pnfs: call layoutcommit after flushing inode's data to disk.
Subsystems (2)
PATCH 1/1 pnfs: call layoutcommit after flushing inode's data to disk.
--- fs/nfs/write.c | 8 +++++++- 1 files changed, 7 insertions(+), 1 deletions(-)
fs/nfs/write.c +7 -1
@@ -1558,6 +1558,7 @@ int nfs_write_inode(struct inode *inode, struct writeback_control *wbc)*/ int nfs_wb_all (struct inode *inode) { +int ret;struct writeback_control wbc = { .sync_mode =WB_SYNC_ALL , .nr_to_write =LONG_MAX ,
@@ -1565,7 +1566,12 @@ int nfs_wb_all(struct inode *inode) .range_end = LLONG_MAX, }; -return sync_inode (inode, &wbc); + ret =sync_inode (inode, &wbc); +#ifdef CONFIG_NFS_V4_1 +if (!ret &&NFS_I (inode)->layoutcommit_ctx) + ret =pnfs_layoutcommit_inode (inode,1 ); +#endif +return ret; }int nfs_wb_page_cancel (struct inode *inode,struct page *page)
Signature · 1 line
1.6.3.3
This is for a bug introduced to 2.6.34. In 2.6.32 and 2.6.33 we call layoutcommit innfs_sync_mapping_wait(), but in 2.6.34 we use sync_inode() to sync inode's data, so
the layoutcommit code is gone.
BTW: In current code, layoutcommit_ctx will increase refcount of nfs_inode's ctx, so iflayoutcommit_ctx is not NULL, we could not reach nfs4_close_context =2E.. --> __nfs_close().
So pnfs_layoutcommit_inode() in __nfs_close() will not be called in whatever situation.
Why we have to use nfs_inode's ctx as layoutcommit_ctx, since we only need its rpc_creds
actually?
fs/nfs/write.c int nfs_write_inode(struct inode *inode, struct=
Quoted 10 lines · Tao Guo
Signed-off-by Tao Guo <[email protected]>---
=C2=A0fs/nfs/write.c | =C2=A0 =C2=A08 +++++++- =C2=A01 files changed, 7 insertions(+), 1 deletions(-)fs/nfs/write.c +0
-0@@ -1558,6 +1558,7 @@ int nfs_write_inode(struct inode *inode, struct=
writeback_control *wbc)
Quoted 6 lines
=C2=A0*/
=C2=A0intnfs_wb_all(struct inode *inode)
=C2=A0{
+ =C2=A0 =C2=A0 =C2=A0 int ret;
=C2=A0 =C2=A0 =C2=A0 =C2=A0structwriteback_controlwbc =3D { =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0.sync_mode =3D=
WB_SYNC_ALL,
Quoted 1 line
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0.nr_to_write =3D=
LONG_MAX,
Quoted 2 lines
fs/nfs/write.c +0
-0@@ -1565,7 +1566,12 @@ int nfs_wb_all(struct inode *inode)=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0.range_end =3D=
LLONG_MAX,
Quoted 7 lines
=C2=A0 =C2=A0 =C2=A0 =C2=A0};
- =C2=A0 =C2=A0 =C2=A0 return
sync_inode(inode, &wbc);
+ =C2=A0 =C2=A0 =C2=A0 ret =3Dsync_inode(inode, &wbc); +#ifdefCONFIG_NFS_V4_1
+ =C2=A0 =C2=A0 =C2=A0 if (!ret && NFS_I(inode)->layoutcommit_ctx)
+ =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ret =3D pnfs_layou=
tcommit_inode(inode, 1);
Quoted 10 lines
+#endif
+ =C2=A0 =C2=A0 =C2=A0 return ret;
=C2=A0}=C2=A0int
nfs_wb_page_cancel(struct inode *inode, struct page *page)--
1.6.3.3--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" =
in
Quoted 2 lines
the body of a message to [email protected]
More majordomo info at =C2=A0http://vger.kernel.org/majordomo-info.ht=
ml
Quoted 1 line
--=20
tao.
Quoted 6 lines · Tao Guo
This is for a bug introduced to 2.6.34. In 2.6.32 and 2.6.33 we call layoutcommit in
nfs_sync_mapping_wait(), but in 2.6.34 we usesync_inode()to sync inode's data, so
the layoutcommit code is gone.
Quoted 9 lines
BTW: In current code,
layoutcommit_ctxwill increase refcount of nfs_inode's ctx, so iflayoutcommit_ctxis not NULL, we could not reachnfs4_close_context... -->__nfs_close().
Sopnfs_layoutcommit_inode()in__nfs_close()will not be called in whatever situation.
Why we have to use nfs_inode's ctx aslayoutcommit_ctx, since we only need itsrpc_creds
actually?
I have a patch to be submitted today that replaces the nfs_open_context with an rpc_cred.
-->Andy
Quoted 51 lines · Tao Guo
Quoted 40 lines
Signed-off-by Tao Guo <[email protected]>--- fs/nfs/write.c | 8 +++++++- 1 files changed, 7 insertions(+), 1 deletions(-)fs/nfs/write.c +1
-0@@ -1558,6 +1558,7 @@ int nfs_write_inode(struct inode *inode,struct writeback_control *wbc) */ int nfs_wb_all (struct inode *inode) { +int ret;struct writeback_control wbc = { .sync_mode =WB_SYNC_ALL ,.nr_to_write =
LONG_MAX,fs/nfs/write.c +6
-2@@ -1565,7 +1566,12 @@ int nfs_wb_all(struct inode *inode) .range_end = LLONG_MAX, }; -return sync_inode (inode, &wbc); + ret =sync_inode (inode, &wbc); +#ifdef CONFIG_NFS_V4_1 +if (!ret &&NFS_I (inode)->layoutcommit_ctx) + ret =pnfs_layoutcommit_inode (inode,1 ); +#endif +return ret; }int nfs_wb_page_cancel (struct inode *inode,struct page *page) --1.6.3.3
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.htmlSignature · 5 lines
tao. -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html