+
+ data->cd_data_version = op_data->op_data_version;
+ data->cd_fid = op_data->op_fid2;
+}
+
void mdc_rename_pack(struct ptlrpc_request *req, struct md_op_data *op_data,
const char *old, size_t oldlen,
const char *new, size_t newlen)
@@ -404,6 +429,15 @@ void mdc_rename_pack(struct ptlrpc_request *req, struct md_op_data *op_data,
if (new)
mdc_pack_name(req, &RMF_SYMTGT, new, newlen);
+
+ if (op_data->op_cli_flags & CLI_MIGRATE &&
+ op_data->op_bias & MDS_RENAME_MIGRATE) {
+ struct mdt_ioepoch *epoch;
+
+ mdc_intent_close_pack(req, op_data);
+ epoch = req_capsule_client_get(&req->rq_pill, &RMF_MDT_EPOCH);
+ mdc_ioepoch_pack(epoch, op_data);
+ }
}
void mdc_getattr_pack(struct ptlrpc_request *req, __u64 valid, u32 flags,
@@ -430,31 +464,6 @@ void mdc_getattr_pack(struct ptlrpc_request *req, __u64 valid, u32 flags,
op_data->op_namelen);
}
-static void mdc_intent_close_pack(struct ptlrpc_request *req,
- struct md_op_data *op_data)
-{
- enum mds_op_bias bias = op_data->op_bias;
- struct close_data *data;
- struct ldlm_lock *lock;
-
- if (!(bias & (MDS_HSM_RELEASE | MDS_CLOSE_LAYOUT_SWAP |
- MDS_RENAME_MIGRATE)))
- return;
-
- data = req_capsule_client_get(&req->rq_pill, &RMF_CLOSE_DATA);
- LASSERT(data);
-
- lock = ldlm_handle2lock(&op_data->op_lease_handle);
- if (lock) {
- data->cd_handle = lock->l_remote_handle;
- LDLM_LOCK_PUT(lock);
- }
- ldlm_cli_cancel(&op_data->op_lease_handle, LCF_LOCAL);
-
- data->cd_data_version = op_data->op_data_version;
- data->cd_fid = op_data->op_fid2;
-}
-
void mdc_close_pack(struct ptlrpc_request *req, struct md_op_data *op_data)
{
struct mdt_ioepoch *epoch;
diff --git a/drivers/staging/lustre/lustre/mdc/mdc_reint.c b/drivers/staging/lustre/lustre/mdc/mdc_reint.c
index 5119588..07b1684 100644
--- a/drivers/staging/lustre/lustre/mdc/mdc_reint.c
+++ b/drivers/staging/lustre/lustre/mdc/mdc_reint.c
@@ -366,7 +366,8 @@ int mdc_rename(struct obd_export *exp, struct md_op_data *op_data,
MDS_INODELOCK_FULL);
req = ptlrpc_request_alloc(class_exp2cliimp(exp),
- &RQF_MDS_REINT_RENAME);
+ op_data->op_cli_flags & CLI_MIGRATE ?
+ &RQF_MDS_REINT_MIGRATE : &RQF_MDS_REINT_RENAME);
if (!req) {
ldlm_lock_list_put(&cancels, l_bl_ast, count);
return -ENOMEM;
@@ -382,6 +383,23 @@ int mdc_rename(struct obd_export *exp, struct md_op_data *op_data,
return rc;
}
+ if (op_data->op_cli_flags & CLI_MIGRATE && op_data->op_data) {
+ struct md_open_data *mod = op_data->op_data;
+
+ LASSERTF(mod->mod_open_req &&
+ mod->mod_open_req->rq_type != LI_POISON,
+ "POISONED open %p!\n", mod->mod_open_req);
+
+ DEBUG_REQ(D_HA, mod->mod_open_req, "matched open");
+ /*
+ * We no longer want to preserve this open for replay even
+ * though the open was committed. b=3632, b=3633
+ */
+ spin_lock(&mod->mod_open_req->rq_lock);
+ mod->mod_open_req->rq_replay = 0;
+ spin_unlock(&mod->mod_open_req->rq_lock);
+ }
+
if (exp_connect_cancelset(exp) && req)
ldlm_cli_cancel_list(&cancels, count, req, 0);
diff --git a/drivers/staging/lustre/lustre/ptlrpc/layout.c b/drivers/staging/lustre/lustre/ptlrpc/layout.c
index 31aa58e..fd976f9 100644
--- a/drivers/staging/lustre/lustre/ptlrpc/layout.c
+++ b/drivers/staging/lustre/lustre/ptlrpc/layout.c
@@ -257,6 +257,18 @@
&RMF_DLM_REQ
};
+static const struct req_msg_field *mds_reint_migrate_client[] = {
+ &RMF_PTLRPC_BODY,
+ &RMF_REC_REINT,
+ &RMF_CAPA1,
+ &RMF_CAPA2,
+ &RMF_NAME,
+ &RMF_SYMTGT,
+ &RMF_DLM_REQ,
+ &RMF_MDT_EPOCH,
+ &RMF_CLOSE_DATA
+};
+
static const struct req_msg_field *mds_last_unlink_server[] = {
&RMF_PTLRPC_BODY,
&RMF_MDT_BODY,
@@ -678,6 +690,7 @@
&RQF_MDS_REINT_UNLINK,
&RQF_MDS_REINT_LINK,
&RQF_MDS_REINT_RENAME,
+ &RQF_MDS_REINT_MIGRATE,
&RQF_MDS_REINT_SETATTR,
&RQF_MDS_REINT_SETXATTR,
&RQF_MDS_QUOTACTL,
@@ -1254,6 +1267,11 @@ struct req_format RQF_MDS_REINT_RENAME =
mds_last_unlink_server);
EXPORT_SYMBOL(RQF_MDS_REINT_RENAME);
+struct req_format RQF_MDS_REINT_MIGRATE =
+ DEFINE_REQ_FMT0("MDS_REINT_MIGRATE", mds_reint_migrate_client,
+ mds_last_unlink_server);
+EXPORT_SYMBOL(RQF_MDS_REINT_MIGRATE);
+
struct req_format RQF_MDS_REINT_SETATTR =
DEFINE_REQ_FMT0("MDS_REINT_SETATTR",
mds_reint_setattr_client, mds_setattr_server);
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 04/22] staging: lustre: osc: handle osc eviction correctly
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (2 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 03/22] staging: lustre: mdt: race between open and migrate James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-05 20:55 ` Dan Carpenter
2016-12-03 0:53 ` [PATCH 05/22] staging: lustre: lmv: remove nlink check in lmv_revalidate_slaves James Simmons
` (18 subsequent siblings)
22 siblings, 1 reply; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Jinshan Xiong, James Simmons
From: Jinshan Xiong <jinshan.xiong@intel.com>
Cleanup everything if an OSC is being evicted.
Group lock is not well supported yet.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6271
Reviewed-on: http://review.whamcloud.com/14989
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/ldlm/ldlm_lock.c | 3 +-
drivers/staging/lustre/lustre/osc/osc_cache.c | 25 ++++----
.../staging/lustre/lustre/osc/osc_cl_internal.h | 11 ++-
drivers/staging/lustre/lustre/osc/osc_internal.h | 2 +
drivers/staging/lustre/lustre/osc/osc_io.c | 60 +++++++++++++++----
drivers/staging/lustre/lustre/osc/osc_object.c | 19 ++++++
drivers/staging/lustre/lustre/osc/osc_request.c | 42 +++++++++++--
7 files changed, 124 insertions(+), 38 deletions(-)
diff --git a/drivers/staging/lustre/lustre/ldlm/ldlm_lock.c b/drivers/staging/lustre/lustre/ldlm/ldlm_lock.c
index d03e6d4..c1665f9 100644
--- a/drivers/staging/lustre/lustre/ldlm/ldlm_lock.c
+++ b/drivers/staging/lustre/lustre/ldlm/ldlm_lock.c
@@ -1130,8 +1130,7 @@ static int lock_matches(struct ldlm_lock *lock, struct lock_match_data *data)
if (!data->lmd_unref && LDLM_HAVE_MASK(lock, GONE))
return INTERVAL_ITER_CONT;
- if ((data->lmd_flags & LDLM_FL_LOCAL_ONLY) &&
- !ldlm_is_local(lock))
+ if (!equi(data->lmd_flags & LDLM_FL_LOCAL_ONLY, ldlm_is_local(lock)))
return INTERVAL_ITER_CONT;
if (data->lmd_flags & LDLM_FL_TEST_LOCK) {
diff --git a/drivers/staging/lustre/lustre/osc/osc_cache.c b/drivers/staging/lustre/lustre/osc/osc_cache.c
index b0f030c..76942cc 100644
--- a/drivers/staging/lustre/lustre/osc/osc_cache.c
+++ b/drivers/staging/lustre/lustre/osc/osc_cache.c
@@ -247,7 +247,7 @@ static int osc_extent_sanity_check0(struct osc_extent *ext,
goto out;
}
- if (ext->oe_dlmlock) {
+ if (ext->oe_dlmlock && !ldlm_is_failed(ext->oe_dlmlock)) {
struct ldlm_extent *extent;
extent = &ext->oe_dlmlock->l_policy_data.l_extent;
@@ -2710,8 +2710,8 @@ int osc_queue_sync_pages(const struct lu_env *env, struct osc_object *obj,
/**
* Called by osc_io_setattr_start() to freeze and destroy covering extents.
*/
-int osc_cache_truncate_start(const struct lu_env *env, struct osc_io *oio,
- struct osc_object *obj, __u64 size)
+int osc_cache_truncate_start(const struct lu_env *env, struct osc_object *obj,
+ u64 size, struct osc_extent **extp)
{
struct client_obd *cli = osc_cli(obj);
struct osc_extent *ext;
@@ -2808,9 +2808,11 @@ int osc_cache_truncate_start(const struct lu_env *env, struct osc_io *oio,
/* we need to hold this extent in OES_TRUNC state so
* that no writeback will happen. This is to avoid
* BUG 17397.
+ * Only partial truncate can reach here, if @size is
+ * not zero, the caller should provide a valid @extp.
*/
- LASSERT(!oio->oi_trunc);
- oio->oi_trunc = osc_extent_get(ext);
+ LASSERT(!*extp);
+ *extp = osc_extent_get(ext);
OSC_EXTENT_DUMP(D_CACHE, ext,
"trunc at %llu\n", size);
}
@@ -2836,13 +2838,10 @@ int osc_cache_truncate_start(const struct lu_env *env, struct osc_io *oio,
/**
* Called after osc_io_setattr_end to add oio->oi_trunc back to cache.
*/
-void osc_cache_truncate_end(const struct lu_env *env, struct osc_io *oio,
- struct osc_object *obj)
+void osc_cache_truncate_end(const struct lu_env *env, struct osc_extent *ext)
{
- struct osc_extent *ext = oio->oi_trunc;
-
- oio->oi_trunc = NULL;
if (ext) {
+ struct osc_object *obj = ext->oe_obj;
bool unplug = false;
EASSERT(ext->oe_nr_pages > 0, ext);
@@ -3183,8 +3182,10 @@ static int discard_cb(const struct lu_env *env, struct cl_io *io,
/* page is top page. */
info->oti_next_index = osc_index(ops) + 1;
if (cl_page_own(env, io, page) == 0) {
- KLASSERT(ergo(page->cp_type == CPT_CACHEABLE,
- !PageDirty(cl_page_vmpage(page))));
+ if (!ergo(page->cp_type == CPT_CACHEABLE,
+ !PageDirty(cl_page_vmpage(page))))
+ CL_PAGE_DEBUG(D_ERROR, env, page,
+ "discard dirty page?\n");
/* discard the page */
cl_page_discard(env, io, page);
diff --git a/drivers/staging/lustre/lustre/osc/osc_cl_internal.h b/drivers/staging/lustre/lustre/osc/osc_cl_internal.h
index cce55a9..c09ab97 100644
--- a/drivers/staging/lustre/lustre/osc/osc_cl_internal.h
+++ b/drivers/staging/lustre/lustre/osc/osc_cl_internal.h
@@ -159,6 +159,10 @@ struct osc_object {
/* Protect osc_lock this osc_object has */
spinlock_t oo_ol_spin;
struct list_head oo_ol_list;
+
+ /** number of active IOs of this object */
+ atomic_t oo_nr_ios;
+ wait_queue_head_t oo_io_waitq;
};
static inline void osc_object_lock(struct osc_object *obj)
@@ -399,10 +403,9 @@ int osc_flush_async_page(const struct lu_env *env, struct cl_io *io,
struct osc_page *ops);
int osc_queue_sync_pages(const struct lu_env *env, struct osc_object *obj,
struct list_head *list, int cmd, int brw_flags);
-int osc_cache_truncate_start(const struct lu_env *env, struct osc_io *oio,
- struct osc_object *obj, __u64 size);
-void osc_cache_truncate_end(const struct lu_env *env, struct osc_io *oio,
- struct osc_object *obj);
+int osc_cache_truncate_start(const struct lu_env *env, struct osc_object *obj,
+ u64 size, struct osc_extent **extp);
+void osc_cache_truncate_end(const struct lu_env *env, struct osc_extent *ext);
int osc_cache_writeback_range(const struct lu_env *env, struct osc_object *obj,
pgoff_t start, pgoff_t end, int hp, int discard);
int osc_cache_wait_range(const struct lu_env *env, struct osc_object *obj,
diff --git a/drivers/staging/lustre/lustre/osc/osc_internal.h b/drivers/staging/lustre/lustre/osc/osc_internal.h
index 688783d..d9837b7 100644
--- a/drivers/staging/lustre/lustre/osc/osc_internal.h
+++ b/drivers/staging/lustre/lustre/osc/osc_internal.h
@@ -218,4 +218,6 @@ struct ldlm_lock *osc_dlmlock_at_pgoff(const struct lu_env *env,
struct osc_object *obj, pgoff_t index,
enum osc_dap_flags flags);
+int osc_object_invalidate(const struct lu_env *env, struct osc_object *osc);
+
#endif /* OSC_INTERNAL_H */
diff --git a/drivers/staging/lustre/lustre/osc/osc_io.c b/drivers/staging/lustre/lustre/osc/osc_io.c
index 439ba64..369761f 100644
--- a/drivers/staging/lustre/lustre/osc/osc_io.c
+++ b/drivers/staging/lustre/lustre/osc/osc_io.c
@@ -329,8 +329,25 @@ static int osc_io_commit_async(const struct lu_env *env,
return result;
}
-static int osc_io_rw_iter_init(const struct lu_env *env,
- const struct cl_io_slice *ios)
+static int osc_io_iter_init(const struct lu_env *env,
+ const struct cl_io_slice *ios)
+{
+ struct osc_object *osc = cl2osc(ios->cis_obj);
+ struct obd_import *imp = osc_cli(osc)->cl_import;
+ int rc = -EIO;
+
+ spin_lock(&imp->imp_lock);
+ if (likely(!imp->imp_invalid)) {
+ atomic_inc(&osc->oo_nr_ios);
+ rc = 0;
+ }
+ spin_unlock(&imp->imp_lock);
+
+ return rc;
+}
+
+static int osc_io_write_iter_init(const struct lu_env *env,
+ const struct cl_io_slice *ios)
{
struct cl_io *io = ios->cis_io;
struct osc_io *oio = osc_env_io(env);
@@ -341,7 +358,7 @@ static int osc_io_rw_iter_init(const struct lu_env *env,
unsigned long max_pages;
if (cl_io_is_append(io))
- return 0;
+ return osc_io_iter_init(env, ios);
npages = io->u.ci_rw.crw_count >> PAGE_SHIFT;
if (io->u.ci_rw.crw_pos & ~PAGE_MASK)
@@ -373,11 +390,21 @@ static int osc_io_rw_iter_init(const struct lu_env *env,
(void)ptlrpcd_queue_work(cli->cl_lru_work);
}
- return 0;
+ return osc_io_iter_init(env, ios);
+}
+
+static void osc_io_iter_fini(const struct lu_env *env,
+ const struct cl_io_slice *ios)
+{
+ struct osc_object *osc = cl2osc(ios->cis_obj);
+
+ LASSERT(atomic_read(&osc->oo_nr_ios) > 0);
+ if (atomic_dec_and_test(&osc->oo_nr_ios))
+ wake_up_all(&osc->oo_io_waitq);
}
-static void osc_io_rw_iter_fini(const struct lu_env *env,
- const struct cl_io_slice *ios)
+static void osc_io_write_iter_fini(const struct lu_env *env,
+ const struct cl_io_slice *ios)
{
struct osc_io *oio = osc_env_io(env);
struct osc_object *osc = cl2osc(ios->cis_obj);
@@ -388,6 +415,8 @@ static void osc_io_rw_iter_fini(const struct lu_env *env,
oio->oi_lru_reserved = 0;
}
oio->oi_write_osclock = NULL;
+
+ osc_io_iter_fini(env, ios);
}
static int osc_io_fault_start(const struct lu_env *env,
@@ -478,7 +507,8 @@ static int osc_io_setattr_start(const struct lu_env *env,
/* truncate cache dirty pages first */
if (cl_io_is_trunc(io))
- result = osc_cache_truncate_start(env, oio, cl2osc(obj), size);
+ result = osc_cache_truncate_start(env, cl2osc(obj), size,
+ &oio->oi_trunc);
if (result == 0 && oio->oi_lockless == 0) {
cl_object_attr_lock(obj);
@@ -588,10 +618,8 @@ static void osc_io_setattr_end(const struct lu_env *env,
__u64 size = io->u.ci_setattr.sa_attr.lvb_size;
osc_trunc_check(env, io, oio, size);
- if (oio->oi_trunc) {
- osc_cache_truncate_end(env, oio, cl2osc(obj));
- oio->oi_trunc = NULL;
- }
+ osc_cache_truncate_end(env, oio->oi_trunc);
+ oio->oi_trunc = NULL;
}
}
@@ -831,17 +859,21 @@ static void osc_io_end(const struct lu_env *env,
static const struct cl_io_operations osc_io_ops = {
.op = {
[CIT_READ] = {
+ .cio_iter_init = osc_io_iter_init,
+ .cio_iter_fini = osc_io_iter_fini,
.cio_start = osc_io_read_start,
.cio_fini = osc_io_fini
},
[CIT_WRITE] = {
- .cio_iter_init = osc_io_rw_iter_init,
- .cio_iter_fini = osc_io_rw_iter_fini,
+ .cio_iter_init = osc_io_write_iter_init,
+ .cio_iter_fini = osc_io_write_iter_fini,
.cio_start = osc_io_write_start,
.cio_end = osc_io_end,
.cio_fini = osc_io_fini
},
[CIT_SETATTR] = {
+ .cio_iter_init = osc_io_iter_init,
+ .cio_iter_fini = osc_io_iter_fini,
.cio_start = osc_io_setattr_start,
.cio_end = osc_io_setattr_end
},
@@ -850,6 +882,8 @@ static void osc_io_end(const struct lu_env *env,
.cio_end = osc_io_data_version_end,
},
[CIT_FAULT] = {
+ .cio_iter_init = osc_io_iter_init,
+ .cio_iter_fini = osc_io_iter_fini,
.cio_start = osc_io_fault_start,
.cio_end = osc_io_end,
.cio_fini = osc_io_fini
diff --git a/drivers/staging/lustre/lustre/osc/osc_object.c b/drivers/staging/lustre/lustre/osc/osc_object.c
index e0c3324..d3e5ca7 100644
--- a/drivers/staging/lustre/lustre/osc/osc_object.c
+++ b/drivers/staging/lustre/lustre/osc/osc_object.c
@@ -78,6 +78,9 @@ static int osc_object_init(const struct lu_env *env, struct lu_object *obj,
INIT_LIST_HEAD(&osc->oo_write_item);
INIT_LIST_HEAD(&osc->oo_read_item);
+ atomic_set(&osc->oo_nr_ios, 0);
+ init_waitqueue_head(&osc->oo_io_waitq);
+
osc->oo_root.rb_node = NULL;
INIT_LIST_HEAD(&osc->oo_hp_exts);
INIT_LIST_HEAD(&osc->oo_urgent_exts);
@@ -112,6 +115,7 @@ static void osc_object_free(const struct lu_env *env, struct lu_object *obj)
LASSERT(atomic_read(&osc->oo_nr_reads) == 0);
LASSERT(atomic_read(&osc->oo_nr_writes) == 0);
LASSERT(list_empty(&osc->oo_ol_list));
+ LASSERT(!atomic_read(&osc->oo_nr_ios));
lu_object_fini(obj);
kmem_cache_free(osc_object_kmem, osc);
@@ -444,4 +448,19 @@ struct lu_object *osc_object_alloc(const struct lu_env *env,
return obj;
}
+int osc_object_invalidate(const struct lu_env *env, struct osc_object *osc)
+{
+ struct l_wait_info lwi = { 0 };
+
+ CDEBUG(D_INODE, "Invalidate osc object: %p, # of active IOs: %d\n",
+ osc, atomic_read(&osc->oo_nr_ios));
+
+ l_wait_event(osc->oo_io_waitq, !atomic_read(&osc->oo_nr_ios), &lwi);
+
+ /* Discard all pages of this object. */
+ osc_cache_truncate_start(env, osc, 0, NULL);
+
+ return 0;
+}
+
/** @} osc */
diff --git a/drivers/staging/lustre/lustre/osc/osc_request.c b/drivers/staging/lustre/lustre/osc/osc_request.c
index 5197768..bc698d3 100644
--- a/drivers/staging/lustre/lustre/osc/osc_request.c
+++ b/drivers/staging/lustre/lustre/osc/osc_request.c
@@ -2501,6 +2501,33 @@ static int osc_disconnect(struct obd_export *exp)
return rc;
}
+static int osc_ldlm_resource_invalidate(struct cfs_hash *hs,
+ struct cfs_hash_bd *bd,
+ struct hlist_node *hnode, void *arg)
+{
+ struct ldlm_resource *res = cfs_hash_object(hs, hnode);
+ struct osc_object *osc = NULL;
+ struct lu_env *env = arg;
+ struct ldlm_lock *lock;
+
+ lock_res(res);
+ list_for_each_entry(lock, &res->lr_granted, l_res_link) {
+ if (lock->l_ast_data && !osc) {
+ osc = lock->l_ast_data;
+ cl_object_get(osc2cl(osc));
+ }
+ lock->l_ast_data = NULL;
+ }
+ unlock_res(res);
+
+ if (osc) {
+ osc_object_invalidate(env, osc);
+ cl_object_put(env, osc2cl(osc));
+ }
+
+ return 0;
+}
+
static int osc_import_event(struct obd_device *obd,
struct obd_import *imp,
enum obd_import_event event)
@@ -2528,17 +2555,18 @@ static int osc_import_event(struct obd_device *obd,
struct lu_env *env;
int refcheck;
+ ldlm_namespace_cleanup(ns, LDLM_FL_LOCAL_ONLY);
+
env = cl_env_get(&refcheck);
if (!IS_ERR(env)) {
- /* Reset grants */
- cli = &obd->u.cli;
- /* all pages go to failing rpcs due to the invalid
- * import
- */
- osc_io_unplug(env, cli, NULL);
+ osc_io_unplug(env, &obd->u.cli, NULL);
- ldlm_namespace_cleanup(ns, LDLM_FL_LOCAL_ONLY);
+ cfs_hash_for_each_nolock(ns->ns_rs_hash,
+ osc_ldlm_resource_invalidate,
+ env, 0);
cl_env_put(env, &refcheck);
+
+ ldlm_namespace_cleanup(ns, LDLM_FL_LOCAL_ONLY);
} else {
rc = PTR_ERR(env);
}
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* Re: [PATCH 04/22] staging: lustre: osc: handle osc eviction correctly
2016-12-03 0:53 ` [PATCH 04/22] staging: lustre: osc: handle osc eviction correctly James Simmons
@ 2016-12-05 20:55 ` Dan Carpenter
2016-12-05 23:03 ` Oleg Drokin
0 siblings, 1 reply; 29+ messages in thread
From: Dan Carpenter @ 2016-12-05 20:55 UTC (permalink / raw)
To: James Simmons
Cc: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin,
Linux Kernel Mailing List, Jinshan Xiong,
Lustre Development List
On Fri, Dec 02, 2016 at 07:53:11PM -0500, James Simmons wrote:
> @@ -3183,8 +3182,10 @@ static int discard_cb(const struct lu_env *env, struct cl_io *io,
> /* page is top page. */
> info->oti_next_index = osc_index(ops) + 1;
> if (cl_page_own(env, io, page) == 0) {
> - KLASSERT(ergo(page->cp_type == CPT_CACHEABLE,
> - !PageDirty(cl_page_vmpage(page))));
> + if (!ergo(page->cp_type == CPT_CACHEABLE,
> + !PageDirty(cl_page_vmpage(page))))
> + CL_PAGE_DEBUG(D_ERROR, env, page,
> + "discard dirty page?\n");
I don't understand the point of the ergo macro. There are way too many
double negatives (some of them hidden for my small brain). How is that
simpler than just writing it out:
if (page->cp_type == CPT_CACHEABLE &&
PageDirty(cl_page_vmpage(page))
CL_PAGE_DEBUG(D_ERROR, env, page, "discard dirty page?\n");
regards,
dan carpenter
^ permalink raw reply [flat|nested] 29+ messages in thread* Re: [PATCH 04/22] staging: lustre: osc: handle osc eviction correctly
2016-12-05 20:55 ` Dan Carpenter
@ 2016-12-05 23:03 ` Oleg Drokin
2016-12-07 23:16 ` James Simmons
0 siblings, 1 reply; 29+ messages in thread
From: Oleg Drokin @ 2016-12-05 23:03 UTC (permalink / raw)
To: Dan Carpenter
Cc: James Simmons, Greg Kroah-Hartman, devel, Andreas Dilger,
Linux Kernel Mailing List, Jinshan Xiong,
Lustre Development List
On Dec 5, 2016, at 3:55 PM, Dan Carpenter wrote:
> On Fri, Dec 02, 2016 at 07:53:11PM -0500, James Simmons wrote:
>> @@ -3183,8 +3182,10 @@ static int discard_cb(const struct lu_env *env, struct cl_io *io,
>> /* page is top page. */
>> info->oti_next_index = osc_index(ops) + 1;
>> if (cl_page_own(env, io, page) == 0) {
>> - KLASSERT(ergo(page->cp_type == CPT_CACHEABLE,
>> - !PageDirty(cl_page_vmpage(page))));
>> + if (!ergo(page->cp_type == CPT_CACHEABLE,
>> + !PageDirty(cl_page_vmpage(page))))
>> + CL_PAGE_DEBUG(D_ERROR, env, page,
>> + "discard dirty page?\n");
>
>
> I don't understand the point of the ergo macro. There are way too many
> double negatives (some of them hidden for my small brain). How is that
> simpler than just writing it out:
>
> if (page->cp_type == CPT_CACHEABLE &&
> PageDirty(cl_page_vmpage(page))
> CL_PAGE_DEBUG(D_ERROR, env, page, "discard dirty page?\n");
I guess it makes it sound chic or something?
I am not a huge fan of it either, esp. in a case like this, though
it might be somewhat more convenient in assertions (where this is converted from).
^ permalink raw reply [flat|nested] 29+ messages in thread* Re: [PATCH 04/22] staging: lustre: osc: handle osc eviction correctly
2016-12-05 23:03 ` Oleg Drokin
@ 2016-12-07 23:16 ` James Simmons
0 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-07 23:16 UTC (permalink / raw)
To: Oleg Drokin
Cc: Dan Carpenter, Greg Kroah-Hartman, devel, Andreas Dilger,
Linux Kernel Mailing List, Jinshan Xiong,
Lustre Development List
> On Dec 5, 2016, at 3:55 PM, Dan Carpenter wrote:
>
> > On Fri, Dec 02, 2016 at 07:53:11PM -0500, James Simmons wrote:
> >> @@ -3183,8 +3182,10 @@ static int discard_cb(const struct lu_env *env, struct cl_io *io,
> >> /* page is top page. */
> >> info->oti_next_index = osc_index(ops) + 1;
> >> if (cl_page_own(env, io, page) == 0) {
> >> - KLASSERT(ergo(page->cp_type == CPT_CACHEABLE,
> >> - !PageDirty(cl_page_vmpage(page))));
> >> + if (!ergo(page->cp_type == CPT_CACHEABLE,
> >> + !PageDirty(cl_page_vmpage(page))))
> >> + CL_PAGE_DEBUG(D_ERROR, env, page,
> >> + "discard dirty page?\n");
> >
> >
> > I don't understand the point of the ergo macro. There are way too many
> > double negatives (some of them hidden for my small brain). How is that
> > simpler than just writing it out:
> >
> > if (page->cp_type == CPT_CACHEABLE &&
> > PageDirty(cl_page_vmpage(page))
> > CL_PAGE_DEBUG(D_ERROR, env, page, "discard dirty page?\n");
>
> I guess it makes it sound chic or something?
> I am not a huge fan of it either, esp. in a case like this, though
> it might be somewhat more convenient in assertions (where this is converted from).
Not a fan either. Resubmitted patch with ergo removed.
^ permalink raw reply [flat|nested] 29+ messages in thread
* [PATCH 05/22] staging: lustre: lmv: remove nlink check in lmv_revalidate_slaves
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (3 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 04/22] staging: lustre: osc: handle osc eviction correctly James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-05 20:57 ` Dan Carpenter
2016-12-03 0:53 ` [PATCH 06/22] staging: lustre: llog: reset llog bitmap James Simmons
` (17 subsequent siblings)
22 siblings, 1 reply; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, wang di,
James Simmons
From: wang di <di.wang@intel.com>
Remove nlink < 2 check in lmv_revalidate_slaves, because
after nlink reaches to LDISKFS_LINK_MAX (65000), the inode
nlink will be set to 1.
Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6984
Reviewed-on: http://review.whamcloud.com/16490
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Jian Yu <jian.yu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/lmv/lmv_intent.c | 16 +---------------
1 files changed, 1 insertions(+), 15 deletions(-)
diff --git a/drivers/staging/lustre/lustre/lmv/lmv_intent.c b/drivers/staging/lustre/lustre/lmv/lmv_intent.c
index b1071cf..aa42066 100644
--- a/drivers/staging/lustre/lustre/lmv/lmv_intent.c
+++ b/drivers/staging/lustre/lustre/lmv/lmv_intent.c
@@ -220,21 +220,7 @@ int lmv_revalidate_slaves(struct obd_export *exp,
/* refresh slave from server */
body = req_capsule_server_get(&req->rq_pill,
&RMF_MDT_BODY);
- LASSERT(body);
-
- if (unlikely(body->mbo_nlink < 2)) {
- /*
- * If this is bad stripe, most likely due
- * to the race between close(unlink) and
- * getattr, let's return -EONENT, so llite
- * will revalidate the dentry see
- * ll_inode_revalidate_fini()
- */
- CDEBUG(D_INODE, "%s: nlink %d < 2 corrupt stripe %d "DFID":" DFID"\n",
- obd->obd_name, body->mbo_nlink, i,
- PFID(&lsm->lsm_md_oinfo[i].lmo_fid),
- PFID(&lsm->lsm_md_oinfo[0].lmo_fid));
-
+ if (!body) {
if (it.it_lock_mode && lockh) {
ldlm_lock_decref(lockh, it.it_lock_mode);
it.it_lock_mode = 0;
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* Re: [PATCH 05/22] staging: lustre: lmv: remove nlink check in lmv_revalidate_slaves
2016-12-03 0:53 ` [PATCH 05/22] staging: lustre: lmv: remove nlink check in lmv_revalidate_slaves James Simmons
@ 2016-12-05 20:57 ` Dan Carpenter
0 siblings, 0 replies; 29+ messages in thread
From: Dan Carpenter @ 2016-12-05 20:57 UTC (permalink / raw)
To: James Simmons
Cc: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin, wang di,
Linux Kernel Mailing List, Lustre Development List
On Fri, Dec 02, 2016 at 07:53:12PM -0500, James Simmons wrote:
> From: wang di <di.wang@intel.com>
>
> Remove nlink < 2 check in lmv_revalidate_slaves, because
> after nlink reaches to LDISKFS_LINK_MAX (65000), the inode
> nlink will be set to 1.
>
I don't understand how this changelog matches the patch...
The changelog says we're removing a check but really this is a NULL
dereference fix, right? We're not really removing a check at all, just
changing it for something else. We do remove a printk, though.
regards,
dan carpenter
^ permalink raw reply [flat|nested] 29+ messages in thread
* [PATCH 06/22] staging: lustre: llog: reset llog bitmap
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (4 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 05/22] staging: lustre: lmv: remove nlink check in lmv_revalidate_slaves James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 07/22] staging: lustre: obdclass: lu_site_purge() to handle purge-all James Simmons
` (16 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, wang di,
James Simmons
From: wang di <di.wang@intel.com>
Once update request fails due to eviction or other failures,
all of update request in the sending list should return fail,
because after the failure, the update log in the following
request will have wrong llog bitmap.
Signed-off-by: wang di <di.wang@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7039
Reviewed-on: http://review.whamcloud.com/16969
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.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>
---
drivers/staging/lustre/lustre/obdclass/llog.c | 7 +++++++
1 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/drivers/staging/lustre/lustre/obdclass/llog.c b/drivers/staging/lustre/lustre/obdclass/llog.c
index ae63047..65e6ce6 100644
--- a/drivers/staging/lustre/lustre/obdclass/llog.c
+++ b/drivers/staging/lustre/lustre/obdclass/llog.c
@@ -114,6 +114,7 @@ static int llog_read_header(const struct lu_env *env,
rc = lop->lop_read_header(env, handle);
if (rc == LLOG_EEMPTY) {
struct llog_log_hdr *llh = handle->lgh_hdr;
+ size_t len;
/* lrh_len should be initialized in llog_init_handle */
handle->lgh_last_idx = 0; /* header is record with index 0 */
@@ -127,6 +128,12 @@ static int llog_read_header(const struct lu_env *env,
memcpy(&llh->llh_tgtuuid, uuid,
sizeof(llh->llh_tgtuuid));
llh->llh_bitmap_offset = offsetof(typeof(*llh), llh_bitmap);
+ /*
+ * Since update llog header might also call this function,
+ * let's reset the bitmap to 0 here
+ */
+ len = llh->llh_hdr.lrh_len - llh->llh_bitmap_offset;
+ memset(LLOG_HDR_BITMAP(llh), 0, len - sizeof(llh->llh_tail));
ext2_set_bit(0, LLOG_HDR_BITMAP(llh));
LLOG_HDR_TAIL(llh)->lrt_len = llh->llh_hdr.lrh_len;
LLOG_HDR_TAIL(llh)->lrt_index = llh->llh_hdr.lrh_index;
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 07/22] staging: lustre: obdclass: lu_site_purge() to handle purge-all
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (5 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 06/22] staging: lustre: llog: reset llog bitmap James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 08/22] staging: lustre: clio: revise read ahead algorithm James Simmons
` (15 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Alex Zhuravlev, James Simmons
From: Alex Zhuravlev <alexey.zhuravlev@intel.com>
if the callers wants to purge all objects, then scanning
should start from the first bucket.
Signed-off-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7038
Reviewed-on: http://review.whamcloud.com/18505
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Faccini Bruno <bruno.faccini@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/obdclass/lu_object.c | 5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/staging/lustre/lustre/obdclass/lu_object.c b/drivers/staging/lustre/lustre/obdclass/lu_object.c
index 43868ed..a02aaa3 100644
--- a/drivers/staging/lustre/lustre/obdclass/lu_object.c
+++ b/drivers/staging/lustre/lustre/obdclass/lu_object.c
@@ -338,7 +338,7 @@ int lu_site_purge(const struct lu_env *env, struct lu_site *s, int nr)
struct cfs_hash_bd bd2;
struct list_head dispose;
int did_sth;
- unsigned int start;
+ unsigned int start = 0;
int count;
int bnr;
unsigned int i;
@@ -351,7 +351,8 @@ int lu_site_purge(const struct lu_env *env, struct lu_site *s, int nr)
* Under LRU list lock, scan LRU list and move unreferenced objects to
* the dispose list, removing them from LRU and hash table.
*/
- start = s->ls_purge_start;
+ if (nr != ~0)
+ start = s->ls_purge_start;
bnr = (nr == ~0) ? -1 : nr / (int)CFS_HASH_NBKT(s->ls_obj_hash) + 1;
again:
/*
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 08/22] staging: lustre: clio: revise read ahead algorithm
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (6 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 07/22] staging: lustre: obdclass: lu_site_purge() to handle purge-all James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 09/22] staging: lustre: llite: Add client mount opt to ignore suppress_pings James Simmons
` (14 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Jinshan Xiong, James Simmons
From: Jinshan Xiong <jinshan.xiong@intel.com>
ras_window_len should only be updated in ras_update() by read
pattern and it can't be adjusted in ll_readahead() at all;
ras_consecutive_pages is used to detect read pattern from
mmap. It will be used to increase read ahead window length
gradually.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-5505
Reviewed-on: http://review.whamcloud.com/11528
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
.../staging/lustre/lustre/llite/llite_internal.h | 5 +-
drivers/staging/lustre/lustre/llite/rw.c | 71 +++++++++++---------
2 files changed, 43 insertions(+), 33 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/llite_internal.h b/drivers/staging/lustre/lustre/llite/llite_internal.h
index ae0bb09..e37ba1f 100644
--- a/drivers/staging/lustre/lustre/llite/llite_internal.h
+++ b/drivers/staging/lustre/lustre/llite/llite_internal.h
@@ -1005,8 +1005,11 @@ int ll_xattr_list(struct inode *inode, const char *name, int type,
*/
int cl_sb_init(struct super_block *sb);
int cl_sb_fini(struct super_block *sb);
-void ll_io_init(struct cl_io *io, const struct file *file, int write);
+enum ras_update_flags {
+ LL_RAS_HIT = 0x1,
+ LL_RAS_MMAP = 0x2
+};
void ll_ra_count_put(struct ll_sb_info *sbi, unsigned long len);
void ll_ra_stats_inc(struct inode *inode, enum ra_stat which);
diff --git a/drivers/staging/lustre/lustre/llite/rw.c b/drivers/staging/lustre/lustre/llite/rw.c
index e34017d..e2d5e75 100644
--- a/drivers/staging/lustre/lustre/llite/rw.c
+++ b/drivers/staging/lustre/lustre/llite/rw.c
@@ -457,30 +457,25 @@ static int ll_readahead(const struct lu_env *env, struct cl_io *io,
spin_lock(&ras->ras_lock);
- /* Enlarge the RA window to encompass the full read */
- if (vio->vui_ra_valid &&
- ras->ras_window_start + ras->ras_window_len <
- vio->vui_ra_start + vio->vui_ra_count) {
- ras->ras_window_len = vio->vui_ra_start + vio->vui_ra_count -
- ras->ras_window_start;
- }
+ /**
+ * Note: other thread might rollback the ras_next_readahead,
+ * if it can not get the full size of prepared pages, see the
+ * end of this function. For stride read ahead, it needs to
+ * make sure the offset is no less than ras_stride_offset,
+ * so that stride read ahead can work correctly.
+ */
+ if (stride_io_mode(ras))
+ start = max(ras->ras_next_readahead, ras->ras_stride_offset);
+ else
+ start = ras->ras_next_readahead;
- /* Reserve a part of the read-ahead window that we'll be issuing */
- if (ras->ras_window_len > 0) {
- /*
- * Note: other thread might rollback the ras_next_readahead,
- * if it can not get the full size of prepared pages, see the
- * end of this function. For stride read ahead, it needs to
- * make sure the offset is no less than ras_stride_offset,
- * so that stride read ahead can work correctly.
- */
- if (stride_io_mode(ras))
- start = max(ras->ras_next_readahead,
- ras->ras_stride_offset);
- else
- start = ras->ras_next_readahead;
+ if (ras->ras_window_len > 0)
end = ras->ras_window_start + ras->ras_window_len - 1;
- }
+
+ /* Enlarge the RA window to encompass the full read */
+ if (vio->vui_ra_valid &&
+ end < vio->vui_ra_start + vio->vui_ra_count - 1)
+ end = vio->vui_ra_start + vio->vui_ra_count - 1;
if (end != 0) {
unsigned long rpc_boundary;
@@ -602,7 +597,7 @@ static void ras_reset(struct inode *inode, struct ll_readahead_state *ras,
ras->ras_consecutive_pages = 0;
ras->ras_window_len = 0;
ras_set_start(inode, ras, index);
- ras->ras_next_readahead = max(ras->ras_window_start, index);
+ ras->ras_next_readahead = max(ras->ras_window_start, index + 1);
RAS_CDEBUG(ras);
}
@@ -733,10 +728,11 @@ static void ras_increase_window(struct inode *inode,
static void ras_update(struct ll_sb_info *sbi, struct inode *inode,
struct ll_readahead_state *ras, unsigned long index,
- unsigned int hit)
+ enum ras_update_flags flags)
{
struct ll_ra_info *ra = &sbi->ll_ra_info;
int zero = 0, stride_detect = 0, ra_miss = 0;
+ bool hit = flags & LL_RAS_HIT;
spin_lock(&ras->ras_lock);
@@ -766,7 +762,7 @@ static void ras_update(struct ll_sb_info *sbi, struct inode *inode,
* to for subsequent IO. The mmap case does not increment
* ras_requests and thus can never trigger this behavior.
*/
- if (ras->ras_requests == 2 && !ras->ras_request_index) {
+ if (ras->ras_requests >= 2 && !ras->ras_request_index) {
__u64 kms_pages;
kms_pages = (i_size_read(inode) + PAGE_SIZE - 1) >>
@@ -778,8 +774,7 @@ static void ras_update(struct ll_sb_info *sbi, struct inode *inode,
if (kms_pages &&
kms_pages <= ra->ra_max_read_ahead_whole_pages) {
ras->ras_window_start = 0;
- ras->ras_last_readpage = 0;
- ras->ras_next_readahead = 0;
+ ras->ras_next_readahead = index + 1;
ras->ras_window_len = min(ra->ra_max_pages_per_file,
ra->ra_max_read_ahead_whole_pages);
goto out_unlock;
@@ -867,8 +862,13 @@ static void ras_update(struct ll_sb_info *sbi, struct inode *inode,
/* Trigger RA in the mmap case where ras_consecutive_requests
* is not incremented and thus can't be used to trigger RA
*/
- if (!ras->ras_window_len && ras->ras_consecutive_pages == 4) {
- ras->ras_window_len = RAS_INCREASE_STEP(inode);
+ if (ras->ras_consecutive_pages >= 4 && flags & LL_RAS_MMAP) {
+ ras_increase_window(inode, ras, ra);
+ /*
+ * reset consecutive pages so that the readahead window can
+ * grow gradually.
+ */
+ ras->ras_consecutive_pages = 0;
goto out_unlock;
}
@@ -1101,9 +1101,16 @@ static int ll_io_read_page(const struct lu_env *env, struct cl_io *io,
vpg = cl2vvp_page(cl_object_page_slice(page->cp_obj, page));
if (sbi->ll_ra_info.ra_max_pages_per_file > 0 &&
- sbi->ll_ra_info.ra_max_pages > 0)
- ras_update(sbi, inode, ras, vvp_index(vpg),
- vpg->vpg_defer_uptodate);
+ sbi->ll_ra_info.ra_max_pages > 0) {
+ struct vvp_io *vio = vvp_env_io(env);
+ enum ras_update_flags flags = 0;
+
+ if (vpg->vpg_defer_uptodate)
+ flags |= LL_RAS_HIT;
+ if (!vio->vui_ra_valid)
+ flags |= LL_RAS_MMAP;
+ ras_update(sbi, inode, ras, vvp_index(vpg), flags);
+ }
if (vpg->vpg_defer_uptodate) {
vpg->vpg_ra_used = 1;
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 09/22] staging: lustre: llite: Add client mount opt to ignore suppress_pings
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (7 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 08/22] staging: lustre: clio: revise read ahead algorithm James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 10/22] staging: lustre: obdclass: limit lu_site hash table size on clients James Simmons
` (13 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Wally Wang,
James Simmons
From: Wally Wang <wang@cray.com>
When Lustre servers enable 'suppress_pings', all clients will stop
pinging. However, some clients may not have external mechanism
to notify Lustre servers for node death and therefore need to
preserve the Lustre ping.
This patch provides a mount option 'always_ping' so that the
client will not stop pinging even if the server has enabled
'suppress_pings'.
Signed-off-by: Wally Wang <wang@cray.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6391
Reviewed-on: http://review.whamcloud.com/14127
Reviewed-by: Li Wei <wei.g.li@intel.com>
Reviewed-by: Chris Horn <hornc@cray.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>
---
.../staging/lustre/lustre/llite/llite_internal.h | 3 +++
drivers/staging/lustre/lustre/llite/llite_lib.c | 16 ++++++++++++++++
2 files changed, 19 insertions(+), 0 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/llite_internal.h b/drivers/staging/lustre/lustre/llite/llite_internal.h
index e37ba1f..2f46d47 100644
--- a/drivers/staging/lustre/lustre/llite/llite_internal.h
+++ b/drivers/staging/lustre/lustre/llite/llite_internal.h
@@ -391,6 +391,8 @@ enum stats_track_type {
#define LL_SBI_USER_FID2PATH 0x40000 /* allow fid2path by unprivileged users */
#define LL_SBI_XATTR_CACHE 0x80000 /* support for xattr cache */
#define LL_SBI_NOROOTSQUASH 0x100000 /* do not apply root squash */
+#define LL_SBI_ALWAYS_PING 0x200000 /* always ping even if server
+ * suppress_pings */
#define LL_SBI_FLAGS { \
"nolck", \
@@ -414,6 +416,7 @@ enum stats_track_type {
"user_fid2path",\
"xattr_cache", \
"norootsquash", \
+ "always_ping", \
}
/*
diff --git a/drivers/staging/lustre/lustre/llite/llite_lib.c b/drivers/staging/lustre/lustre/llite/llite_lib.c
index 2a51efa..25f5aed 100644
--- a/drivers/staging/lustre/lustre/llite/llite_lib.c
+++ b/drivers/staging/lustre/lustre/llite/llite_lib.c
@@ -224,6 +224,10 @@ static int client_common_fill_super(struct super_block *sb, char *md, char *dt,
/* real client */
data->ocd_connect_flags |= OBD_CONNECT_REAL;
+ /* always ping even if server suppress_pings */
+ if (sbi->ll_flags & LL_SBI_ALWAYS_PING)
+ data->ocd_connect_flags &= ~OBD_CONNECT_PINGLESS;
+
data->ocd_brw_size = MD_MAX_BRW_SIZE;
err = obd_connect(NULL, &sbi->ll_md_exp, obd, &sbi->ll_sb_uuid,
@@ -373,6 +377,10 @@ static int client_common_fill_super(struct super_block *sb, char *md, char *dt,
data->ocd_connect_flags |= OBD_CONNECT_LRU_RESIZE;
+ /* always ping even if server suppress_pings */
+ if (sbi->ll_flags & LL_SBI_ALWAYS_PING)
+ data->ocd_connect_flags &= ~OBD_CONNECT_PINGLESS;
+
CDEBUG(D_RPCTRACE, "ocd_connect_flags: %#llx ocd_version: %d ocd_grant: %d\n",
data->ocd_connect_flags,
data->ocd_version, data->ocd_grant);
@@ -788,6 +796,11 @@ static int ll_options(char *options, int *flags)
*flags &= ~tmp;
goto next;
}
+ tmp = ll_set_opt("always_ping", s1, LL_SBI_ALWAYS_PING);
+ if (tmp) {
+ *flags |= tmp;
+ goto next;
+ }
LCONSOLE_ERROR_MSG(0x152, "Unknown option '%s', won't mount.\n",
s1);
return -EINVAL;
@@ -2361,6 +2374,9 @@ int ll_show_options(struct seq_file *seq, struct dentry *dentry)
if (sbi->ll_flags & LL_SBI_USER_FID2PATH)
seq_puts(seq, ",user_fid2path");
+ if (sbi->ll_flags & LL_SBI_ALWAYS_PING)
+ seq_puts(seq, ",always_ping");
+
return 0;
}
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 10/22] staging: lustre: obdclass: limit lu_site hash table size on clients
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (8 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 09/22] staging: lustre: llite: Add client mount opt to ignore suppress_pings James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 11/22] staging: lustre: mdt: fail FMODE_WRITE open if the client is read only James Simmons
` (12 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Li Dongyang,
James Simmons
From: Li Dongyang <dongyang.li@anu.edu.au>
Allocating a big hash table using the current formula
does not really work for clients. We will create new
hash table for each mount on a single client which is
a lot of memory more than expected.
This patch limits the hash table up to 8M for clients,
which has 524288 entries.
Signed-off-by: Li Dongyang <dongyang.li@anu.edu.au>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7689
Reviewed-on: http://review.whamcloud.com/18048
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/obdclass/lu_object.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/drivers/staging/lustre/lustre/obdclass/lu_object.c b/drivers/staging/lustre/lustre/obdclass/lu_object.c
index a02aaa3..80e0984 100644
--- a/drivers/staging/lustre/lustre/obdclass/lu_object.c
+++ b/drivers/staging/lustre/lustre/obdclass/lu_object.c
@@ -68,6 +68,7 @@ enum {
#define LU_SITE_BITS_MIN 12
#define LU_SITE_BITS_MAX 24
+#define LU_SITE_BITS_MAX_CL 19
/**
* total 256 buckets, we don't want too many buckets because:
* - consume too much memory
@@ -878,6 +879,9 @@ static unsigned long lu_htable_order(struct lu_device *top)
unsigned long cache_size;
unsigned long bits;
+ if (!strcmp(top->ld_type->ldt_name, LUSTRE_VVP_NAME))
+ bits_max = LU_SITE_BITS_MAX_CL;
+
/*
* Calculate hash table size, assuming that we want reasonable
* performance when 20% of total memory is occupied by cache of
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 11/22] staging: lustre: mdt: fail FMODE_WRITE open if the client is read only
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (9 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 10/22] staging: lustre: obdclass: limit lu_site hash table size on clients James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 12/22] staging: lustre: libcfs: report hnode value for cfs_hash_putref James Simmons
` (11 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Li Dongyang,
James Simmons
From: Li Dongyang <dongyang.li@anu.edu.au>
O_WRONLY/O_RDWR open on a file will get EROFS on a read only client,
but the rpc gets sent to the mdt anyway.
mdt will increase the mot_write_count of the mdt object, blocking
subsequent FMODE_EXEC open to the same file.
This patch makes sure we fail the FMODE_WRITE open with EROFS on the
client straight away without sending the rpc to mdt.
Signed-off-by: Li Dongyang <dongyang.li@anu.edu.au>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7727
Reviewed-on: http://review.whamcloud.com/18242
Reviewed-by: Ian Costello <icostello@ddn.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Reviewed-by: Yang Sheng <yang.sheng@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/llite/namei.c | 4 ++++
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/namei.c b/drivers/staging/lustre/lustre/llite/namei.c
index b07079c..9426759 100644
--- a/drivers/staging/lustre/lustre/llite/namei.c
+++ b/drivers/staging/lustre/lustre/llite/namei.c
@@ -572,6 +572,10 @@ static int ll_lookup_it_finish(struct ptlrpc_request *request,
}
}
+ if (it->it_op & IT_OPEN && it->it_flags & FMODE_WRITE &&
+ dentry->d_sb->s_flags & MS_RDONLY)
+ return ERR_PTR(-EROFS);
+
if (it->it_op & IT_CREAT)
opc = LUSTRE_OPC_CREATE;
else
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 12/22] staging: lustre: libcfs: report hnode value for cfs_hash_putref
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (10 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 11/22] staging: lustre: mdt: fail FMODE_WRITE open if the client is read only James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 13/22] staging: lustre: statahead: set sai_index_wait with lli_sa_lock held James Simmons
` (10 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Yang Sheng,
James Simmons
From: Yang Sheng <yang.sheng@intel.com>
Add more debugging info.
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7084
Reviewed-on: http://review.whamcloud.com/17673
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Mike Pershin <mike.pershin@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lnet/libcfs/hash.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/staging/lustre/lnet/libcfs/hash.c b/drivers/staging/lustre/lnet/libcfs/hash.c
index b90dfb9..c93c59d 100644
--- a/drivers/staging/lustre/lnet/libcfs/hash.c
+++ b/drivers/staging/lustre/lnet/libcfs/hash.c
@@ -1215,7 +1215,7 @@ void cfs_hash_putref(struct cfs_hash *hs)
struct cfs_hash_bd bds[2];
int bits = 0;
- LASSERT(hlist_unhashed(hnode));
+ LASSERTF(hlist_unhashed(hnode), "hnode = %p\n", hnode);
cfs_hash_lock(hs, 0);
cfs_hash_dual_bd_get_and_lock(hs, key, bds, 1);
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 13/22] staging: lustre: statahead: set sai_index_wait with lli_sa_lock held
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (11 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 12/22] staging: lustre: libcfs: report hnode value for cfs_hash_putref James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 14/22] staging: lustre: obd: add callback for llog_cat_process_or_fork James Simmons
` (9 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Fan Yong,
Li Xi, James Simmons
From: Fan Yong <fan.yong@intel.com>
It is the sponsor thread of the statahead thread to update the
sai::sai_index_wait. Originally, it didn't hold the lli_sa_lock
when did that. Becuase of out-of-order execution others may miss
to wakeup such thread.
On the other hand, if the statahead RPC gets failure, it should
wakeup the sponsor thread, not the statahead thread.
Signed-off-by: Li Xi <lixi@ddn.com>
Signed-off-by: Fan Yong <fan.yong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7828
Reviewed-on: http://review.whamcloud.com/18499
Reviewed-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/llite/statahead.c | 18 +++++++++++-------
1 files changed, 11 insertions(+), 7 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/statahead.c b/drivers/staging/lustre/lustre/llite/statahead.c
index b43955f..4769a22 100644
--- a/drivers/staging/lustre/lustre/llite/statahead.c
+++ b/drivers/staging/lustre/lustre/llite/statahead.c
@@ -659,8 +659,8 @@ static int ll_statahead_interpret(struct ptlrpc_request *req,
struct ll_inode_info *lli = ll_i2info(dir);
struct ll_statahead_info *sai = lli->lli_sai;
struct sa_entry *entry = (struct sa_entry *)minfo->mi_cbdata;
+ wait_queue_head_t *waitq = NULL;
__u64 handle = 0;
- bool wakeup;
if (it_disposition(it, DISP_LOOKUP_NEG))
rc = -ENOENT;
@@ -693,7 +693,8 @@ static int ll_statahead_interpret(struct ptlrpc_request *req,
spin_lock(&lli->lli_sa_lock);
if (rc) {
- wakeup = __sa_make_ready(sai, entry, rc);
+ if (__sa_make_ready(sai, entry, rc))
+ waitq = &sai->sai_waitq;
} else {
entry->se_minfo = minfo;
entry->se_req = ptlrpc_request_addref(req);
@@ -704,13 +705,15 @@ static int ll_statahead_interpret(struct ptlrpc_request *req,
* with parent's lock held, for example: unlink.
*/
entry->se_handle = handle;
- wakeup = !sa_has_callback(sai);
+ if (!sa_has_callback(sai))
+ waitq = &sai->sai_thread.t_ctl_waitq;
+
list_add_tail(&entry->se_list, &sai->sai_interim_entries);
}
sai->sai_replied++;
- if (wakeup)
- wake_up(&sai->sai_thread.t_ctl_waitq);
+ if (waitq)
+ wake_up(waitq);
spin_unlock(&lli->lli_sa_lock);
return rc;
@@ -1397,10 +1400,10 @@ static int revalidate_statahead_dentry(struct inode *dir,
struct dentry **dentryp,
bool unplug)
{
+ struct ll_inode_info *lli = ll_i2info(dir);
struct sa_entry *entry = NULL;
struct l_wait_info lwi = { 0 };
struct ll_dentry_data *ldd;
- struct ll_inode_info *lli;
int rc = 0;
if ((*dentryp)->d_name.name[0] == '.') {
@@ -1446,7 +1449,9 @@ static int revalidate_statahead_dentry(struct inode *dir,
sa_handle_callback(sai);
if (!sa_ready(entry)) {
+ spin_lock(&lli->lli_sa_lock);
sai->sai_index_wait = entry->se_index;
+ spin_unlock(&lli->lli_sa_lock);
lwi = LWI_TIMEOUT_INTR(cfs_time_seconds(30), NULL,
LWI_ON_SIGNAL_NOOP, NULL);
rc = l_wait_event(sai->sai_waitq, sa_ready(entry), &lwi);
@@ -1514,7 +1519,6 @@ static int revalidate_statahead_dentry(struct inode *dir,
* dentry_may_statahead().
*/
ldd = ll_d2d(*dentryp);
- lli = ll_i2info(dir);
/* ldd can be NULL if llite lookup failed. */
if (ldd)
ldd->lld_sa_generation = lli->lli_sa_generation;
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 14/22] staging: lustre: obd: add callback for llog_cat_process_or_fork
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (12 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 13/22] staging: lustre: statahead: set sai_index_wait with lli_sa_lock held James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-06 9:59 ` Greg Kroah-Hartman
2016-12-03 0:53 ` [PATCH 15/22] staging: lustre: rpc: increase bulk size James Simmons
` (8 subsequent siblings)
22 siblings, 1 reply; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Alexander Boyko, James Simmons
From: Alexander Boyko <alexander.boyko@seagate.com>
Currently llog_process_or_fork() is hard coded to
always pass the function pointer llog_cat_process_cb().
Change llog_cat_process_or_fork() to pass in any
function pointer which will allow us more options
for llog_cat callback routines in the future.
Signed-off-by: Alexander Boyko <alexander.boyko@seagate.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7156
Seagate-bug-id: MRP-2383
Reviewed-on: http://review.whamcloud.com/16416
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Nathaniel Clark <nathaniel.l.clark@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/obdclass/llog_cat.c | 16 +++++++---------
1 files changed, 7 insertions(+), 9 deletions(-)
diff --git a/drivers/staging/lustre/lustre/obdclass/llog_cat.c b/drivers/staging/lustre/lustre/obdclass/llog_cat.c
index ce8e2f6..de9d01d 100644
--- a/drivers/staging/lustre/lustre/obdclass/llog_cat.c
+++ b/drivers/staging/lustre/lustre/obdclass/llog_cat.c
@@ -188,7 +188,8 @@ static int llog_cat_process_cb(const struct lu_env *env,
static int llog_cat_process_or_fork(const struct lu_env *env,
struct llog_handle *cat_llh,
- llog_cb_t cb, void *data, int startcat,
+ llog_cb_t cat_cb, llog_cb_t cb,
+ void *data, int startcat,
int startidx, bool fork)
{
struct llog_process_data d;
@@ -209,18 +210,15 @@ static int llog_cat_process_or_fork(const struct lu_env *env,
cd.lpcd_first_idx = llh->llh_cat_idx;
cd.lpcd_last_idx = 0;
- rc = llog_process_or_fork(env, cat_llh, llog_cat_process_cb,
- &d, &cd, fork);
+ rc = llog_process_or_fork(env, cat_llh, cat_cb, &d, &cd, fork);
if (rc != 0)
return rc;
cd.lpcd_first_idx = 0;
cd.lpcd_last_idx = cat_llh->lgh_last_idx;
- rc = llog_process_or_fork(env, cat_llh, llog_cat_process_cb,
- &d, &cd, fork);
+ rc = llog_process_or_fork(env, cat_llh, cat_cb, &d, &cd, fork);
} else {
- rc = llog_process_or_fork(env, cat_llh, llog_cat_process_cb,
- &d, NULL, fork);
+ rc = llog_process_or_fork(env, cat_llh, cat_cb, &d, NULL, fork);
}
return rc;
@@ -229,7 +227,7 @@ static int llog_cat_process_or_fork(const struct lu_env *env,
int llog_cat_process(const struct lu_env *env, struct llog_handle *cat_llh,
llog_cb_t cb, void *data, int startcat, int startidx)
{
- return llog_cat_process_or_fork(env, cat_llh, cb, data, startcat,
- startidx, false);
+ return llog_cat_process_or_fork(env, cat_llh, llog_cat_process_cb, cb,
+ data, startcat, startidx, false);
}
EXPORT_SYMBOL(llog_cat_process);
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* Re: [PATCH 14/22] staging: lustre: obd: add callback for llog_cat_process_or_fork
2016-12-03 0:53 ` [PATCH 14/22] staging: lustre: obd: add callback for llog_cat_process_or_fork James Simmons
@ 2016-12-06 9:59 ` Greg Kroah-Hartman
0 siblings, 0 replies; 29+ messages in thread
From: Greg Kroah-Hartman @ 2016-12-06 9:59 UTC (permalink / raw)
To: James Simmons
Cc: devel, Andreas Dilger, Oleg Drokin, Linux Kernel Mailing List,
Alexander Boyko, Lustre Development List
On Fri, Dec 02, 2016 at 07:53:21PM -0500, James Simmons wrote:
> From: Alexander Boyko <alexander.boyko@seagate.com>
>
> Currently llog_process_or_fork() is hard coded to
> always pass the function pointer llog_cat_process_cb().
> Change llog_cat_process_or_fork() to pass in any
> function pointer which will allow us more options
> for llog_cat callback routines in the future.
Don't change this until you have an actual user for this. Adding
infrastructure that is not used is not ok for a staging driver, or
really any kernel driver at all.
sorry,
greg k-h
^ permalink raw reply [flat|nested] 29+ messages in thread
* [PATCH 15/22] staging: lustre: rpc: increase bulk size
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (13 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 14/22] staging: lustre: obd: add callback for llog_cat_process_or_fork James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 16/22] staging: lustre: llite: Invoke file_update_time in page_mkwrite James Simmons
` (7 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Jinshan Xiong, Gu Zheng, James Simmons
From: Jinshan Xiong <jinshan.xiong@intel.com>
To make the ptlrpc be able to size 16MB IO
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: Gu Zheng <gzheng@ddn.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7990
Reviewed-on: http://review.whamcloud.com/19366
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
.../lustre/lustre/include/lustre/lustre_idl.h | 6 +++++-
drivers/staging/lustre/lustre/include/lustre_net.h | 8 ++++++--
drivers/staging/lustre/lustre/ptlrpc/wiretest.c | 2 ++
3 files changed, 13 insertions(+), 3 deletions(-)
diff --git a/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h b/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
index b9f333b..20f55b7 100644
--- a/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
+++ b/drivers/staging/lustre/lustre/include/lustre/lustre_idl.h
@@ -1683,8 +1683,12 @@ struct obd_ioobj {
__u32 ioo_bufcnt; /* number of niobufs for this object */
};
+/*
+ * NOTE: IOOBJ_MAX_BRW_BITS defines the _offset_ of the max_brw field in
+ * ioo_max_brw, NOT the maximum number of bits in PTLRPC_BULK_OPS_BITS.
+ * That said, ioo_max_brw is a 32-bit field so the limit is also 16 bits.
+ */
#define IOOBJ_MAX_BRW_BITS 16
-#define IOOBJ_TYPE_MASK ((1U << IOOBJ_MAX_BRW_BITS) - 1)
#define ioobj_max_brw_get(ioo) (((ioo)->ioo_max_brw >> IOOBJ_MAX_BRW_BITS) + 1)
#define ioobj_max_brw_set(ioo, num) \
do { (ioo)->ioo_max_brw = ((num) - 1) << IOOBJ_MAX_BRW_BITS; } while (0)
diff --git a/drivers/staging/lustre/lustre/include/lustre_net.h b/drivers/staging/lustre/lustre/include/lustre_net.h
index 2be135d..411eb0d 100644
--- a/drivers/staging/lustre/lustre/include/lustre_net.h
+++ b/drivers/staging/lustre/lustre/include/lustre_net.h
@@ -69,13 +69,17 @@
#define PTLRPC_MD_OPTIONS 0
/**
- * Max # of bulk operations in one request.
+ * log2 max # of bulk operations in one request: 2=4MB/RPC, 5=32MB/RPC, ...
* In order for the client and server to properly negotiate the maximum
* possible transfer size, PTLRPC_BULK_OPS_COUNT must be a power-of-two
* value. The client is free to limit the actual RPC size for any bulk
* transfer via cl_max_pages_per_rpc to some non-power-of-two value.
+ * NOTE: This is limited to 16 (=64GB RPCs) by IOOBJ_MAX_BRW_BITS.
*/
-#define PTLRPC_BULK_OPS_BITS 2
+#define PTLRPC_BULK_OPS_BITS 4
+#if PTLRPC_BULK_OPS_BITS > 16
+#error "More than 65536 BRW RPCs not allowed by IOOBJ_MAX_BRW_BITS."
+#endif
#define PTLRPC_BULK_OPS_COUNT (1U << PTLRPC_BULK_OPS_BITS)
/**
* PTLRPC_BULK_OPS_MASK is for the convenience of the client only, and
diff --git a/drivers/staging/lustre/lustre/ptlrpc/wiretest.c b/drivers/staging/lustre/lustre/ptlrpc/wiretest.c
index f50f487..e12eb83 100644
--- a/drivers/staging/lustre/lustre/ptlrpc/wiretest.c
+++ b/drivers/staging/lustre/lustre/ptlrpc/wiretest.c
@@ -1576,6 +1576,8 @@ void lustre_assert_wire_constants(void)
(long long)(int)offsetof(struct obd_ioobj, ioo_bufcnt));
LASSERTF((int)sizeof(((struct obd_ioobj *)0)->ioo_bufcnt) == 4, "found %lld\n",
(long long)(int)sizeof(((struct obd_ioobj *)0)->ioo_bufcnt));
+ LASSERTF(IOOBJ_MAX_BRW_BITS == 16, "found %lld\n",
+ (long long)IOOBJ_MAX_BRW_BITS);
/* Checks for union lquota_id */
LASSERTF((int)sizeof(union lquota_id) == 16, "found %lld\n",
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 16/22] staging: lustre: llite: Invoke file_update_time in page_mkwrite
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (14 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 15/22] staging: lustre: rpc: increase bulk size James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 17/22] staging: lustre: clio: remove mtime check in vvp_io_fault_start() James Simmons
` (6 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Yang Sheng,
James Simmons
From: Yang Sheng <yang.sheng@intel.com>
Only update file times if page_mkwrite is not set. So we
need call file_update_time by ourselves.
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-1118
Reviewed-on: http://review.whamcloud.com/18683
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Niu Yawei <yawei.niu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/llite/llite_mmap.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/llite_mmap.c b/drivers/staging/lustre/lustre/llite/llite_mmap.c
index f50b6c1..ee01f20 100644
--- a/drivers/staging/lustre/lustre/llite/llite_mmap.c
+++ b/drivers/staging/lustre/lustre/llite/llite_mmap.c
@@ -369,6 +369,7 @@ static int ll_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
bool retry;
int result;
+ file_update_time(vma->vm_file);
do {
retry = false;
result = ll_page_mkwrite0(vma, vmf->page, &retry);
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 17/22] staging: lustre: clio: remove mtime check in vvp_io_fault_start()
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (15 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 16/22] staging: lustre: llite: Invoke file_update_time in page_mkwrite James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 18/22] staging: lustre: import: don't reconnect during connect interpret James Simmons
` (5 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Bobi Jam,
James Simmons
From: Bobi Jam <bobijam.xu@intel.com>
In fault IO initialization, inode's mtime is saved, and after
getting locks, when the IO is about to start, vvp_io_fault_start()
checks the mtime's intactness.
It's a false alarm, since the timestamp from MDS could be stale,
we maintain mtime mainly on OST objects, and if the check in
vvp_io_fault_start() happens before mtime on OST objects are merged,
it will get wrong timestamp from the inode, even the timestamp it
fetched in vvp_io_fault_init() could be wrong in the first place.
This patch remove the mtime check in vvp_io_fault_start().
Signed-off-by: Bobi Jam <bobijam.xu@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7198
Reviewed-on: http://review.whamcloud.com/19162
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: Jinshan Xiong <jinshan.xiong@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/llite/vvp_io.c | 6 ------
1 files changed, 0 insertions(+), 6 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/vvp_io.c b/drivers/staging/lustre/lustre/llite/vvp_io.c
index 114906d..0b6d388 100644
--- a/drivers/staging/lustre/lustre/llite/vvp_io.c
+++ b/drivers/staging/lustre/lustre/llite/vvp_io.c
@@ -1064,12 +1064,6 @@ static int vvp_io_fault_start(const struct lu_env *env,
loff_t size;
pgoff_t last_index;
- if (fio->ft_executable &&
- inode->i_mtime.tv_sec != vio->u.fault.ft_mtime)
- CWARN("binary "DFID
- " changed while waiting for the page fault lock\n",
- PFID(lu_object_fid(&obj->co_lu)));
-
down_read(&lli->lli_trunc_sem);
/* offset of the last byte on the page */
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 18/22] staging: lustre: import: don't reconnect during connect interpret
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (16 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 17/22] staging: lustre: clio: remove mtime check in vvp_io_fault_start() James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 19/22] staging: lustre: llite: ll_dir_ioctl cleanup of redundant comparisons James Simmons
` (4 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Mikhal Pershin, James Simmons
From: Mikhal Pershin <mike.pershin@intel.com>
The import connect flags might be cleared by ptlrpc_connect_import()
wrongly if there is still connect interpret function is running.
Use imp_connected boolean variable to indicate that we are still
interpretting connect reply and don't try to reconnect until it ends.
Signed-off-by: Mikhal Pershin <mike.pershin@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-7558
Reviewed-on: http://review.whamcloud.com/19312
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
.../staging/lustre/lustre/include/lustre_import.h | 4 +++-
drivers/staging/lustre/lustre/ptlrpc/import.c | 16 +++++++++++++++-
2 files changed, 18 insertions(+), 2 deletions(-)
diff --git a/drivers/staging/lustre/lustre/include/lustre_import.h b/drivers/staging/lustre/lustre/include/lustre_import.h
index 4499c69..f0c931c 100644
--- a/drivers/staging/lustre/lustre/include/lustre_import.h
+++ b/drivers/staging/lustre/lustre/include/lustre_import.h
@@ -299,7 +299,9 @@ struct obd_import {
*/
imp_force_reconnect:1,
/* import has tried to connect with server */
- imp_connect_tried:1;
+ imp_connect_tried:1,
+ /* connected but not FULL yet */
+ imp_connected:1;
__u32 imp_connect_op;
struct obd_connect_data imp_connect_data;
__u64 imp_connect_flags_orig;
diff --git a/drivers/staging/lustre/lustre/ptlrpc/import.c b/drivers/staging/lustre/lustre/ptlrpc/import.c
index 66f5b49..e828019 100644
--- a/drivers/staging/lustre/lustre/ptlrpc/import.c
+++ b/drivers/staging/lustre/lustre/ptlrpc/import.c
@@ -622,7 +622,8 @@ int ptlrpc_connect_import(struct obd_import *imp)
spin_unlock(&imp->imp_lock);
CERROR("already connected\n");
return 0;
- } else if (imp->imp_state == LUSTRE_IMP_CONNECTING) {
+ } else if (imp->imp_state == LUSTRE_IMP_CONNECTING ||
+ imp->imp_connected) {
spin_unlock(&imp->imp_lock);
CERROR("already connecting\n");
return -EALREADY;
@@ -971,6 +972,13 @@ static int ptlrpc_connect_interpret(const struct lu_env *env,
ptlrpc_maybe_ping_import_soon(imp);
goto out;
}
+
+ /*
+ * LU-7558: indicate that we are interpretting connect reply,
+ * pltrpc_connect_import() will not try to reconnect until
+ * interpret will finish.
+ */
+ imp->imp_connected = 1;
spin_unlock(&imp->imp_lock);
LASSERT(imp->imp_conn_current);
@@ -1194,12 +1202,18 @@ static int ptlrpc_connect_interpret(const struct lu_env *env,
obd2cli_tgt(imp->imp_obd),
imp->imp_connection->c_remote_uuid.uuid);
ptlrpc_connect_import(imp);
+ spin_lock(&imp->imp_lock);
+ imp->imp_connected = 0;
imp->imp_connect_tried = 1;
+ spin_unlock(&imp->imp_lock);
return 0;
}
out:
+ spin_lock(&imp->imp_lock);
+ imp->imp_connected = 0;
imp->imp_connect_tried = 1;
+ spin_unlock(&imp->imp_lock);
if (rc != 0) {
IMPORT_SET_STATE(imp, LUSTRE_IMP_DISCON);
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 19/22] staging: lustre: llite: ll_dir_ioctl cleanup of redundant comparisons
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (17 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 18/22] staging: lustre: import: don't reconnect during connect interpret James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 20/22] staging: lustre: osc: set lock data for readahead lock James Simmons
` (3 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Parinay Kondekar, James Simmons
From: Parinay Kondekar <parinay.kondekar@seagate.com>
In ll_dir_ioctl() two identical comparisions are present for
return code (rc) of ll_dir_getstripe(). This patch removes
the other inside if( ) condition which is not necessary.
Signed-off-by: Parinay Kondekar <parinay.kondekar@seagate.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6512
Reviewed-on: http://review.whamcloud.com/18027
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/llite/dir.c | 3 ---
1 files changed, 0 insertions(+), 3 deletions(-)
diff --git a/drivers/staging/lustre/lustre/llite/dir.c b/drivers/staging/lustre/lustre/llite/dir.c
index 351e900..3dc7344 100644
--- a/drivers/staging/lustre/lustre/llite/dir.c
+++ b/drivers/staging/lustre/lustre/llite/dir.c
@@ -1227,9 +1227,6 @@ static long ll_dir_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
/* Get default LMV EA */
if (lum.lum_magic == LMV_USER_MAGIC) {
- if (rc)
- goto finish_req;
-
if (lmmsize > sizeof(*ulmv)) {
rc = -EINVAL;
goto finish_req;
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 20/22] staging: lustre: osc: set lock data for readahead lock
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (18 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 19/22] staging: lustre: llite: ll_dir_ioctl cleanup of redundant comparisons James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 21/22] staging: lustre: remove set but unused variables James Simmons
` (2 subsequent siblings)
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Jinshan Xiong, James Simmons
From: Jinshan Xiong <jinshan.xiong@intel.com>
If osc_io_readahead() finds a lock that belongs to the previous
instance of osc_object, the lock data pointer will be null. It has
to instantiate with new instance otherwise those pages won't be
destroyed at lock cancel, and then finally hit the assertion in
osc_req_attr_set().
This patch revised dlmlock_at_pgoff() to call osc_match_base() to
find caching locks for readahead. And new osc_object will be set
to the lock if it doesn't have one yet.
Signed-off-by: Jinshan Xiong <jinshan.xiong@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8005
Reviewed-on: http://review.whamcloud.com/19453
Reviewed-by: Bobi Jam <bobijam@hotmail.com>
Reviewed-by: John L. Hammond <john.hammond@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lustre/osc/osc_io.c | 1 +
drivers/staging/lustre/lustre/osc/osc_lock.c | 7 +--
drivers/staging/lustre/lustre/osc/osc_request.c | 49 ++++++----------------
3 files changed, 18 insertions(+), 39 deletions(-)
diff --git a/drivers/staging/lustre/lustre/osc/osc_io.c b/drivers/staging/lustre/lustre/osc/osc_io.c
index 369761f..d96f4f2 100644
--- a/drivers/staging/lustre/lustre/osc/osc_io.c
+++ b/drivers/staging/lustre/lustre/osc/osc_io.c
@@ -88,6 +88,7 @@ static int osc_io_read_ahead(const struct lu_env *env,
dlmlock = osc_dlmlock_at_pgoff(env, osc, start, 0);
if (dlmlock) {
+ LASSERT(dlmlock->l_ast_data == osc);
if (dlmlock->l_req_mode != LCK_PR) {
struct lustre_handle lockh;
diff --git a/drivers/staging/lustre/lustre/osc/osc_lock.c b/drivers/staging/lustre/lustre/osc/osc_lock.c
index 130460d..001fe75 100644
--- a/drivers/staging/lustre/lustre/osc/osc_lock.c
+++ b/drivers/staging/lustre/lustre/osc/osc_lock.c
@@ -1205,10 +1205,9 @@ struct ldlm_lock *osc_dlmlock_at_pgoff(const struct lu_env *env,
* with a uniq gid and it conflicts with all other lock modes too
*/
again:
- mode = ldlm_lock_match(osc_export(obj)->exp_obd->obd_namespace,
- flags, resname, LDLM_EXTENT, policy,
- LCK_PR | LCK_PW | LCK_GROUP, &lockh,
- dap_flags & OSC_DAP_FL_CANCELING);
+ mode = osc_match_base(osc_export(obj), resname, LDLM_EXTENT, policy,
+ LCK_PR | LCK_PW | LCK_GROUP, &flags, obj, &lockh,
+ dap_flags & OSC_DAP_FL_CANCELING);
if (mode != 0) {
lock = ldlm_handle2lock(&lockh);
/* RACE: the lock is cancelled so let's try again */
diff --git a/drivers/staging/lustre/lustre/osc/osc_request.c b/drivers/staging/lustre/lustre/osc/osc_request.c
index bc698d3..0977127 100644
--- a/drivers/staging/lustre/lustre/osc/osc_request.c
+++ b/drivers/staging/lustre/lustre/osc/osc_request.c
@@ -1813,16 +1813,11 @@ int osc_build_rpc(const struct lu_env *env, struct client_obd *cli,
return rc;
}
-static int osc_set_lock_data_with_check(struct ldlm_lock *lock,
- struct ldlm_enqueue_info *einfo)
+static int osc_set_lock_data(struct ldlm_lock *lock, void *data)
{
- void *data = einfo->ei_cbdata;
int set = 0;
- LASSERT(lock->l_blocking_ast == einfo->ei_cb_bl);
- LASSERT(lock->l_resource->lr_type == einfo->ei_type);
- LASSERT(lock->l_completion_ast == einfo->ei_cb_cp);
- LASSERT(lock->l_glimpse_ast == einfo->ei_cb_gl);
+ LASSERT(lock);
lock_res_and_lock(lock);
@@ -1836,21 +1831,6 @@ static int osc_set_lock_data_with_check(struct ldlm_lock *lock,
return set;
}
-static int osc_set_data_with_check(struct lustre_handle *lockh,
- struct ldlm_enqueue_info *einfo)
-{
- struct ldlm_lock *lock = ldlm_handle2lock(lockh);
- int set = 0;
-
- if (lock) {
- set = osc_set_lock_data_with_check(lock, einfo);
- LDLM_LOCK_PUT(lock);
- } else
- CERROR("lockh %p, data %p - client evicted?\n",
- lockh, einfo->ei_cbdata);
- return set;
-}
-
static int osc_enqueue_fini(struct ptlrpc_request *req,
osc_enqueue_upcall_f upcall, void *cookie,
struct lustre_handle *lockh, enum ldlm_mode mode,
@@ -2016,7 +1996,7 @@ int osc_enqueue_base(struct obd_export *exp, struct ldlm_res_id *res_id,
ldlm_lock_decref(&lockh, mode);
LDLM_LOCK_PUT(matched);
return -ECANCELED;
- } else if (osc_set_lock_data_with_check(matched, einfo)) {
+ } else if (osc_set_lock_data(matched, einfo->ei_cbdata)) {
*flags |= LDLM_FL_LVB_READY;
/* We already have a lock, and it's referenced. */
(*upcall)(cookie, &lockh, ELDLM_LOCK_MATCHED);
@@ -2128,19 +2108,18 @@ int osc_match_base(struct obd_export *exp, struct ldlm_res_id *res_id,
rc |= LCK_PW;
rc = ldlm_lock_match(obd->obd_namespace, lflags,
res_id, type, policy, rc, lockh, unref);
- if (rc) {
- if (data) {
- if (!osc_set_data_with_check(lockh, data)) {
- if (!(lflags & LDLM_FL_TEST_LOCK))
- ldlm_lock_decref(lockh, rc);
- return 0;
- }
- }
- if (!(lflags & LDLM_FL_TEST_LOCK) && mode != rc) {
- ldlm_lock_addref(lockh, LCK_PR);
- ldlm_lock_decref(lockh, LCK_PW);
- }
+ if (!rc || lflags & LDLM_FL_TEST_LOCK)
return rc;
+
+ if (data) {
+ struct ldlm_lock *lock = ldlm_handle2lock(lockh);
+
+ LASSERT(lock);
+ if (!osc_set_lock_data(lock, data)) {
+ ldlm_lock_decref(lockh, rc);
+ rc = 0;
+ }
+ LDLM_LOCK_PUT(lock);
}
return rc;
}
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 21/22] staging: lustre: remove set but unused variables
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (19 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 20/22] staging: lustre: osc: set lock data for readahead lock James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-03 0:53 ` [PATCH 22/22] staging: lustre: libcfs: remove lnet upcall code James Simmons
2016-12-06 10:00 ` [PATCH 00/22] Next batch of missing work for upstream client Greg Kroah-Hartman
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List, Yang Sheng,
James Simmons
From: Yang Sheng <yang.sheng@intel.com>
Remove set but unused variables in nidstring.c
and osc_request.c as reported by make W=1.
Signed-off-by: Yang Sheng <yang.sheng@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8378
Reviewed-on: http://review.whamcloud.com/23221
Reviewed-by: Bob Glossman <bob.glossman@intel.com>
Reviewed-by: Emoly Liu <emoly.liu@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
drivers/staging/lustre/lnet/lnet/nidstrings.c | 2 --
drivers/staging/lustre/lustre/osc/osc_request.c | 3 +--
2 files changed, 1 insertions(+), 4 deletions(-)
diff --git a/drivers/staging/lustre/lnet/lnet/nidstrings.c b/drivers/staging/lustre/lnet/lnet/nidstrings.c
index e1f1ca8..a9fe3e6 100644
--- a/drivers/staging/lustre/lnet/lnet/nidstrings.c
+++ b/drivers/staging/lustre/lnet/lnet/nidstrings.c
@@ -247,10 +247,8 @@ struct addrrange {
{
struct cfs_lstr addrrange;
struct cfs_lstr net;
- struct cfs_lstr tmp;
struct nidrange *nr;
- tmp = *src;
if (!cfs_gettok(src, '@', &addrrange))
goto failed;
diff --git a/drivers/staging/lustre/lustre/osc/osc_request.c b/drivers/staging/lustre/lustre/osc/osc_request.c
index 0977127..0c50225 100644
--- a/drivers/staging/lustre/lustre/osc/osc_request.c
+++ b/drivers/staging/lustre/lustre/osc/osc_request.c
@@ -933,7 +933,6 @@ static u32 osc_checksum_bulk(int nob, u32 pg_count,
int i = 0;
struct cfs_crypto_hash_desc *hdesc;
unsigned int bufsize;
- int err;
unsigned char cfs_alg = cksum_obd2cfs(cksum_type);
LASSERT(pg_count > 0);
@@ -975,7 +974,7 @@ static u32 osc_checksum_bulk(int nob, u32 pg_count,
}
bufsize = sizeof(cksum);
- err = cfs_crypto_hash_final(hdesc, (unsigned char *)&cksum, &bufsize);
+ cfs_crypto_hash_final(hdesc, (unsigned char *)&cksum, &bufsize);
/* For sending we only compute the wrong checksum instead
* of corrupting the data so it is still correct on a redo
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* [PATCH 22/22] staging: lustre: libcfs: remove lnet upcall code
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (20 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 21/22] staging: lustre: remove set but unused variables James Simmons
@ 2016-12-03 0:53 ` James Simmons
2016-12-06 10:00 ` [PATCH 00/22] Next batch of missing work for upstream client Greg Kroah-Hartman
22 siblings, 0 replies; 29+ messages in thread
From: James Simmons @ 2016-12-03 0:53 UTC (permalink / raw)
To: Greg Kroah-Hartman, devel, Andreas Dilger, Oleg Drokin
Cc: Linux Kernel Mailing List, Lustre Development List,
Alexander Zarochentsev, James Simmons
From: Alexander Zarochentsev <alexander.zarochentsev@seagate.com>
Removing lnet upcall infrastructure completely
as nobody uses it anymore. The upcall causes a delay
before calling BUG() and might even cause a hang
making getting a crash dump unreliable or containing
outdated info.
Signed-off-by: Alexander Zarochentsev <alexander.zarochentsev@seagate.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-8418
Seagate-bug-id: MRP-2939
Reviewed-on: http://review.whamcloud.com/21440
Reviewed-by: Andreas Dilger <andreas.dilger@intel.com>
Reviewed-by: James Simmons <uja.ornl@yahoo.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
---
.../staging/lustre/include/linux/libcfs/libcfs.h | 1 -
.../lustre/include/linux/libcfs/libcfs_private.h | 2 -
.../staging/lustre/lnet/libcfs/linux/linux-debug.c | 54 --------------------
drivers/staging/lustre/lnet/libcfs/module.c | 8 ---
4 files changed, 0 insertions(+), 65 deletions(-)
diff --git a/drivers/staging/lustre/include/linux/libcfs/libcfs.h b/drivers/staging/lustre/include/linux/libcfs/libcfs.h
index 70eb08e..cc2c0e9 100644
--- a/drivers/staging/lustre/include/linux/libcfs/libcfs.h
+++ b/drivers/staging/lustre/include/linux/libcfs/libcfs.h
@@ -125,7 +125,6 @@ int libcfs_ioctl_getdata(struct libcfs_ioctl_hdr **hdr_pp,
/**
* The path of debug log dump upcall script.
*/
-extern char lnet_upcall[1024];
extern char lnet_debug_log_upcall[1024];
extern struct cfs_wi_sched *cfs_sched_rehash;
diff --git a/drivers/staging/lustre/include/linux/libcfs/libcfs_private.h b/drivers/staging/lustre/include/linux/libcfs/libcfs_private.h
index 41a651f..aab15d8 100644
--- a/drivers/staging/lustre/include/linux/libcfs/libcfs_private.h
+++ b/drivers/staging/lustre/include/linux/libcfs/libcfs_private.h
@@ -169,8 +169,6 @@
#define ntohs(x) ___ntohs(x)
#endif
-void libcfs_run_upcall(char **argv);
-void libcfs_run_lbug_upcall(struct libcfs_debug_msg_data *msg);
void libcfs_debug_dumplog(void);
int libcfs_debug_init(unsigned long bufsize);
int libcfs_debug_cleanup(void);
diff --git a/drivers/staging/lustre/lnet/libcfs/linux/linux-debug.c b/drivers/staging/lustre/lnet/libcfs/linux/linux-debug.c
index d8a9894..39a72e3 100644
--- a/drivers/staging/lustre/lnet/libcfs/linux/linux-debug.c
+++ b/drivers/staging/lustre/lnet/libcfs/linux/linux-debug.c
@@ -57,7 +57,6 @@
#include <linux/kallsyms.h>
-char lnet_upcall[1024] = "/usr/lib/lustre/lnet_upcall";
char lnet_debug_log_upcall[1024] = "/usr/lib/lustre/lnet_debug_log_upcall";
/**
@@ -92,58 +91,6 @@ void libcfs_run_debug_log_upcall(char *file)
}
}
-void libcfs_run_upcall(char **argv)
-{
- int rc;
- int argc;
- static const char * const envp[] = {
- "HOME=/",
- "PATH=/sbin:/bin:/usr/sbin:/usr/bin",
- NULL
- };
-
- argv[0] = lnet_upcall;
- argc = 1;
- while (argv[argc])
- argc++;
-
- LASSERT(argc >= 2);
-
- rc = call_usermodehelper(argv[0], argv, (char **)envp, 1);
- if (rc < 0 && rc != -ENOENT) {
- CERROR("Error %d invoking LNET upcall %s %s%s%s%s%s%s%s%s; check /sys/kernel/debug/lnet/upcall\n",
- rc, argv[0], argv[1],
- argc < 3 ? "" : ",", argc < 3 ? "" : argv[2],
- argc < 4 ? "" : ",", argc < 4 ? "" : argv[3],
- argc < 5 ? "" : ",", argc < 5 ? "" : argv[4],
- argc < 6 ? "" : ",...");
- } else {
- CDEBUG(D_HA, "Invoked LNET upcall %s %s%s%s%s%s%s%s%s\n",
- argv[0], argv[1],
- argc < 3 ? "" : ",", argc < 3 ? "" : argv[2],
- argc < 4 ? "" : ",", argc < 4 ? "" : argv[3],
- argc < 5 ? "" : ",", argc < 5 ? "" : argv[4],
- argc < 6 ? "" : ",...");
- }
-}
-
-void libcfs_run_lbug_upcall(struct libcfs_debug_msg_data *msgdata)
-{
- char *argv[6];
- char buf[32];
-
- snprintf(buf, sizeof(buf), "%d", msgdata->msg_line);
-
- argv[1] = "LBUG";
- argv[2] = (char *)msgdata->msg_file;
- argv[3] = (char *)msgdata->msg_fn;
- argv[4] = buf;
- argv[5] = NULL;
-
- libcfs_run_upcall(argv);
-}
-EXPORT_SYMBOL(libcfs_run_lbug_upcall);
-
/* coverity[+kill] */
void __noreturn lbug_with_loc(struct libcfs_debug_msg_data *msgdata)
{
@@ -158,7 +105,6 @@ void __noreturn lbug_with_loc(struct libcfs_debug_msg_data *msgdata)
dump_stack();
if (!libcfs_panic_on_lbug)
libcfs_debug_dumplog();
- libcfs_run_lbug_upcall(msgdata);
if (libcfs_panic_on_lbug)
panic("LBUG");
set_task_state(current, TASK_UNINTERRUPTIBLE);
diff --git a/drivers/staging/lustre/lnet/libcfs/module.c b/drivers/staging/lustre/lnet/libcfs/module.c
index 5884a33..161e042 100644
--- a/drivers/staging/lustre/lnet/libcfs/module.c
+++ b/drivers/staging/lustre/lnet/libcfs/module.c
@@ -365,14 +365,6 @@ static int proc_cpt_table(struct ctl_table *table, int write,
.mode = 0444,
.proc_handler = &proc_cpt_table,
},
-
- {
- .procname = "upcall",
- .data = lnet_upcall,
- .maxlen = sizeof(lnet_upcall),
- .mode = 0644,
- .proc_handler = &proc_dostring,
- },
{
.procname = "debug_log_upcall",
.data = lnet_debug_log_upcall,
--
1.7.1
^ permalink raw reply related [flat|nested] 29+ messages in thread* Re: [PATCH 00/22] Next batch of missing work for upstream client
2016-12-03 0:53 [PATCH 00/22] Next batch of missing work for upstream client James Simmons
` (21 preceding siblings ...)
2016-12-03 0:53 ` [PATCH 22/22] staging: lustre: libcfs: remove lnet upcall code James Simmons
@ 2016-12-06 10:00 ` Greg Kroah-Hartman
22 siblings, 0 replies; 29+ messages in thread
From: Greg Kroah-Hartman @ 2016-12-06 10:00 UTC (permalink / raw)
To: James Simmons
Cc: devel, Andreas Dilger, Oleg Drokin, Linux Kernel Mailing List,
Lustre Development List
On Fri, Dec 02, 2016 at 07:53:07PM -0500, James Simmons wrote:
> Batch of various fixes and clean ups missing in the upstream client.
> Only one smaller batch of patches left to sync lustre 2.8.0 version.
> These patches are independent of each other so they can be landed
> in any order.
I've applied most of these now, please fix up the remaining ones and
resend.
thanks,
greg k-h
^ permalink raw reply [flat|nested] 29+ messages in thread