xfs: LARP - clean up xfs_attr_set_iter state machine
v2 cover xfs: LARP - clean up xfs_attr_set_iter state machine
HI folks,
Version 2 is here and working better. The first 11 patches are largely the same as the first version, just we a couple of bugs fixes which caused a couple patches to be reworked to fix conflicts. These patches set up the xfs_attr_set_iter() state machine to be more readable and maintainable. It largely jsut worked with LARP=0, but failed on the first recovery when LARP was enabled.
I realised that recovery wasn't setting the initial state properly, and then started digging into the recovery code. At this point, I realised the recovery code didn't work correctly in all cases and could often leave unremovable incomplete attrs sitting around. The issues are larger documented ini patch 16 so I won't go over them here, just read that patch.
However, to get, the whole replace operation for LARP=1 needed to change. Luckily, that turned out to be pretty simple because it was largely already broken down into component operations in the state machine. hence I just needed to add new "remove" initial states to the set_iter state machine, and that allowed the new algorithm to function.
Then I realised that I'd just implemented the remove_iter algorithm in the set_iter state machine, so I removed the remove_iter code and it's states altogether and just pointed remove ops at the set-iter remove initial states. The code now uses the XFS_DA_OP_RENAME flag to determine if it should follow up an add or remove with a remove or add, and it all largely jsut works. All runtime algorithms run throught he same state machine just with different initial states and state progressions.
And with patch 16, attr item log recovery uses that same state machine, too. It has a few quirks that need to be handled, so I added the XFS_DA_OP_RECOVERY flag to allow the right thing to be done with the INCOMPLETE flag deep in the guts of the attr lookup code. And so recovery with LARP=1 now seems to mostly work.
The current state is that larp=0 seems to pass attr group and recoverloop group testing just fine. My current build is about 10 loops in without a failure. With larp=1, attr group passes jsut fine, but I'm getting random recovery failures with recoveryloop testing that aren't immediately obvious as being attribute intent recvoery failures. (e.g. things like unlinked inode recovery finding inodes on the unlinked list that have nlink != 0). Hence while it mostly works, I'm sure there are still some bugs lurking there.
There's also a fair bit of cleanup needed of the last couple of patches. There's some quirks in the state change code when completing an op and determining whether we need to run the second half of a replace op that need cleaning up, and the last patch needs further splitting up and it changes stuff all over the place.
However, if you want to get an idea of what the code looks like, see if it runs and breaks on your machines and maybe find some bugs in it, help is welcome!
Cheers,
Dave.
PATCH 1/16 xfs: avoid empty xattr transaction when attrs are inline
From: Dave Chinner <[email protected]>
generic/642 triggered a reproducable assert failure in xlog_cil_commit() that resulted from a xfs_attr_set() committing an empty but dirty transaction. When the CIL is empty and this occurs, xlog_cil_commit() tries a background push and this triggers a "pushing an empty CIL" assert.
XFS: Assertion failed: !list_empty(&cil->xc_cil), file: fs/xfs/xfs_log_cil.c, line: 1274
Call Trace: <TASK> xlog_cil_commit+0xa5a/0xad0 __xfs_trans_commit+0xb8/0x330 xfs_trans_commit+0x10/0x20 xfs_attr_set+0x3e2/0x4c0 xfs_xattr_set+0x8d/0xe0 __vfs_setxattr+0x6b/0x90 __vfs_setxattr_noperm+0x76/0x220 __vfs_setxattr_locked+0xdf/0x100 vfs_setxattr+0x94/0x170 setxattr+0x110/0x200 path_setxattr+0xbf/0xe0 __x64_sys_setxattr+0x2b/0x30 do_syscall_64+0x35/0x80
The problem is related to the breakdown of attribute addition in xfs_attr_set_iter() and how it is called from deferred operations. When we have a pure leaf xattr insert, we add the xattr to the leaf and set the next state to XFS_DAS_FOUND_LBLK and return -EAGAIN. This requeues the xattr defered work, rolls the transaction and runs xfs_attr_set_iter() again. This then checks the xattr for being remote (it's not) and whether a replace op is being done (this is a create op) and if neither are true it returns without having done anything.
xfs_xattri_finish_update() then unconditionally sets the transaction dirty, and the deferops finishes and returns to __xfs_trans_commit() which sees the transaction dirty and tries to commit it by calling xlog_cil_commit(). The transaction is empty, and then the assert fires if this happens when the CIL is empty.
This patch addresses the structure of xfs_attr_set_iter() that requires re-entry on leaf add even when nothing will be done. This gets rid of the trailing empty transaction and so doesn't trigger the XFS_TRANS_DIRTY assignment in xfs_xattri_finish_update() incorrectly. Addressing that is for a different patch.
--- fs/xfs/libxfs/xfs_attr.c | 39 +++++++++++++++++++-------------------- 1 file changed, 19 insertions(+), 20 deletions(-)
fs/xfs/libxfs/xfs_attr.c +19 -20
@@ -315,6 +315,7 @@ xfs_attr_leaf_addname( {struct xfs_da_args *args = attr->xattri_da_args;struct xfs_inode *dp = args->dp; +enum xfs_delattr_state next_state =XFS_DAS_UNINIT ;int error;if (xfs_attr_is_leaf (dp)) {
@@ -335,37 +336,35 @@ xfs_attr_leaf_addname(* when we come back, we'll be a node, so we'll fall * down into the node handling code below */ -trace_xfs_attr_set_iter_return ( - attr->xattri_dela_state, args->dp); -return -EAGAIN ; + error = -EAGAIN ; + goto out; } - -if (error) -return error; - - attr->xattri_dela_state =XFS_DAS_FOUND_LBLK ; + next_state =XFS_DAS_FOUND_LBLK ; }else { error =xfs_attr_node_addname_find_attr (attr);if (error)return error; + next_state =XFS_DAS_FOUND_NBLK ; error =xfs_attr_node_addname (attr); -if (error) -return error; - -/* -* If addname was successful, and we dont need to alloc or -* remove anymore blks, we're done. -*/ -if (!args->rmtblkno && - !(args->op_flags &XFS_DA_OP_RENAME )) -return 0 ; + } +if (error) +return error; - attr->xattri_dela_state =XFS_DAS_FOUND_NBLK ; +/* +* We need to commit and roll if we need to allocate remote xattr blocks +* or perform more xattr manipulations. Otherwise there is nothing more +* to do and we can return success. +*/ +if (args->rmtblkno || + (args->op_flags &XFS_DA_OP_RENAME )) { + attr->xattri_dela_state = next_state; + error = -EAGAIN ; } +out:trace_xfs_attr_leaf_addname_return (attr->xattri_dela_state, args->dp); -return -EAGAIN ; +return error; }/*
Signature · 1 line
2.35.1
Re: [PATCH 01/16] xfs: avoid empty xattr transaction when attrs are inline
Quoted 42 lines · Dave Chinner
From: Dave Chinner <[email protected]>
generic/642 triggered a reproducable assert failure in
xlog_cil_commit()that resulted from axfs_attr_set()committing an empty but dirty transaction. When the CIL is empty and this occurs,xlog_cil_commit()tries a background push and this triggers a "pushing an empty CIL" assert.XFS: Assertion failed: !list_empty(&cil->xc_cil), file:
fs/xfs/xfs_log_cil.c, line: 1274Call Trace: <TASK> xlog_cil_commit+0xa5a/0xad0 __xfs_trans_commit+0xb8/0x330 xfs_trans_commit+0x10/0x20 xfs_attr_set+0x3e2/0x4c0 xfs_xattr_set+0x8d/0xe0 __vfs_setxattr+0x6b/0x90 __vfs_setxattr_noperm+0x76/0x220 __vfs_setxattr_locked+0xdf/0x100 vfs_setxattr+0x94/0x170 setxattr+0x110/0x200 path_setxattr+0xbf/0xe0 __x64_sys_setxattr+0x2b/0x30 do_syscall_64+0x35/0x80The problem is related to the breakdown of attribute addition in
xfs_attr_set_iter()and how it is called from deferred operations. When we have a pure leaf xattr insert, we add the xattr to the leaf and set the next state toXFS_DAS_FOUND_LBLKand return-EAGAIN. This requeues the xattr defered work, rolls the transaction and runsxfs_attr_set_iter()again. This then checks the xattr for being remote (it's not) and whether a replace op is being done (this is a create op) and if neither are true it returns without having done anything.
xfs_xattri_finish_update()then unconditionally sets the transaction dirty, and the deferops finishes and returns to__xfs_trans_commit()which sees the transaction dirty and tries to commit it by callingxlog_cil_commit(). The transaction is empty, and then the assert fires if this happens when the CIL is empty.
Quoted 7 lines
This patch addresses the structure of
xfs_attr_set_iter()that requires re-entry on leaf add even when nothing will be done. This gets rid of the trailing empty transaction and so doesn't trigger theXFS_TRANS_DIRTYassignment inxfs_xattri_finish_update()incorrectly. Addressing that is for a different patch.Signed-off-by Dave Chinner <[email protected]>
Ok makes sense, thx for catching this!
Quoted 78 lines
--- fs/xfs/libxfs/xfs_attr.c | 39 +++++++++++++++++++-------------------(quoted diff) +0
-1- 1 file changed, 19 insertions(+), 20 deletions(-)fs/xfs/libxfs/xfs_attr.c +1
-0@@ -315,6 +315,7 @@ xfs_attr_leaf_addname( {struct xfs_da_args *args = attr->xattri_da_args;struct xfs_inode *dp = args->dp; +enum xfs_delattr_state next_state =XFS_DAS_UNINIT ;int error;if (xfs_attr_is_leaf (dp)) {@@ -335,37 +336,35 @@ xfs_attr_leaf_addname(* when we come back, we'll be a node, so we'll fall
fs/xfs/libxfs/xfs_attr.c +4
-13* down into the node handling code below */ -trace_xfs_attr_set_iter_return ( - attr->xattri_dela_state, args->dp); -return -EAGAIN ; + error = -EAGAIN ; + goto out; } - -if (error) -return error; - - attr->xattri_dela_state =XFS_DAS_FOUND_LBLK ; + next_state =XFS_DAS_FOUND_LBLK ; }else { error =xfs_attr_node_addname_find_attr (attr);if (error)return error; + next_state =XFS_DAS_FOUND_NBLK ; error =xfs_attr_node_addname (attr); -if (error) -return error; - -/* -* If addname was successful, and we dont need to alloc or
fs/xfs/libxfs/xfs_attr.c +5
-6-* remove anymore blks, we're done. -*/ -if (!args->rmtblkno && - !(args->op_flags &XFS_DA_OP_RENAME )) -return 0 ; + } +if (error) +return error; - attr->xattri_dela_state =XFS_DAS_FOUND_NBLK ; +/* +* We need to commit and roll if we need to allocate remote xattr blocks
fs/xfs/libxfs/xfs_attr.c +1
-0+* or perform more xattr manipulations. Otherwise there is nothing more
fs/xfs/libxfs/xfs_attr.c +7
-0+* to do and we can return success. +*/ +if (args->rmtblkno || + (args->op_flags &XFS_DA_OP_RENAME )) { + attr->xattri_dela_state = next_state; + error = -EAGAIN ; } +out: trace_xfs_attr_leaf_addname_return(attr->xattri_dela_state,
args->dp);fs/xfs/libxfs/xfs_attr.c +1
-1-return -EAGAIN ; +return error; }/*
PATCH 2/16 xfs: initialise attrd item to zero
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
On the first allocation of a attrd item, xfs_trans_add_item() fires an assert like so:
XFS (pmem0): EXPERIMENTAL logged extended attributes feature added. Use at your own risk!
XFS: Assertion failed: !test_bit(XFS_LI_DIRTY, &lip->li_flags), file: fs/xfs/xfs_trans.c, line: 683
------------[ cut here ]------------ kernel BUG at fs/xfs/xfs_message.c:102! Call Trace: <TASK> xfs_trans_add_item+0x17e/0x190 xfs_trans_get_attrd+0x67/0x90 xfs_attr_create_done+0x13/0x20 xfs_defer_finish_noroll+0x100/0x690 __xfs_trans_commit+0x144/0x330 xfs_trans_commit+0x10/0x20 xfs_attr_set+0x3e2/0x4c0 xfs_initxattrs+0xaa/0xe0 security_inode_init_security+0xb0/0x130 xfs_init_security+0x18/0x20 xfs_generic_create+0x13a/0x340 xfs_vn_create+0x17/0x20 path_openat+0xff3/0x12f0 do_filp_open+0xb2/0x150
The attrd log item is allocated via kmem_cache_alloc, and xfs_log_item_init() does not zero the entire log item structure - it assumes that the structure is already all zeros as it only initialises non-zero fields. Fix the attr items to be allocated via the *zalloc methods.
--- fs/xfs/xfs_attr_item.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/xfs/xfs_attr_item.c +1 -1
@@ -725,7 +725,7 @@ xfs_trans_get_attrd(struct xfs_trans *tp,ASSERT (tp !=NULL ); - attrdp =kmem_cache_alloc (xfs_attrd_cache,GFP_NOFS | __GFP_NOFAIL); + attrdp =kmem_cache_zalloc (xfs_attrd_cache,GFP_NOFS | __GFP_NOFAIL);xfs_log_item_init (tp->t_mountp, &attrdp->attrd_item,XFS_LI_ATTRD , &xfs_attrd_item_ops);
Signature · 1 line
2.35.1
Re: [PATCH 02/16] xfs: initialise attrd item to zero
Quoted 28 lines · Dave Chinner
From: Dave Chinner <[email protected]>
On the first allocation of a attrd item,
xfs_trans_add_item()fires an assert like so:XFS (pmem0): EXPERIMENTAL logged extended attributes feature added.
Use at your own risk!
XFS: Assertion failed: !test_bit(XFS_LI_DIRTY,&lip->li_flags),
file:fs/xfs/xfs_trans.c, line: 683------------[ cut here ]------------ kernel BUG at fs/xfs/xfs_message.c:102! Call Trace: <TASK> xfs_trans_add_item+0x17e/0x190 xfs_trans_get_attrd+0x67/0x90 xfs_attr_create_done+0x13/0x20 xfs_defer_finish_noroll+0x100/0x690 __xfs_trans_commit+0x144/0x330 xfs_trans_commit+0x10/0x20 xfs_attr_set+0x3e2/0x4c0 xfs_initxattrs+0xaa/0xe0 security_inode_init_security+0xb0/0x130 xfs_init_security+0x18/0x20 xfs_generic_create+0x13a/0x340 xfs_vn_create+0x17/0x20 path_openat+0xff3/0x12f0 do_filp_open+0xb2/0x150
Quoted 7 lines
The attrd log item is allocated via
kmem_cache_alloc, andxfs_log_item_init()does not zero the entire log item structure - it assumes that the structure is already all zeros as it only initialises non-zero fields. Fix the attr items to be allocated via the *zalloc methods.Signed-off-by Dave Chinner <[email protected]>
Ok, looks fine
Quoted 21 lines
--- fs/xfs/xfs_attr_item.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)fs/xfs/xfs_attr_item.c +1
-1@@ -725,7 +725,7 @@ xfs_trans_get_attrd(struct xfs_trans *tp,ASSERT (tp !=NULL ); - attrdp =kmem_cache_alloc (xfs_attrd_cache,GFP_NOFS | __GFP_NOFAIL); + attrdp =kmem_cache_zalloc (xfs_attrd_cache,GFP_NOFS | __GFP_NOFAIL);xfs_log_item_init(tp->t_mountp,
&attrdp->attrd_item,XFS_LI_ATTRD,
&xfs_attrd_item_ops);
PATCH 3/16 xfs: make xattri_leaf_bp more useful
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
We currently set it and hold it when converting from short to leaf form, then release it only to immediately look it back up again to do the leaf insert.
Do a bit of refactoring to xfs_attr_leaf_try_add() to avoid this messy handling of the newly allocated leaf buffer.
--- fs/xfs/libxfs/xfs_attr.c | 50 +++++++++++++++++++++++++--------------- 1 file changed, 32 insertions(+), 18 deletions(-)
fs/xfs/libxfs/xfs_attr.c +32 -18
@@ -319,7 +319,15 @@ xfs_attr_leaf_addname(int error;if (xfs_attr_is_leaf (dp)) { + +/* +* Use the leaf buffer we may already hold locked as a result of +* a sf-to-leaf conversion. The held buffer is no longer valid +* after this call, regardless of the result. +*/ error =xfs_attr_leaf_try_add (args, attr->xattri_leaf_bp); + attr->xattri_leaf_bp =NULL ; +if (error == -ENOSPC ) { error =xfs_attr3_leaf_to_node (args);if (error)
@@ -341,6 +349,8 @@ xfs_attr_leaf_addname( } next_state =XFS_DAS_FOUND_LBLK ; }else { +ASSERT (!attr->xattri_leaf_bp); + error =xfs_attr_node_addname_find_attr (attr);if (error)return error;
@@ -396,12 +406,6 @@ xfs_attr_set_iter(*/ if (xfs_attr_is_shortform (dp))return xfs_attr_sf_addname (attr); -if (attr->xattri_leaf_bp !=NULL ) { -xfs_trans_bhold_release (args->trans, - attr->xattri_leaf_bp); - attr->xattri_leaf_bp =NULL ; - } -return xfs_attr_leaf_addname (attr);case XFS_DAS_FOUND_LBLK :
@@ -992,18 +996,31 @@ xfs_attr_leaf_try_add(struct xfs_da_args *args, struct xfs_buf *bp) { - int retval; + int error;/* -* Look up the given attribute in the leaf block. Figure out if -* the given flags produce an error or call for an atomic rename. +* If the caller provided a buffer to us, it is locked and held in +* the transaction because it just did a shortform to leaf conversion. +* Hence we don't need to read it again. Otherwise read in the leaf +* buffer. */ - retval =xfs_attr_leaf_hasname (args, &bp); -if (retval != -ENOATTR && retval != -EEXIST ) -return retval; -if (retval == -ENOATTR && (args->attr_flags &XATTR_REPLACE )) +if (bp) { +xfs_trans_bhold_release (args->trans, bp); + }else { + error =xfs_attr3_leaf_read (args->trans, args->dp,0 , &bp); +if (error) +return error; + } + +/* +* Look up the xattr name to set the insertion point for the new xattr. +*/ + error =xfs_attr3_leaf_lookup_int (bp, args); +if (error != -ENOATTR && error != -EEXIST ) goto out_brelse; -if (retval == -EEXIST ) { +if (error == -ENOATTR && (args->attr_flags &XATTR_REPLACE )) + goto out_brelse; +if (error == -EEXIST ) {if (args->attr_flags &XATTR_CREATE ) goto out_brelse;
@@ -1023,14 +1040,11 @@ xfs_attr_leaf_try_add( args->rmtvaluelen =0 ; } -/* -* Add the attribute to the leaf block -*/ return xfs_attr3_leaf_add (bp, args); out_brelse:xfs_trans_brelse (args->trans, bp); -return retval; +return error; }/*
Signature · 1 line
2.35.1
Re: [PATCH 03/16] xfs: make xattri_leaf_bp more useful
Quoted 10 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We currently set it and hold it when converting from short to leaf form, then release it only to immediately look it back up again to do the leaf insert.
Do a bit of refactoring to
xfs_attr_leaf_try_add()to avoid this messy handling of the newly allocated leaf buffer.Signed-off-by Dave Chinner <[email protected]>
Ok, I think that looks good
Quoted 113 lines
--- fs/xfs/libxfs/xfs_attr.c | 50 +++++++++++++++++++++++++------------- -- 1 file changed, 32 insertions(+), 18 deletions(-)fs/xfs/libxfs/xfs_attr.c +3
-0@@ -319,7 +319,15 @@ xfs_attr_leaf_addname(int error;if (xfs_attr_is_leaf (dp)) { + +/* +* Use the leaf buffer we may already hold locked as a result of
fs/xfs/libxfs/xfs_attr.c +1
-0+* a sf-to-leaf conversion. The held buffer is no longer valid
fs/xfs/libxfs/xfs_attr.c +2
-0+* after this call, regardless of the result. +*/ error = xfs_attr_leaf_try_add(args, attr-Quoted 1 line
xattri_leaf_bp);fs/xfs/libxfs/xfs_attr.c +5
-8+ attr->xattri_leaf_bp =NULL ; +if (error == -ENOSPC ) { error =xfs_attr3_leaf_to_node (args);if (error)@@ -341,6 +349,8 @@ xfs_attr_leaf_addname( } next_state =XFS_DAS_FOUND_LBLK ; }else { +ASSERT (!attr->xattri_leaf_bp); + error =xfs_attr_node_addname_find_attr (attr);if (error)return error;@@ -396,12 +406,6 @@ xfs_attr_set_iter(*/ if (xfs_attr_is_shortform (dp))return xfs_attr_sf_addname (attr); -if (attr->xattri_leaf_bp !=NULL ) { -xfs_trans_bhold_release (args->trans, - attr->xattri_leaf_bp); - attr->xattri_leaf_bp =NULL ; - } -return xfs_attr_leaf_addname (attr);case XFS_DAS_FOUND_LBLK :@@ -992,18 +996,31 @@ xfs_attr_leaf_try_add(struct xfs_da_args *args, struct xfs_buf *bp) { - int retval; + int error;/* -* Look up the given attribute in the leaf block. Figure out if
fs/xfs/libxfs/xfs_attr.c +0
-1-* the given flags produce an error or call for an atomic rename.
fs/xfs/libxfs/xfs_attr.c +1
-0+* If the caller provided a buffer to us, it is locked and held in
fs/xfs/libxfs/xfs_attr.c +1
-0+* the transaction because it just did a shortform to leaf conversion.
fs/xfs/libxfs/xfs_attr.c +1
-0+* Hence we don't need to read it again. Otherwise read in the leaf
fs/xfs/libxfs/xfs_attr.c +5
-4+* buffer. */ - retval =xfs_attr_leaf_hasname (args, &bp); -if (retval != -ENOATTR && retval != -EEXIST ) -return retval; -if (retval == -ENOATTR && (args->attr_flags &XATTR_REPLACE )) +if (bp) { +xfs_trans_bhold_release (args->trans, bp); + }else { + error = xfs_attr3_leaf_read(args->trans, args->dp, 0,&bp);
fs/xfs/libxfs/xfs_attr.c +6
-0+if (error) +return error; + } + +/* +* Look up the xattr name to set the insertion point for the new xattr.
fs/xfs/libxfs/xfs_attr.c +7
-5+*/ + error =xfs_attr3_leaf_lookup_int (bp, args); +if (error != -ENOATTR && error != -EEXIST ) goto out_brelse; -if (retval == -EEXIST ) { +if (error == -ENOATTR && (args->attr_flags &XATTR_REPLACE )) + goto out_brelse; +if (error == -EEXIST ) {if (args->attr_flags &XATTR_CREATE ) goto out_brelse;@@ -1023,14 +1040,11 @@ xfs_attr_leaf_try_add( args->rmtvaluelen =0 ; } -/* -* Add the attribute to the leaf block -*/ return xfs_attr3_leaf_add (bp, args); out_brelse:xfs_trans_brelse (args->trans, bp); -return retval; +return error; }/*
PATCH 4/16 xfs: separate out initial attr_set states
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
We current use XFS_DAS_UNINIT for several steps in the attr_set state machine. We use it for setting shortform xattrs, converting from shortform to leaf, leaf add, leaf-to-node and leaf add. All of these things are essentially known before we start the state machine iterating, so we really should separate them out:
XFS_DAS_SF_ADD:
- tries to do a shortform add
- on success -> done
- on ENOSPC converts to leaf, -> XFS_DAS_LEAF_ADD
- on error, dies.
XFS_DAS_LEAF_ADD:
- tries to do leaf add
- on success:
- inline attr -> done
- remote xattr || REPLACE -> XFS_DAS_FOUND_LBLK
- on ENOSPC converts to node, -> XFS_DAS_NODE_ADD
- on error, dies
XFS_DAS_NODE_ADD:
- tries to do node add
- on success:
- inline attr -> done
- remote xattr || REPLACE -> XFS_DAS_FOUND_NBLK
- on error, dies
This makes it easier to understand how the state machine starts up and sets us up on the path to further state machine simplifications.
This also converts the DAS state tracepoints to use strings rather than numbers, as converting between enums and numbers requires manual counting rather than just reading the name.
This also introduces a XFS_DAS_DONE state so that we can trace successful operation completions easily.
--- fs/xfs/libxfs/xfs_attr.c | 151 ++++++++++++++++++++++----------------- fs/xfs/libxfs/xfs_attr.h | 49 +++++++++---- fs/xfs/xfs_trace.h | 22 +++++- 3 files changed, 140 insertions(+), 82 deletions(-)
fs/xfs/libxfs/xfs_attr.c +84 -67
@@ -59,7 +59,7 @@ STATIC int xfs_attr_leaf_try_add(struct xfs_da_args *args, struct xfs_buf *bp);*/ STATIC int xfs_attr_node_get(xfs_da_args_t *args); STATIC void xfs_attr_restore_rmt_blk(struct xfs_da_args *args); -STATIC int xfs_attr_node_addname(struct xfs_attr_item *attr); +static int xfs_attr_node_try_addname (struct xfs_attr_item *attr); STATIC int xfs_attr_node_addname_find_attr(struct xfs_attr_item *attr); STATIC int xfs_attr_node_addname_clear_incomplete(struct xfs_attr_item *attr); STATIC int xfs_attr_node_hasname(xfs_da_args_t *args,
@@ -224,6 +224,11 @@ xfs_init_attr_trans( } } +/* +* Add an attr to a shortform fork. If there is no space, +* xfs_attr_shortform_addname() will convert to leaf format and return -ENOSPC. +* to use. +*/ STATIC int xfs_attr_try_sf_addname( struct xfs_inode *dp,
@@ -268,7 +273,7 @@ xfs_attr_is_shortform( ip->i_afp->if_nextents == 0); } -STATIC int +static int xfs_attr_sf_addname (struct xfs_attr_item *attr) {
@@ -276,14 +281,12 @@ xfs_attr_sf_addname(struct xfs_inode *dp = args->dp;int error =0 ; -/* -* Try to add the attr to the attribute list in the inode. -*/ error =xfs_attr_try_sf_addname (dp, args); - -/* Should only be 0, -EEXIST or -ENOSPC */ -if (error != -ENOSPC ) -return error; +if (error != -ENOSPC ) { +ASSERT (!error || error == -EEXIST ); + attr->xattri_dela_state =XFS_DAS_DONE ; + goto out; + }/* * It won't fit in the shortform, transform to a leaf block. GROT:
@@ -299,64 +302,42 @@ xfs_attr_sf_addname(* with the write verifier. */ xfs_trans_bhold (args->trans, attr->xattri_leaf_bp); - -/* -* We're still in XFS_DAS_UNINIT state here. We've converted -* the attr fork to leaf format and will restart with the leaf -* add. -*/ -trace_xfs_attr_sf_addname_return (XFS_DAS_UNINIT , args->dp); -return -EAGAIN ; + attr->xattri_dela_state =XFS_DAS_LEAF_ADD ; + error = -EAGAIN ; +out: +trace_xfs_attr_sf_addname_return (attr->xattri_dela_state, args->dp); +return error; } -STATIC int +static int xfs_attr_leaf_addname (struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args; -struct xfs_inode *dp = args->dp; -enum xfs_delattr_state next_state =XFS_DAS_UNINIT ;int error; -if (xfs_attr_is_leaf (dp)) { +ASSERT (xfs_attr_is_leaf (args->dp)); -/* -* Use the leaf buffer we may already hold locked as a result of -* a sf-to-leaf conversion. The held buffer is no longer valid -* after this call, regardless of the result. -*/ - error =xfs_attr_leaf_try_add (args, attr->xattri_leaf_bp); - attr->xattri_leaf_bp =NULL ; - -if (error == -ENOSPC ) { - error =xfs_attr3_leaf_to_node (args); -if (error) -return error; - -/* -* Finish any deferred work items and roll the -* transaction once more. The goal here is to call -* node_addname with the inode and transaction in the -* same state (inode locked and joined, transaction -* clean) no matter how we got to this step. -* -* At this point, we are still in XFS_DAS_UNINIT, but -* when we come back, we'll be a node, so we'll fall -* down into the node handling code below -*/ - error = -EAGAIN ; - goto out; - } - next_state =XFS_DAS_FOUND_LBLK ; - }else { -ASSERT (!attr->xattri_leaf_bp); +/* +* Use the leaf buffer we may already hold locked as a result of +* a sf-to-leaf conversion. The held buffer is no longer valid +* after this call, regardless of the result. +*/ + error =xfs_attr_leaf_try_add (args, attr->xattri_leaf_bp); + attr->xattri_leaf_bp =NULL ; - error =xfs_attr_node_addname_find_attr (attr); +if (error == -ENOSPC ) { + error =xfs_attr3_leaf_to_node (args);if (error)return error; - next_state =XFS_DAS_FOUND_NBLK ; - error =xfs_attr_node_addname (attr); +/* +* We're not in leaf format anymore, so roll the transaction and +* retry the add to the newly allocated node block. +*/ + attr->xattri_dela_state =XFS_DAS_NODE_ADD ; + error = -EAGAIN ; + goto out; }if (error)return error;
@@ -368,15 +349,46 @@ xfs_attr_leaf_addname(*/ if (args->rmtblkno || (args->op_flags &XFS_DA_OP_RENAME )) { - attr->xattri_dela_state = next_state; + attr->xattri_dela_state =XFS_DAS_FOUND_LBLK ; error = -EAGAIN ; + }else { + attr->xattri_dela_state =XFS_DAS_DONE ; } - out:trace_xfs_attr_leaf_addname_return (attr->xattri_dela_state, args->dp);return error; } +static int +xfs_attr_node_addname ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +int error; + +ASSERT (!attr->xattri_leaf_bp); + + error =xfs_attr_node_addname_find_attr (attr); +if (error) +return error; + + error =xfs_attr_node_try_addname (attr); +if (error) +return error; + +if (args->rmtblkno || + (args->op_flags &XFS_DA_OP_RENAME )) { + attr->xattri_dela_state =XFS_DAS_FOUND_NBLK ; + error = -EAGAIN ; + }else { + attr->xattri_dela_state =XFS_DAS_DONE ; + } + +trace_xfs_attr_node_addname_return (attr->xattri_dela_state, args->dp); +return error; +} + +/* * Set the attribute specified in @args. * This routine is meant to function as a delayed operation, and may return
@@ -397,16 +409,14 @@ xfs_attr_set_iter(/* State machine switch */ switch (attr->xattri_dela_state) {case XFS_DAS_UNINIT : -/* -* If the fork is shortform, attempt to add the attr. If there -* is no space, this converts to leaf format and returns -* -EAGAIN with the leaf buffer held across the roll. The caller -* will deal with a transaction roll error, but otherwise -* release the hold once we return with a clean transaction. -*/ -if (xfs_attr_is_shortform (dp)) -return xfs_attr_sf_addname (attr); +ASSERT (0 ); +return -EFSCORRUPTED ; +case XFS_DAS_SF_ADD : +return xfs_attr_sf_addname (attr); +case XFS_DAS_LEAF_ADD :return xfs_attr_leaf_addname (attr); +case XFS_DAS_NODE_ADD : +return xfs_attr_node_addname (attr);case XFS_DAS_FOUND_LBLK :/*
@@ -874,6 +884,13 @@ xfs_attr_set_deferred(if (error)return error; +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_ADD ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_ADD ; +else + new->xattri_dela_state =XFS_DAS_NODE_ADD ; +xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list);return 0 ;
@@ -1233,8 +1250,8 @@ xfs_attr_node_addname_find_attr(* to handle this, and recall the function until a successful error code is *returned. */ -STATIC int -xfs_attr_node_addname ( +static int +xfs_attr_node_try_addname (struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args;
fs/xfs/libxfs/xfs_attr.h +36 -13
57 lines not on this page · raw message
fs/xfs/xfs_trace.h +20 -2
@@ -4098,6 +4098,23 @@ DEFINE_ICLOG_EVENT(xlog_iclog_want_sync);DEFINE_ICLOG_EVENT (xlog_iclog_wait_on);DEFINE_ICLOG_EVENT (xlog_iclog_write); +TRACE_DEFINE_ENUM (XFS_DAS_UNINIT ); +TRACE_DEFINE_ENUM (XFS_DAS_SF_ADD ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ADD ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ADD ); +TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK ); +TRACE_DEFINE_ENUM (XFS_DAS_FOUND_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_FOUND_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); +TRACE_DEFINE_ENUM (XFS_DAS_ALLOC_NODE ); +TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG ); + DECLARE_EVENT_CLASS(xfs_das_state_class, TP_PROTO(int das, struct xfs_inode *ip), TP_ARGS(das, ip),
@@ -4109,8 +4126,9 @@ DECLARE_EVENT_CLASS(xfs_das_state_class, __entry->das = das; __entry->ino = ip->i_ino; ), - TP_printk("state change %d ino 0x%llx", - __entry->das, __entry->ino) + TP_printk("state change %s ino 0x%llx", + __print_symbolic(__entry->das, XFS_DAS_STRINGS), + __entry->ino) )#define DEFINE_DAS_STATE_EVENT (name) \
Signature · 1 line
2.35.1
Re: [PATCH 04/16] xfs: separate out initial attr_set states
Quoted 288 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We current use
XFS_DAS_UNINITfor several steps in theattr_setstate machine. We use it for setting shortform xattrs, converting from shortform to leaf, leaf add, leaf-to-node and leaf add. All of these things are essentially known before we start the state machine iterating, so we really should separate them out:
XFS_DAS_SF_ADD:
- tries to do a shortform add
- on success -> done
- on ENOSPC converts to leaf, ->XFS_DAS_LEAF_ADD
- on error, dies.
XFS_DAS_LEAF_ADD:
- tries to do leaf add
- on success:
- inline attr -> done
- remote xattr || REPLACE ->XFS_DAS_FOUND_LBLK
- on ENOSPC converts to node, ->XFS_DAS_NODE_ADD
- on error, dies
XFS_DAS_NODE_ADD:
- tries to do node add
- on success:
- inline attr -> done
- remote xattr || REPLACE ->XFS_DAS_FOUND_NBLK
- on error, diesThis makes it easier to understand how the state machine starts up and sets us up on the path to further state machine simplifications.
This also converts the DAS state tracepoints to use strings rather than numbers, as converting between enums and numbers requires manual counting rather than just reading the name.
This also introduces a
XFS_DAS_DONEstate so that we can trace successful operation completions easily.Signed-off-by Dave Chinner <[email protected]>--- fs/xfs/libxfs/xfs_attr.c | 151 ++++++++++++++++++++++--------------- -- fs/xfs/libxfs/xfs_attr.h | 49 +++++++++---- fs/xfs/xfs_trace.h | 22 +++++- 3 files changed, 140 insertions(+), 82 deletions(-)fs/xfs/libxfs/xfs_attr.c +1
-1@@ -59,7 +59,7 @@ STATIC int xfs_attr_leaf_try_add(struct xfs_da_args*args, struct xfs_buf *bp); */ STATIC int xfs_attr_node_get(xfs_da_args_t *args); STATIC void xfs_attr_restore_rmt_blk(struct xfs_da_args *args); -STATIC int xfs_attr_node_addname(struct xfs_attr_item *attr); +static int xfs_attr_node_try_addname (struct xfs_attr_item *attr); STATIC int xfs_attr_node_addname_find_attr(struct xfs_attr_item *attr);STATIC int xfs_attr_node_addname_clear_incomplete(struct
xfs_attr_item*attr);
STATIC int xfs_attr_node_hasname(xfs_da_args_t *args,fs/xfs/libxfs/xfs_attr.c +3
-0@@ -224,6 +224,11 @@ xfs_init_attr_trans( } } +/* +* Add an attr to a shortform fork. If there is no space, +* xfs_attr_shortform_addname() will convert to leaf format and return
-ENOSPC.
+ * to use.
+ */
STATIC int
xfs_attr_try_sf_addname(fs/xfs/libxfs/xfs_attr.c +6
-8struct xfs_inode *dp,@@ -268,7 +273,7 @@ xfs_attr_is_shortform( ip->i_afp->if_nextents == 0); } -STATIC int +static int xfs_attr_sf_addname (struct xfs_attr_item *attr) {@@ -276,14 +281,12 @@ xfs_attr_sf_addname(struct xfs_inode *dp = args->dp;int error =0 ; -/* -* Try to add the attr to the attribute list in the inode. -*/ error =xfs_attr_try_sf_addname (dp, args); - -/* Should only be 0, -EEXIST or -ENOSPC */ -if (error != -ENOSPC ) -return error; +if (error != -ENOSPC ) { +ASSERT (!error || error == -EEXIST ); + attr->xattri_dela_state =XFS_DAS_DONE ; + goto out; + }/* * It won't fit in the shortform, transform to a leaf block. GROT:
fs/xfs/libxfs/xfs_attr.c +7
-14@@ -299,64 +302,42 @@ xfs_attr_sf_addname(* with the write verifier. */ xfs_trans_bhold (args->trans, attr->xattri_leaf_bp); - -/* -* We're still in XFS_DAS_UNINIT state here. We've converted -* the attr fork to leaf format and will restart with the leaf -* add. -*/ -trace_xfs_attr_sf_addname_return (XFS_DAS_UNINIT , args->dp); -return -EAGAIN ; + attr->xattri_dela_state =XFS_DAS_LEAF_ADD ; + error = -EAGAIN ; +out: +trace_xfs_attr_sf_addname_return (attr->xattri_dela_state, args- >dp); +return error; } -STATIC int +static int xfs_attr_leaf_addname (struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args; -struct xfs_inode *dp = args->dp; -enum xfs_delattr_state next_state =XFS_DAS_UNINIT ;int error; -if (xfs_attr_is_leaf (dp)) { +ASSERT (xfs_attr_is_leaf (args->dp)); -/* -* Use the leaf buffer we may already hold locked as a result of
fs/xfs/libxfs/xfs_attr.c +0
-1-* a sf-to-leaf conversion. The held buffer is no longer valid
fs/xfs/libxfs/xfs_attr.c +0
-3-* after this call, regardless of the result. -*/ - error = xfs_attr_leaf_try_add(args, attr-Quoted 1 line
xattri_leaf_bp);fs/xfs/libxfs/xfs_attr.c +0
-10- attr->xattri_leaf_bp =NULL ; - -if (error == -ENOSPC ) { - error =xfs_attr3_leaf_to_node (args); -if (error) -return error; - -/* -* Finish any deferred work items and roll the -* transaction once more. The goal here is to call
fs/xfs/libxfs/xfs_attr.c +0
-1-* node_addname with the inode and transaction in the
fs/xfs/libxfs/xfs_attr.c +0
-1-* same state (inode locked and joined, transaction
fs/xfs/libxfs/xfs_attr.c +0
-3-* clean) no matter how we got to this step. -* -* At this point, we are still in
XFS_DAS_UNINIT, butfs/xfs/libxfs/xfs_attr.c +0
-1-* when we come back, we'll be a node, so we'll fall
fs/xfs/libxfs/xfs_attr.c +2
-8-* down into the node handling code below -*/ - error = -EAGAIN ; - goto out; - } - next_state =XFS_DAS_FOUND_LBLK ; - }else { -ASSERT (!attr->xattri_leaf_bp); +/* +* Use the leaf buffer we may already hold locked as a result of
fs/xfs/libxfs/xfs_attr.c +9
-3+* a sf-to-leaf conversion. The held buffer is no longer valid +* after this call, regardless of the result. +*/ + error =xfs_attr_leaf_try_add (args, attr->xattri_leaf_bp); + attr->xattri_leaf_bp =NULL ; - error =xfs_attr_node_addname_find_attr (attr); +if (error == -ENOSPC ) { + error =xfs_attr3_leaf_to_node (args);if (error)return error; - next_state =XFS_DAS_FOUND_NBLK ; - error =xfs_attr_node_addname (attr); +/* +* We're not in leaf format anymore, so roll the transaction and
fs/xfs/libxfs/xfs_attr.c +38
-2+* retry the add to the newly allocated node block. +*/ + attr->xattri_dela_state =XFS_DAS_NODE_ADD ; + error = -EAGAIN ; + goto out; }if (error)return error;@@ -368,15 +349,46 @@ xfs_attr_leaf_addname(*/ if (args->rmtblkno || (args->op_flags &XFS_DA_OP_RENAME )) { - attr->xattri_dela_state = next_state; + attr->xattri_dela_state =XFS_DAS_FOUND_LBLK ; error = -EAGAIN ; + }else { + attr->xattri_dela_state =XFS_DAS_DONE ; } - out:trace_xfs_attr_leaf_addname_return (attr->xattri_dela_state, args->dp);return error; } +static int +xfs_attr_node_addname ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +int error; + +ASSERT (!attr->xattri_leaf_bp); + + error =xfs_attr_node_addname_find_attr (attr); +if (error) +return error; + + error =xfs_attr_node_try_addname (attr); +if (error) +return error; + +if (args->rmtblkno || + (args->op_flags &XFS_DA_OP_RENAME )) { + attr->xattri_dela_state =XFS_DAS_FOUND_NBLK ; + error = -EAGAIN ; + }else { + attr->xattri_dela_state =XFS_DAS_DONE ; + } + +trace_xfs_attr_node_addname_return (attr->xattri_dela_state, args->dp); +return error; +} + +/* * Set the attribute specified in @args.
* This routine is meant to function as a delayed operation, and may
returnfs/xfs/libxfs/xfs_attr.c +0
-2@@ -397,16 +409,14 @@ xfs_attr_set_iter(/* State machine switch */ switch (attr->xattri_dela_state) {case XFS_DAS_UNINIT : -/* -* If the fork is shortform, attempt to add the attr. If there
fs/xfs/libxfs/xfs_attr.c +0
-1-* is no space, this converts to leaf format and returns
fs/xfs/libxfs/xfs_attr.c +0
-1-* -EAGAIN with the leaf buffer held across the roll. The caller
fs/xfs/libxfs/xfs_attr.c +0
-1-* will deal with a transaction roll error, but otherwise
fs/xfs/libxfs/xfs_attr.c +0
-1-* release the hold once we return with a clean transaction.
fs/xfs/libxfs/xfs_attr.c +7
-3-*/ -if (xfs_attr_is_shortform (dp)) -return xfs_attr_sf_addname (attr); +ASSERT (0 ); +return -EFSCORRUPTED ; +case XFS_DAS_SF_ADD : +return xfs_attr_sf_addname (attr); +case XFS_DAS_LEAF_ADD :return xfs_attr_leaf_addname (attr); +case XFS_DAS_NODE_ADD : +return xfs_attr_node_addname (attr);case XFS_DAS_FOUND_LBLK :/*
fs/xfs/libxfs/xfs_attr.c xfs_attr_set_deferred( · line 893
Quoted 10 lines
fs/xfs/libxfs/xfs_attr.c +7
-0@@ -874,6 +884,13 @@ xfs_attr_set_deferred(if (error)return error; +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_ADD ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_ADD ; +else + new->xattri_dela_state =XFS_DAS_NODE_ADD ; +
Mmmm, I dont know about putting this part here, because the xfs_attr_*_deferred routines do not get called during a replay, so this initial state config would get missed. If you scoot it up into the xfs_attr_item_init call just a few lines up, then things should be fine since both code path start with that. Rest looks ok though.
Allison
Quoted 135 lines
fs/xfs/libxfs/xfs_attr.c +0
-0xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new- >xattri_list);fs/xfs/libxfs/xfs_attr.c +0
-0return 0 ;@@ -1233,8 +1250,8 @@ xfs_attr_node_addname_find_attr(* to handle this, and recall the function until a successful error code is
*returned.
*/fs/xfs/libxfs/xfs_attr.c +2
-2-STATIC int -xfs_attr_node_addname ( +static int +xfs_attr_node_try_addname (struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args;fs/xfs/libxfs/xfs_attr.h +0
-7@@ -443,21 +443,44 @@ struct xfs_attr_list_context {* to where it was and resume executing where it left off. */ enum xfs_delattr_state { -XFS_DAS_UNINIT =0 ,/* No state has been set yet */ -XFS_DAS_RMTBLK ,/* Removing remote blks */ -XFS_DAS_RM_NAME ,/* Remove attr name */ -XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ -XFS_DAS_FOUND_LBLK ,/* We found leaf blk for attr */ -XFS_DAS_FOUND_NBLK ,/* We found node blk for attr */ -XFS_DAS_FLIP_LFLAG ,/* Flipped leaf INCOMPLETE attr flag */
fs/xfs/libxfs/xfs_attr.h +0
-3-XFS_DAS_RM_LBLK ,/* A rename is removing leaf blocks */ -XFS_DAS_RD_LEAF ,/* Read in the new leaf */ - XFS_DAS_ALLOC_NODE, /* We are allocating nodeblocks */
fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_FLIP_NFLAG, /* Flipped node INCOMPLETE attrflag */
fs/xfs/libxfs/xfs_attr.h +1
-2-XFS_DAS_RM_NBLK ,/* A rename is removing node blocks */ - XFS_DAS_CLR_FLAG, /* Clear incomplete flag */ + XFS_DAS_UNINIT = 0, /* No state has been set yet*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_SF_ADD, /* Initial shortform set iterstate */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_LEAF_ADD, /* Initial leaf form set iter state*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_ADD, /* Initial node form set iter state*/
fs/xfs/libxfs/xfs_attr.h +3
-0+XFS_DAS_RMTBLK ,/* Removing remote blks */ +XFS_DAS_RM_NAME ,/* Remove attr name */ + XFS_DAS_RM_SHRINK, /* We are shrinking the tree*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_FOUND_LBLK, /* We found leaf blk for attr*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_FOUND_NBLK, /* We found node blk for attr*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_FLIP_LFLAG, /* Flipped leaf INCOMPLETEattr flag */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_RM_LBLK, /* A rename is removing leaf blocks*/
fs/xfs/libxfs/xfs_attr.h +2
-0+XFS_DAS_RD_LEAF ,/* Read in the new leaf */ + XFS_DAS_ALLOC_NODE, /* We are allocating nodeblocks */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_FLIP_NFLAG, /* Flipped node INCOMPLETEattr flag */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_RM_NBLK, /* A rename is removing node blocks*/
fs/xfs/libxfs/xfs_attr.h +21
-0+XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ + XFS_DAS_DONE, /* finished operation */ }; +#define XFS_DAS_STRINGS \ + { XFS_DAS_UNINIT, "XFS_DAS_UNINIT" }, \ + { XFS_DAS_SF_ADD, "XFS_DAS_SF_ADD" }, \ + { XFS_DAS_LEAF_ADD, "XFS_DAS_LEAF_ADD" }, \ + { XFS_DAS_NODE_ADD, "XFS_DAS_NODE_ADD" }, \ + { XFS_DAS_RMTBLK, "XFS_DAS_RMTBLK" }, \ + { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ + { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ + { XFS_DAS_FOUND_LBLK, "XFS_DAS_FOUND_LBLK" }, \ + { XFS_DAS_FOUND_NBLK, "XFS_DAS_FOUND_NBLK" }, \ + { XFS_DAS_FLIP_LFLAG, "XFS_DAS_FLIP_LFLAG" }, \ + { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ + { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ + { XFS_DAS_ALLOC_NODE, "XFS_DAS_ALLOC_NODE" }, \ + { XFS_DAS_FLIP_NFLAG, "XFS_DAS_FLIP_NFLAG" }, \ + { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ + { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ + { XFS_DAS_DONE, "XFS_DAS_DONE" } +/* * Defines for xfs_attr_item.xattri_flags */ fs/xfs/xfs_trace.h +20
-2@@ -4098,6 +4098,23 @@ DEFINE_ICLOG_EVENT(xlog_iclog_want_sync);DEFINE_ICLOG_EVENT (xlog_iclog_wait_on);DEFINE_ICLOG_EVENT (xlog_iclog_write); +TRACE_DEFINE_ENUM (XFS_DAS_UNINIT ); +TRACE_DEFINE_ENUM (XFS_DAS_SF_ADD ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ADD ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ADD ); +TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK ); +TRACE_DEFINE_ENUM (XFS_DAS_FOUND_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_FOUND_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); +TRACE_DEFINE_ENUM (XFS_DAS_ALLOC_NODE ); +TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG ); + DECLARE_EVENT_CLASS(xfs_das_state_class, TP_PROTO(int das, struct xfs_inode *ip), TP_ARGS(das, ip),@@ -4109,8 +4126,9 @@ DECLARE_EVENT_CLASS(xfs_das_state_class, __entry->das = das; __entry->ino = ip->i_ino; ), - TP_printk("state change %d ino 0x%llx", - __entry->das, __entry->ino) + TP_printk("state change %s ino 0x%llx", + __print_symbolic(__entry->das, XFS_DAS_STRINGS), + __entry->ino) )#define DEFINE_DAS_STATE_EVENT (name) \
Re: [PATCH 04/16] xfs: separate out initial attr_set states
Quoted 29 lines · Alli
Quoted 28 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We current use
XFS_DAS_UNINITfor several steps in theattr_setstate machine. We use it for setting shortform xattrs, converting from shortform to leaf, leaf add, leaf-to-node and leaf add. All of these things are essentially known before we start the state machine iterating, so we really should separate them out:
XFS_DAS_SF_ADD:
- tries to do a shortform add
- on success -> done
- on ENOSPC converts to leaf, ->XFS_DAS_LEAF_ADD
- on error, dies.
XFS_DAS_LEAF_ADD:
- tries to do leaf add
- on success:
- inline attr -> done
- remote xattr || REPLACE ->XFS_DAS_FOUND_LBLK
- on ENOSPC converts to node, ->XFS_DAS_NODE_ADD
- on error, dies
XFS_DAS_NODE_ADD:
- tries to do node add
- on success:
- inline attr -> done
- remote xattr || REPLACE ->XFS_DAS_FOUND_NBLK
- on error, dies
....
Quoted 4 lines
Quoted 4 lines
fs/xfs/libxfs/xfs_attr.c +0
-0@@ -874,6 +884,13 @@ xfs_attr_set_deferred(if (error)return error;
Quoted 12 lines
fs/xfs/libxfs/xfs_attr.c xfs_attr_set_deferred( · line 893
Quoted 7 lines
fs/xfs/libxfs/xfs_attr.c +7
-0+if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_ADD ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_ADD ; +else + new->xattri_dela_state =XFS_DAS_NODE_ADD ; +Mmmm, I dont know about putting this part here, because the xfs_attr_*_deferred routines do not get called during a replay, so this initial state config would get missed. If you scoot it up into the
xfs_attr_item_initcall just a few lines up, then things should be fine since both code path start with that. Rest looks ok though.
Yeah, recovery gets fixed up later on in the patchset. I managed to get this through several rounds of fstests auto group and several iterations of recoveryloop testing before I actually had recovery hit the ASSERT(state != XFS_DAS_UNINIT) case....
I'm just getting back to this series now, so I'll look to clean this aspect of it up.
Cheers,
Dave.
Signature · 2 lines
Dave Chinner [email protected]
PATCH 5/16 xfs: kill XFS_DAC_LEAF_ADDNAME_INIT
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
We re-enter the XFS_DAS_FOUND_LBLK state when we have to allocate multiple extents for a remote xattr. We currently have a flag called XFS_DAC_LEAF_ADDNAME_INIT to avoid running the remote attr hole finding code more than once.
However, for the node format tree, we have a separate state for this so we never reenter the state machine at XFS_DAS_FOUND_NBLK and so it does not need a special flag to skip over the remote attr hold finding code.
Convert the leaf block code to use the same state machine as the node blocks and kill the XFS_DAC_LEAF_ADDNAME_INIT flag.
This further points out that this "ALLOC" state is only traversed if we have remote xattrs or we are doing a rename operation. Rename both the leaf and node alloc states to _ALLOC_RMT to indicate they are iterating to do allocation of remote xattr blocks.
--- fs/xfs/libxfs/xfs_attr.c | 45 ++++++++++++++++++++-------------------- fs/xfs/libxfs/xfs_attr.h | 6 ++++-- fs/xfs/xfs_trace.h | 3 ++- 3 files changed, 29 insertions(+), 25 deletions(-)
fs/xfs/libxfs/xfs_attr.c +23 -22
@@ -419,40 +419,41 @@ xfs_attr_set_iter(return xfs_attr_node_addname (attr);case XFS_DAS_FOUND_LBLK : +/* +* Find space for remote blocks and fall into the allocation +* state. +*/ +if (args->rmtblkno >0 ) { + error =xfs_attr_rmtval_find_space (attr); +if (error) +return error; + } + attr->xattri_dela_state =XFS_DAS_LEAF_ALLOC_RMT ; + fallthrough; +case XFS_DAS_LEAF_ALLOC_RMT : +/* * If there was an out-of-line value, allocate the blocks we * identified for its storage and copy the value. This is done * after we create the attribute so that we don't overflow the * maximum size of a transaction and/or hit a deadlock. */ - -/* Open coded xfs_attr_rmtval_set without trans handling */ -if ((attr->xattri_flags &XFS_DAC_LEAF_ADDNAME_INIT ) ==0 ) { - attr->xattri_flags |=XFS_DAC_LEAF_ADDNAME_INIT ; -if (args->rmtblkno >0 ) { - error =xfs_attr_rmtval_find_space (attr); +if (args->rmtblkno >0 ) { +if (attr->xattri_blkcnt >0 ) { + error =xfs_attr_rmtval_set_blk (attr);if (error)return error; +trace_xfs_attr_set_iter_return ( + attr->xattri_dela_state, + args->dp); +return -EAGAIN ; } - } -/* -* Repeat allocating remote blocks for the attr value until -* blkcnt drops to zero. -*/ -if (attr->xattri_blkcnt >0 ) { - error =xfs_attr_rmtval_set_blk (attr); + error =xfs_attr_rmtval_set_value (args);if (error)return error; -trace_xfs_attr_set_iter_return (attr->xattri_dela_state, - args->dp); -return -EAGAIN ; } - error =xfs_attr_rmtval_set_value (args); -if (error) -return error; -/* * If this is not a rename, clear the incomplete flag and we're * done.
@@ -547,15 +548,15 @@ xfs_attr_set_iter(return error; } + attr->xattri_dela_state =XFS_DAS_NODE_ALLOC_RMT ; fallthrough; -case XFS_DAS_ALLOC_NODE : +case XFS_DAS_NODE_ALLOC_RMT :/* * If there was an out-of-line value, allocate the blocks we * identified for its storage and copy the value. This is done * after we create the attribute so that we don't overflow the * maximum size of a transaction and/or hit a deadlock. */ - attr->xattri_dela_state =XFS_DAS_ALLOC_NODE ;if (args->rmtblkno >0 ) {if (attr->xattri_blkcnt >0 ) { error =xfs_attr_rmtval_set_blk (attr);
fs/xfs/libxfs/xfs_attr.h +4 -2
@@ -451,11 +451,12 @@ enum xfs_delattr_state {XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ XFS_DAS_FOUND_LBLK ,/* We found leaf blk for attr */ +XFS_DAS_LEAF_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_FOUND_NBLK ,/* We found node blk for attr */ +XFS_DAS_NODE_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_FLIP_LFLAG ,/* Flipped leaf INCOMPLETE attr flag */ XFS_DAS_RM_LBLK ,/* A rename is removing leaf blocks */ XFS_DAS_RD_LEAF ,/* Read in the new leaf */ -XFS_DAS_ALLOC_NODE ,/* We are allocating node blocks */ XFS_DAS_FLIP_NFLAG ,/* Flipped node INCOMPLETE attr flag */ XFS_DAS_RM_NBLK ,/* A rename is removing node blocks */ XFS_DAS_CLR_FLAG, /* Clear incomplete flag */
@@ -471,11 +472,12 @@ enum xfs_delattr_state { { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ { XFS_DAS_FOUND_LBLK, "XFS_DAS_FOUND_LBLK" }, \ + { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ { XFS_DAS_FOUND_NBLK, "XFS_DAS_FOUND_NBLK" }, \ + { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ { XFS_DAS_FLIP_LFLAG, "XFS_DAS_FLIP_LFLAG" }, \ { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ - { XFS_DAS_ALLOC_NODE, "XFS_DAS_ALLOC_NODE" }, \ { XFS_DAS_FLIP_NFLAG, "XFS_DAS_FLIP_NFLAG" }, \ { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \
fs/xfs/xfs_trace.h +2 -1
@@ -4106,11 +4106,12 @@ TRACE_DEFINE_ENUM(XFS_DAS_RMTBLK);TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME );TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK );TRACE_DEFINE_ENUM (XFS_DAS_FOUND_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_FOUND_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); -TRACE_DEFINE_ENUM (XFS_DAS_ALLOC_NODE );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG );
Signature · 1 line
2.35.1
Re: [PATCH 05/16] xfs: kill XFS_DAC_LEAF_ADDNAME_INIT
Quoted 15 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We re-enter the
XFS_DAS_FOUND_LBLKstate when we have to allocate multiple extents for a remote xattr. We currently have a flag calledXFS_DAC_LEAF_ADDNAME_INITto avoid running the remote attr hole finding code more than once.However, for the node format tree, we have a separate state for this so we never reenter the state machine at
XFS_DAS_FOUND_NBLKand so it does not need a special flag to skip over the remote attr hold finding code.Convert the leaf block code to use the same state machine as the node blocks and kill the
XFS_DAC_LEAF_ADDNAME_INITflag.
Quoted 6 lines
This further points out that this "ALLOC" state is only traversed if we have remote xattrs or we are doing a rename operation. Rename both the leaf and node alloc states to
_ALLOC_RMTto indicate they are iterating to do allocation of remote xattr blocks.Signed-off-by Dave Chinner <[email protected]>
Looks ok to me
Quoted 166 lines
--- fs/xfs/libxfs/xfs_attr.c | 45 ++++++++++++++++++++---------------- ---- fs/xfs/libxfs/xfs_attr.h | 6 ++++-- fs/xfs/xfs_trace.h | 3 ++- 3 files changed, 29 insertions(+), 25 deletions(-)fs/xfs/libxfs/xfs_attr.c +2
-0@@ -419,40 +419,41 @@ xfs_attr_set_iter(return xfs_attr_node_addname (attr);case XFS_DAS_FOUND_LBLK : +/* +* Find space for remote blocks and fall into the allocation
fs/xfs/libxfs/xfs_attr.c +11
-0+* state. +*/ +if (args->rmtblkno >0 ) { + error =xfs_attr_rmtval_find_space (attr); +if (error) +return error; + } + attr->xattri_dela_state =XFS_DAS_LEAF_ALLOC_RMT ; + fallthrough; +case XFS_DAS_LEAF_ALLOC_RMT : +/* * If there was an out-of-line value, allocate the blocks we
* identified for its storage and copy the value. This
is done
* after we create the attribute so that we don't
overflow thefs/xfs/libxfs/xfs_attr.c +0
-2* maximum size of a transaction and/or hit a deadlock. */ - -/* Open coded xfs_attr_rmtval_set without trans handling */
fs/xfs/libxfs/xfs_attr.c +0
-1- if ((attr->xattri_flags & XFS_DAC_LEAF_ADDNAME_INIT) ==0) {
fs/xfs/libxfs/xfs_attr.c +0
-1- attr->xattri_flags |=
XFS_DAC_LEAF_ADDNAME_INIT;fs/xfs/libxfs/xfs_attr.c +0
-2-if (args->rmtblkno >0 ) { - error =
xfs_attr_rmtval_find_space(attr);fs/xfs/libxfs/xfs_attr.c +5
-0+if (args->rmtblkno >0 ) { +if (attr->xattri_blkcnt >0 ) { + error =xfs_attr_rmtval_set_blk (attr);if (error)return error; + trace_xfs_attr_set_iter_return( + attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +2
-3+ args->dp); +return -EAGAIN ; } - } -/* -* Repeat allocating remote blocks for the attr value until
fs/xfs/libxfs/xfs_attr.c +1
-5-* blkcnt drops to zero. -*/ -if (attr->xattri_blkcnt >0 ) { - error =xfs_attr_rmtval_set_blk (attr); + error =xfs_attr_rmtval_set_value (args);if (error)return error; - trace_xfs_attr_set_iter_return(attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +0
-6- args->dp); -return -EAGAIN ; } - error =xfs_attr_rmtval_set_value (args); -if (error) -return error; -/* * If this is not a rename, clear the incomplete flag and we're
fs/xfs/libxfs/xfs_attr.c +2
-1* done. @@ -547,15 +548,15 @@ xfs_attr_set_iter(return error; } + attr->xattri_dela_state =XFS_DAS_NODE_ALLOC_RMT ; fallthrough; -case XFS_DAS_ALLOC_NODE : +case XFS_DAS_NODE_ALLOC_RMT :/* * If there was an out-of-line value, allocate the blocks we
* identified for its storage and copy the value. This
is done
* after we create the attribute so that we don't
overflow thefs/xfs/libxfs/xfs_attr.c +0
-1* maximum size of a transaction and/or hit a deadlock. */ - attr->xattri_dela_state =XFS_DAS_ALLOC_NODE ;if (args->rmtblkno >0 ) {if (attr->xattri_blkcnt >0 ) { error =xfs_attr_rmtval_set_blk (attr);fs/xfs/libxfs/xfs_attr.h +1
-0@@ -451,11 +451,12 @@ enum xfs_delattr_state {XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ XFS_DAS_FOUND_LBLK ,/* We found leaf blk for attr */ + XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remoteblocks */
XFS_DAS_FOUND_NBLK, /* We found node blk for attr
*/fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remoteblocks */
XFS_DAS_FLIP_LFLAG, /* Flipped leaf INCOMPLETE
attr flag */
XFS_DAS_RM_LBLK, /* A rename is removing leaf blocks
*/fs/xfs/libxfs/xfs_attr.h +0
-1XFS_DAS_RD_LEAF, /* Read in the new leaf */ - XFS_DAS_ALLOC_NODE, /* We are allocating nodeblocks */
XFS_DAS_FLIP_NFLAG, /* Flipped node INCOMPLETE
attr flag */
XFS_DAS_RM_NBLK, /* A rename is removing node blocks
*/fs/xfs/libxfs/xfs_attr.h +2
-1XFS_DAS_CLR_FLAG, /* Clear incomplete flag */@@ -471,11 +472,12 @@ enum xfs_delattr_state { { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ { XFS_DAS_FOUND_LBLK, "XFS_DAS_FOUND_LBLK" }, \ + { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ { XFS_DAS_FOUND_NBLK, "XFS_DAS_FOUND_NBLK" }, \ + { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ { XFS_DAS_FLIP_LFLAG, "XFS_DAS_FLIP_LFLAG" }, \ { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ - { XFS_DAS_ALLOC_NODE, "XFS_DAS_ALLOC_NODE" }, \ { XFS_DAS_FLIP_NFLAG, "XFS_DAS_FLIP_NFLAG" }, \ { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \fs/xfs/xfs_trace.h +2
-1@@ -4106,11 +4106,12 @@ TRACE_DEFINE_ENUM(XFS_DAS_RMTBLK);TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME );TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK );TRACE_DEFINE_ENUM (XFS_DAS_FOUND_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_FOUND_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); -TRACE_DEFINE_ENUM (XFS_DAS_ALLOC_NODE );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG );
PATCH 6/16 xfs: consolidate leaf/node states in xfs_attr_set_iter
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
The operations performed from XFS_DAS_FOUND_LBLK through to XFS_DAS_RM_LBLK are now identical to XFS_DAS_FOUND_NBLK through to XFS_DAS_RM_NBLK. We can collapse these down into a single set of code.
To do this, define the states that leaf and node run through as separate sets of sequential states. Then as we move to the next state, we can use increments rather than specific state assignments to move through the states. This means the state progression is set by the initial state that enters the series and we don't need to duplicate the code anymore.
At the exit point of the series we need to select the correct leaf or node state, but that can also be done by state increment rather than assignment.
--- fs/xfs/libxfs/xfs_attr.c | 127 ++++++--------------------------------- fs/xfs/libxfs/xfs_attr.h | 9 ++- 2 files changed, 27 insertions(+), 109 deletions(-)
fs/xfs/libxfs/xfs_attr.c +20 -107
@@ -407,6 +407,7 @@ xfs_attr_set_iter(struct xfs_mount *mp = args->dp->i_mount;/* State machine switch */ +next_state:switch (attr->xattri_dela_state) {case XFS_DAS_UNINIT :ASSERT (0 );
@@ -419,6 +420,7 @@ xfs_attr_set_iter(return xfs_attr_node_addname (attr);case XFS_DAS_FOUND_LBLK : +case XFS_DAS_FOUND_NBLK :/* * Find space for remote blocks and fall into the allocation * state.
@@ -428,9 +430,10 @@ xfs_attr_set_iter(if (error)return error; } - attr->xattri_dela_state =XFS_DAS_LEAF_ALLOC_RMT ; + attr->xattri_dela_state++; fallthrough;case XFS_DAS_LEAF_ALLOC_RMT : +case XFS_DAS_NODE_ALLOC_RMT :/* * If there was an out-of-line value, allocate the blocks we
@@ -479,16 +482,18 @@ xfs_attr_set_iter(return error;/* * Commit the flag value change and start the next trans -* in series. +* in series at FLIP_FLAG. */ - attr->xattri_dela_state =XFS_DAS_FLIP_LFLAG ; + attr->xattri_dela_state++;trace_xfs_attr_set_iter_return (attr->xattri_dela_state, args->dp);return -EAGAIN ; } + attr->xattri_dela_state++; fallthrough;case XFS_DAS_FLIP_LFLAG : +case XFS_DAS_FLIP_NFLAG :/* * Dismantle the "old" attribute/value pair by removing a * "remote" value (if it exists).
@@ -498,10 +503,10 @@ xfs_attr_set_iter(if (error)return error; + attr->xattri_dela_state++; fallthrough;case XFS_DAS_RM_LBLK : -/* Set state in case xfs_attr_rmtval_remove returns -EAGAIN */ - attr->xattri_dela_state =XFS_DAS_RM_LBLK ; +case XFS_DAS_RM_NBLK :if (args->rmtblkno) { error =xfs_attr_rmtval_remove (attr);if (error == -EAGAIN )
@@ -516,7 +521,16 @@ xfs_attr_set_iter(return -EAGAIN ; } - fallthrough; +/* +* This is the end of the shared leaf/node sequence. We need +* to continue at the next state in the sequence, but we can't +* easily just fall through. So we increment to the next state +* and then jump back to switch statement to evaluate the next +* state correctly. +*/ + attr->xattri_dela_state++; + goto next_state; +case XFS_DAS_RD_LEAF :/* * This is the last step for leaf format. Read the block with
@@ -537,106 +551,6 @@ xfs_attr_set_iter(return error; -case XFS_DAS_FOUND_NBLK : -/* -* Find space for remote blocks and fall into the allocation -* state. -*/ -if (args->rmtblkno >0 ) { - error =xfs_attr_rmtval_find_space (attr); -if (error) -return error; - } - - attr->xattri_dela_state =XFS_DAS_NODE_ALLOC_RMT ; - fallthrough; -case XFS_DAS_NODE_ALLOC_RMT : -/* -* If there was an out-of-line value, allocate the blocks we -* identified for its storage and copy the value. This is done -* after we create the attribute so that we don't overflow the -* maximum size of a transaction and/or hit a deadlock. -*/ -if (args->rmtblkno >0 ) { -if (attr->xattri_blkcnt >0 ) { - error =xfs_attr_rmtval_set_blk (attr); -if (error) -return error; -trace_xfs_attr_set_iter_return ( - attr->xattri_dela_state, args->dp); -return -EAGAIN ; - } - - error =xfs_attr_rmtval_set_value (args); -if (error) -return error; - } - -/* -* If this was not a rename, clear the incomplete flag and we're -* done. -*/ -if (!(args->op_flags &XFS_DA_OP_RENAME )) { -if (args->rmtblkno >0 ) - error =xfs_attr3_leaf_clearflag (args); - goto out; - } - -/* -* If this is an atomic rename operation, we must "flip" the -* incomplete flags on the "new" and "old" attribute/value pairs -* so that one disappears and one appears atomically. Then we -* must remove the "old" attribute/value pair. -* -* In a separate transaction, set the incomplete flag on the -* "old" attr and clear the incomplete flag on the "new" attr. -*/ -if (!xfs_has_larp (mp)) { - error =xfs_attr3_leaf_flipflags (args); -if (error) - goto out; -/* -* Commit the flag value change and start the next trans -* in series -*/ - attr->xattri_dela_state =XFS_DAS_FLIP_NFLAG ; -trace_xfs_attr_set_iter_return (attr->xattri_dela_state, - args->dp); -return -EAGAIN ; - } - - fallthrough; -case XFS_DAS_FLIP_NFLAG : -/* -* Dismantle the "old" attribute/value pair by removing a -* "remote" value (if it exists). -*/ -xfs_attr_restore_rmt_blk (args); - - error =xfs_attr_rmtval_invalidate (args); -if (error) -return error; - - fallthrough; -case XFS_DAS_RM_NBLK : -/* Set state in case xfs_attr_rmtval_remove returns -EAGAIN */ - attr->xattri_dela_state =XFS_DAS_RM_NBLK ; -if (args->rmtblkno) { - error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) -trace_xfs_attr_set_iter_return ( - attr->xattri_dela_state, args->dp); - -if (error) -return error; - - attr->xattri_dela_state =XFS_DAS_CLR_FLAG ; -trace_xfs_attr_set_iter_return (attr->xattri_dela_state, - args->dp); -return -EAGAIN ; - } - - fallthrough;case XFS_DAS_CLR_FLAG :/* * The last state for node format. Look up the old attr and
@@ -648,7 +562,6 @@ xfs_attr_set_iter(ASSERT (0 );break ; } -out:return error; }
fs/xfs/libxfs/xfs_attr.h +7 -2
@@ -450,16 +450,21 @@ enum xfs_delattr_state {XFS_DAS_RMTBLK ,/* Removing remote blks */ XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ + +/* Leaf state set sequence */ XFS_DAS_FOUND_LBLK ,/* We found leaf blk for attr */ XFS_DAS_LEAF_ALLOC_RMT ,/* We are allocating remote blocks */ -XFS_DAS_FOUND_NBLK ,/* We found node blk for attr */ -XFS_DAS_NODE_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_FLIP_LFLAG ,/* Flipped leaf INCOMPLETE attr flag */ XFS_DAS_RM_LBLK ,/* A rename is removing leaf blocks */ XFS_DAS_RD_LEAF ,/* Read in the new leaf */ + +/* Node state set sequence, must match leaf state above */ +XFS_DAS_FOUND_NBLK ,/* We found node blk for attr */ +XFS_DAS_NODE_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_FLIP_NFLAG ,/* Flipped node INCOMPLETE attr flag */ XFS_DAS_RM_NBLK ,/* A rename is removing node blocks */ XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ + XFS_DAS_DONE, /* finished operation */ };
Signature · 1 line
2.35.1
Re: [PATCH 06/16] xfs: consolidate leaf/node states in xfs_attr_set_iter
Quoted 14 lines · Dave Chinner
From: Dave Chinner <[email protected]>
The operations performed from
XFS_DAS_FOUND_LBLKthrough toXFS_DAS_RM_LBLKare now identical toXFS_DAS_FOUND_NBLKthrough toXFS_DAS_RM_NBLK. We can collapse these down into a single set of code.To do this, define the states that leaf and node run through as separate sets of sequential states. Then as we move to the next state, we can use increments rather than specific state assignments to move through the states. This means the state progression is set by the initial state that enters the series and we don't need to duplicate the code anymore.
Quoted 5 lines
At the exit point of the series we need to select the correct leaf or node state, but that can also be done by state increment rather than assignment.
Signed-off-by Dave Chinner <[email protected]>
This is kind of loopy, but I can follow it. As long as folks are ok with it, I think it's fine, it does collapse down a lot of duplicate code.
Quoted 273 lines
--- fs/xfs/libxfs/xfs_attr.c | 127 ++++++------------------------------- -- fs/xfs/libxfs/xfs_attr.h | 9 ++- 2 files changed, 27 insertions(+), 109 deletions(-)fs/xfs/libxfs/xfs_attr.c +2
-0@@ -407,6 +407,7 @@ xfs_attr_set_iter(struct xfs_mount *mp = args->dp->i_mount;/* State machine switch */ +next_state:switch (attr->xattri_dela_state) {case XFS_DAS_UNINIT :ASSERT (0 );@@ -419,6 +420,7 @@ xfs_attr_set_iter(return xfs_attr_node_addname (attr);case XFS_DAS_FOUND_LBLK : +case XFS_DAS_FOUND_NBLK :/* * Find space for remote blocks and fall into the allocation
fs/xfs/libxfs/xfs_attr.c +2
-1* state. @@ -428,9 +430,10 @@ xfs_attr_set_iter(if (error)return error; } - attr->xattri_dela_state =XFS_DAS_LEAF_ALLOC_RMT ; + attr->xattri_dela_state++; fallthrough;case XFS_DAS_LEAF_ALLOC_RMT : +case XFS_DAS_NODE_ALLOC_RMT :/* * If there was an out-of-line value, allocate the blocks we
fs/xfs/libxfs/xfs_attr.c +0
-0@@ -479,16 +482,18 @@ xfs_attr_set_iter(return error;/* * Commit the flag value change and start the next trans
fs/xfs/libxfs/xfs_attr.c +2
-2-* in series. +* in series at FLIP_FLAG. */ - attr->xattri_dela_state =XFS_DAS_FLIP_LFLAG ; + attr->xattri_dela_state++; trace_xfs_attr_set_iter_return(attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +2
-0args->dp);return -EAGAIN ; } + attr->xattri_dela_state++; fallthrough;case XFS_DAS_FLIP_LFLAG : +case XFS_DAS_FLIP_NFLAG :/* * Dismantle the "old" attribute/value pair by removing a
fs/xfs/libxfs/xfs_attr.c +4
-4* "remote" value (if it exists). @@ -498,10 +503,10 @@ xfs_attr_set_iter(if (error)return error; + attr->xattri_dela_state++; fallthrough;case XFS_DAS_RM_LBLK : -/* Set state in case xfs_attr_rmtval_remove returns -EAGAIN */ - attr->xattri_dela_state =XFS_DAS_RM_LBLK ; +case XFS_DAS_RM_NBLK :if (args->rmtblkno) { error =xfs_attr_rmtval_remove (attr);if (error == -EAGAIN )@@ -516,7 +521,16 @@ xfs_attr_set_iter(return -EAGAIN ; } - fallthrough; +/* +* This is the end of the shared leaf/node sequence. We need
fs/xfs/libxfs/xfs_attr.c +1
-0+* to continue at the next state in the sequence, but we can't
fs/xfs/libxfs/xfs_attr.c +1
-0+* easily just fall through. So we increment to the next state
fs/xfs/libxfs/xfs_attr.c +1
-0+* and then jump back to switch statement to evaluate the next
fs/xfs/libxfs/xfs_attr.c +5
-0+* state correctly. +*/ + attr->xattri_dela_state++; + goto next_state; +case XFS_DAS_RD_LEAF :/* * This is the last step for leaf format. Read the block with
fs/xfs/libxfs/xfs_attr.c +0
-3@@ -537,106 +551,6 @@ xfs_attr_set_iter(return error; -case XFS_DAS_FOUND_NBLK : -/* -* Find space for remote blocks and fall into the allocation
fs/xfs/libxfs/xfs_attr.c +0
-13-* state. -*/ -if (args->rmtblkno >0 ) { - error =xfs_attr_rmtval_find_space (attr); -if (error) -return error; - } - - attr->xattri_dela_state =XFS_DAS_NODE_ALLOC_RMT ; - fallthrough; -case XFS_DAS_NODE_ALLOC_RMT : -/* -* If there was an out-of-line value, allocate the blocks we
fs/xfs/libxfs/xfs_attr.c +0
-1-* identified for its storage and copy the value. This is done
fs/xfs/libxfs/xfs_attr.c +0
-1-* after we create the attribute so that we don't overflow the
fs/xfs/libxfs/xfs_attr.c +0
-9-* maximum size of a transaction and/or hit a deadlock. -*/ -if (args->rmtblkno >0 ) { -if (attr->xattri_blkcnt >0 ) { - error =xfs_attr_rmtval_set_blk (attr); -if (error) -return error; - trace_xfs_attr_set_iter_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +0
-10-return -EAGAIN ; - } - - error =xfs_attr_rmtval_set_value (args); -if (error) -return error; - } - -/* -* If this was not a rename, clear the incomplete flag and we're
fs/xfs/libxfs/xfs_attr.c +0
-10-* done. -*/ -if (!(args->op_flags &XFS_DA_OP_RENAME )) { -if (args->rmtblkno >0 ) - error =xfs_attr3_leaf_clearflag (args); - goto out; - } - -/* -* If this is an atomic rename operation, we must "flip" the
fs/xfs/libxfs/xfs_attr.c +0
-1-* incomplete flags on the "new" and "old" attribute/value pairs
fs/xfs/libxfs/xfs_attr.c +0
-1-* so that one disappears and one appears atomically. Then we
fs/xfs/libxfs/xfs_attr.c +0
-3-* must remove the "old" attribute/value pair. -* -* In a separate transaction, set the incomplete flag on the
fs/xfs/libxfs/xfs_attr.c +0
-1-* "old" attr and clear the incomplete flag on the "new" attr.
fs/xfs/libxfs/xfs_attr.c +0
-7-*/ -if (!xfs_has_larp (mp)) { - error =xfs_attr3_leaf_flipflags (args); -if (error) - goto out; -/* -* Commit the flag value change and start the next trans
fs/xfs/libxfs/xfs_attr.c +0
-4-* in series -*/ - attr->xattri_dela_state =XFS_DAS_FLIP_NFLAG ; - trace_xfs_attr_set_iter_return(attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +0
-8- args->dp); -return -EAGAIN ; - } - - fallthrough; -case XFS_DAS_FLIP_NFLAG : -/* -* Dismantle the "old" attribute/value pair by removing a
fs/xfs/libxfs/xfs_attr.c +0
-18-* "remote" value (if it exists). -*/ -xfs_attr_restore_rmt_blk (args); - - error =xfs_attr_rmtval_invalidate (args); -if (error) -return error; - - fallthrough; -case XFS_DAS_RM_NBLK : -/* Set state in case xfs_attr_rmtval_remove returns -EAGAIN */ - attr->xattri_dela_state =XFS_DAS_RM_NBLK ; -if (args->rmtblkno) { - error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) - trace_xfs_attr_set_iter_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +0
-6- -if (error) -return error; - - attr->xattri_dela_state =XFS_DAS_CLR_FLAG ; - trace_xfs_attr_set_iter_return(attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +0
-5- args->dp); -return -EAGAIN ; - } - - fallthrough;case XFS_DAS_CLR_FLAG :/* * The last state for node format. Look up the old attr and
fs/xfs/libxfs/xfs_attr.c +0
-1@@ -648,7 +562,6 @@ xfs_attr_set_iter(ASSERT (0 );break ; } -out:return error; }fs/xfs/libxfs/xfs_attr.h +2
-0@@ -450,16 +450,21 @@ enum xfs_delattr_state {XFS_DAS_RMTBLK ,/* Removing remote blks */ XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ + +/* Leaf state set sequence */ XFS_DAS_FOUND_LBLK ,/* We found leaf blk for attr */ XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remoteblocks */
fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_FOUND_NBLK, /* We found node blk for attr*/
fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remoteblocks */
XFS_DAS_FLIP_LFLAG, /* Flipped leaf INCOMPLETE
attr flag */
XFS_DAS_RM_LBLK, /* A rename is removing leaf blocks
*/fs/xfs/libxfs/xfs_attr.h +3
-0XFS_DAS_RD_LEAF ,/* Read in the new leaf */ + +/* Node state set sequence, must match leaf state above */ + XFS_DAS_FOUND_NBLK, /* We found node blk for attr*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remoteblocks */
XFS_DAS_FLIP_NFLAG, /* Flipped node INCOMPLETE
attr flag */
XFS_DAS_RM_NBLK, /* A rename is removing node blocks
*/fs/xfs/libxfs/xfs_attr.h +1
-0XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ + XFS_DAS_DONE, /* finished operation */ };
PATCH 7/16 xfs: split remote attr setting out from replace path
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
When we set a new xattr, we have three exit paths:
1. nothign else to do
2. allocate and set the remote xattr value
3. perform the rest of a replace operation
Currently we push both 2 and 3 into the same state, regardless of whether we just set a remote attribute or not. Once we've set the remote xattr, we have two exit states:
1. nothign else to do
2. perform the rest of a replace operation
Hence we can split the remote xattr allocation and setting into their own states and factor it out of xfs_attr_set_iter() to further clean up the state machine and the implementation of the state machine.
--- fs/xfs/libxfs/xfs_attr.c | 113 +++++++++++++++++++++------------------ fs/xfs/libxfs/xfs_attr.h | 14 +++-- fs/xfs/xfs_trace.h | 9 ++-- 3 files changed, 77 insertions(+), 59 deletions(-)
fs/xfs/libxfs/xfs_attr.c +62 -51
@@ -347,9 +347,11 @@ xfs_attr_leaf_addname(* or perform more xattr manipulations. Otherwise there is nothing more * to do and we can return success. */ -if (args->rmtblkno || - (args->op_flags &XFS_DA_OP_RENAME )) { - attr->xattri_dela_state =XFS_DAS_FOUND_LBLK ; +if (args->rmtblkno) { + attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; + error = -EAGAIN ; + }else if (args->op_flags &XFS_DA_OP_RENAME ) { + attr->xattri_dela_state =XFS_DAS_LEAF_REPLACE ; error = -EAGAIN ; }else { attr->xattri_dela_state =XFS_DAS_DONE ;
@@ -376,9 +378,11 @@ xfs_attr_node_addname(if (error)return error; -if (args->rmtblkno || - (args->op_flags &XFS_DA_OP_RENAME )) { - attr->xattri_dela_state =XFS_DAS_FOUND_NBLK ; +if (args->rmtblkno) { + attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; + error = -EAGAIN ; + }else if (args->op_flags &XFS_DA_OP_RENAME ) { + attr->xattri_dela_state =XFS_DAS_NODE_REPLACE ; error = -EAGAIN ; }else { attr->xattri_dela_state =XFS_DAS_DONE ;
@@ -388,6 +392,40 @@ xfs_attr_node_addname(return error; } +static int +xfs_attr_rmtval_alloc ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +int error =0 ; + +/* +* If there was an out-of-line value, allocate the blocks we +* identified for its storage and copy the value. This is done +* after we create the attribute so that we don't overflow the +* maximum size of a transaction and/or hit a deadlock. +*/ +if (attr->xattri_blkcnt >0 ) { + error =xfs_attr_rmtval_set_blk (attr); +if (error) +return error; + error = -EAGAIN ; + goto out; + } + + error =xfs_attr_rmtval_set_value (args); +if (error) +return error; + +/* If this is not a rename, clear the incomplete flag and we're done. */ +if (!(args->op_flags &XFS_DA_OP_RENAME )) { + error =xfs_attr3_leaf_clearflag (args); + attr->xattri_dela_state =XFS_DAS_DONE ; + } +out: +trace_xfs_attr_rmtval_alloc (attr->xattri_dela_state, args->dp); +return error; +}/* * Set the attribute specified in @args.
@@ -419,54 +457,26 @@ xfs_attr_set_iter(case XFS_DAS_NODE_ADD :return xfs_attr_node_addname (attr); -case XFS_DAS_FOUND_LBLK : -case XFS_DAS_FOUND_NBLK : -/* -* Find space for remote blocks and fall into the allocation -* state. -*/ -if (args->rmtblkno >0 ) { - error =xfs_attr_rmtval_find_space (attr); -if (error) -return error; - } +case XFS_DAS_LEAF_SET_RMT : +case XFS_DAS_NODE_SET_RMT : + error =xfs_attr_rmtval_find_space (attr); +if (error) +return error; attr->xattri_dela_state++; fallthrough; +case XFS_DAS_LEAF_ALLOC_RMT :case XFS_DAS_NODE_ALLOC_RMT : - -/* -* If there was an out-of-line value, allocate the blocks we -* identified for its storage and copy the value. This is done -* after we create the attribute so that we don't overflow the -* maximum size of a transaction and/or hit a deadlock. -*/ -if (args->rmtblkno >0 ) { -if (attr->xattri_blkcnt >0 ) { - error =xfs_attr_rmtval_set_blk (attr); -if (error) -return error; -trace_xfs_attr_set_iter_return ( - attr->xattri_dela_state, - args->dp); -return -EAGAIN ; - } - - error =xfs_attr_rmtval_set_value (args); -if (error) -return error; - } - -/* -* If this is not a rename, clear the incomplete flag and we're -* done. -*/ -if (!(args->op_flags &XFS_DA_OP_RENAME )) { -if (args->rmtblkno >0 ) - error =xfs_attr3_leaf_clearflag (args); + error =xfs_attr_rmtval_alloc (attr); +if (error)return error; - } +if (attr->xattri_dela_state ==XFS_DAS_DONE ) +break ; + attr->xattri_dela_state++; + fallthrough; +case XFS_DAS_LEAF_REPLACE : +case XFS_DAS_NODE_REPLACE :/* * If this is an atomic rename operation, we must "flip" the * incomplete flags on the "new" and "old" attribute/value pairs
@@ -484,10 +494,9 @@ xfs_attr_set_iter(* Commit the flag value change and start the next trans * in series at FLIP_FLAG. */ + error = -EAGAIN ; attr->xattri_dela_state++; -trace_xfs_attr_set_iter_return (attr->xattri_dela_state, - args->dp); -return -EAGAIN ; +break ; } attr->xattri_dela_state++;
@@ -562,6 +571,8 @@ xfs_attr_set_iter(ASSERT (0 );break ; } + +trace_xfs_attr_set_iter_return (attr->xattri_dela_state, args->dp);return error; }
fs/xfs/libxfs/xfs_attr.h +9 -5
@@ -452,15 +452,17 @@ enum xfs_delattr_state {XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ /* Leaf state set sequence */ -XFS_DAS_FOUND_LBLK ,/* We found leaf blk for attr */ +XFS_DAS_LEAF_SET_RMT ,/* set a remote xattr from a leaf */ XFS_DAS_LEAF_ALLOC_RMT ,/* We are allocating remote blocks */ +XFS_DAS_LEAF_REPLACE ,/* Perform replace ops on a leaf */ XFS_DAS_FLIP_LFLAG ,/* Flipped leaf INCOMPLETE attr flag */ XFS_DAS_RM_LBLK ,/* A rename is removing leaf blocks */ XFS_DAS_RD_LEAF ,/* Read in the new leaf */ /* Node state set sequence, must match leaf state above */ -XFS_DAS_FOUND_NBLK ,/* We found node blk for attr */ +XFS_DAS_NODE_SET_RMT ,/* set a remote xattr from a node */ XFS_DAS_NODE_ALLOC_RMT ,/* We are allocating remote blocks */ +XFS_DAS_NODE_REPLACE ,/* Perform replace ops on a node */ XFS_DAS_FLIP_NFLAG ,/* Flipped node INCOMPLETE attr flag */ XFS_DAS_RM_NBLK ,/* A rename is removing node blocks */ XFS_DAS_CLR_FLAG, /* Clear incomplete flag */
@@ -476,13 +478,15 @@ enum xfs_delattr_state { { XFS_DAS_RMTBLK, "XFS_DAS_RMTBLK" }, \ { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ - { XFS_DAS_FOUND_LBLK, "XFS_DAS_FOUND_LBLK" }, \ + { XFS_DAS_LEAF_SET_RMT, "XFS_DAS_LEAF_SET_RMT" }, \ { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ - { XFS_DAS_FOUND_NBLK, "XFS_DAS_FOUND_NBLK" }, \ - { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ + { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ { XFS_DAS_FLIP_LFLAG, "XFS_DAS_FLIP_LFLAG" }, \ { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ + { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ + { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ + { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ { XFS_DAS_FLIP_NFLAG, "XFS_DAS_FLIP_NFLAG" }, \ { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \
fs/xfs/xfs_trace.h +6 -3
@@ -4105,13 +4105,15 @@ TRACE_DEFINE_ENUM(XFS_DAS_NODE_ADD);TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK );TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME );TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK ); -TRACE_DEFINE_ENUM (XFS_DAS_FOUND_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_FOUND_NBLK ); -TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG );
@@ -4141,6 +4143,7 @@ DEFINE_DAS_STATE_EVENT(xfs_attr_set_iter_return);DEFINE_DAS_STATE_EVENT (xfs_attr_leaf_addname_return);DEFINE_DAS_STATE_EVENT (xfs_attr_node_addname_return);DEFINE_DAS_STATE_EVENT (xfs_attr_remove_iter_return); +DEFINE_DAS_STATE_EVENT (xfs_attr_rmtval_alloc);DEFINE_DAS_STATE_EVENT (xfs_attr_rmtval_remove_return); TRACE_EVENT(xfs_force_shutdown,
Signature · 1 line
2.35.1
Re: [PATCH 07/16] xfs: split remote attr setting out from replace path
Quoted 12 lines · Dave Chinner
From: Dave Chinner <[email protected]>
When we set a new xattr, we have three exit paths:
1. nothign else to do
2. allocate and set the remote xattr value
3. perform the rest of a replace operationCurrently we push both 2 and 3 into the same state, regardless of whether we just set a remote attribute or not. Once we've set the remote xattr, we have two exit states:
Quoted 9 lines
1. nothign else to do
2. perform the rest of a replace operationHence we can split the remote xattr allocation and setting into their own states and factor it out of
xfs_attr_set_iter()to further clean up the state machine and the implementation of the state machine.
Looks like it should retain the same code flow. I can follow it, but I feel like I'm the only one reviewing this? As long as folks are ok with the new code flow, I'm fine with it.
Quoted 275 lines
Signed-off-by Dave Chinner <[email protected]>--- fs/xfs/libxfs/xfs_attr.c | 113 +++++++++++++++++++++-------------- ---- fs/xfs/libxfs/xfs_attr.h | 14 +++-- fs/xfs/xfs_trace.h | 9 ++-- 3 files changed, 77 insertions(+), 59 deletions(-)fs/xfs/libxfs/xfs_attr.c +0
-0@@ -347,9 +347,11 @@ xfs_attr_leaf_addname(* or perform more xattr manipulations. Otherwise there is nothing more
fs/xfs/libxfs/xfs_attr.c +36
-6* to do and we can return success. */ -if (args->rmtblkno || - (args->op_flags &XFS_DA_OP_RENAME )) { - attr->xattri_dela_state =XFS_DAS_FOUND_LBLK ; +if (args->rmtblkno) { + attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; + error = -EAGAIN ; + }else if (args->op_flags &XFS_DA_OP_RENAME ) { + attr->xattri_dela_state =XFS_DAS_LEAF_REPLACE ; error = -EAGAIN ; }else { attr->xattri_dela_state =XFS_DAS_DONE ;@@ -376,9 +378,11 @@ xfs_attr_node_addname(if (error)return error; -if (args->rmtblkno || - (args->op_flags &XFS_DA_OP_RENAME )) { - attr->xattri_dela_state =XFS_DAS_FOUND_NBLK ; +if (args->rmtblkno) { + attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; + error = -EAGAIN ; + }else if (args->op_flags &XFS_DA_OP_RENAME ) { + attr->xattri_dela_state =XFS_DAS_NODE_REPLACE ; error = -EAGAIN ; }else { attr->xattri_dela_state =XFS_DAS_DONE ;@@ -388,6 +392,40 @@ xfs_attr_node_addname(return error; } +static int +xfs_attr_rmtval_alloc ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +int error =0 ; + +/* +* If there was an out-of-line value, allocate the blocks we +* identified for its storage and copy the value. This is done +* after we create the attribute so that we don't overflow the +* maximum size of a transaction and/or hit a deadlock. +*/ +if (attr->xattri_blkcnt >0 ) { + error =xfs_attr_rmtval_set_blk (attr); +if (error) +return error; + error = -EAGAIN ; + goto out; + } + + error =xfs_attr_rmtval_set_value (args); +if (error) +return error; + +/* If this is not a rename, clear the incomplete flag and we're done. */
fs/xfs/libxfs/xfs_attr.c +8
-4+if (!(args->op_flags &XFS_DA_OP_RENAME )) { + error =xfs_attr3_leaf_clearflag (args); + attr->xattri_dela_state =XFS_DAS_DONE ; + } +out: +trace_xfs_attr_rmtval_alloc (attr->xattri_dela_state, args->dp); +return error; +}/* * Set the attribute specified in @args. @@ -419,54 +457,26 @@ xfs_attr_set_iter(case XFS_DAS_NODE_ADD :return xfs_attr_node_addname (attr); -case XFS_DAS_FOUND_LBLK : -case XFS_DAS_FOUND_NBLK : -/* -* Find space for remote blocks and fall into the allocation
fs/xfs/libxfs/xfs_attr.c +6
-10-* state. -*/ -if (args->rmtblkno >0 ) { - error =xfs_attr_rmtval_find_space (attr); -if (error) -return error; - } +case XFS_DAS_LEAF_SET_RMT : +case XFS_DAS_NODE_SET_RMT : + error =xfs_attr_rmtval_find_space (attr); +if (error) +return error; attr->xattri_dela_state++; fallthrough; +case XFS_DAS_LEAF_ALLOC_RMT :case XFS_DAS_NODE_ALLOC_RMT : - -/* -* If there was an out-of-line value, allocate the blocks we
fs/xfs/libxfs/xfs_attr.c +0
-1-* identified for its storage and copy the value. This is done
fs/xfs/libxfs/xfs_attr.c +0
-1-* after we create the attribute so that we don't overflow the
fs/xfs/libxfs/xfs_attr.c +0
-9-* maximum size of a transaction and/or hit a deadlock. -*/ -if (args->rmtblkno >0 ) { -if (attr->xattri_blkcnt >0 ) { - error =xfs_attr_rmtval_set_blk (attr); -if (error) -return error; - trace_xfs_attr_set_iter_return( - attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +0
-11- args->dp); -return -EAGAIN ; - } - - error =xfs_attr_rmtval_set_value (args); -if (error) -return error; - } - -/* -* If this is not a rename, clear the incomplete flag and we're
fs/xfs/libxfs/xfs_attr.c +8
-6-* done. -*/ -if (!(args->op_flags &XFS_DA_OP_RENAME )) { -if (args->rmtblkno >0 ) - error =xfs_attr3_leaf_clearflag (args); + error =xfs_attr_rmtval_alloc (attr); +if (error)return error; - } +if (attr->xattri_dela_state ==XFS_DAS_DONE ) +break ; + attr->xattri_dela_state++; + fallthrough; +case XFS_DAS_LEAF_REPLACE : +case XFS_DAS_NODE_REPLACE :/* * If this is an atomic rename operation, we must "flip" the
* incomplete flags on the "new" and "old"
attribute/value pairsfs/xfs/libxfs/xfs_attr.c +0
-0@@ -484,10 +494,9 @@ xfs_attr_set_iter(* Commit the flag value change and start the next trans
fs/xfs/libxfs/xfs_attr.c +1
-1* in series at FLIP_FLAG. */ + error = -EAGAIN ; attr->xattri_dela_state++; - trace_xfs_attr_set_iter_return(attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +3
-2- args->dp); -return -EAGAIN ; +break ; } attr->xattri_dela_state++;@@ -562,6 +571,8 @@ xfs_attr_set_iter(ASSERT (0 );break ; } + +trace_xfs_attr_set_iter_return (attr->xattri_dela_state, args- >dp);return error; }fs/xfs/libxfs/xfs_attr.h +1
-1@@ -452,15 +452,17 @@ enum xfs_delattr_state {XFS_DAS_RM_SHRINK, /* We are shrinking the tree */ /* Leaf state set sequence */ - XFS_DAS_FOUND_LBLK, /* We found leaf blk for attr */ + XFS_DAS_LEAF_SET_RMT, /* set a remote xattr from aleaf */
XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remote
blocks */fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_LEAF_REPLACE, /* Perform replace ops on aleaf */
XFS_DAS_FLIP_LFLAG, /* Flipped leaf INCOMPLETE
attr flag */
XFS_DAS_RM_LBLK, /* A rename is removing leaf blocks
*/fs/xfs/libxfs/xfs_attr.h +0
-1XFS_DAS_RD_LEAF, /* Read in the new leaf *//* Node state set sequence, must match leaf state above */ - XFS_DAS_FOUND_NBLK, /* We found node blk for attr*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_SET_RMT, /* set a remote xattr from anode */
XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remote
blocks */fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_REPLACE, /* Perform replace ops on anode */
XFS_DAS_FLIP_NFLAG, /* Flipped node INCOMPLETE
attr flag */
XFS_DAS_RM_NBLK, /* A rename is removing node blocks
*/fs/xfs/libxfs/xfs_attr.h +5
-3XFS_DAS_CLR_FLAG, /* Clear incomplete flag */@@ -476,13 +478,15 @@ enum xfs_delattr_state { { XFS_DAS_RMTBLK, "XFS_DAS_RMTBLK" }, \ { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ - { XFS_DAS_FOUND_LBLK, "XFS_DAS_FOUND_LBLK" }, \ + { XFS_DAS_LEAF_SET_RMT, "XFS_DAS_LEAF_SET_RMT" }, \ { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ - { XFS_DAS_FOUND_NBLK, "XFS_DAS_FOUND_NBLK" }, \ - { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ + { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ { XFS_DAS_FLIP_LFLAG, "XFS_DAS_FLIP_LFLAG" }, \ { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ + { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ + { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ + { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ { XFS_DAS_FLIP_NFLAG, "XFS_DAS_FLIP_NFLAG" }, \ { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \fs/xfs/xfs_trace.h +6
-3@@ -4105,13 +4105,15 @@ TRACE_DEFINE_ENUM(XFS_DAS_NODE_ADD);TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK );TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME );TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK ); -TRACE_DEFINE_ENUM (XFS_DAS_FOUND_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_FOUND_NBLK ); -TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG );TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG );@@ -4141,6 +4143,7 @@DEFINE_DAS_STATE_EVENT (xfs_attr_set_iter_return);DEFINE_DAS_STATE_EVENT (xfs_attr_leaf_addname_return);DEFINE_DAS_STATE_EVENT (xfs_attr_node_addname_return);DEFINE_DAS_STATE_EVENT (xfs_attr_remove_iter_return); +DEFINE_DAS_STATE_EVENT (xfs_attr_rmtval_alloc);DEFINE_DAS_STATE_EVENT (xfs_attr_rmtval_remove_return);TRACE_EVENT(xfs_force_shutdown,
PATCH 8/16 xfs: XFS_DAS_LEAF_REPLACE state only needed if !LARP
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
We can skip the REPLACE state when LARP is enabled, but that means the XFS_DAS_FLIP_LFLAG state is now poorly named - it indicates something that has been done rather than what the state is going to do. Rename it to "REMOVE_OLD" to indicate that we are now going to perform removal of the old attr.
--- fs/xfs/libxfs/xfs_attr.c | 79 ++++++++++++++++++++++++++-------------- fs/xfs/libxfs/xfs_attr.h | 44 +++++++++++----------- fs/xfs/xfs_trace.h | 4 +- 3 files changed, 75 insertions(+), 52 deletions(-)
fs/xfs/libxfs/xfs_attr.c +51 -28
@@ -309,6 +309,23 @@ xfs_attr_sf_addname(return error; } +/* +* When we bump the state to REPLACE, we may actually need to skip over the +* state. When LARP mode is enabled, we don't need to run the atomic flags flip, +* so we skip straight over the REPLACE state and go on to REMOVE_OLD. +*/ +static void +xfs_attr_dela_state_set_replace ( +struct xfs_attr_item *attr, +enum xfs_delattr_state replace) +{ +struct xfs_da_args *args = attr->xattri_da_args; + + attr->xattri_dela_state = replace; +if (xfs_has_larp (args->dp->i_mount)) + attr->xattri_dela_state++; +} +static int xfs_attr_leaf_addname (struct xfs_attr_item *attr)
@@ -351,7 +368,7 @@ xfs_attr_leaf_addname( attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; error = -EAGAIN ; }else if (args->op_flags &XFS_DA_OP_RENAME ) { - attr->xattri_dela_state =XFS_DAS_LEAF_REPLACE ; +xfs_attr_dela_state_set_replace (attr,XFS_DAS_LEAF_REPLACE ); error = -EAGAIN ; }else { attr->xattri_dela_state =XFS_DAS_DONE ;
@@ -382,7 +399,7 @@ xfs_attr_node_addname( attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; error = -EAGAIN ; }else if (args->op_flags &XFS_DA_OP_RENAME ) { - attr->xattri_dela_state =XFS_DAS_NODE_REPLACE ; +xfs_attr_dela_state_set_replace (attr,XFS_DAS_NODE_REPLACE ); error = -EAGAIN ; }else { attr->xattri_dela_state =XFS_DAS_DONE ;
@@ -421,6 +438,13 @@ xfs_attr_rmtval_alloc(if (!(args->op_flags &XFS_DA_OP_RENAME )) { error =xfs_attr3_leaf_clearflag (args); attr->xattri_dela_state =XFS_DAS_DONE ; + }else { +/* +* We are running a REPLACE operation, so we need to bump the +* state to the step in that operation. +*/ + attr->xattri_dela_state++; +xfs_attr_dela_state_set_replace (attr, attr->xattri_dela_state); } out:trace_xfs_attr_rmtval_alloc (attr->xattri_dela_state, args->dp);
@@ -442,7 +466,6 @@ xfs_attr_set_iter(struct xfs_inode *dp = args->dp;struct xfs_buf *bp =NULL ;int forkoff, error =0 ; -struct xfs_mount *mp = args->dp->i_mount;/* State machine switch */ next_state:
@@ -470,39 +493,39 @@ xfs_attr_set_iter( error =xfs_attr_rmtval_alloc (attr);if (error)return error; +/* +* If there is still more to allocate we need to roll the +* transaction so we have a full transaction reservation for +* the next allocation. +*/ +if (attr->xattri_blkcnt >0 ) +break ;if (attr->xattri_dela_state ==XFS_DAS_DONE )break ; - attr->xattri_dela_state++; - fallthrough; + + goto next_state;case XFS_DAS_LEAF_REPLACE :case XFS_DAS_NODE_REPLACE :/* -* If this is an atomic rename operation, we must "flip" the -* incomplete flags on the "new" and "old" attribute/value pairs -* so that one disappears and one appears atomically. Then we -* must remove the "old" attribute/value pair. -* -* In a separate transaction, set the incomplete flag on the -* "old" attr and clear the incomplete flag on the "new" attr. +* We must "flip" the incomplete flags on the "new" and "old" +* attribute/value pairs so that one disappears and one appears +* atomically. Then we must remove the "old" attribute/value +* pair. */ -if (!xfs_has_larp (mp)) { - error =xfs_attr3_leaf_flipflags (args); -if (error) -return error; -/* -* Commit the flag value change and start the next trans -* in series at FLIP_FLAG. -*/ - error = -EAGAIN ; - attr->xattri_dela_state++; -break ; - } - + error =xfs_attr3_leaf_flipflags (args); +if (error) +return error; +/* +* Commit the flag value change and start the next trans +* in series at REMOVE_OLD. +*/ + error = -EAGAIN ; attr->xattri_dela_state++; - fallthrough; -case XFS_DAS_FLIP_LFLAG : -case XFS_DAS_FLIP_NFLAG : +break ; + +case XFS_DAS_LEAF_REMOVE_OLD : +case XFS_DAS_NODE_REMOVE_OLD :/* * Dismantle the "old" attribute/value pair by removing a * "remote" value (if it exists).
fs/xfs/libxfs/xfs_attr.h +22 -22
62 lines not on this page · raw message
fs/xfs/xfs_trace.h +2 -2
@@ -4108,13 +4108,13 @@ TRACE_DEFINE_ENUM(XFS_DAS_RM_SHRINK);TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE ); -TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF );TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE ); -TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG );
Signature · 1 line
2.35.1
Re: [PATCH 08/16] xfs: XFS_DAS_LEAF_REPLACE state only needed if !LARP
Quoted 32 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We can skip the REPLACE state when LARP is enabled, but that means the
XFS_DAS_FLIP_LFLAGstate is now poorly named - it indicates something that has been done rather than what the state is going to do. Rename it to "REMOVE_OLD" to indicate that we are now going to perform removal of the old attr.Signed-off-by Dave Chinner <[email protected]>--- fs/xfs/libxfs/xfs_attr.c | 79 ++++++++++++++++++++++++++---------- ---- fs/xfs/libxfs/xfs_attr.h | 44 +++++++++++----------- fs/xfs/xfs_trace.h | 4 +- 3 files changed, 75 insertions(+), 52 deletions(-)fs/xfs/libxfs/xfs_attr.c +2
-0@@ -309,6 +309,23 @@ xfs_attr_sf_addname(return error; } +/* +* When we bump the state to REPLACE, we may actually need to skip over the
+ * state. When LARP mode is enabled, we don't need to run the atomic flags flip,
+ * so we skip straight over the REPLACE state and go on toREMOVE_OLD.
+ */
fs/xfs/libxfs/xfs_attr.c xfs_attr_sf_addname( · line 323
Quoted 7 lines
fs/xfs/libxfs/xfs_attr.c +7
-0+static void +xfs_attr_dela_state_set_replace ( +struct xfs_attr_item *attr, +enum xfs_delattr_state replace) +{ +struct xfs_da_args *args = attr->xattri_da_args; +
ASSERT(replace == XFS_DAS_LEAF_REPLACE || replace == XFS_DAS_NODE_REPLACE)
?
Quoted 61 lines
fs/xfs/libxfs/xfs_attr.c +6
-1+ attr->xattri_dela_state = replace; +if (xfs_has_larp (args->dp->i_mount)) + attr->xattri_dela_state++; +} +static int xfs_attr_leaf_addname (struct xfs_attr_item *attr)@@ -351,7 +368,7 @@ xfs_attr_leaf_addname( attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; error = -EAGAIN ; }else if (args->op_flags &XFS_DA_OP_RENAME ) { - attr->xattri_dela_state =XFS_DAS_LEAF_REPLACE ; +xfs_attr_dela_state_set_replace (attr,XFS_DAS_LEAF_REPLACE ); error = -EAGAIN ; }else {fs/xfs/libxfs/xfs_attr.c +1
-1attr->xattri_dela_state =XFS_DAS_DONE ;@@ -382,7 +399,7 @@ xfs_attr_node_addname( attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; error = -EAGAIN ; }else if (args->op_flags &XFS_DA_OP_RENAME ) { - attr->xattri_dela_state =XFS_DAS_NODE_REPLACE ; +xfs_attr_dela_state_set_replace (attr,XFS_DAS_NODE_REPLACE ); error = -EAGAIN ; }else {fs/xfs/libxfs/xfs_attr.c +3
-0attr->xattri_dela_state =XFS_DAS_DONE ;@@ -421,6 +438,13 @@ xfs_attr_rmtval_alloc(if (!(args->op_flags &XFS_DA_OP_RENAME )) { error =xfs_attr3_leaf_clearflag (args); attr->xattri_dela_state =XFS_DAS_DONE ; + }else { +/* +* We are running a REPLACE operation, so we need to bump the
fs/xfs/libxfs/xfs_attr.c +4
-0+* state to the step in that operation. +*/ + attr->xattri_dela_state++; + xfs_attr_dela_state_set_replace(attr, attr-Quoted 1 line
xattri_dela_state);fs/xfs/libxfs/xfs_attr.c +2
-1} out:trace_xfs_attr_rmtval_alloc (attr->xattri_dela_state, args->dp);@@ -442,7 +466,6 @@ xfs_attr_set_iter(struct xfs_inode *dp = args->dp;struct xfs_buf *bp =NULL ;int forkoff, error =0 ; -struct xfs_mount *mp = args->dp->i_mount;/* State machine switch */ next_state:@@ -470,39 +493,39 @@ xfs_attr_set_iter( error =xfs_attr_rmtval_alloc (attr);if (error)return error; +/* +* If there is still more to allocate we need to roll the
fs/xfs/libxfs/xfs_attr.c +1
-0+* transaction so we have a full transaction reservation for
fs/xfs/libxfs/xfs_attr.c xfs_attr_set_iter( · line 502
Quoted 4 lines
fs/xfs/libxfs/xfs_attr.c +4
-0+* the next allocation. +*/ +if (attr->xattri_blkcnt >0 ) +break ;
Hmm, if attr->xattri_blkcnt > 0, then xfs_attr_rmtval_alloc returned -EAGAIN, and we would have bailed out by now, so I dont think this part ever gets executed. Though I guess it doesnt hurt anything either?
Rest looks ok though.
Allison
Quoted 168 lines
fs/xfs/libxfs/xfs_attr.c +2
-3if (attr->xattri_dela_state ==XFS_DAS_DONE )break ; - attr->xattri_dela_state++; - fallthrough; + + goto next_state;case XFS_DAS_LEAF_REPLACE :case XFS_DAS_NODE_REPLACE :/* -* If this is an atomic rename operation, we must "flip" the
fs/xfs/libxfs/xfs_attr.c +0
-1-* incomplete flags on the "new" and "old" attribute/value pairs
fs/xfs/libxfs/xfs_attr.c +0
-1-* so that one disappears and one appears atomically. Then we
fs/xfs/libxfs/xfs_attr.c +0
-3-* must remove the "old" attribute/value pair. -* -* In a separate transaction, set the incomplete flag on the
fs/xfs/libxfs/xfs_attr.c +0
-1-* "old" attr and clear the incomplete flag on the "new" attr.
fs/xfs/libxfs/xfs_attr.c +1
-0+* We must "flip" the incomplete flags on the "new" and "old"
fs/xfs/libxfs/xfs_attr.c +1
-0+* attribute/value pairs so that one disappears and one appears
fs/xfs/libxfs/xfs_attr.c +1
-0+* atomically. Then we must remove the "old" attribute/value
fs/xfs/libxfs/xfs_attr.c +1
-6+* pair. */ -if (!xfs_has_larp (mp)) { - error =xfs_attr3_leaf_flipflags (args); -if (error) -return error; -/* -* Commit the flag value change and start the next trans
fs/xfs/libxfs/xfs_attr.c +5
-7-* in series at FLIP_FLAG. -*/ - error = -EAGAIN ; - attr->xattri_dela_state++; -break ; - } - + error =xfs_attr3_leaf_flipflags (args); +if (error) +return error; +/* +* Commit the flag value change and start the next trans
fs/xfs/libxfs/xfs_attr.c +7
-3+* in series at REMOVE_OLD. +*/ + error = -EAGAIN ; attr->xattri_dela_state++; - fallthrough; -case XFS_DAS_FLIP_LFLAG : -case XFS_DAS_FLIP_NFLAG : +break ; + +case XFS_DAS_LEAF_REMOVE_OLD : +case XFS_DAS_NODE_REMOVE_OLD :/* * Dismantle the "old" attribute/value pair by removing a
fs/xfs/libxfs/xfs_attr.c +0
-0* "remote" value (if it exists). fs/xfs/libxfs/xfs_attr.h +0
-0@@ -455,7 +455,7 @@ enum xfs_delattr_state { XFS_DAS_LEAF_SET_RMT, /* set a remote xattr from aleaf */
XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_LEAF_REPLACE, /* Perform replace ops on a
leaf */fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_FLIP_LFLAG, /* Flipped leaf INCOMPLETEattr flag */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_LEAF_REMOVE_OLD, /* Start removing old attr from leaf*/
XFS_DAS_RM_LBLK, /* A rename is removing leaf blocks
*/fs/xfs/libxfs/xfs_attr.h +0
-0XFS_DAS_RD_LEAF, /* Read in the new leaf */@@ -463,7 +463,7 @@ enum xfs_delattr_state { XFS_DAS_NODE_SET_RMT, /* set a remote xattr from anode */
XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_NODE_REPLACE, /* Perform replace ops on a
node */fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_FLIP_NFLAG, /* Flipped node INCOMPLETEattr flag */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_REMOVE_OLD, /* Start removing old attr from node*/
XFS_DAS_RM_NBLK, /* A rename is removing node blocks
*/fs/xfs/libxfs/xfs_attr.h +20
-20XFS_DAS_CLR_FLAG, /* Clear incomplete flag */@@ -471,26 +471,26 @@ enum xfs_delattr_state { };#define XFS_DAS_STRINGS \ - { XFS_DAS_UNINIT, "XFS_DAS_UNINIT" }, \ - { XFS_DAS_SF_ADD, "XFS_DAS_SF_ADD" }, \ - { XFS_DAS_LEAF_ADD, "XFS_DAS_LEAF_ADD" }, \ - { XFS_DAS_NODE_ADD, "XFS_DAS_NODE_ADD" }, \ - { XFS_DAS_RMTBLK, "XFS_DAS_RMTBLK" }, \ - { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ - { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ - { XFS_DAS_LEAF_SET_RMT, "XFS_DAS_LEAF_SET_RMT" }, \ - { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ - { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ - { XFS_DAS_FLIP_LFLAG, "XFS_DAS_FLIP_LFLAG" }, \ - { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ - { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ - { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ - { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ - { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ - { XFS_DAS_FLIP_NFLAG, "XFS_DAS_FLIP_NFLAG" }, \ - { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ - { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ - { XFS_DAS_DONE, "XFS_DAS_DONE" } + { XFS_DAS_UNINIT, "XFS_DAS_UNINIT" }, \ + { XFS_DAS_SF_ADD, "XFS_DAS_SF_ADD" }, \ + { XFS_DAS_LEAF_ADD, "XFS_DAS_LEAF_ADD" }, \ + { XFS_DAS_NODE_ADD, "XFS_DAS_NODE_ADD" }, \ + { XFS_DAS_RMTBLK, "XFS_DAS_RMTBLK" }, \ + { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ + { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ + { XFS_DAS_LEAF_SET_RMT, "XFS_DAS_LEAF_SET_RMT" }, \ + { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ + { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ + { XFS_DAS_LEAF_REMOVE_OLD, "XFS_DAS_LEAF_REMOVE_OLD" }, \ + { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ + { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ + { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ + { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ + { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ + { XFS_DAS_NODE_REMOVE_OLD, "XFS_DAS_NODE_REMOVE_OLD" }, \ + { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ + { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ + { XFS_DAS_DONE, "XFS_DAS_DONE" }/*
* Defines for xfs_attr_item.xattri_flagsfs/xfs/xfs_trace.h +2
-2@@ -4108,13 +4108,13 @@ TRACE_DEFINE_ENUM(XFS_DAS_RM_SHRINK);TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE ); -TRACE_DEFINE_ENUM (XFS_DAS_FLIP_LFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF );TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE ); -TRACE_DEFINE_ENUM (XFS_DAS_FLIP_NFLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG );
PATCH 9/16 xfs: remote xattr removal in xfs_attr_set_iter() is conditional
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
We may not have a remote value for the old xattr we have to remove, so skip over the remote value removal states and go straight to the xattr name removal in the leaf/node block.
--- fs/xfs/libxfs/xfs_attr.c | 59 ++++++++++++++++++++-------------------- fs/xfs/libxfs/xfs_attr.h | 8 +++--- fs/xfs/xfs_trace.h | 4 +-- 3 files changed, 36 insertions(+), 35 deletions(-)
fs/xfs/libxfs/xfs_attr.c +30 -29
@@ -510,15 +510,14 @@ xfs_attr_set_iter(/* * We must "flip" the incomplete flags on the "new" and "old" * attribute/value pairs so that one disappears and one appears -* atomically. Then we must remove the "old" attribute/value -* pair. +* atomically. */ error =xfs_attr3_leaf_flipflags (args);if (error)return error;/* -* Commit the flag value change and start the next trans -* in series at REMOVE_OLD. +* We must commit the flag value change now to make it atomic +* and then we can start the next trans in series at REMOVE_OLD. */ error = -EAGAIN ; attr->xattri_dela_state++;
@@ -527,41 +526,43 @@ xfs_attr_set_iter(case XFS_DAS_LEAF_REMOVE_OLD :case XFS_DAS_NODE_REMOVE_OLD :/* -* Dismantle the "old" attribute/value pair by removing a -* "remote" value (if it exists). +* If we have a remote attr, start the process of removing it +* by invalidating any cached buffers. +* +* If we don't have a remote attr, we skip the remote block +* removal state altogether with a second state increment. */ xfs_attr_restore_rmt_blk (args); - error =xfs_attr_rmtval_invalidate (args); -if (error) -return error; - - attr->xattri_dela_state++; - fallthrough; -case XFS_DAS_RM_LBLK : -case XFS_DAS_RM_NBLK :if (args->rmtblkno) { - error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) -trace_xfs_attr_set_iter_return ( - attr->xattri_dela_state, args->dp); + error =xfs_attr_rmtval_invalidate (args);if (error)return error; - - attr->xattri_dela_state =XFS_DAS_RD_LEAF ; -trace_xfs_attr_set_iter_return (attr->xattri_dela_state, - args->dp); -return -EAGAIN ; + }else { + attr->xattri_dela_state++; } + attr->xattri_dela_state++; + goto next_state; + +case XFS_DAS_LEAF_REMOVE_RMT : +case XFS_DAS_NODE_REMOVE_RMT : + error =xfs_attr_rmtval_remove (attr); +if (error == -EAGAIN ) +break ; +if (error) +return error; +/* -* This is the end of the shared leaf/node sequence. We need -* to continue at the next state in the sequence, but we can't -* easily just fall through. So we increment to the next state -* and then jump back to switch statement to evaluate the next -* state correctly. +* We've finished removing the remote attr blocks, so commit the +* transaction and move on to removing the attr name from the +* leaf/node block. Removing the attr might require a full +* transaction reservation for btree block freeing, so we +* can't do that in the same transaction where we removed the +* remote attr blocks. */ + error = -EAGAIN ; attr->xattri_dela_state++; - goto next_state; +break ;case XFS_DAS_RD_LEAF :/*
fs/xfs/libxfs/xfs_attr.h +4 -4
@@ -456,7 +456,7 @@ enum xfs_delattr_state {XFS_DAS_LEAF_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_LEAF_REPLACE ,/* Perform replace ops on a leaf */ XFS_DAS_LEAF_REMOVE_OLD ,/* Start removing old attr from leaf */ -XFS_DAS_RM_LBLK ,/* A rename is removing leaf blocks */ +XFS_DAS_LEAF_REMOVE_RMT ,/* A rename is removing remote blocks */ XFS_DAS_RD_LEAF, /* Read in the new leaf *//* Node state set sequence, must match leaf state above */
@@ -464,7 +464,7 @@ enum xfs_delattr_state {XFS_DAS_NODE_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_NODE_REPLACE ,/* Perform replace ops on a node */ XFS_DAS_NODE_REMOVE_OLD ,/* Start removing old attr from node */ -XFS_DAS_RM_NBLK ,/* A rename is removing node blocks */ +XFS_DAS_NODE_REMOVE_RMT ,/* A rename is removing remote blocks */ XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ XFS_DAS_DONE, /* finished operation */
@@ -482,13 +482,13 @@ enum xfs_delattr_state { { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ { XFS_DAS_LEAF_REMOVE_OLD, "XFS_DAS_LEAF_REMOVE_OLD" }, \ - { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ + { XFS_DAS_LEAF_REMOVE_RMT, "XFS_DAS_LEAF_REMOVE_RMT" }, \ { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ { XFS_DAS_NODE_REMOVE_OLD, "XFS_DAS_NODE_REMOVE_OLD" }, \ - { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ + { XFS_DAS_NODE_REMOVE_RMT, "XFS_DAS_NODE_REMOVE_RMT" }, \ { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ { XFS_DAS_DONE, "XFS_DAS_DONE" }
fs/xfs/xfs_trace.h +2 -2
@@ -4109,13 +4109,13 @@ TRACE_DEFINE_ENUM(XFS_DAS_LEAF_SET_RMT);TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_OLD ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_RMT );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF );TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_OLD ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_RMT );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG ); DECLARE_EVENT_CLASS(xfs_das_state_class,
Signature · 1 line
2.35.1
Re: [PATCH 09/16] xfs: remote xattr removal in xfs_attr_set_iter() is conditional
Quoted 7 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We may not have a remote value for the old xattr we have to remove, so skip over the remote value removal states and go straight to the xattr name removal in the leaf/node block.
Signed-off-by Dave Chinner <[email protected]>
Ok, I can follow it. I may try and think of a new way to try and illustrate it though as it is not nearly as linear as it used to be, but I don't think that's a bad thing.
Quoted 192 lines
--- fs/xfs/libxfs/xfs_attr.c | 59 ++++++++++++++++++++---------------- ---- fs/xfs/libxfs/xfs_attr.h | 8 +++--- fs/xfs/xfs_trace.h | 4 +-- 3 files changed, 36 insertions(+), 35 deletions(-)fs/xfs/libxfs/xfs_attr.c +0
-0@@ -510,15 +510,14 @@ xfs_attr_set_iter(/* * We must "flip" the incomplete flags on the "new" and "old"
* attribute/value pairs so that one disappears and one
appearsfs/xfs/libxfs/xfs_attr.c +0
-1-* atomically. Then we must remove the "old" attribute/value
fs/xfs/libxfs/xfs_attr.c +1
-2-* pair. +* atomically. */ error =xfs_attr3_leaf_flipflags (args);if (error)return error;/* -* Commit the flag value change and start the next trans
fs/xfs/libxfs/xfs_attr.c +1
-1-* in series at REMOVE_OLD. +* We must commit the flag value change now to make it atomic
fs/xfs/libxfs/xfs_attr.c +1
-0+* and then we can start the next trans in series at
REMOVE_OLD.fs/xfs/libxfs/xfs_attr.c +0
-1*/ error = -EAGAIN ; attr->xattri_dela_state++;@@ -527,41 +526,43 @@ xfs_attr_set_iter(case XFS_DAS_LEAF_REMOVE_OLD :case XFS_DAS_NODE_REMOVE_OLD :/* -* Dismantle the "old" attribute/value pair by removing a
fs/xfs/libxfs/xfs_attr.c +1
-1-* "remote" value (if it exists). +* If we have a remote attr, start the process of removing it
fs/xfs/libxfs/xfs_attr.c +3
-0+* by invalidating any cached buffers. +* +* If we don't have a remote attr, we skip the remote block
fs/xfs/libxfs/xfs_attr.c +1
-0+* removal state altogether with a second state increment.
fs/xfs/libxfs/xfs_attr.c +0
-12*/ xfs_attr_restore_rmt_blk (args); - error =xfs_attr_rmtval_invalidate (args); -if (error) -return error; - - attr->xattri_dela_state++; - fallthrough; -case XFS_DAS_RM_LBLK : -case XFS_DAS_RM_NBLK :if (args->rmtblkno) { - error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) - trace_xfs_attr_set_iter_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +1
-3+ error =xfs_attr_rmtval_invalidate (args);if (error)return error; - - attr->xattri_dela_state =XFS_DAS_RD_LEAF ; - trace_xfs_attr_set_iter_return(attr-Quoted 1 line
xattri_dela_state,fs/xfs/libxfs/xfs_attr.c +13
-3- args->dp); -return -EAGAIN ; + }else { + attr->xattri_dela_state++; } + attr->xattri_dela_state++; + goto next_state; + +case XFS_DAS_LEAF_REMOVE_RMT : +case XFS_DAS_NODE_REMOVE_RMT : + error =xfs_attr_rmtval_remove (attr); +if (error == -EAGAIN ) +break ; +if (error) +return error; +/* -* This is the end of the shared leaf/node sequence. We need
fs/xfs/libxfs/xfs_attr.c +0
-1-* to continue at the next state in the sequence, but we can't
fs/xfs/libxfs/xfs_attr.c +0
-1-* easily just fall through. So we increment to the next state
fs/xfs/libxfs/xfs_attr.c +0
-1-* and then jump back to switch statement to evaluate the next
fs/xfs/libxfs/xfs_attr.c +1
-1-* state correctly. +* We've finished removing the remote attr blocks, so commit the
fs/xfs/libxfs/xfs_attr.c +1
-0+* transaction and move on to removing the attr name from the
fs/xfs/libxfs/xfs_attr.c +1
-0+* leaf/node block. Removing the attr might require a full
fs/xfs/libxfs/xfs_attr.c +1
-0+* transaction reservation for btree block freeing, so we
fs/xfs/libxfs/xfs_attr.c +1
-0+* can't do that in the same transaction where we removed the
fs/xfs/libxfs/xfs_attr.c +3
-1+* remote attr blocks. */ + error = -EAGAIN ; attr->xattri_dela_state++; - goto next_state; +break ;case XFS_DAS_RD_LEAF :/* fs/xfs/libxfs/xfs_attr.h +0
-0@@ -456,7 +456,7 @@ enum xfs_delattr_state { XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remoteblocks */
XFS_DAS_LEAF_REPLACE, /* Perform replace ops on a
leaf */
XFS_DAS_LEAF_REMOVE_OLD, /* Start removing old attr from leaf
*/fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_RM_LBLK, /* A rename is removing leaf blocks*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_LEAF_REMOVE_RMT, /* A rename is removing remote blocks*/
fs/xfs/libxfs/xfs_attr.h +0
-0XFS_DAS_RD_LEAF, /* Read in the new leaf *//* Node state set sequence, must match leaf state above */ @@ -464,7 +464,7 @@ enum xfs_delattr_state { XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remoteblocks */
XFS_DAS_NODE_REPLACE, /* Perform replace ops on a
node */
XFS_DAS_NODE_REMOVE_OLD, /* Start removing old attr from node
*/fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_RM_NBLK, /* A rename is removing node blocks*/
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_REMOVE_RMT, /* A rename is removing remote blocks*/
fs/xfs/libxfs/xfs_attr.h +2
-2XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ XFS_DAS_DONE, /* finished operation */@@ -482,13 +482,13 @@ enum xfs_delattr_state { { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ { XFS_DAS_LEAF_REMOVE_OLD, "XFS_DAS_LEAF_REMOVE_OLD" }, \ - { XFS_DAS_RM_LBLK, "XFS_DAS_RM_LBLK" }, \ + { XFS_DAS_LEAF_REMOVE_RMT, "XFS_DAS_LEAF_REMOVE_RMT" }, \ { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ { XFS_DAS_NODE_REMOVE_OLD, "XFS_DAS_NODE_REMOVE_OLD" }, \ - { XFS_DAS_RM_NBLK, "XFS_DAS_RM_NBLK" }, \ + { XFS_DAS_NODE_REMOVE_RMT, "XFS_DAS_NODE_REMOVE_RMT" },\
fs/xfs/libxfs/xfs_attr.h +0
-0{ XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ { XFS_DAS_DONE, "XFS_DAS_DONE" }fs/xfs/xfs_trace.h +2
-2@@ -4109,13 +4109,13 @@ TRACE_DEFINE_ENUM(XFS_DAS_LEAF_SET_RMT);TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_OLD ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_LBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_RMT );TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF );TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_OLD ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_NBLK ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_RMT );TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG ); DECLARE_EVENT_CLASS(xfs_das_state_class,
PATCH 10/16 xfs: clean up final attr removal in xfs_attr_set_iter
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
Clean up the final leaf/node states in xfs_attr_set_iter() to further simplify the highe level state machine and to set the completion state correctly.
--- fs/xfs/libxfs/xfs_attr.c | 72 +++++++++++++++++++++++----------------- fs/xfs/libxfs/xfs_attr.h | 12 +++---- fs/xfs/xfs_trace.h | 5 +-- 3 files changed, 50 insertions(+), 39 deletions(-)
fs/xfs/libxfs/xfs_attr.c +41 -31
@@ -61,7 +61,7 @@ STATIC int xfs_attr_node_get(xfs_da_args_t *args); STATIC void xfs_attr_restore_rmt_blk(struct xfs_da_args *args);static int xfs_attr_node_try_addname (struct xfs_attr_item *attr); STATIC int xfs_attr_node_addname_find_attr(struct xfs_attr_item *attr); -STATIC int xfs_attr_node_addname_clear_incomplete(struct xfs_attr_item *attr); +STATIC int xfs_attr_node_remove_attr(struct xfs_attr_item *attr); STATIC int xfs_attr_node_hasname(xfs_da_args_t *args,struct xfs_da_state **state); STATIC int xfs_attr_fillstate(xfs_da_state_t *state);
@@ -451,6 +451,36 @@ xfs_attr_rmtval_alloc(return error; } +/* +* Remove the original attr we have just replaced. This is dependent on the +* original lookup and insert placing the old attr in args->blkno/args->index +* and the new attr in args->blkno2/args->index2. +*/ +static int +xfs_attr_leaf_remove_attr ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +struct xfs_inode *dp = args->dp; +struct xfs_buf *bp =NULL ; +int forkoff; +int error; + + error =xfs_attr3_leaf_read (args->trans, args->dp, args->blkno, + &bp); +if (error) +return error; + +xfs_attr3_leaf_remove (bp, args); + + forkoff =xfs_attr_shortform_allfit (bp, dp); +if (forkoff) + error =xfs_attr3_leaf_to_shortform (bp, args, forkoff); +/* bp is gone due to xfs_da_shrink_inode */ + +return error; +} +/* * Set the attribute specified in @args. * This routine is meant to function as a delayed operation, and may return
@@ -463,9 +493,7 @@ xfs_attr_set_iter( struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args; -struct xfs_inode *dp = args->dp; -struct xfs_buf *bp =NULL ; -int forkoff, error =0 ; +int error =0 ;/* State machine switch */ next_state:
@@ -564,32 +592,14 @@ xfs_attr_set_iter( attr->xattri_dela_state++;break ; -case XFS_DAS_RD_LEAF : -/* -* This is the last step for leaf format. Read the block with -* the old attr, remove the old attr, check for shortform -* conversion and return. -*/ - error =xfs_attr3_leaf_read (args->trans, args->dp, args->blkno, - &bp); -if (error) -return error; - -xfs_attr3_leaf_remove (bp, args); - - forkoff =xfs_attr_shortform_allfit (bp, dp); -if (forkoff) - error =xfs_attr3_leaf_to_shortform (bp, args, forkoff); -/* bp is gone due to xfs_da_shrink_inode */ - -return error; +case XFS_DAS_LEAF_REMOVE_ATTR : + error =xfs_attr_leaf_remove_attr (attr); + attr->xattri_dela_state =XFS_DAS_DONE ; +break ; -case XFS_DAS_CLR_FLAG : -/* -* The last state for node format. Look up the old attr and -* remove it. -*/ - error =xfs_attr_node_addname_clear_incomplete (attr); +case XFS_DAS_NODE_REMOVE_ATTR : + error =xfs_attr_node_remove_attr (attr); + attr->xattri_dela_state =XFS_DAS_DONE ;break ;default :ASSERT (0 );
@@ -1262,8 +1272,8 @@ xfs_attr_node_try_addname( } -STATIC int -xfs_attr_node_addname_clear_incomplete ( +static int +xfs_attr_node_remove_attr (struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args;
fs/xfs/libxfs/xfs_attr.h +6 -6
@@ -451,21 +451,21 @@ enum xfs_delattr_state {XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ -/* Leaf state set sequence */ +/* Leaf state set/replace sequence */ XFS_DAS_LEAF_SET_RMT ,/* set a remote xattr from a leaf */ XFS_DAS_LEAF_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_LEAF_REPLACE ,/* Perform replace ops on a leaf */ XFS_DAS_LEAF_REMOVE_OLD ,/* Start removing old attr from leaf */ XFS_DAS_LEAF_REMOVE_RMT ,/* A rename is removing remote blocks */ -XFS_DAS_RD_LEAF ,/* Read in the new leaf */ +XFS_DAS_LEAF_REMOVE_ATTR ,/* Remove the old attr from a leaf */ -/* Node state set sequence, must match leaf state above */ +/* Node state set/replace sequence, must match leaf state above */ XFS_DAS_NODE_SET_RMT ,/* set a remote xattr from a node */ XFS_DAS_NODE_ALLOC_RMT ,/* We are allocating remote blocks */ XFS_DAS_NODE_REPLACE ,/* Perform replace ops on a node */ XFS_DAS_NODE_REMOVE_OLD ,/* Start removing old attr from node */ XFS_DAS_NODE_REMOVE_RMT ,/* A rename is removing remote blocks */ -XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ +XFS_DAS_NODE_REMOVE_ATTR ,/* Remove the old attr from a node */ XFS_DAS_DONE, /* finished operation */ };
@@ -483,13 +483,13 @@ enum xfs_delattr_state { { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ { XFS_DAS_LEAF_REMOVE_OLD, "XFS_DAS_LEAF_REMOVE_OLD" }, \ { XFS_DAS_LEAF_REMOVE_RMT, "XFS_DAS_LEAF_REMOVE_RMT" }, \ - { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ + { XFS_DAS_LEAF_REMOVE_ATTR, "XFS_DAS_LEAF_REMOVE_ATTR" }, \ { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ { XFS_DAS_NODE_REMOVE_OLD, "XFS_DAS_NODE_REMOVE_OLD" }, \ { XFS_DAS_NODE_REMOVE_RMT, "XFS_DAS_NODE_REMOVE_RMT" }, \ - { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ + { XFS_DAS_NODE_REMOVE_ATTR, "XFS_DAS_NODE_REMOVE_ATTR" }, \ { XFS_DAS_DONE, "XFS_DAS_DONE" }/*
fs/xfs/xfs_trace.h +3 -2
@@ -4110,13 +4110,14 @@ TRACE_DEFINE_ENUM(XFS_DAS_LEAF_ALLOC_RMT);TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_ATTR );TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_ATTR ); +TRACE_DEFINE_ENUM (XFS_DAS_DONE ); DECLARE_EVENT_CLASS(xfs_das_state_class, TP_PROTO(int das, struct xfs_inode *ip),
Signature · 1 line
2.35.1
Re: [PATCH 10/16] xfs: clean up final attr removal in xfs_attr_set_iter
Quoted 7 lines · Dave Chinner
From: Dave Chinner <[email protected]>
Clean up the final leaf/node states in
xfs_attr_set_iter()to further simplify the highe level state machine and to set the completion state correctly.Signed-off-by Dave Chinner <[email protected]>
Ok, I can follow it
Quoted 214 lines
--- fs/xfs/libxfs/xfs_attr.c | 72 +++++++++++++++++++++++--------------- -- fs/xfs/libxfs/xfs_attr.h | 12 +++---- fs/xfs/xfs_trace.h | 5 +-- 3 files changed, 50 insertions(+), 39 deletions(-)fs/xfs/libxfs/xfs_attr.c +1
-1@@ -61,7 +61,7 @@ STATIC int xfs_attr_node_get(xfs_da_args_t *args); STATIC void xfs_attr_restore_rmt_blk(struct xfs_da_args *args);static int xfs_attr_node_try_addname (struct xfs_attr_item *attr); STATIC int xfs_attr_node_addname_find_attr(struct xfs_attr_item *attr); -STATIC int xfs_attr_node_addname_clear_incomplete(struct xfs_attr_item *attr); +STATIC int xfs_attr_node_remove_attr(struct xfs_attr_item *attr); STATIC int xfs_attr_node_hasname(xfs_da_args_t *args,fs/xfs/libxfs/xfs_attr.c +2
-0struct xfs_da_state **state); STATIC int xfs_attr_fillstate(xfs_da_state_t *state);@@ -451,6 +451,36 @@ xfs_attr_rmtval_alloc(return error; } +/* +* Remove the original attr we have just replaced. This is dependent on the
+ * original lookup and insert placing the old attr in args-Quoted 1 line
blkno/args->index
+ * and the new attr in args->blkno2/args->index2.
+ */fs/xfs/libxfs/xfs_attr.c +25
-0+static int +xfs_attr_leaf_remove_attr ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +struct xfs_inode *dp = args->dp; +struct xfs_buf *bp =NULL ; +int forkoff; +int error; + + error =xfs_attr3_leaf_read (args->trans, args->dp, args->blkno, + &bp); +if (error) +return error; + +xfs_attr3_leaf_remove (bp, args); + + forkoff =xfs_attr_shortform_allfit (bp, dp); +if (forkoff) + error =xfs_attr3_leaf_to_shortform (bp, args, forkoff); +/* bp is gone due to xfs_da_shrink_inode */ + +return error; +} +/* * Set the attribute specified in @args. * This routine is meant to function as a delayed operation, and may return
fs/xfs/libxfs/xfs_attr.c +1
-6@@ -463,9 +493,7 @@ xfs_attr_set_iter( struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args; -struct xfs_inode *dp = args->dp; -struct xfs_buf *bp =NULL ; -int forkoff, error =0 ; +int error =0 ;/* State machine switch */ next_state:@@ -564,32 +592,14 @@ xfs_attr_set_iter( attr->xattri_dela_state++;break ; -case XFS_DAS_RD_LEAF : -/* -* This is the last step for leaf format. Read the block with
fs/xfs/libxfs/xfs_attr.c +0
-1-* the old attr, remove the old attr, check for shortform
fs/xfs/libxfs/xfs_attr.c +0
-3-* conversion and return. -*/ - error = xfs_attr3_leaf_read(args->trans, args->dp,
args->blkno,fs/xfs/libxfs/xfs_attr.c +0
-9- &bp); -if (error) -return error; - -xfs_attr3_leaf_remove (bp, args); - - forkoff =xfs_attr_shortform_allfit (bp, dp); -if (forkoff) - error = xfs_attr3_leaf_to_shortform(bp, args,forkoff);
fs/xfs/libxfs/xfs_attr.c +4
-6-/* bp is gone due to xfs_da_shrink_inode */ - -return error; +case XFS_DAS_LEAF_REMOVE_ATTR : + error =xfs_attr_leaf_remove_attr (attr); + attr->xattri_dela_state =XFS_DAS_DONE ; +break ; -case XFS_DAS_CLR_FLAG : -/* -* The last state for node format. Look up the old attr and
fs/xfs/libxfs/xfs_attr.c +5
-5-* remove it. -*/ - error =xfs_attr_node_addname_clear_incomplete (attr); +case XFS_DAS_NODE_REMOVE_ATTR : + error =xfs_attr_node_remove_attr (attr); + attr->xattri_dela_state =XFS_DAS_DONE ;break ;default :ASSERT (0 );@@ -1262,8 +1272,8 @@ xfs_attr_node_try_addname( } -STATIC int -xfs_attr_node_addname_clear_incomplete ( +static int +xfs_attr_node_remove_attr (struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args;fs/xfs/libxfs/xfs_attr.h +1
-1@@ -451,21 +451,21 @@ enum xfs_delattr_state {XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ -/* Leaf state set sequence */ +/* Leaf state set/replace sequence */ XFS_DAS_LEAF_SET_RMT, /* set a remote xattr from aleaf */
XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_LEAF_REPLACE, /* Perform replace ops on a
leaf */
XFS_DAS_LEAF_REMOVE_OLD, /* Start removing old attr from leaf
*/
XFS_DAS_LEAF_REMOVE_RMT, /* A rename is removing remote blocks
*/fs/xfs/libxfs/xfs_attr.h +2
-2- XFS_DAS_RD_LEAF, /* Read in the new leaf */ + XFS_DAS_LEAF_REMOVE_ATTR, /* Remove the old attr from a leaf */ -/* Node state set sequence, must match leaf state above */ +/* Node state set/replace sequence, must match leaf state above */
XFS_DAS_NODE_SET_RMT, /* set a remote xattr from a
node */
XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_NODE_REPLACE, /* Perform replace ops on a
node */
XFS_DAS_NODE_REMOVE_OLD, /* Start removing old attr from node
*/
XFS_DAS_NODE_REMOVE_RMT, /* A rename is removing remote blocks
*/fs/xfs/libxfs/xfs_attr.h +2
-2-XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ +XFS_DAS_NODE_REMOVE_ATTR ,/* Remove the old attr from a node */ XFS_DAS_DONE, /* finished operation */ };@@ -483,13 +483,13 @@ enum xfs_delattr_state { { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ { XFS_DAS_LEAF_REMOVE_OLD, "XFS_DAS_LEAF_REMOVE_OLD" }, \ { XFS_DAS_LEAF_REMOVE_RMT, "XFS_DAS_LEAF_REMOVE_RMT" }, \ - { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ + { XFS_DAS_LEAF_REMOVE_ATTR, "XFS_DAS_LEAF_REMOVE_ATTR" }, \ { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ { XFS_DAS_NODE_REMOVE_OLD, "XFS_DAS_NODE_REMOVE_OLD" }, \ { XFS_DAS_NODE_REMOVE_RMT, "XFS_DAS_NODE_REMOVE_RMT" },\
fs/xfs/libxfs/xfs_attr.h +1
-1- { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ + { XFS_DAS_NODE_REMOVE_ATTR, "XFS_DAS_NODE_REMOVE_ATTR" },\
fs/xfs/libxfs/xfs_attr.h +0
-0{ XFS_DAS_DONE, "XFS_DAS_DONE" }/* fs/xfs/xfs_trace.h +3
-2@@ -4110,13 +4110,14 @@ TRACE_DEFINE_ENUM(XFS_DAS_LEAF_ALLOC_RMT);TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_ATTR );TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_ATTR ); +TRACE_DEFINE_ENUM (XFS_DAS_DONE ); DECLARE_EVENT_CLASS(xfs_das_state_class, TP_PROTO(int das, struct xfs_inode *ip),
Re: [PATCH 10/16] xfs: clean up final attr removal in xfs_attr_set_iter
In reply to Dave ChinnerQuoted 7 lines · Dave Chinner
From: Dave Chinner <[email protected]>
Clean up the final leaf/node states in
xfs_attr_set_iter()to further simplify the highe level state machine and to set the completion state correctly.Signed-off-by Dave Chinner <[email protected]>
Ok, looks ok to me
Quoted 214 lines
--- fs/xfs/libxfs/xfs_attr.c | 72 +++++++++++++++++++++++--------------- -- fs/xfs/libxfs/xfs_attr.h | 12 +++---- fs/xfs/xfs_trace.h | 5 +-- 3 files changed, 50 insertions(+), 39 deletions(-)fs/xfs/libxfs/xfs_attr.c +1
-1@@ -61,7 +61,7 @@ STATIC int xfs_attr_node_get(xfs_da_args_t *args); STATIC void xfs_attr_restore_rmt_blk(struct xfs_da_args *args);static int xfs_attr_node_try_addname (struct xfs_attr_item *attr); STATIC int xfs_attr_node_addname_find_attr(struct xfs_attr_item *attr); -STATIC int xfs_attr_node_addname_clear_incomplete(struct xfs_attr_item *attr); +STATIC int xfs_attr_node_remove_attr(struct xfs_attr_item *attr); STATIC int xfs_attr_node_hasname(xfs_da_args_t *args,fs/xfs/libxfs/xfs_attr.c +2
-0struct xfs_da_state **state); STATIC int xfs_attr_fillstate(xfs_da_state_t *state);@@ -451,6 +451,36 @@ xfs_attr_rmtval_alloc(return error; } +/* +* Remove the original attr we have just replaced. This is dependent on the
+ * original lookup and insert placing the old attr in args-Quoted 1 line
blkno/args->index
+ * and the new attr in args->blkno2/args->index2.
+ */fs/xfs/libxfs/xfs_attr.c +25
-0+static int +xfs_attr_leaf_remove_attr ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +struct xfs_inode *dp = args->dp; +struct xfs_buf *bp =NULL ; +int forkoff; +int error; + + error =xfs_attr3_leaf_read (args->trans, args->dp, args->blkno, + &bp); +if (error) +return error; + +xfs_attr3_leaf_remove (bp, args); + + forkoff =xfs_attr_shortform_allfit (bp, dp); +if (forkoff) + error =xfs_attr3_leaf_to_shortform (bp, args, forkoff); +/* bp is gone due to xfs_da_shrink_inode */ + +return error; +} +/* * Set the attribute specified in @args. * This routine is meant to function as a delayed operation, and may return
fs/xfs/libxfs/xfs_attr.c +1
-6@@ -463,9 +493,7 @@ xfs_attr_set_iter( struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args; -struct xfs_inode *dp = args->dp; -struct xfs_buf *bp =NULL ; -int forkoff, error =0 ; +int error =0 ;/* State machine switch */ next_state:@@ -564,32 +592,14 @@ xfs_attr_set_iter( attr->xattri_dela_state++;break ; -case XFS_DAS_RD_LEAF : -/* -* This is the last step for leaf format. Read the block with
fs/xfs/libxfs/xfs_attr.c +0
-1-* the old attr, remove the old attr, check for shortform
fs/xfs/libxfs/xfs_attr.c +0
-3-* conversion and return. -*/ - error = xfs_attr3_leaf_read(args->trans, args->dp,
args->blkno,fs/xfs/libxfs/xfs_attr.c +0
-9- &bp); -if (error) -return error; - -xfs_attr3_leaf_remove (bp, args); - - forkoff =xfs_attr_shortform_allfit (bp, dp); -if (forkoff) - error = xfs_attr3_leaf_to_shortform(bp, args,forkoff);
fs/xfs/libxfs/xfs_attr.c +4
-6-/* bp is gone due to xfs_da_shrink_inode */ - -return error; +case XFS_DAS_LEAF_REMOVE_ATTR : + error =xfs_attr_leaf_remove_attr (attr); + attr->xattri_dela_state =XFS_DAS_DONE ; +break ; -case XFS_DAS_CLR_FLAG : -/* -* The last state for node format. Look up the old attr and
fs/xfs/libxfs/xfs_attr.c +5
-5-* remove it. -*/ - error =xfs_attr_node_addname_clear_incomplete (attr); +case XFS_DAS_NODE_REMOVE_ATTR : + error =xfs_attr_node_remove_attr (attr); + attr->xattri_dela_state =XFS_DAS_DONE ;break ;default :ASSERT (0 );@@ -1262,8 +1272,8 @@ xfs_attr_node_try_addname( } -STATIC int -xfs_attr_node_addname_clear_incomplete ( +static int +xfs_attr_node_remove_attr (struct xfs_attr_item *attr) {struct xfs_da_args *args = attr->xattri_da_args;fs/xfs/libxfs/xfs_attr.h +1
-1@@ -451,21 +451,21 @@ enum xfs_delattr_state {XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ -/* Leaf state set sequence */ +/* Leaf state set/replace sequence */ XFS_DAS_LEAF_SET_RMT, /* set a remote xattr from aleaf */
XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_LEAF_REPLACE, /* Perform replace ops on a
leaf */
XFS_DAS_LEAF_REMOVE_OLD, /* Start removing old attr from leaf
*/
XFS_DAS_LEAF_REMOVE_RMT, /* A rename is removing remote blocks
*/fs/xfs/libxfs/xfs_attr.h +2
-2- XFS_DAS_RD_LEAF, /* Read in the new leaf */ + XFS_DAS_LEAF_REMOVE_ATTR, /* Remove the old attr from a leaf */ -/* Node state set sequence, must match leaf state above */ +/* Node state set/replace sequence, must match leaf state above */
XFS_DAS_NODE_SET_RMT, /* set a remote xattr from a
node */
XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_NODE_REPLACE, /* Perform replace ops on a
node */
XFS_DAS_NODE_REMOVE_OLD, /* Start removing old attr from node
*/
XFS_DAS_NODE_REMOVE_RMT, /* A rename is removing remote blocks
*/fs/xfs/libxfs/xfs_attr.h +2
-2-XFS_DAS_CLR_FLAG ,/* Clear incomplete flag */ +XFS_DAS_NODE_REMOVE_ATTR ,/* Remove the old attr from a node */ XFS_DAS_DONE, /* finished operation */ };@@ -483,13 +483,13 @@ enum xfs_delattr_state { { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \ { XFS_DAS_LEAF_REMOVE_OLD, "XFS_DAS_LEAF_REMOVE_OLD" }, \ { XFS_DAS_LEAF_REMOVE_RMT, "XFS_DAS_LEAF_REMOVE_RMT" }, \ - { XFS_DAS_RD_LEAF, "XFS_DAS_RD_LEAF" }, \ + { XFS_DAS_LEAF_REMOVE_ATTR, "XFS_DAS_LEAF_REMOVE_ATTR" }, \ { XFS_DAS_NODE_SET_RMT, "XFS_DAS_NODE_SET_RMT" }, \ { XFS_DAS_NODE_ALLOC_RMT, "XFS_DAS_NODE_ALLOC_RMT" }, \ { XFS_DAS_NODE_REPLACE, "XFS_DAS_NODE_REPLACE" }, \ { XFS_DAS_NODE_REMOVE_OLD, "XFS_DAS_NODE_REMOVE_OLD" }, \ { XFS_DAS_NODE_REMOVE_RMT, "XFS_DAS_NODE_REMOVE_RMT" },\
fs/xfs/libxfs/xfs_attr.h +1
-1- { XFS_DAS_CLR_FLAG, "XFS_DAS_CLR_FLAG" }, \ + { XFS_DAS_NODE_REMOVE_ATTR, "XFS_DAS_NODE_REMOVE_ATTR" },\
fs/xfs/libxfs/xfs_attr.h +0
-0{ XFS_DAS_DONE, "XFS_DAS_DONE" }/* fs/xfs/xfs_trace.h +3
-2@@ -4110,13 +4110,14 @@ TRACE_DEFINE_ENUM(XFS_DAS_LEAF_ALLOC_RMT);TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_RD_LEAF ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE_ATTR );TRACE_DEFINE_ENUM (XFS_DAS_NODE_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REPLACE );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_OLD );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_RMT ); -TRACE_DEFINE_ENUM (XFS_DAS_CLR_FLAG ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE_ATTR ); +TRACE_DEFINE_ENUM (XFS_DAS_DONE ); DECLARE_EVENT_CLASS(xfs_das_state_class, TP_PROTO(int das, struct xfs_inode *ip),
PATCH 11/16 xfs: xfs_attr_set_iter() does not need to return EAGAIN
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
Now that the full xfs_attr_set_iter() state machine always terminates with either the state being XFS_DAS_DONE on success or an error on failure, we can get rid of the need for it to return -EAGAIN whenever it needs to roll the transaction before running the next state.
That is, we don't need to spray -EAGAIN return states everywhere, the caller just check the state machine state for completion to determine what action should be taken next. This greatly simplifies the code within the state machine implementation as it now only has to handle 0 for success or -errno for error and it doesn't need to tell the caller to retry.
--- fs/xfs/libxfs/xfs_attr.c | 83 +++++++++++++++++----------------------- fs/xfs/xfs_attr_item.c | 2 + 2 files changed, 37 insertions(+), 48 deletions(-)
fs/xfs/libxfs/xfs_attr.c +35 -48
@@ -303,7 +303,6 @@ xfs_attr_sf_addname(*/ xfs_trans_bhold (args->trans, attr->xattri_leaf_bp); attr->xattri_dela_state =XFS_DAS_LEAF_ADD ; - error = -EAGAIN ; out:trace_xfs_attr_sf_addname_return (attr->xattri_dela_state, args->dp);return error;
@@ -353,7 +352,6 @@ xfs_attr_leaf_addname(* retry the add to the newly allocated node block. */ attr->xattri_dela_state =XFS_DAS_NODE_ADD ; - error = -EAGAIN ; goto out; }if (error)
@@ -364,20 +362,24 @@ xfs_attr_leaf_addname(* or perform more xattr manipulations. Otherwise there is nothing more * to do and we can return success. */ -if (args->rmtblkno) { +if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; - error = -EAGAIN ; - }else if (args->op_flags &XFS_DA_OP_RENAME ) { +else if (args->op_flags &XFS_DA_OP_RENAME )xfs_attr_dela_state_set_replace (attr,XFS_DAS_LEAF_REPLACE ); - error = -EAGAIN ; - }else { +else attr->xattri_dela_state =XFS_DAS_DONE ; - } out:trace_xfs_attr_leaf_addname_return (attr->xattri_dela_state, args->dp);return error; } +/* +* Add an entry to a node format attr tree. +* +* Note that we might still have a leaf here - xfs_attr_is_leaf() cannot tell +* the difference between leaf + remote attr blocks and a node format tree, +* so we may still end up having to convert from leaf to node format here. +*/ static int xfs_attr_node_addname (struct xfs_attr_item *attr)
@@ -392,19 +394,26 @@ xfs_attr_node_addname(return error; error =xfs_attr_node_try_addname (attr); +if (error == -ENOSPC ) { + error =xfs_attr3_leaf_to_node (args); +if (error) +return error; +/* +* No state change, we really are in node form now +* but we need the transaction rolled to continue. +*/ + goto out; + }if (error)return error; -if (args->rmtblkno) { +if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; - error = -EAGAIN ; - }else if (args->op_flags &XFS_DA_OP_RENAME ) { +else if (args->op_flags &XFS_DA_OP_RENAME )xfs_attr_dela_state_set_replace (attr,XFS_DAS_NODE_REPLACE ); - error = -EAGAIN ; - }else { +else attr->xattri_dela_state =XFS_DAS_DONE ; - } - +out:trace_xfs_attr_node_addname_return (attr->xattri_dela_state, args->dp);return error; }
@@ -426,7 +435,6 @@ xfs_attr_rmtval_alloc( error =xfs_attr_rmtval_set_blk (attr);if (error)return error; - error = -EAGAIN ; goto out; }
@@ -482,11 +490,12 @@ xfs_attr_leaf_remove_attr( }/* -* Set the attribute specified in @args. -* This routine is meant to function as a delayed operation, and may return -* -EAGAIN when the transaction needs to be rolled. Calling functions will need -* to handle this, and recall the function until a successful error code is -* returned. +* Run the attribute operation specified in @attr. +* +* This routine is meant to function as a delayed operation and will set the +* state to XFS_DAS_DONE when the operation is complete. Calling functions will +* need to handle this, and recall the function until either an error or +* XFS_DAS_DONE is detected. */ int xfs_attr_set_iter(
@@ -547,7 +556,6 @@ xfs_attr_set_iter(* We must commit the flag value change now to make it atomic * and then we can start the next trans in series at REMOVE_OLD. */ - error = -EAGAIN ; attr->xattri_dela_state++;break ;
@@ -575,8 +583,10 @@ xfs_attr_set_iter(case XFS_DAS_LEAF_REMOVE_RMT :case XFS_DAS_NODE_REMOVE_RMT : error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) +if (error == -EAGAIN ) { + error =0 ;break ; + }if (error)return error;
@@ -588,7 +598,6 @@ xfs_attr_set_iter(* can't do that in the same transaction where we removed the * remote attr blocks. */ - error = -EAGAIN ; attr->xattri_dela_state++;break ;
@@ -1200,14 +1209,6 @@ xfs_attr_node_addname_find_attr(* This will involve walking down the Btree, and may involve splitting * leaf nodes and even splitting intermediate nodes up to and including * the root node (a special case of an intermediate node). -* -* "Remote" attribute values confuse the issue and atomic rename operations -* add a whole extra layer of confusion on top of that. -* -* This routine is meant to function as a delayed operation, and may return -* -EAGAIN when the transaction needs to be rolled. Calling functions will need -* to handle this, and recall the function until a successful error code is -*returned. */ static int xfs_attr_node_try_addname(
@@ -1229,24 +1230,10 @@ xfs_attr_node_try_addname(/* * Its really a single leaf node, but it had * out-of-line values so it looked like it *might* -* have been a b-tree. +* have been a b-tree. Let the caller deal with this. */ xfs_da_state_free (state); - state =NULL ; - error =xfs_attr3_leaf_to_node (args); -if (error) - goto out; - -/* -* Now that we have converted the leaf to a node, we can -* roll the transaction, and try xfs_attr3_leaf_add -* again on re-entry. No need to set dela_state to do -* this. dela_state is still unset by this function at -* this point. -*/ -trace_xfs_attr_node_addname_return ( - attr->xattri_dela_state, args->dp); -return -EAGAIN ; +return -ENOSPC ; }/*
fs/xfs/xfs_attr_item.c +2 -0
@@ -314,6 +314,8 @@ xfs_xattri_finish_update(switch (op) {case XFS_ATTR_OP_FLAGS_SET : error =xfs_attr_set_iter (attr); +if (!error && attr->xattri_dela_state !=XFS_DAS_DONE ) + error = -EAGAIN ;break ;case XFS_ATTR_OP_FLAGS_REMOVE :ASSERT (XFS_IFORK_Q (args->dp));
Signature · 1 line
2.35.1
Re: [PATCH 11/16] xfs: xfs_attr_set_iter() does not need to return EAGAIN
Quoted 8 lines · Dave Chinner
From: Dave Chinner <[email protected]>
Now that the full
xfs_attr_set_iter()state machine always terminates with either the state beingXFS_DAS_DONEon success or an error on failure, we can get rid of the need for it to return-EAGAINwhenever it needs to roll the transaction before running the next state.
Quoted 8 lines
That is, we don't need to spray
-EAGAINreturn states everywhere, the caller just check the state machine state for completion to determine what action should be taken next. This greatly simplifies the code within the state machine implementation as it now only has to handle 0 for success or -errno for error and it doesn't need to tell the caller to retry.Signed-off-by Dave Chinner <[email protected]>
Looks ok
Quoted 231 lines
--- fs/xfs/libxfs/xfs_attr.c | 83 +++++++++++++++++--------------------- -- fs/xfs/xfs_attr_item.c | 2 + 2 files changed, 37 insertions(+), 48 deletions(-)fs/xfs/libxfs/xfs_attr.c +0
-1@@ -303,7 +303,6 @@ xfs_attr_sf_addname(*/ xfs_trans_bhold (args->trans, attr->xattri_leaf_bp); attr->xattri_dela_state =XFS_DAS_LEAF_ADD ; - error = -EAGAIN ; out:trace_xfs_attr_sf_addname_return (attr->xattri_dela_state, args- >dp);fs/xfs/libxfs/xfs_attr.c +0
-1return error;@@ -353,7 +352,6 @@ xfs_attr_leaf_addname(* retry the add to the newly allocated node block. */ attr->xattri_dela_state =XFS_DAS_NODE_ADD ; - error = -EAGAIN ; goto out; }if (error)@@ -364,20 +362,24 @@ xfs_attr_leaf_addname(* or perform more xattr manipulations. Otherwise there is nothing more
fs/xfs/libxfs/xfs_attr.c +2
-3* to do and we can return success. */ -if (args->rmtblkno) { +if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; - error = -EAGAIN ; - }else if (args->op_flags &XFS_DA_OP_RENAME ) { +else if (args->op_flags &XFS_DA_OP_RENAME ) xfs_attr_dela_state_set_replace(attr,
XFS_DAS_LEAF_REPLACE);fs/xfs/libxfs/xfs_attr.c +1
-3- error = -EAGAIN ; - }else { +else attr->xattri_dela_state =XFS_DAS_DONE ; - } out: trace_xfs_attr_leaf_addname_return(attr->xattri_dela_state,
args->dp);fs/xfs/libxfs/xfs_attr.c +4
-0return error; } +/* +* Add an entry to a node format attr tree. +* +* Note that we might still have a leaf here - xfs_attr_is_leaf() cannot tell
+ * the difference between leaf + remote attr blocks and a node format tree,
+ * so we may still end up having to convert from leaf to node format here.
+ */
static int
xfs_attr_node_addname(fs/xfs/libxfs/xfs_attr.c +14
-7struct xfs_attr_item *attr)@@ -392,19 +394,26 @@ xfs_attr_node_addname(return error; error =xfs_attr_node_try_addname (attr); +if (error == -ENOSPC ) { + error =xfs_attr3_leaf_to_node (args); +if (error) +return error; +/* +* No state change, we really are in node form now +* but we need the transaction rolled to continue. +*/ + goto out; + }if (error)return error; -if (args->rmtblkno) { +if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; - error = -EAGAIN ; - }else if (args->op_flags &XFS_DA_OP_RENAME ) { +else if (args->op_flags &XFS_DA_OP_RENAME )xfs_attr_dela_state_set_replace (attr,XFS_DAS_NODE_REPLACE ); - error = -EAGAIN ; - }else { +else attr->xattri_dela_state =XFS_DAS_DONE ; - } - +out:trace_xfs_attr_node_addname_return (attr->xattri_dela_state, args->dp);fs/xfs/libxfs/xfs_attr.c +0
-3return error; }@@ -426,7 +435,6 @@ xfs_attr_rmtval_alloc( error =xfs_attr_rmtval_set_blk (attr);if (error)return error; - error = -EAGAIN ; goto out; }@@ -482,11 +490,12 @@ xfs_attr_leaf_remove_attr( }/* -* Set the attribute specified in @args. -* This routine is meant to function as a delayed operation, and may return
- *-EAGAINwhen the transaction needs to be rolled. Calling functions will need
- * to handle this, and recall the function until a successful error code is
- * returned.
+ * Run the attribute operation specified in @attr.
+ *
+ * This routine is meant to function as a delayed operation and will set the
+ * state toXFS_DAS_DONEwhen the operation is complete. Calling functions will
+ * need to handle this, and recall the function until either an error or
+ *XFS_DAS_DONEis detected.
*/
int
xfs_attr_set_iter(fs/xfs/libxfs/xfs_attr.c +0
-0@@ -547,7 +556,6 @@ xfs_attr_set_iter(* We must commit the flag value change now to make it atomic
* and then we can start the next trans in series atREMOVE_OLD.fs/xfs/libxfs/xfs_attr.c +3
-2*/ - error = -EAGAIN ; attr->xattri_dela_state++;break ;@@ -575,8 +583,10 @@ xfs_attr_set_iter(case XFS_DAS_LEAF_REMOVE_RMT :case XFS_DAS_NODE_REMOVE_RMT : error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) +if (error == -EAGAIN ) { + error =0 ;break ; + }if (error)return error;@@ -588,7 +598,6 @@ xfs_attr_set_iter(* can't do that in the same transaction where we removed the
fs/xfs/libxfs/xfs_attr.c +0
-1* remote attr blocks. */ - error = -EAGAIN ; attr->xattri_dela_state++;break ;@@ -1200,14 +1209,6 @@ xfs_attr_node_addname_find_attr(* This will involve walking down the Btree, and may involve splitting
* leaf nodes and even splitting intermediate nodes up to and
including
* the root node (a special case of an intermediate node).
- *
- * "Remote" attribute values confuse the issue and atomic rename operations
- * add a whole extra layer of confusion on top of that.
- *
- * This routine is meant to function as a delayed operation, and may return
- *-EAGAINwhen the transaction needs to be rolled. Calling functions will need
- * to handle this, and recall the function until a successful error code is
- *returned.
*/
static int
xfs_attr_node_try_addname(fs/xfs/libxfs/xfs_attr.c +1
-1@@ -1229,24 +1230,10 @@ xfs_attr_node_try_addname(/* * Its really a single leaf node, but it had * out-of-line values so it looked like it *might* -* have been a b-tree. +* have been a b-tree. Let the caller deal with this.
fs/xfs/libxfs/xfs_attr.c +0
-7*/ xfs_da_state_free (state); - state =NULL ; - error =xfs_attr3_leaf_to_node (args); -if (error) - goto out; - -/* -* Now that we have converted the leaf to a node, we can
fs/xfs/libxfs/xfs_attr.c +0
-1-* roll the transaction, and try
xfs_attr3_leaf_addfs/xfs/libxfs/xfs_attr.c +0
-1-* again on re-entry. No need to set
dela_stateto dofs/xfs/libxfs/xfs_attr.c +0
-1-* this. dela_state is still unset by this function at
fs/xfs/libxfs/xfs_attr.c +0
-4-* this point. -*/ - trace_xfs_attr_node_addname_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +1
-1-return -EAGAIN ; +return -ENOSPC ; }/* fs/xfs/xfs_attr_item.c +2
-0@@ -314,6 +314,8 @@ xfs_xattri_finish_update(switch (op) {case XFS_ATTR_OP_FLAGS_SET : error =xfs_attr_set_iter (attr); +if (!error && attr->xattri_dela_state !=XFS_DAS_DONE ) + error = -EAGAIN ;break ;case XFS_ATTR_OP_FLAGS_REMOVE :ASSERT (XFS_IFORK_Q (args->dp));
PATCH 12/16 xfs: introduce attr remove initial states into xfs_attr_set_iter
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
Ground work to enable safe recovery of replace operations.
--- fs/xfs/libxfs/xfs_attr.c | 139 ++++++++++++++++++++++----------------- fs/xfs/libxfs/xfs_attr.h | 4 ++ fs/xfs/xfs_trace.h | 3 + 3 files changed, 84 insertions(+), 62 deletions(-)
fs/xfs/libxfs/xfs_attr.c +77 -62
@@ -459,6 +459,68 @@ xfs_attr_rmtval_alloc(return error; } +/* +* Mark an attribute entry INCOMPLETE and save pointers to the relevant buffers +* for later deletion of the entry. +*/ +static int +xfs_attr_leaf_mark_incomplete ( +struct xfs_da_args *args, +struct xfs_da_state *state) +{ +int error; + +/* +* Fill in disk block numbers in the state structure +* so that we can get the buffers back after we commit +* several transactions in the following calls. +*/ + error =xfs_attr_fillstate (state); +if (error) +return error; + +/* +* Mark the attribute as INCOMPLETE +*/ +return xfs_attr3_leaf_setflag (args); +} + +/* +* Initial setup for xfs_attr_node_removename. Make sure the attr is there and +* the blocks are valid. Attr keys with remote blocks will be marked +* incomplete. +*/ +static +int xfs_attr_node_removename_setup ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; +struct xfs_da_state **state = &attr->xattri_da_state; +int error; + + error =xfs_attr_node_hasname (args, state); +if (error != -EEXIST ) + goto out; + error =0 ; + +ASSERT ((*state)->path.blk[(*state)->path.active -1 ].bp !=NULL ); +ASSERT ((*state)->path.blk[(*state)->path.active -1 ].magic == +XFS_ATTR_LEAF_MAGIC ); + +if (args->rmtblkno >0 ) { + error =xfs_attr_leaf_mark_incomplete (args, *state); +if (error) + goto out; + + error =xfs_attr_rmtval_invalidate (args); + } +out: +if (error) +xfs_da_state_free (*state); + +return error; +} +/* * Remove the original attr we have just replaced. This is dependent on the * original lookup and insert placing the old attr in args->blkno/args->index
@@ -517,6 +579,21 @@ xfs_attr_set_iter(case XFS_DAS_NODE_ADD :return xfs_attr_node_addname (attr); +case XFS_DAS_SF_REMOVE : + attr->xattri_dela_state =XFS_DAS_DONE ; +return xfs_attr_sf_removename (args); +case XFS_DAS_LEAF_REMOVE : + attr->xattri_dela_state =XFS_DAS_DONE ; +return xfs_attr_leaf_removename (args); +case XFS_DAS_NODE_REMOVE : + error =xfs_attr_node_removename_setup (attr); +if (error) +return error; + attr->xattri_dela_state =XFS_DAS_NODE_REMOVE_RMT ; +if (args->rmtblkno ==0 ) + attr->xattri_dela_state++; +break ; +case XFS_DAS_LEAF_SET_RMT :case XFS_DAS_NODE_SET_RMT : error =xfs_attr_rmtval_find_space (attr);
@@ -1334,68 +1411,6 @@ xfs_attr_node_shrink(return error; } -/* -* Mark an attribute entry INCOMPLETE and save pointers to the relevant buffers -* for later deletion of the entry. -*/ -STATIC int -xfs_attr_leaf_mark_incomplete ( -struct xfs_da_args *args, -struct xfs_da_state *state) -{ -int error; - -/* -* Fill in disk block numbers in the state structure -* so that we can get the buffers back after we commit -* several transactions in the following calls. -*/ - error =xfs_attr_fillstate (state); -if (error) -return error; - -/* -* Mark the attribute as INCOMPLETE -*/ -return xfs_attr3_leaf_setflag (args); -} - -/* -* Initial setup for xfs_attr_node_removename. Make sure the attr is there and -* the blocks are valid. Attr keys with remote blocks will be marked -* incomplete. -*/ -STATIC -int xfs_attr_node_removename_setup( -struct xfs_attr_item *attr) -{ -struct xfs_da_args *args = attr->xattri_da_args; -struct xfs_da_state **state = &attr->xattri_da_state; -int error; - - error =xfs_attr_node_hasname (args, state); -if (error != -EEXIST ) - goto out; - error =0 ; - -ASSERT ((*state)->path.blk[(*state)->path.active -1 ].bp !=NULL ); -ASSERT ((*state)->path.blk[(*state)->path.active -1 ].magic == -XFS_ATTR_LEAF_MAGIC ); - -if (args->rmtblkno >0 ) { - error =xfs_attr_leaf_mark_incomplete (args, *state); -if (error) - goto out; - - error =xfs_attr_rmtval_invalidate (args); - } -out: -if (error) -xfs_da_state_free (*state); - -return error; -} - STATIC int xfs_attr_node_removename( struct xfs_da_args *args,
fs/xfs/libxfs/xfs_attr.h +4 -0
@@ -451,6 +451,10 @@ enum xfs_delattr_state {XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ +XFS_DAS_SF_REMOVE ,/* Initial shortform set iter state */ +XFS_DAS_LEAF_REMOVE ,/* Initial leaf form set iter state */ +XFS_DAS_NODE_REMOVE ,/* Initial node form set iter state */ +/* Leaf state set/replace sequence */ XFS_DAS_LEAF_SET_RMT ,/* set a remote xattr from a leaf */ XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remote blocks */
fs/xfs/xfs_trace.h +3 -0
@@ -4105,6 +4105,9 @@ TRACE_DEFINE_ENUM(XFS_DAS_NODE_ADD);TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK );TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME );TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK ); +TRACE_DEFINE_ENUM (XFS_DAS_SF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );
Signature · 1 line
2.35.1
Re: [PATCH 12/16] xfs: introduce attr remove initial states into xfs_attr_set_iter
Quoted 101 lines · Dave Chinner
From: Dave Chinner <[email protected]>
Ground work to enable safe recovery of replace operations.
Signed-off-by Dave Chinner <[email protected]>--- fs/xfs/libxfs/xfs_attr.c | 139 ++++++++++++++++++++++--------------- -- fs/xfs/libxfs/xfs_attr.h | 4 ++ fs/xfs/xfs_trace.h | 3 + 3 files changed, 84 insertions(+), 62 deletions(-)fs/xfs/libxfs/xfs_attr.c +2
-0@@ -459,6 +459,68 @@ xfs_attr_rmtval_alloc(return error; } +/* +* Mark an attribute entry INCOMPLETE and save pointers to the relevant buffers
+ * for later deletion of the entry.
+ */fs/xfs/libxfs/xfs_attr.c +24
-0+static int +xfs_attr_leaf_mark_incomplete ( +struct xfs_da_args *args, +struct xfs_da_state *state) +{ +int error; + +/* +* Fill in disk block numbers in the state structure +* so that we can get the buffers back after we commit +* several transactions in the following calls. +*/ + error =xfs_attr_fillstate (state); +if (error) +return error; + +/* +* Mark the attribute as INCOMPLETE +*/ +return xfs_attr3_leaf_setflag (args); +} + +/* +* Initial setup for xfs_attr_node_removename. Make sure the attr is there and
+ * the blocks are valid. Attr keys with remote blocks will be marked
+ * incomplete.
+ */fs/xfs/libxfs/xfs_attr.c +6
-0+static +int xfs_attr_node_removename_setup ( +struct xfs_attr_item *attr) +{ +struct xfs_da_args *args = attr->xattri_da_args; + struct xfs_da_state **state = &attr-Quoted 1 line
xattri_da_state;fs/xfs/libxfs/xfs_attr.c +8
-0+int error; + + error =xfs_attr_node_hasname (args, state); +if (error != -EEXIST ) + goto out; + error =0 ; + + ASSERT((*state)->path.blk[(*state)->path.active - 1].bp !=NULL);
fs/xfs/libxfs/xfs_attr.c +17
-0+ASSERT ((*state)->path.blk[(*state)->path.active -1 ].magic == +XFS_ATTR_LEAF_MAGIC ); + +if (args->rmtblkno >0 ) { + error =xfs_attr_leaf_mark_incomplete (args, *state); +if (error) + goto out; + + error =xfs_attr_rmtval_invalidate (args); + } +out: +if (error) +xfs_da_state_free (*state); + +return error; +} +/* * Remove the original attr we have just replaced. This is dependent on the
* original lookup and insert placing the old attr in args-Quoted 1 line
blkno/args->index
fs/xfs/libxfs/xfs_attr.c +5
-0@@ -517,6 +579,21 @@ xfs_attr_set_iter(case XFS_DAS_NODE_ADD :return xfs_attr_node_addname (attr); +case XFS_DAS_SF_REMOVE : + attr->xattri_dela_state =XFS_DAS_DONE ; +return xfs_attr_sf_removename (args); +case XFS_DAS_LEAF_REMOVE : + attr->xattri_dela_state =XFS_DAS_DONE ;
fs/xfs/libxfs/xfs_attr.c xfs_attr_set_iter( · line 596
Quoted 10 lines
fs/xfs/libxfs/xfs_attr.c +10
-0+return xfs_attr_leaf_removename (args); +case XFS_DAS_NODE_REMOVE : + error =xfs_attr_node_removename_setup (attr); +if (error) +return error; + attr->xattri_dela_state =XFS_DAS_NODE_REMOVE_RMT ; +if (args->rmtblkno ==0 ) + attr->xattri_dela_state++; +break ; +
Ok... it took me a little bit to understand why this was here, but it
makes sense after having skipped ahead to the next patch. We're combining the add/remove code paths to better manage rename for journal replays. Probably just add a blurb at the top?
"We need to merge the add and remove code paths to enable safe recovery of replace operations. Hoist the initial remove states from xfs_attr_remove_iter into xfs_attr_set_iter. We will make use of them in the next patches."
I think the rest looks ok though.
Quoted 112 lines
fs/xfs/libxfs/xfs_attr.c +0
-2case XFS_DAS_LEAF_SET_RMT :case XFS_DAS_NODE_SET_RMT : error =xfs_attr_rmtval_find_space (attr);@@ -1334,68 +1411,6 @@ xfs_attr_node_shrink(return error; } -/* -* Mark an attribute entry INCOMPLETE and save pointers to the relevant buffers
- * for later deletion of the entry.
- */fs/xfs/libxfs/xfs_attr.c +0
-24-STATIC int -xfs_attr_leaf_mark_incomplete ( -struct xfs_da_args *args, -struct xfs_da_state *state) -{ -int error; - -/* -* Fill in disk block numbers in the state structure -* so that we can get the buffers back after we commit -* several transactions in the following calls. -*/ - error =xfs_attr_fillstate (state); -if (error) -return error; - -/* -* Mark the attribute as INCOMPLETE -*/ -return xfs_attr3_leaf_setflag (args); -} - -/* -* Initial setup for xfs_attr_node_removename. Make sure the attr is there and
- * the blocks are valid. Attr keys with remote blocks will be marked
- * incomplete.
- */fs/xfs/libxfs/xfs_attr.c +0
-6-STATIC -int xfs_attr_node_removename_setup( -struct xfs_attr_item *attr) -{ -struct xfs_da_args *args = attr->xattri_da_args; - struct xfs_da_state **state = &attr-Quoted 1 line
xattri_da_state;fs/xfs/libxfs/xfs_attr.c +0
-8-int error; - - error =xfs_attr_node_hasname (args, state); -if (error != -EEXIST ) - goto out; - error =0 ; - - ASSERT((*state)->path.blk[(*state)->path.active - 1].bp !=NULL);
fs/xfs/libxfs/xfs_attr.c +0
-17-ASSERT ((*state)->path.blk[(*state)->path.active -1 ].magic == -XFS_ATTR_LEAF_MAGIC ); - -if (args->rmtblkno >0 ) { - error =xfs_attr_leaf_mark_incomplete (args, *state); -if (error) - goto out; - - error =xfs_attr_rmtval_invalidate (args); - } -out: -if (error) -xfs_da_state_free (*state); - -return error; -} - STATIC int xfs_attr_node_removename( struct xfs_da_args *args,fs/xfs/libxfs/xfs_attr.h +1
-0@@ -451,6 +451,10 @@ enum xfs_delattr_state {XFS_DAS_RM_NAME ,/* Remove attr name */ XFS_DAS_RM_SHRINK ,/* We are shrinking the tree */ + XFS_DAS_SF_REMOVE, /* Initial shortform set iterstate */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_LEAF_REMOVE, /* Initial leaf form set iterstate */
fs/xfs/libxfs/xfs_attr.h +1
-0+ XFS_DAS_NODE_REMOVE, /* Initial node form set iterstate */
fs/xfs/libxfs/xfs_attr.h +1
-0+/* Leaf state set/replace sequence */ XFS_DAS_LEAF_SET_RMT, /* set a remote xattr from aleaf */
XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remote
blocks */fs/xfs/xfs_trace.h +3
-0@@ -4105,6 +4105,9 @@ TRACE_DEFINE_ENUM(XFS_DAS_NODE_ADD);TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK );TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME );TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK ); +TRACE_DEFINE_ENUM (XFS_DAS_SF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REPLACE );
PATCH 13/16 xfs: switch attr remove to xfs_attri_set_iter
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
Now that xfs_attri_set_iter() has initial states for removing attributes, switch the pure attribute removal code over to using it. This requires attrs being removed to always be marked as INCOMPLETE before we start the removal due to the fact we look up the attr to remove again in xfs_attr_node_remove_attr().
Note: this drops the fillstate/refillstate optimisations from the remove path that avoid having to look up the path again after setting the incomplete flag and removeing remote attrs. Restoring that optimisation to this path is future Dave's problem.
--- fs/xfs/libxfs/xfs_attr.c | 26 +++++++++++++++----------- fs/xfs/xfs_attr_item.c | 27 ++++++--------------------- 2 files changed, 21 insertions(+), 32 deletions(-)
fs/xfs/libxfs/xfs_attr.c +15 -11
@@ -507,13 +507,11 @@ int xfs_attr_node_removename_setup(ASSERT ((*state)->path.blk[(*state)->path.active -1 ].magic ==XFS_ATTR_LEAF_MAGIC ); -if (args->rmtblkno >0 ) { - error =xfs_attr_leaf_mark_incomplete (args, *state); -if (error) - goto out; - + error =xfs_attr_leaf_mark_incomplete (args, *state); +if (error) + goto out; +if (args->rmtblkno >0 ) error =xfs_attr_rmtval_invalidate (args); - } out:if (error)xfs_da_state_free (*state);
@@ -954,6 +952,13 @@ xfs_attr_remove_deferred(if (error)return error; +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_REMOVE ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; +else + new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; +xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list);return 0 ;
@@ -1342,16 +1347,15 @@ xfs_attr_node_remove_attr( {struct xfs_da_args *args = attr->xattri_da_args;struct xfs_da_state *state =NULL ; -struct xfs_mount *mp = args->dp->i_mount;int retval =0 ;int error =0 ;/* -* Re-find the "old" attribute entry after any split ops. The INCOMPLETE -* flag means that we will find the "old" attr, not the "new" one. +* The attr we are removing has already been marked incomplete, so +* we need to set the filter appropriately to re-find the "old" +* attribute entry after any split ops. */ -if (!xfs_has_larp (mp)) - args->attr_filter |=XFS_ATTR_INCOMPLETE ; + args->attr_filter |=XFS_ATTR_INCOMPLETE ; state =xfs_da_state_alloc (args); state->inleaf =0 ; error =xfs_da3_node_lookup_int (state, &retval);
fs/xfs/xfs_attr_item.c +6 -21
@@ -298,12 +298,9 @@ xfs_attrd_item_release( STATIC intxfs_xattri_finish_update (struct xfs_attr_item *attr, -struct xfs_attrd_log_item *attrdp, -uint32_t op_flags) +struct xfs_attrd_log_item *attrdp) {struct xfs_da_args *args = attr->xattri_da_args; -unsigned int op = op_flags & -XFS_ATTR_OP_FLAGS_TYPE_MASK ;int error;if (XFS_TEST_ERROR (false, args->dp->i_mount,XFS_ERRTAG_LARP )) {
@@ -311,20 +308,9 @@ xfs_xattri_finish_update( goto out; } -switch (op) { -case XFS_ATTR_OP_FLAGS_SET : - error =xfs_attr_set_iter (attr); -if (!error && attr->xattri_dela_state !=XFS_DAS_DONE ) - error = -EAGAIN ; -break ; -case XFS_ATTR_OP_FLAGS_REMOVE : -ASSERT (XFS_IFORK_Q (args->dp)); - error =xfs_attr_remove_iter (attr); -break ; -default : - error = -EFSCORRUPTED ; -break ; - } + error =xfs_attr_set_iter (attr); +if (!error && attr->xattri_dela_state !=XFS_DAS_DONE ) + error = -EAGAIN ; out:/*
@@ -435,8 +421,7 @@ xfs_attr_finish_item(*/ attr->xattri_da_args->trans = tp; - error =xfs_xattri_finish_update (attr, done_item, - attr->xattri_op_flags); + error =xfs_xattri_finish_update (attr, done_item);if (error != -EAGAIN )kmem_free (attr);
@@ -586,7 +571,7 @@ xfs_attri_item_recover(xfs_ilock (ip,XFS_ILOCK_EXCL );xfs_trans_ijoin (tp, ip,0 ); - ret =xfs_xattri_finish_update (attr, done_item, attrp->alfi_op_flags); + ret =xfs_xattri_finish_update (attr, done_item);if (ret == -EAGAIN ) {/* There's more work to do, so add it to this transaction */ xfs_defer_add (tp,XFS_DEFER_OPS_TYPE_ATTR , &attr->xattri_list);
Signature · 1 line
2.35.1
Re: [PATCH 13/16] xfs: switch attr remove to xfs_attri_set_iter
Quoted 42 lines · Dave Chinner
From: Dave Chinner <[email protected]>
Now that
xfs_attri_set_iter()has initial states for removing attributes, switch the pure attribute removal code over to using it. This requires attrs being removed to always be marked as INCOMPLETE before we start the removal due to the fact we look up the attr to remove again inxfs_attr_node_remove_attr().Note: this drops the fillstate/refillstate optimisations from the remove path that avoid having to look up the path again after setting the incomplete flag and removeing remote attrs. Restoring that optimisation to this path is future Dave's problem.
Signed-off-by Dave Chinner <[email protected]>--- fs/xfs/libxfs/xfs_attr.c | 26 +++++++++++++++----------- fs/xfs/xfs_attr_item.c | 27 ++++++--------------------- 2 files changed, 21 insertions(+), 32 deletions(-)fs/xfs/libxfs/xfs_attr.c +4
-6@@ -507,13 +507,11 @@ int xfs_attr_node_removename_setup(ASSERT ((*state)->path.blk[(*state)->path.active -1 ].magic ==XFS_ATTR_LEAF_MAGIC ); -if (args->rmtblkno >0 ) { - error =xfs_attr_leaf_mark_incomplete (args, *state); -if (error) - goto out; - + error =xfs_attr_leaf_mark_incomplete (args, *state); +if (error) + goto out; +if (args->rmtblkno >0 ) error =xfs_attr_rmtval_invalidate (args); - } out:if (error)xfs_da_state_free (*state);
fs/xfs/libxfs/xfs_attr.c xfs_attr_remove_deferred( · line 961
Quoted 10 lines
fs/xfs/libxfs/xfs_attr.c +7
-0@@ -954,6 +952,13 @@ xfs_attr_remove_deferred(if (error)return error; +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_REMOVE ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; +else + new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; +
Mmmm, same issue here as in patch 4, this initial state configs would get missed during a replay since these routines are only used in the delayed attr code path, not the replay code path.
Quoted 60 lines
fs/xfs/libxfs/xfs_attr.c +0
-0xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new- >xattri_list);fs/xfs/libxfs/xfs_attr.c +0
-2return 0 ;@@ -1342,16 +1347,15 @@ xfs_attr_node_remove_attr( {struct xfs_da_args *args = attr->xattri_da_args;struct xfs_da_state *state =NULL ; -struct xfs_mount *mp = args->dp->i_mount;int retval =0 ;int error =0 ;/* -* Re-find the "old" attribute entry after any split ops. The INCOMPLETE
fs/xfs/libxfs/xfs_attr.c +0
-1-* flag means that we will find the "old" attr, not the "new" one.
fs/xfs/libxfs/xfs_attr.c +1
-0+* The attr we are removing has already been marked incomplete, so
fs/xfs/libxfs/xfs_attr.c +3
-2+* we need to set the filter appropriately to re-find the "old" +* attribute entry after any split ops. */ -if (!xfs_has_larp (mp)) - args->attr_filter |=XFS_ATTR_INCOMPLETE ; + args->attr_filter |=XFS_ATTR_INCOMPLETE ; state =xfs_da_state_alloc (args); state->inleaf =0 ; error =xfs_da3_node_lookup_int (state, &retval);fs/xfs/xfs_attr_item.c +1
-4@@ -298,12 +298,9 @@ xfs_attrd_item_release( STATIC intxfs_xattri_finish_update (struct xfs_attr_item *attr, -struct xfs_attrd_log_item *attrdp, -uint32_t op_flags) +struct xfs_attrd_log_item *attrdp) {struct xfs_da_args *args = attr->xattri_da_args; -unsigned int op = op_flags & -XFS_ATTR_OP_FLAGS_TYPE_MAS K ;int error;if (
XFS_TEST_ERROR(false, args->dp->i_mount, XFS_ERRTAG_LARP))
{fs/xfs/xfs_attr_item.c +0
-8@@ -311,20 +308,9 @@ xfs_xattri_finish_update( goto out; } -switch (op) { -case XFS_ATTR_OP_FLAGS_SET : - error =xfs_attr_set_iter (attr); -if (!error && attr->xattri_dela_state !=XFS_DAS_DONE ) - error = -EAGAIN ; -break ; -case XFS_ATTR_OP_FLAGS_REMOVE : -ASSERT (XFS_IFORK_Q (args->dp));
fs/xfs/xfs_attr_item.c xfs_xattri_finish_update( · line 314
Quoted 10 lines
fs/xfs/xfs_attr_item.c +3
-6- error =xfs_attr_remove_iter (attr); -break ; -default : - error = -EFSCORRUPTED ; -break ; - } + error =xfs_attr_set_iter (attr); +if (!error && attr->xattri_dela_state !=XFS_DAS_DONE ) + error = -EAGAIN ;
The concern I have here is that op_flags is recorded and recovered from the log item (see xfs_attri_item_recover). Where as xattri_dela_state is not. What ever value was set in attr before the shut down would not be there upon recovery, and with out op_flag we wont know how to configure it.
I think maybe what you meant to do is log the state? We need to get into xfs_attri_log_format and turn alfi_op_flags into
a alfi_dela_state. I think that would sew together what you are trying to do here?
Allison
Quoted 24 lines
fs/xfs/xfs_attr_item.c +2
-3out:/* @@ -435,8 +421,7 @@ xfs_attr_finish_item(*/ attr->xattri_da_args->trans = tp; - error =xfs_xattri_finish_update (attr, done_item, - attr->xattri_op_flags); + error =xfs_xattri_finish_update (attr, done_item);if (error != -EAGAIN )kmem_free (attr);@@ -586,7 +571,7 @@ xfs_attri_item_recover(xfs_ilock (ip,XFS_ILOCK_EXCL );xfs_trans_ijoin (tp, ip,0 ); - ret =xfs_xattri_finish_update (attr, done_item, attrp- >alfi_op_flags); + ret =xfs_xattri_finish_update (attr, done_item);if (ret == -EAGAIN ) {/* There's more work to do, so add it to this transaction */
xfs_defer_add(tp,XFS_DEFER_OPS_TYPE_ATTR, &attr-Quoted 1 line
xattri_list);
Re: [PATCH 13/16] xfs: switch attr remove to xfs_attri_set_iter
Quoted 46 lines · Alli
Quoted 45 lines · Dave Chinner
From: Dave Chinner <[email protected]>
Now that
xfs_attri_set_iter()has initial states for removing attributes, switch the pure attribute removal code over to using it. This requires attrs being removed to always be marked as INCOMPLETE before we start the removal due to the fact we look up the attr to remove again inxfs_attr_node_remove_attr().Note: this drops the fillstate/refillstate optimisations from the remove path that avoid having to look up the path again after setting the incomplete flag and removeing remote attrs. Restoring that optimisation to this path is future Dave's problem.
Signed-off-by Dave Chinner <[email protected]>--- fs/xfs/libxfs/xfs_attr.c | 26 +++++++++++++++----------- fs/xfs/xfs_attr_item.c | 27 ++++++--------------------- 2 files changed, 21 insertions(+), 32 deletions(-)fs/xfs/libxfs/xfs_attr.c +4
-6@@ -507,13 +507,11 @@ int xfs_attr_node_removename_setup(ASSERT ((*state)->path.blk[(*state)->path.active -1 ].magic ==XFS_ATTR_LEAF_MAGIC ); -if (args->rmtblkno >0 ) { - error =xfs_attr_leaf_mark_incomplete (args, *state); -if (error) - goto out; - + error =xfs_attr_leaf_mark_incomplete (args, *state); +if (error) + goto out; +if (args->rmtblkno >0 ) error =xfs_attr_rmtval_invalidate (args); - } out:if (error)xfs_da_state_free (*state);@@ -954,6 +952,13 @@ xfs_attr_remove_deferred(if (error)return error;
Quoted 10 lines
fs/xfs/libxfs/xfs_attr.c xfs_attr_remove_deferred( · line 961
Quoted 7 lines
fs/xfs/libxfs/xfs_attr.c +7
-0+if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_REMOVE ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; +else + new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; +Mmmm, same issue here as in patch 4, this initial state configs would get missed during a replay since these routines are only used in the delayed attr code path, not the replay code path.
*nod*
I'm working on fixing this right now.
Quoted 20 lines
Quoted 20 lines
fs/xfs/libxfs/xfs_attr.c +2
-14@@ -311,20 +308,9 @@ xfs_xattri_finish_update( goto out; } -switch (op) { -case XFS_ATTR_OP_FLAGS_SET : - error =xfs_attr_set_iter (attr); -if (!error && attr->xattri_dela_state !=XFS_DAS_DONE ) - error = -EAGAIN ; -break ; -case XFS_ATTR_OP_FLAGS_REMOVE : -ASSERT (XFS_IFORK_Q (args->dp)); - error =xfs_attr_remove_iter (attr); -break ; -default : - error = -EFSCORRUPTED ; -break ; - } + error =xfs_attr_set_iter (attr); +if (!error && attr->xattri_dela_state !=XFS_DAS_DONE )
Quoted 8 lines
fs/xfs/libxfs/xfs_attr.c xfs_xattri_finish_update( · line 314
Quoted 2 lines
fs/xfs/libxfs/xfs_attr.c +1
-0+ error = -EAGAIN ;The concern I have here is that
op_flagsis recorded and recovered from the log item (seexfs_attri_item_recover). Where asxattri_dela_stateis not. What ever value was set in attr before the shut down would not be there upon recovery, and with outop_flagwe wont know how to configure it.
Right, that's the real problem with the existing operational order and intent contents - what is on disk and/or in the journal is not consistent and is dependent on the state at which the operation failed. hence to recover from this, we'd need to push the current state into the intents as well.
That's pretty messy and nasty, and I'm trying to avoid that. This is the reason why I've reworking the way the logged attribute is logged and recovered in this series.
When we are doing a pure attr create, we have to consider:
- shortform create is a single transaction operation and so never leaves anything behind for recovery to clean up.
- leaf and node insertion of inline attributes are single transaction operations and never leave anything for recovery to clean up
- short-form to leaf and leaf to node operations roll the transaction without changing the attr contents, so if we crash after those conversions are committed, recovery only needs to create the new attr. IOWs, nothing to clean up before we run the create replay.
- creating a remote attr sets the INCOMPLETE flag on the new attr in when the name is inserted into the btree, and it is removed when the remote attr creation is complete. Hence there's a transient state in the journal where the attr is present but INCOMPLETE.
This last state is the problem - if recovery does not remove this INCOMPLETE xattr, or it does not restart the recovery from the exact point it failed, we will leave stale INCOMPLETE xattrs in the btree whenever we recover from a crash. That leaves us with two choices; either we:
- put a whole lot more state into the intent to enable exact continuation of the operation (logging state and remote attr extent target); or
- we just remove the INCOMPLETE xattr and run a normal create operation to recreate the entire xattr.
When we are doing a replace, we have similar state based problems - did we crash during create or removal? IF we are doing the create first, then we can crash with both a new incomplete and an old valid xattr of the given name. ANd after we flip the flags over, we have a new valid and an old incomplete old xattr, and like the create state we can't tell which is which.
Now what do we do in recovery - which one is the one we have to remove? Or do we have to remove both? We ahve the same problem as a pure create - we don't know what to do without knowing the exact state of the operation.
However, if we change the replace operation to do things in a different order because we have an intent that stores the new attr {name, value} pair, we can avoid this whole problem. THat isL
- log the new attr intent atomically with marking the existing attr INCOMPLETE.
- remove the old INCOMPLETE attr
- insert the new attr as per the pure create above
If we crash at any point during this operation, we will only ever see a single INCOMPLETE entry under the name of the new attr, regardless of whether we failed during the remove of the old xattr of the create of the new xattr. And because of the intents, we'll always have the new xattr name+val at recovery time.
Pure remove has the same problem - partial remove still needs recovery to clean up, but we don't want partially removed xattrs to ever be visible to userspace. Hence the logged remove operation is:
- log the new attr intent atomically with marking the existing attr INCOMPLETE.
- remove the old INCOMPLETE attr
And now the recovery operation is simply "remove the INCOMPLETE xattr under the logged name".
IOWs, we now have the same recovery path for all three logged xattr operations:
- remove the INCOMPLETE xattr under the logged name if it exists to return the attr fork to a known good state.
- for set/replace, create the new xattr that was logged in the intent.
To return to the original question, this means all recovery needs to know is whether we are recovering a SET/REPLACE or a REMOVE operation along with the logged xattr name/val pair from the intent. We don't need to know what state we crashed in, nor do we need to know partial setup/teardown target/state in the journal because all we ever need to do to get back to a known good state is teardown the existing INCOMPLETE xattr....
Cheers,
Dave.
Signature · 2 lines
Dave Chinner [email protected]
PATCH 14/16 xfs: remove xfs_attri_remove_iter
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
xfs_attri_remove_iter is not used anymore, so remove it and all the infrastructure it uses and is needed to drive it.
--- fs/xfs/libxfs/xfs_attr.c | 210 ++++----------------------------------- 1 file changed, 18 insertions(+), 192 deletions(-)
fs/xfs/libxfs/xfs_attr.c +18 -192
@@ -1340,6 +1340,24 @@ xfs_attr_node_try_addname(return error; } +static int +xfs_attr_node_removename ( +struct xfs_da_args *args, +struct xfs_da_state *state) +{ +struct xfs_da_state_blk *blk; +int retval; + +/* +* Remove the name and update the hashvals in the tree. +*/ + blk = &state->path.blk[state->path.active-1 ]; +ASSERT (blk->magic ==XFS_ATTR_LEAF_MAGIC ); + retval =xfs_attr3_leaf_remove (blk->bp, args); +xfs_da3_fixhashpath (state, &state->path); + +return retval; +}static int xfs_attr_node_remove_attr(
@@ -1382,198 +1400,6 @@ xfs_attr_node_remove_attr(return retval; } -/* -* Shrink an attribute from leaf to shortform -*/ -STATIC int -xfs_attr_node_shrink ( -struct xfs_da_args *args, -struct xfs_da_state *state) -{ -struct xfs_inode *dp = args->dp; -int error, forkoff; -struct xfs_buf *bp; - -/* -* Have to get rid of the copy of this dabuf in the state. -*/ -ASSERT (state->path.active ==1 ); -ASSERT (state->path.blk[0 ].bp); - state->path.blk[0 ].bp =NULL ; - - error =xfs_attr3_leaf_read (args->trans, args->dp,0 , &bp); -if (error) -return error; - - forkoff =xfs_attr_shortform_allfit (bp, dp); -if (forkoff) { - error =xfs_attr3_leaf_to_shortform (bp, args, forkoff); -/* bp is gone due to xfs_da_shrink_inode */ - }else -xfs_trans_brelse (args->trans, bp); - -return error; -} - -STATIC int -xfs_attr_node_removename ( -struct xfs_da_args *args, -struct xfs_da_state *state) -{ -struct xfs_da_state_blk *blk; -int retval; - -/* -* Remove the name and update the hashvals in the tree. -*/ - blk = &state->path.blk[state->path.active-1 ]; -ASSERT (blk->magic ==XFS_ATTR_LEAF_MAGIC ); - retval =xfs_attr3_leaf_remove (blk->bp, args); -xfs_da3_fixhashpath (state, &state->path); - -return retval; -} - -/* -* Remove the attribute specified in @args. -* -* This will involve walking down the Btree, and may involve joining -* leaf nodes and even joining intermediate nodes up to and including -* the root node (a special case of an intermediate node). -* -* This routine is meant to function as either an in-line or delayed operation, -* and may return -EAGAIN when the transaction needs to be rolled. Calling -* functions will need to handle this, and call the function until a -* successful error code is returned. -*/ -int -xfs_attr_remove_iter ( -struct xfs_attr_item *attr) -{ -struct xfs_da_args *args = attr->xattri_da_args; -struct xfs_da_state *state = attr->xattri_da_state; -int retval, error =0 ; -struct xfs_inode *dp = args->dp; - -trace_xfs_attr_node_removename (args); - -switch (attr->xattri_dela_state) { -case XFS_DAS_UNINIT : -if (!xfs_inode_hasattr (dp)) -return -ENOATTR ; - -/* -* Shortform or leaf formats don't require transaction rolls and -* thus state transitions. Call the right helper and return. -*/ -if (dp->i_afp->if_format ==XFS_DINODE_FMT_LOCAL ) -return xfs_attr_sf_removename (args); - -if (xfs_attr_is_leaf (dp)) -return xfs_attr_leaf_removename (args); - -/* -* Node format may require transaction rolls. Set up the -* state context and fall into the state machine. -*/ -if (!attr->xattri_da_state) { - error =xfs_attr_node_removename_setup (attr); -if (error) -return error; - state = attr->xattri_da_state; - } - - fallthrough; -case XFS_DAS_RMTBLK : - attr->xattri_dela_state =XFS_DAS_RMTBLK ; - -/* -* If there is an out-of-line value, de-allocate the blocks. -* This is done before we remove the attribute so that we don't -* overflow the maximum size of a transaction and/or hit a -* deadlock. -*/ -if (args->rmtblkno >0 ) { -/* -* May return -EAGAIN. Roll and repeat until all remote -* blocks are removed. -*/ - error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) { -trace_xfs_attr_remove_iter_return ( - attr->xattri_dela_state, args->dp); -return error; - }else if (error) { - goto out; - } - -/* -* Refill the state structure with buffers (the prior -* calls released our buffers) and close out this -* transaction before proceeding. -*/ -ASSERT (args->rmtblkno ==0 ); - error =xfs_attr_refillstate (state); -if (error) - goto out; - - attr->xattri_dela_state =XFS_DAS_RM_NAME ; -trace_xfs_attr_remove_iter_return ( - attr->xattri_dela_state, args->dp); -return -EAGAIN ; - } - - fallthrough; -case XFS_DAS_RM_NAME : -/* -* If we came here fresh from a transaction roll, reattach all -* the buffers to the current transaction. -*/ -if (attr->xattri_dela_state ==XFS_DAS_RM_NAME ) { - error =xfs_attr_refillstate (state); -if (error) - goto out; - } - - retval =xfs_attr_node_removename (args, state); - -/* -* Check to see if the tree needs to be collapsed. If so, roll -* the transacton and fall into the shrink state. -*/ -if (retval && (state->path.active >1 )) { - error =xfs_da3_join (state); -if (error) - goto out; - - attr->xattri_dela_state =XFS_DAS_RM_SHRINK ; -trace_xfs_attr_remove_iter_return ( - attr->xattri_dela_state, args->dp); -return -EAGAIN ; - } - - fallthrough; -case XFS_DAS_RM_SHRINK : -/* -* If the result is small enough, push it all into the inode. -* This is our final state so it's safe to return a dirty -* transaction. -*/ -if (xfs_attr_is_leaf (dp)) - error =xfs_attr_node_shrink (args, state); -ASSERT (error != -EAGAIN ); -break ; -default : -ASSERT (0 ); - error = -EINVAL ; - goto out; - } -out: -if (state) -xfs_da_state_free (state); -return error; -} -/* * Fill in the disk block numbers in the state structure for the buffers * that are attached to the state structure.
Signature · 1 line
2.35.1
Re: [PATCH 14/16] xfs: remove xfs_attri_remove_iter
Quoted 186 lines · Dave Chinner
From: Dave Chinner <[email protected]>
xfs_attri_remove_iteris not used anymore, so remove it and all the infrastructure it uses and is needed to drive it.Signed-off-by Dave Chinner <[email protected]>--- fs/xfs/libxfs/xfs_attr.c | 210 ++++--------------------------------- -- 1 file changed, 18 insertions(+), 192 deletions(-)fs/xfs/libxfs/xfs_attr.c +18
-57@@ -1340,6 +1340,24 @@ xfs_attr_node_try_addname(return error; } +static int +xfs_attr_node_removename ( +struct xfs_da_args *args, +struct xfs_da_state *state) +{ +struct xfs_da_state_blk *blk; +int retval; + +/* +* Remove the name and update the hashvals in the tree. +*/ + blk = &state->path.blk[state->path.active-1 ]; +ASSERT (blk->magic ==XFS_ATTR_LEAF_MAGIC ); + retval =xfs_attr3_leaf_remove (blk->bp, args); +xfs_da3_fixhashpath (state, &state->path); + +return retval; +}static int xfs_attr_node_remove_attr(@@ -1382,198 +1400,6 @@ xfs_attr_node_remove_attr(return retval; } -/* -* Shrink an attribute from leaf to shortform -*/ -STATIC int -xfs_attr_node_shrink ( -struct xfs_da_args *args, -struct xfs_da_state *state) -{ -struct xfs_inode *dp = args->dp; -int error, forkoff; -struct xfs_buf *bp; - -/* -* Have to get rid of the copy of this dabuf in the state. -*/ -ASSERT (state->path.active ==1 ); -ASSERT (state->path.blk[0 ].bp); - state->path.blk[0 ].bp =NULL ; - - error =xfs_attr3_leaf_read (args->trans, args->dp,0 , &bp); -if (error) -return error; - - forkoff =xfs_attr_shortform_allfit (bp, dp); -if (forkoff) { - error =xfs_attr3_leaf_to_shortform (bp, args, forkoff); -/* bp is gone due to xfs_da_shrink_inode */ - }else -xfs_trans_brelse (args->trans, bp); - -return error; -} - -STATIC int -xfs_attr_node_removename ( -struct xfs_da_args *args, -struct xfs_da_state *state) -{ -struct xfs_da_state_blk *blk; -int retval; - -/* -* Remove the name and update the hashvals in the tree. -*/ - blk = &state->path.blk[state->path.active-1 ]; -ASSERT (blk->magic ==XFS_ATTR_LEAF_MAGIC ); - retval =xfs_attr3_leaf_remove (blk->bp, args); -xfs_da3_fixhashpath (state, &state->path); - -return retval; -} - -/* -* Remove the attribute specified in @args. -* -* This will involve walking down the Btree, and may involve joining -* leaf nodes and even joining intermediate nodes up to and including
- * the root node (a special case of an intermediate node).
- *
- * This routine is meant to function as either an in-line or delayed operation,
- * and may return-EAGAINwhen the transaction needs to be rolled. Calling
- * functions will need to handle this, and call the function until a
- * successful error code is returned.
- */fs/xfs/libxfs/xfs_attr.c +0
-6-int -xfs_attr_remove_iter ( -struct xfs_attr_item *attr) -{ -struct xfs_da_args *args = attr->xattri_da_args; - struct xfs_da_state *state = attr-Quoted 1 line
xattri_da_state;fs/xfs/libxfs/xfs_attr.c +0
-12-int retval, error =0 ; -struct xfs_inode *dp = args->dp; - -trace_xfs_attr_node_removename (args); - -switch (attr->xattri_dela_state) { -case XFS_DAS_UNINIT : -if (!xfs_inode_hasattr (dp)) -return -ENOATTR ; - -/* -* Shortform or leaf formats don't require transaction rolls and
fs/xfs/libxfs/xfs_attr.c +0
-1-* thus state transitions. Call the right helper and return.
fs/xfs/libxfs/xfs_attr.c +0
-9-*/ -if (dp->i_afp->if_format ==XFS_DINODE_FMT_LOCAL ) -return xfs_attr_sf_removename (args); - -if (xfs_attr_is_leaf (dp)) -return xfs_attr_leaf_removename (args); - -/* -* Node format may require transaction rolls. Set up the
fs/xfs/libxfs/xfs_attr.c +0
-15-* state context and fall into the state machine. -*/ -if (!attr->xattri_da_state) { - error =xfs_attr_node_removename_setup (attr); -if (error) -return error; - state = attr->xattri_da_state; - } - - fallthrough; -case XFS_DAS_RMTBLK : - attr->xattri_dela_state =XFS_DAS_RMTBLK ; - -/* -* If there is an out-of-line value, de-allocate the blocks.
fs/xfs/libxfs/xfs_attr.c +0
-1-* This is done before we remove the attribute so that we don't
fs/xfs/libxfs/xfs_attr.c +0
-1-* overflow the maximum size of a transaction and/or hit a
fs/xfs/libxfs/xfs_attr.c +0
-5-* deadlock. -*/ -if (args->rmtblkno >0 ) { -/* -* May return -EAGAIN. Roll and repeat until all remote
fs/xfs/libxfs/xfs_attr.c +0
-6-* blocks are removed. -*/ - error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) { - trace_xfs_attr_remove_iter_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +0
-7-return error; - }else if (error) { - goto out; - } - -/* -* Refill the state structure with buffers (the prior
fs/xfs/libxfs/xfs_attr.c +0
-1-* calls released our buffers) and close out this
fs/xfs/libxfs/xfs_attr.c xfs_attr_node_remove_attr( · line 1516
Quoted 4 lines
fs/xfs/libxfs/xfs_attr.c +0
-4-* transaction before proceeding. -*/ -ASSERT (args->rmtblkno ==0 ); - error =xfs_attr_refillstate (state);
I think you can remove xfs_attr_refillstate too. I'm getting some compiler gripes about that being declared but not used, and I'm pretty sure this was the last call to it, so probably it can go too. Other than that this patch looks ok.
Quoted 70 lines
fs/xfs/libxfs/xfs_attr.c +0
-6-if (error) - goto out; - - attr->xattri_dela_state =XFS_DAS_RM_NAME ; - trace_xfs_attr_remove_iter_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +0
-7-return -EAGAIN ; - } - - fallthrough; -case XFS_DAS_RM_NAME : -/* -* If we came here fresh from a transaction roll, reattach all
fs/xfs/libxfs/xfs_attr.c +0
-12-* the buffers to the current transaction. -*/ -if (attr->xattri_dela_state ==XFS_DAS_RM_NAME ) { - error =xfs_attr_refillstate (state); -if (error) - goto out; - } - - retval =xfs_attr_node_removename (args, state); - -/* -* Check to see if the tree needs to be collapsed. If so, roll
fs/xfs/libxfs/xfs_attr.c +0
-10-* the transacton and fall into the shrink state. -*/ -if (retval && (state->path.active >1 )) { - error =xfs_da3_join (state); -if (error) - goto out; - - attr->xattri_dela_state =XFS_DAS_RM_SHRINK ; - trace_xfs_attr_remove_iter_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +0
-7-return -EAGAIN ; - } - - fallthrough; -case XFS_DAS_RM_SHRINK : -/* -* If the result is small enough, push it all into the inode.
fs/xfs/libxfs/xfs_attr.c +0
-1-* This is our final state so it's safe to return a dirty
fs/xfs/libxfs/xfs_attr.c +0
-17-* transaction. -*/ -if (xfs_attr_is_leaf (dp)) - error =xfs_attr_node_shrink (args, state); -ASSERT (error != -EAGAIN ); -break ; -default : -ASSERT (0 ); - error = -EINVAL ; - goto out; - } -out: -if (state) -xfs_da_state_free (state); -return error; -} -/* * Fill in the disk block numbers in the state structure for the buffers
* that are attached to the state structure.
Re: [PATCH 14/16] xfs: remove xfs_attri_remove_iter
Quoted 41 lines · Alli
Quoted 40 lines · Dave Chinner
From: Dave Chinner <[email protected]>
xfs_attri_remove_iteris not used anymore, so remove it and all the infrastructure it uses and is needed to drive it.Signed-off-by Dave Chinner <[email protected]>---
fs/xfs/libxfs/xfs_attr.c +0
-5-case XFS_DAS_RMTBLK : - attr->xattri_dela_state =XFS_DAS_RMTBLK ; - -/* -* If there is an out-of-line value, de-allocate the blocks.
(quoted diff) +0
-1- * This is done before we remove the attribute so thatwe don't
(quoted diff) +0
-1- * overflow the maximum size of a transaction and/orhit a
fs/xfs/libxfs/xfs_attr.c +0
-5-* deadlock. -*/ -if (args->rmtblkno >0 ) { -/* -* May return -EAGAIN. Roll and repeat until all remote
fs/xfs/libxfs/xfs_attr.c +0
-6-* blocks are removed. -*/ - error =xfs_attr_rmtval_remove (attr); -if (error == -EAGAIN ) { - trace_xfs_attr_remove_iter_return( - attr->xattri_dela_state, args-Quoted 1 line
dp);
fs/xfs/libxfs/xfs_attr.c +0
-7-return error; - }else if (error) { - goto out; - } - -/* -* Refill the state structure with buffers (the prior
(quoted diff) +0
-1- * calls released our buffers) and close outthis
Quoted 8 lines
fs/xfs/libxfs/xfs_attr.c xfs_attr_node_remove_attr( · line 1516
Quoted 4 lines
fs/xfs/libxfs/xfs_attr.c +0
-4-* transaction before proceeding. -*/ -ASSERT (args->rmtblkno ==0 ); - error =xfs_attr_refillstate (state);I think you can remove
xfs_attr_refillstatetoo. I'm getting some compiler gripes about that being declared but not used, and I'm pretty sure this was the last call to it, so probably it can go too.
I'm going to ifdef it out for now, which will remove the compiler warning. I'll leave a comment explaining why it's been left there unused for the moment. That is a bit of a long story.....
.... so it's Story Time! :)
The xfs_attr_savestate(), xfs_attr_refillstate() pair are for saving the state path cursor that points to the name entry in the btree so we don't have to look it up again after we've removed the remote value blocks.
Removing the remote value extents trashes the state cursor (points to remote value extents that have been removed on return, not the name entry in the btree), so to avoid doing another tree walk we save the disk addresses of the path blocks before remote extent removal and restore them to the state cursor afterwards.
This avoids having to walk the hash tree to find the data block that contains the name entry again to remove that. It's an optimisation that I removed when combining the two state machines because it hurt my brain trying to reconcile the two paths. Hence I ended up using the simpler _set_iter() path that just does the lookup a second time when I combined the state machines rather than the more complex save/restore path in the remove_iter state machine.
I left the code there because I want to re-introduce the optimisation in future - it will benefit both the replace and remove paths - but I want to get the code working correctly first before I worry too much about intricate performance optimisations like this.
Quoted 4 lines
Other
than that this patch looks ok.Reviewed-by Allison Henderson<[email protected]>
Thanks!
-Dave.
Signature · 2 lines
Dave Chinner [email protected]
PATCH 15/16 xfs: split attr replace op setup from create op setup
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
In preparation for having a different replace algorithm when LARP mode is active.
--- fs/xfs/libxfs/xfs_attr.c | 194 ++++++++++++++++++++++----------------- fs/xfs/libxfs/xfs_attr.h | 2 - 2 files changed, 112 insertions(+), 84 deletions(-)
fs/xfs/libxfs/xfs_attr.c +112 -82
@@ -724,6 +724,96 @@ xfs_attr_lookup(return xfs_attr_node_hasname (args,NULL ); } +static int +xfs_attr_item_init ( +struct xfs_da_args *args, +unsigned int op_flags,/* op flag (set or remove) */ +struct xfs_attr_item **attr)/* new xfs_attr_item */ +{ + +struct xfs_attr_item *new; + + new =kmem_zalloc (sizeof (struct xfs_attr_item ),KM_NOFS ); + new->xattri_op_flags = op_flags; + new->xattri_da_args = args; + + *attr = new; +return 0 ; +} + +/* Sets an attribute for an inode as a deferred operation */ +static int +xfs_attr_defer_add ( +struct xfs_da_args *args) +{ +struct xfs_attr_item *new; +int error =0 ; + + error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); +if (error) +return error; + +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_ADD ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_ADD ; +else + new->xattri_dela_state =XFS_DAS_NODE_ADD ; + +xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); + +return 0 ; +} + +/* Sets an attribute for an inode as a deferred operation */ +static int +xfs_attr_defer_replace ( +struct xfs_da_args *args) +{ +struct xfs_attr_item *new; +int error =0 ; + + error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); +if (error) +return error; + +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_ADD ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_ADD ; +else + new->xattri_dela_state =XFS_DAS_NODE_ADD ; + +xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); + +return 0 ; +} + +/* Removes an attribute for an inode as a deferred operation */ +static int +xfs_attr_defer_remove ( +struct xfs_da_args *args) +{ + +struct xfs_attr_item *new; +int error; + + error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_REMOVE , &new); +if (error) +return error; + +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_REMOVE ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; +else + new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; + +xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); + +return 0 ; +} +/* * Note: If args->value is NULL the attribute will be removed, just like the * Linux ->setattr API.
@@ -812,29 +902,35 @@ xfs_attr_set( } error =xfs_attr_lookup (args); -if (args->value) { -if (error == -EEXIST && (args->attr_flags &XATTR_CREATE )) - goto out_trans_cancel; -if (error == -ENOATTR && (args->attr_flags &XATTR_REPLACE )) - goto out_trans_cancel; -if (error != -ENOATTR && error != -EEXIST ) +switch (error) { +case -EEXIST : +/* if no value, we are performing a remove operation */ +if (!args->value) { + error =xfs_attr_defer_remove (args); +break ; + } +/* Pure create fails if the attr already exists */ +if (args->attr_flags &XATTR_CREATE ) goto out_trans_cancel; - error =xfs_attr_set_deferred (args); -if (error) + error =xfs_attr_defer_replace (args); +break ; +case -ENOATTR : +/* Can't remove what isn't there. */ +if (!args->value) goto out_trans_cancel; -/* shortform attribute has already been committed */ -if (!args->trans) - goto out_unlock; - }else { -if (error != -EEXIST ) +/* Pure replace fails if no existing attr to replace. */ +if (args->attr_flags &XATTR_REPLACE ) goto out_trans_cancel; - error =xfs_attr_remove_deferred (args); -if (error) - goto out_trans_cancel; + error =xfs_attr_defer_add (args); +break ; +default : + goto out_trans_cancel; } +if (error) + goto out_trans_cancel;/* * If this is a synchronous mount, make sure that the
@@ -898,72 +994,6 @@ xfs_attrd_destroy_cache(void) xfs_attrd_cache =NULL ; } -STATIC int -xfs_attr_item_init ( -struct xfs_da_args *args, -unsigned int op_flags,/* op flag (set or remove) */ -struct xfs_attr_item **attr)/* new xfs_attr_item */ -{ - -struct xfs_attr_item *new; - - new =kmem_zalloc (sizeof (struct xfs_attr_item ),KM_NOFS ); - new->xattri_op_flags = op_flags; - new->xattri_da_args = args; - - *attr = new; -return 0 ; -} - -/* Sets an attribute for an inode as a deferred operation */ -int -xfs_attr_set_deferred ( -struct xfs_da_args *args) -{ -struct xfs_attr_item *new; -int error =0 ; - - error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); -if (error) -return error; - -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_ADD ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_ADD ; -else - new->xattri_dela_state =XFS_DAS_NODE_ADD ; - -xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); - -return 0 ; -} - -/* Removes an attribute for an inode as a deferred operation */ -int -xfs_attr_remove_deferred ( -struct xfs_da_args *args) -{ - -struct xfs_attr_item *new; -int error; - - error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_REMOVE , &new); -if (error) -return error; - -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_REMOVE ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; -else - new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; - -xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); - -return 0 ; -} -/*======================================================================== * External routines when attribute list is inside the inode *========================================================================*/
fs/xfs/libxfs/xfs_attr.h +0 -2
@@ -559,8 +559,6 @@ bool xfs_attr_namecheck(const void *name, size_t length);int xfs_attr_calc_size (struct xfs_da_args *args,int *local);void xfs_init_attr_trans (struct xfs_da_args *args,struct xfs_trans_res *tres,unsigned int *total); -int xfs_attr_set_deferred (struct xfs_da_args *args); -int xfs_attr_remove_deferred (struct xfs_da_args *args);extern struct kmem_cache *xfs_attri_cache;extern struct kmem_cache *xfs_attrd_cache;
Signature · 1 line
2.35.1
Re: [PATCH 15/16] xfs: split attr replace op setup from create op setup
Quoted 6 lines · Dave Chinner
From: Dave Chinner <[email protected]>
In preparation for having a different replace algorithm when LARP mode is active.
Signed-off-by Dave Chinner <[email protected]>
Looks fine
Quoted 265 lines
--- fs/xfs/libxfs/xfs_attr.c | 194 ++++++++++++++++++++++--------------- -- fs/xfs/libxfs/xfs_attr.h | 2 - 2 files changed, 112 insertions(+), 84 deletions(-)fs/xfs/libxfs/xfs_attr.c +4
-0@@ -724,6 +724,96 @@ xfs_attr_lookup(return xfs_attr_node_hasname (args,NULL ); } +static int +xfs_attr_item_init( + struct xfs_da_args *args, + unsigned int op_flags, /* op flag (set orremove) */
fs/xfs/libxfs/xfs_attr.c +1
-0+ struct xfs_attr_item **attr) /* new xfs_attr_item*/
fs/xfs/libxfs/xfs_attr.c +32
-0+{ + +struct xfs_attr_item *new; + + new =kmem_zalloc (sizeof (struct xfs_attr_item ),KM_NOFS ); + new->xattri_op_flags = op_flags; + new->xattri_da_args = args; + + *attr = new; +return 0 ; +} + +/* Sets an attribute for an inode as a deferred operation */ +static int +xfs_attr_defer_add ( +struct xfs_da_args *args) +{ +struct xfs_attr_item *new; +int error =0 ; + + error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); +if (error) +return error; + +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_ADD ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_ADD ; +else + new->xattri_dela_state =XFS_DAS_NODE_ADD ; + + xfs_defer_add(args->trans, XFS_DEFER_OPS_TYPE_ATTR, &new-Quoted 1 line
xattri_list);fs/xfs/libxfs/xfs_attr.c +24
-0+ +return 0 ; +} + +/* Sets an attribute for an inode as a deferred operation */ +static int +xfs_attr_defer_replace ( +struct xfs_da_args *args) +{ +struct xfs_attr_item *new; +int error =0 ; + + error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); +if (error) +return error; + +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_ADD ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_ADD ; +else + new->xattri_dela_state =XFS_DAS_NODE_ADD ; + + xfs_defer_add(args->trans, XFS_DEFER_OPS_TYPE_ATTR, &new-Quoted 1 line
xattri_list);fs/xfs/libxfs/xfs_attr.c +14
-0+ +return 0 ; +} + +/* Removes an attribute for an inode as a deferred operation */ +static int +xfs_attr_defer_remove ( +struct xfs_da_args *args) +{ + +struct xfs_attr_item *new; +int error; + + error = xfs_attr_item_init(args, XFS_ATTR_OP_FLAGS_REMOVE,&new);
fs/xfs/libxfs/xfs_attr.c +11
-0+if (error) +return error; + +if (xfs_attr_is_shortform (args->dp)) + new->xattri_dela_state =XFS_DAS_SF_REMOVE ; +else if (xfs_attr_is_leaf (args->dp)) + new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; +else + new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; + + xfs_defer_add(args->trans, XFS_DEFER_OPS_TYPE_ATTR, &new-Quoted 1 line
xattri_list);fs/xfs/libxfs/xfs_attr.c +4
-0+ +return 0 ; +} +/* * Note: If args->value is NULL the attribute will be removed, just like the
* Linux ->setattr API.fs/xfs/libxfs/xfs_attr.c +22
-16@@ -812,29 +902,35 @@ xfs_attr_set( } error =xfs_attr_lookup (args); -if (args->value) { -if (error == -EEXIST && (args->attr_flags & XATTR_CREATE)) -goto out_trans_cancel; -if (error == -ENOATTR && (args->attr_flags & XATTR_REPLACE)) -goto out_trans_cancel; -if (error != -ENOATTR && error != -EEXIST) + switch (error) { +case -EEXIST : +/* if no value, we are performing a remove operation */ +if (!args->value) { + error =xfs_attr_defer_remove (args); +break ; + } +/* Pure create fails if the attr already exists */ +if (args->attr_flags &XATTR_CREATE ) goto out_trans_cancel; -error = xfs_attr_set_deferred(args); -if (error) + error =xfs_attr_defer_replace (args); +break ; +case -ENOATTR : +/* Can't remove what isn't there. */ +if (!args->value) goto out_trans_cancel; -/* shortform attribute has already been committed */ -if (!args->trans) - goto out_unlock; - }else { -if (error != -EEXIST ) +/* Pure replace fails if no existing attr to replace. */ +if (args->attr_flags &XATTR_REPLACE ) goto out_trans_cancel; - error =xfs_attr_remove_deferred (args); -if (error) - goto out_trans_cancel; + error =xfs_attr_defer_add (args); +break ; +default : + goto out_trans_cancel; } +if (error) + goto out_trans_cancel;fs/xfs/libxfs/xfs_attr.c +0
-4/* * If this is a synchronous mount, make sure that the @@ -898,72 +994,6 @@ xfs_attrd_destroy_cache(void) xfs_attrd_cache =NULL ; } -STATIC int -xfs_attr_item_init( - struct xfs_da_args *args, - unsigned int op_flags, /* op flag (set orremove) */
fs/xfs/libxfs/xfs_attr.c +0
-1- struct xfs_attr_item **attr) /* new xfs_attr_item*/
fs/xfs/libxfs/xfs_attr.c +0
-32-{ - -struct xfs_attr_item *new; - - new =kmem_zalloc (sizeof (struct xfs_attr_item ),KM_NOFS ); - new->xattri_op_flags = op_flags; - new->xattri_da_args = args; - - *attr = new; -return 0 ; -} - -/* Sets an attribute for an inode as a deferred operation */ -int -xfs_attr_set_deferred ( -struct xfs_da_args *args) -{ -struct xfs_attr_item *new; -int error =0 ; - - error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); -if (error) -return error; - -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_ADD ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_ADD ; -else - new->xattri_dela_state =XFS_DAS_NODE_ADD ; - - xfs_defer_add(args->trans, XFS_DEFER_OPS_TYPE_ATTR, &new-Quoted 1 line
xattri_list);fs/xfs/libxfs/xfs_attr.c +0
-14- -return 0 ; -} - -/* Removes an attribute for an inode as a deferred operation */ -int -xfs_attr_remove_deferred ( -struct xfs_da_args *args) -{ - -struct xfs_attr_item *new; -int error; - - error = xfs_attr_item_init(args, XFS_ATTR_OP_FLAGS_REMOVE,&new);
fs/xfs/libxfs/xfs_attr.c +0
-11-if (error) -return error; - -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_REMOVE ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; -else - new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; - - xfs_defer_add(args->trans, XFS_DEFER_OPS_TYPE_ATTR, &new-Quoted 1 line
xattri_list);fs/xfs/libxfs/xfs_attr.c +0
-4- -return 0 ; -} -/*================================================================== ======
* External routines when attribute list is inside the inode*====================================================================
====*/fs/xfs/libxfs/xfs_attr.h +0
-2@@ -559,8 +559,6 @@ bool xfs_attr_namecheck(const void *name, size_t length);int xfs_attr_calc_size (struct xfs_da_args *args,int *local);void xfs_init_attr_trans (struct xfs_da_args *args,struct xfs_trans_res *tres,unsigned int *total); -int xfs_attr_set_deferred (struct xfs_da_args *args); -int xfs_attr_remove_deferred (struct xfs_da_args *args);extern struct
kmem_cache*xfs_attri_cache;
extern structkmem_cache*xfs_attrd_cache;
PATCH 16/16 xfs: ATTR_REPLACE algorithm with LARP enabled needs rework
In reply to Dave ChinnerFrom: Dave Chinner <[email protected]>
We can't use the same algorithm for replacing an existing attribute when logging attributes. The existing algorithm is essentially:
1. create new attr w/ INCOMPLETE
2. atomically flip INCOMPLETE flags between old + new attribute
3. remove old attr which is marked w/ INCOMPLETE
This algorithm guarantees that we see either the old or new attribute, and if we fail after the atomic flag flip, we don't have to recover the removal of the old attr because we never see INCOMPLETE attributes in lookups.
For logged attributes, however, this does not work. The logged attribute intents do not track the work that has been done as the transaction rolls, and hence the only recovery mechanism we have is "run the replace operation from scratch".
This is further exacerbated by the attempt to avoid needing the INCOMPLETE flag to create an atomic swap. This means we can create a second active attribute of the same name before we remove the original. If we fail at any point after the create but before the removal has completed, we end up with duplicate attributes in the attr btree and recovery only tries to replace one of them.
There are several other failure modes where we can leave partially allocated remote attributes that expose stale data, partially free remote attributes that enable UAF based stale data exposure, etc.
TO fix this, we need a different algorithm for replace operations when LARP is enabled. Luckily, it's not that complex if we take the right first step. That is, the first thing we log is the attri intent with the new name/value pair and mark the old attr as INCOMPLETE in the same transaction.
From there, we then remove the old attr and keep relogging the new name/value in the intent, such that we always know that we have to create the new attr in recovery. Once the old attr is removed, we then run a normal ATTR_CREATE operation relogging the intent as we go. If the new attr is local, then it gets created in a single atomic transaction that also logs the final intent done. If the new attr is remote, the we set INCOMPLETE on the new attr while we allocate and set the remote value, and then we clear the INCOMPLETE flag at in the last transaction taht logs the final intent done.
If we fail at any point in this algorithm, log recovery will always see the same state on disk: the new name/value in the intent, and either an INCOMPLETE attr or no attr in the attr btree. If we find an INCOMPLETE attr, we run the full replace starting with removing the INCOMPLETE attr. If we don't find it, then we simply create the new attr.
Notably, recovery of a failed create that has an INCOMPLETE flag set is now the same - we start with the lookup of the INCOMPLETE attr, and if that exists then we do the full replace recovery process, otherwise we just create the new attr.
Hence changing the way we do the replace operation when LARP is enabled allows us to use the same log recovery algorithm for both the ATTR_CREATE and ATTR_REPLACE operations. This is also the same algorithm we use for runtime ATTR_REPLACE operations (except for the step setting up the initial conditions).
The result is that:
- ATTR_CREATE uses the same algorithm regardless of whether LARP is enabled or not
- ATTR_REPLACE with larp=0 is identical to the old algorithm
- ATTR_REPLACE with larp=1 runs an unmodified attr removal algorithm from the larp=0 code and then runs the unmodified ATTR_CREATE code.
- log recovery when larp=1 runs the same ATTR_REPLACE algorithm as it uses at runtime.
Because the state machine is now quite clean, changing the algorithm is really just a case of changing the initial state and how the states link together for the ATTR_REPLACE case. Hence it's not a huge amoutn of code for what is a fairly substantial rework of the attr logging and recovery algorithm....
--- fs/xfs/libxfs/xfs_attr.c | 137 ++++++++++++++------------------- fs/xfs/libxfs/xfs_attr.h | 101 +++++++++++++++++++----- fs/xfs/libxfs/xfs_attr_leaf.c | 22 ++++-- fs/xfs/libxfs/xfs_da_btree.h | 1 + fs/xfs/libxfs/xfs_defer.c | 2 + fs/xfs/libxfs/xfs_log_format.h | 7 +- fs/xfs/xfs_acl.c | 2 +- fs/xfs/xfs_attr_item.c | 27 ++++++- fs/xfs/xfs_ioctl.c | 2 +- fs/xfs/xfs_trace.h | 10 +-- fs/xfs/xfs_xattr.c | 2 +- 11 files changed, 197 insertions(+), 116 deletions(-)
fs/xfs/libxfs/xfs_attr.c +58 -79
@@ -260,19 +260,6 @@ xfs_attr_try_sf_addname(return error; } -/* -* Check to see if the attr should be upgraded from non-existent or shortform to -* single-leaf-block attribute list. -*/ -static inline bool -xfs_attr_is_shortform ( -struct xfs_inode *ip) -{ -return ip->i_afp->if_format ==XFS_DINODE_FMT_LOCAL || - (ip->i_afp->if_format ==XFS_DINODE_FMT_EXTENTS && - ip->i_afp->if_nextents ==0 ); -} -static int xfs_attr_sf_addname (struct xfs_attr_item *attr)
@@ -309,20 +296,29 @@ xfs_attr_sf_addname( }/* -* When we bump the state to REPLACE, we may actually need to skip over the -* state. When LARP mode is enabled, we don't need to run the atomic flags flip, -* so we skip straight over the REPLACE state and go on to REMOVE_OLD. +* Handle the state change on completion of a multi-state attr operation. +* +* If the XFS_DA_OP_RENAME flag is set, this means the operation was the first +* modification in a attr replace operation and we still have to do the second +* state, indicated by @replace_state. +* +* We consume the XFS_DA_OP_RENAME flag so that when we are called again on +* completion of the second half of the attr replace operation we correctly +* signal that it is done. */ -static void -xfs_attr_dela_state_set_replace ( +static enum xfs_delattr_state +xfs_attr_complete_op (struct xfs_attr_item *attr, -enum xfs_delattr_state replace) +enum xfs_delattr_state replace_state) {struct xfs_da_args *args = attr->xattri_da_args; +bool do_replace = args->op_flags &XFS_DA_OP_RENAME ; - attr->xattri_dela_state = replace; -if (xfs_has_larp (args->dp->i_mount)) - attr->xattri_dela_state++; + args->attr_flags &= ~XATTR_REPLACE ; + args->op_flags &= ~XFS_DA_OP_RENAME ; +if (do_replace) +return replace_state; +return XFS_DAS_DONE ; } static int
@@ -364,10 +360,9 @@ xfs_attr_leaf_addname(*/ if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; -else if (args->op_flags &XFS_DA_OP_RENAME ) -xfs_attr_dela_state_set_replace (attr,XFS_DAS_LEAF_REPLACE );else - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state =xfs_attr_complete_op (attr, +XFS_DAS_LEAF_REPLACE ); out:trace_xfs_attr_leaf_addname_return (attr->xattri_dela_state, args->dp);return error;
@@ -409,10 +404,9 @@ xfs_attr_node_addname(if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; -else if (args->op_flags &XFS_DA_OP_RENAME ) -xfs_attr_dela_state_set_replace (attr,XFS_DAS_NODE_REPLACE );else - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state =xfs_attr_complete_op (attr, +XFS_DAS_NODE_REPLACE ); out:trace_xfs_attr_node_addname_return (attr->xattri_dela_state, args->dp);return error;
@@ -442,18 +436,15 @@ xfs_attr_rmtval_alloc(if (error)return error; -/* If this is not a rename, clear the incomplete flag and we're done. */ -if (!(args->op_flags &XFS_DA_OP_RENAME )) { + attr->xattri_dela_state =xfs_attr_complete_op (attr, + ++attr->xattri_dela_state); +/* +* If we are not doing a rename, we've finished the operation but still +* have to clear the incomplete flag protecting the new attr from +* exposing partially initialised state if we crash during creation. +*/ +if (attr->xattri_dela_state ==XFS_DAS_DONE ) error =xfs_attr3_leaf_clearflag (args); - attr->xattri_dela_state =XFS_DAS_DONE ; - }else { -/* -* We are running a REPLACE operation, so we need to bump the -* state to the step in that operation. -*/ - attr->xattri_dela_state++; -xfs_attr_dela_state_set_replace (attr, attr->xattri_dela_state); - } out:trace_xfs_attr_rmtval_alloc (attr->xattri_dela_state, args->dp);return error;
@@ -578,11 +569,15 @@ xfs_attr_set_iter(return xfs_attr_node_addname (attr);case XFS_DAS_SF_REMOVE : - attr->xattri_dela_state =XFS_DAS_DONE ; -return xfs_attr_sf_removename (args); + error =xfs_attr_sf_removename (args); + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args)); +break ;case XFS_DAS_LEAF_REMOVE : - attr->xattri_dela_state =XFS_DAS_DONE ; -return xfs_attr_leaf_removename (args); + error =xfs_attr_leaf_removename (args); + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args)); +break ;case XFS_DAS_NODE_REMOVE : error =xfs_attr_node_removename_setup (attr);if (error)
@@ -678,12 +673,14 @@ xfs_attr_set_iter(case XFS_DAS_LEAF_REMOVE_ATTR : error =xfs_attr_leaf_remove_attr (attr); - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args));break ;case XFS_DAS_NODE_REMOVE_ATTR : error =xfs_attr_node_remove_attr (attr); - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args));break ;default :ASSERT (0 );
@@ -753,14 +750,9 @@ xfs_attr_defer_add(if (error)return error; -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_ADD ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_ADD ; -else - new->xattri_dela_state =XFS_DAS_NODE_ADD ; - + new->xattri_dela_state =xfs_attr_init_add_state (args);xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); +trace_xfs_attr_defer_add (new->xattri_dela_state, args->dp);return 0 ; }
@@ -773,18 +765,13 @@ xfs_attr_defer_replace(struct xfs_attr_item *new;int error =0 ; - error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); + error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_REPLACE , &new);if (error)return error; -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_ADD ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_ADD ; -else - new->xattri_dela_state =XFS_DAS_NODE_ADD ; - + new->xattri_dela_state =xfs_attr_init_replace_state (args);xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); +trace_xfs_attr_defer_replace (new->xattri_dela_state, args->dp);return 0 ; }
@@ -802,14 +789,9 @@ xfs_attr_defer_remove(if (error)return error; -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_REMOVE ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; -else - new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; - + new->xattri_dela_state =xfs_attr_init_remove_state (args);xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new->xattri_list); +trace_xfs_attr_defer_remove (new->xattri_dela_state, args->dp);return 0 ; }
@@ -1032,6 +1014,7 @@ xfs_attr_shortform_addname(xfs_da_args_t *args)* not being around. */ args->attr_flags &= ~XATTR_REPLACE ; + args->op_flags &= ~XFS_DA_OP_RENAME ; } if (args->namelen >= XFS_ATTR_SF_ENTSIZE_MAX ||
@@ -1122,16 +1105,14 @@ xfs_attr_leaf_try_add( goto out_brelse;trace_xfs_attr_leaf_replace (args); - -/* save the attribute state for later removal*/ - args->op_flags |=XFS_DA_OP_RENAME ;/* an atomic rename */ -xfs_attr_save_rmt_blk (args); +ASSERT (args->op_flags &XFS_DA_OP_RENAME );/* -* clear the remote attr state now that it is saved so that the -* values reflect the state of the attribute we are about to +* Save the existing remote attr state so that the current +* values reflect the state of the new attribute we are about to * add, not the attribute we just found and will remove later. */ +xfs_attr_save_rmt_blk (args); args->rmtblkno =0 ; args->rmtblkcnt =0 ; args->rmtvaluelen =0 ;
@@ -1293,16 +1274,14 @@ xfs_attr_node_addname_find_attr( goto error;trace_xfs_attr_node_replace (args); - -/* save the attribute state for later removal*/ - args->op_flags |=XFS_DA_OP_RENAME ;/* atomic rename op */ -xfs_attr_save_rmt_blk (args); +ASSERT (args->op_flags &XFS_DA_OP_RENAME );/* -* clear the remote attr state now that it is saved so that the -* values reflect the state of the attribute we are about to +* Save the existing remote attr state so that the current +* values reflect the state of the new attribute we are about to * add, not the attribute we just found and will remove later. */ +xfs_attr_save_rmt_blk (args); args->rmtblkno =0 ; args->rmtblkcnt =0 ; args->rmtvaluelen =0 ;
fs/xfs/libxfs/xfs_attr.h +83 -18
131 lines not on this page · raw message
fs/xfs/libxfs/xfs_attr_leaf.c +17 -5
@@ -446,6 +446,14 @@ xfs_attr3_leaf_read(* Namespace helper routines *========================================================================*/ +/* +* If we are in log recovery, then we want the lookup to ignore the INCOMPLETE +* flag on disk - if there's an incomplete attr then recovery needs to tear it +* down. If there's no incomplete attr, then recovery needs to tear that attr +* down to replace it with the attr that has been logged. In this case, the +* INCOMPLETE flag will not be set in attr->attr_filter, but rather +* XFS_DA_OP_RECOVERY will be set in args->op_flags. +*/ static bool xfs_attr_match( struct xfs_da_args *args,
@@ -453,14 +461,18 @@ xfs_attr_match( unsigned char *name, int flags) { +if (args->namelen != namelen)return false;if (memcmp (args->name, name, namelen) !=0 )return false; -/* -* If we are looking for incomplete entries, show only those, else only -* show complete entries. -*/ + +/* Recovery ignores the INCOMPLETE flag. */ +if ((args->op_flags &XFS_DA_OP_RECOVERY ) && + args->attr_filter == (flags &XFS_ATTR_NSP_ONDISK_MASK )) +return true; + +/* All remaining matches need to be filtered by INCOMPLETE state. */ if (args->attr_filter != (flags & (XFS_ATTR_NSP_ONDISK_MASK |XFS_ATTR_INCOMPLETE )))return false;
@@ -819,7 +831,7 @@ xfs_attr_sf_removename( totsize -= size;if (totsize ==sizeof (xfs_attr_sf_hdr_t) &&xfs_has_attr2 (mp) && (dp->i_df.if_format !=XFS_DINODE_FMT_BTREE ) && - !(args->op_flags &XFS_DA_OP_ADDNAME )) { + !(args->op_flags & (XFS_DA_OP_ADDNAME |XFS_DA_OP_RENAME ))) {xfs_attr_fork_remove (dp, args->trans); }else {xfs_idata_realloc (dp, -size,XFS_ATTR_FORK );
fs/xfs/libxfs/xfs_da_btree.h +1 -0
@@ -90,6 +90,7 @@ typedef struct xfs_da_args {#define XFS_DA_OP_OKNOENT (1u << 3)/* lookup op, ENOENT ok, else die */ #define XFS_DA_OP_CILOOKUP (1u << 4)/* lookup returns CI name if found */ #define XFS_DA_OP_NOTIME (1u << 5)/* don't update inode timestamps */ +#define XFS_DA_OP_RECOVERY (1u << 6)/* running a recovery operation */ #define XFS_DA_OP_FLAGS \ { XFS_DA_OP_JUSTCHECK, "JUSTCHECK" }, \
fs/xfs/libxfs/xfs_defer.c +2 -0
@@ -23,6 +23,8 @@#include "xfs_bmap.h" #include "xfs_alloc.h" #include "xfs_buf.h" +#include "xfs_da_format.h" +#include "xfs_da_btree.h" #include "xfs_attr.h" static struct kmem_cache *xfs_defer_pending_cache;
fs/xfs/libxfs/xfs_log_format.h +4 -3
@@ -906,9 +906,10 @@ struct xfs_icreate_log {* Flags for deferred attribute operations. * Upper bits are flags, lower byte is type code */ -#define XFS_ATTR_OP_FLAGS_SET 1/* Set the attribute */ -#define XFS_ATTR_OP_FLAGS_REMOVE 2/* Remove the attribute */ -#define XFS_ATTR_OP_FLAGS_TYPE_MASK 0xFF/* Flags type mask */ +#define XFS_ATTR_OP_FLAGS_SET (1u << 0)/* Set the attribute */ +#define XFS_ATTR_OP_FLAGS_REMOVE (1u << 1)/* Remove the attribute */ +#define XFS_ATTR_OP_FLAGS_REPLACE (1u << 2)/* Replace the attribute */ +#define XFS_ATTR_OP_FLAGS_TYPE_MASK 0xFF/* Flags type mask */ /* * This is the structure used to lay out an attr log item in the
fs/xfs/xfs_acl.c +1 -1
@@ -10,12 +10,12 @@#include "xfs_trans_resv.h" #include "xfs_mount.h" #include "xfs_inode.h" -#include "xfs_attr.h" #include "xfs_trace.h" #include "xfs_error.h" #include "xfs_acl.h" #include "xfs_da_format.h" #include "xfs_da_btree.h" +#include "xfs_attr.h" #include "xfs_trans.h" #include <linux/posix_acl_xattr.h>
fs/xfs/xfs_attr_item.c +24 -3
@@ -490,9 +490,14 @@ xfs_attri_validate(if (attrp->__pad !=0 )return false; -/* alfi_op_flags should be either a set or remove */ -if (op !=XFS_ATTR_OP_FLAGS_SET && op !=XFS_ATTR_OP_FLAGS_REMOVE ) +switch (op) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REMOVE : +case XFS_ATTR_OP_FLAGS_REPLACE : +break ; +default :return false; + }if (attrp->alfi_value_len >XATTR_SIZE_MAX )return false;
@@ -553,11 +558,27 @@ xfs_attri_item_recover( args->namelen = attrp->alfi_name_len; args->hashval =xfs_da_hashname (args->name, args->namelen); args->attr_filter = attrp->alfi_attr_flags; + args->op_flags =XFS_DA_OP_RECOVERY ; -if (attrp->alfi_op_flags ==XFS_ATTR_OP_FLAGS_SET ) { +switch (attr->xattri_op_flags) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REPLACE : args->value = attrip->attri_value; args->valuelen = attrp->alfi_value_len; args->total =xfs_attr_calc_size (args, &local); +if (xfs_inode_hasattr (args->dp)) + attr->xattri_dela_state =xfs_attr_init_replace_state (args); +else + attr->xattri_dela_state =xfs_attr_init_add_state (args); +break ; +case XFS_ATTR_OP_FLAGS_REMOVE : +if (!xfs_inode_hasattr (args->dp)) + goto out; + attr->xattri_dela_state =xfs_attr_init_remove_state (args); +break ; +default : + error = -EFSCORRUPTED ; + goto out; }xfs_init_attr_trans (args, &tres, &total);
fs/xfs/xfs_ioctl.c +1 -1
@@ -15,7 +15,6 @@#include "xfs_iwalk.h" #include "xfs_itable.h" #include "xfs_error.h" -#include "xfs_attr.h" #include "xfs_bmap.h" #include "xfs_bmap_util.h" #include "xfs_fsops.h"
@@ -37,6 +36,7 @@#include "xfs_ioctl.h" #include "xfs_da_format.h" #include "xfs_da_btree.h" +#include "xfs_attr.h" #include <linux/mount.h> #include <linux/namei.h>
fs/xfs/xfs_trace.h +5 -5
@@ -4100,13 +4100,10 @@ DEFINE_ICLOG_EVENT(xlog_iclog_write);TRACE_DEFINE_ENUM (XFS_DAS_UNINIT );TRACE_DEFINE_ENUM (XFS_DAS_SF_ADD ); -TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ADD ); -TRACE_DEFINE_ENUM (XFS_DAS_NODE_ADD ); -TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK );TRACE_DEFINE_ENUM (XFS_DAS_SF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ADD );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ADD );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );
@@ -4142,6 +4139,9 @@ DECLARE_EVENT_CLASS(xfs_das_state_class,DEFINE_EVENT (xfs_das_state_class, name, \ TP_PROTO(int das, struct xfs_inode *ip), \ TP_ARGS(das, ip)) +DEFINE_DAS_STATE_EVENT (xfs_attr_defer_add); +DEFINE_DAS_STATE_EVENT (xfs_attr_defer_replace); +DEFINE_DAS_STATE_EVENT (xfs_attr_defer_remove);DEFINE_DAS_STATE_EVENT (xfs_attr_sf_addname_return);DEFINE_DAS_STATE_EVENT (xfs_attr_set_iter_return);DEFINE_DAS_STATE_EVENT (xfs_attr_leaf_addname_return);
fs/xfs/xfs_xattr.c +1 -1
@@ -12,9 +12,9 @@#include "xfs_trans_resv.h" #include "xfs_mount.h" #include "xfs_inode.h" -#include "xfs_attr.h" #include "xfs_acl.h" #include "xfs_da_btree.h" +#include "xfs_attr.h" #include <linux/posix_acl_xattr.h>
Signature · 1 line
2.35.1
Re: [PATCH 16/16] xfs: ATTR_REPLACE algorithm with LARP enabled needs rework
Quoted 75 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We can't use the same algorithm for replacing an existing attribute when logging attributes. The existing algorithm is essentially:
1. create new attr w/ INCOMPLETE
2. atomically flip INCOMPLETE flags between old + new attribute
3. remove old attr which is marked w/ INCOMPLETEThis algorithm guarantees that we see either the old or new attribute, and if we fail after the atomic flag flip, we don't have to recover the removal of the old attr because we never see INCOMPLETE attributes in lookups.
For logged attributes, however, this does not work. The logged attribute intents do not track the work that has been done as the transaction rolls, and hence the only recovery mechanism we have is "run the replace operation from scratch".
This is further exacerbated by the attempt to avoid needing the INCOMPLETE flag to create an atomic swap. This means we can create a second active attribute of the same name before we remove the original. If we fail at any point after the create but before the removal has completed, we end up with duplicate attributes in the attr btree and recovery only tries to replace one of them.
There are several other failure modes where we can leave partially allocated remote attributes that expose stale data, partially free remote attributes that enable UAF based stale data exposure, etc.
TO fix this, we need a different algorithm for replace operations when LARP is enabled. Luckily, it's not that complex if we take the right first step. That is, the first thing we log is the attri intent with the new name/value pair and mark the old attr as INCOMPLETE in the same transaction.
From there, we then remove the old attr and keep relogging the new name/value in the intent, such that we always know that we have to create the new attr in recovery. Once the old attr is removed, we then run a normal
ATTR_CREATEoperation relogging the intent as we go. If the new attr is local, then it gets created in a single atomic transaction that also logs the final intent done. If the new attr is remote, the we set INCOMPLETE on the new attr while we allocate and set the remote value, and then we clear the INCOMPLETE flag at in the last transaction taht logs the final intent done.If we fail at any point in this algorithm, log recovery will always see the same state on disk: the new name/value in the intent, and either an INCOMPLETE attr or no attr in the attr btree. If we find an INCOMPLETE attr, we run the full replace starting with removing the INCOMPLETE attr. If we don't find it, then we simply create the new attr.
Notably, recovery of a failed create that has an INCOMPLETE flag set is now the same - we start with the lookup of the INCOMPLETE attr, and if that exists then we do the full replace recovery process, otherwise we just create the new attr.
Hence changing the way we do the replace operation when LARP is enabled allows us to use the same log recovery algorithm for both the
ATTR_CREATEandATTR_REPLACEoperations. This is also the same algorithm we use for runtimeATTR_REPLACEoperations (except for the step setting up the initial conditions).The result is that:
-
ATTR_CREATEuses the same algorithm regardless of whether LARP is enabled or not
-ATTR_REPLACEwith larp=0 is identical to the old algorithm
-ATTR_REPLACEwith larp=1 runs an unmodified attr removal algorithm from the larp=0 code and then runs the unmodifiedATTR_CREATEcode.
- log recovery when larp=1 runs the sameATTR_REPLACEalgorithm as it uses at runtime.
Quoted 7 lines
Because the state machine is now quite clean, changing the algorithm is really just a case of changing the initial state and how the states link together for the
ATTR_REPLACEcase. Hence it's not a huge amoutn of code for what is a fairly substantial rework of the attr logging and recovery algorithm....Signed-off-by Dave Chinner <[email protected]>
Ok, I think I was able to follow most of this through. I think it looks mostly good, I did catch one buggly down below....
Quoted 646 lines
--- fs/xfs/libxfs/xfs_attr.c | 137 ++++++++++++++----------------- -- fs/xfs/libxfs/xfs_attr.h | 101 +++++++++++++++++++----- fs/xfs/libxfs/xfs_attr_leaf.c | 22 ++++-- fs/xfs/libxfs/xfs_da_btree.h | 1 + fs/xfs/libxfs/xfs_defer.c | 2 + fs/xfs/libxfs/xfs_log_format.h | 7 +- fs/xfs/xfs_acl.c | 2 +- fs/xfs/xfs_attr_item.c | 27 ++++++- fs/xfs/xfs_ioctl.c | 2 +- fs/xfs/xfs_trace.h | 10 +-- fs/xfs/xfs_xattr.c | 2 +- 11 files changed, 197 insertions(+), 116 deletions(-)fs/xfs/libxfs/xfs_attr.c +0
-2@@ -260,19 +260,6 @@ xfs_attr_try_sf_addname(return error; } -/* -* Check to see if the attr should be upgraded from non-existent or shortform to
- * single-leaf-block attribute list.
- */fs/xfs/libxfs/xfs_attr.c +0
-10-static inline bool -xfs_attr_is_shortform ( -struct xfs_inode *ip) -{ -return ip->i_afp->if_format ==XFS_DINODE_FMT_LOCAL || - (ip->i_afp->if_format ==XFS_DINODE_FMT_EXTENTS && - ip->i_afp->if_nextents ==0 ); -} -static int xfs_attr_sf_addname (struct xfs_attr_item *attr)@@ -309,20 +296,29 @@ xfs_attr_sf_addname( }/* -* When we bump the state to REPLACE, we may actually need to skip over the
- * state. When LARP mode is enabled, we don't need to run the atomic flags flip,
- * so we skip straight over the REPLACE state and go on toREMOVE_OLD.
+ * Handle the state change on completion of a multi-state attr operation.
+ *
+ * If theXFS_DA_OP_RENAMEflag is set, this means the operation was the first
+ * modification in a attr replace operation and we still have to do the second
+ * state, indicated by @replace_state.
+ *
+ * We consume theXFS_DA_OP_RENAMEflag so that when we are called again on
+ * completion of the second half of the attr replace operation we correctly
+ * signal that it is done.
*/fs/xfs/libxfs/xfs_attr.c +4
-3-static void -xfs_attr_dela_state_set_replace ( +static enum xfs_delattr_state +xfs_attr_complete_op (struct xfs_attr_item *attr, -enum xfs_delattr_state replace) +enum xfs_delattr_state replace_state) {struct xfs_da_args *args = attr->xattri_da_args; + bool do_replace = args->op_flags &
XFS_DA_OP_RENAME;fs/xfs/libxfs/xfs_attr.c +7
-6- attr->xattri_dela_state = replace; -if (xfs_has_larp (args->dp->i_mount)) - attr->xattri_dela_state++; + args->attr_flags &= ~XATTR_REPLACE ; + args->op_flags &= ~XFS_DA_OP_RENAME ; +if (do_replace) +return replace_state; +return XFS_DAS_DONE ; } static int@@ -364,10 +360,9 @@ xfs_attr_leaf_addname(*/ if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_LEAF_SET_RMT ; -else if (args->op_flags &XFS_DA_OP_RENAME ) -xfs_attr_dela_state_set_replace (attr, XFS_DAS_LEAF_REPLACE); else - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state = xfs_attr_complete_op(attr, +XFS_DAS_LEAF_RE PLACE); out:trace_xfs_attr_leaf_addname_return(attr->xattri_dela_state,
args->dp);fs/xfs/libxfs/xfs_attr.c +2
-3return error;@@ -409,10 +404,9 @@ xfs_attr_node_addname(if (args->rmtblkno) attr->xattri_dela_state =XFS_DAS_NODE_SET_RMT ; -else if (args->op_flags &XFS_DA_OP_RENAME ) -xfs_attr_dela_state_set_replace (attr, XFS_DAS_NODE_REPLACE); else - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state = xfs_attr_complete_op(attr, +XFS_DAS_NODE_RE PLACE); out:trace_xfs_attr_node_addname_return(attr->xattri_dela_state,
args->dp);fs/xfs/libxfs/xfs_attr.c +0
-1return error;@@ -442,18 +436,15 @@ xfs_attr_rmtval_alloc(if (error)return error; -/* If this is not a rename, clear the incomplete flag and we're done. */
fs/xfs/libxfs/xfs_attr.c +2
-1-if (!(args->op_flags &XFS_DA_OP_RENAME )) { + attr->xattri_dela_state = xfs_attr_complete_op(attr, + ++attr-Quoted 1 line
xattri_dela_state);fs/xfs/libxfs/xfs_attr.c +2
-0+/* +* If we are not doing a rename, we've finished the operation but still
fs/xfs/libxfs/xfs_attr.c +1
-0+* have to clear the incomplete flag protecting the new attr from
fs/xfs/libxfs/xfs_attr.c +1
-0+* exposing partially initialised state if we crash during creation.
fs/xfs/libxfs/xfs_attr.c +2
-4+*/ +if (attr->xattri_dela_state ==XFS_DAS_DONE ) error =xfs_attr3_leaf_clearflag (args); - attr->xattri_dela_state =XFS_DAS_DONE ; - }else { -/* -* We are running a REPLACE operation, so we need to bump the
fs/xfs/libxfs/xfs_attr.c +0
-4-* state to the step in that operation. -*/ - attr->xattri_dela_state++; - xfs_attr_dela_state_set_replace(attr, attr-Quoted 1 line
xattri_dela_state);fs/xfs/libxfs/xfs_attr.c +8
-5- } out:trace_xfs_attr_rmtval_alloc (attr->xattri_dela_state, args->dp);return error;@@ -578,11 +569,15 @@ xfs_attr_set_iter(return xfs_attr_node_addname (attr);case XFS_DAS_SF_REMOVE : - attr->xattri_dela_state =XFS_DAS_DONE ; -return xfs_attr_sf_removename (args); + error =xfs_attr_sf_removename (args); + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args)); +break ;case XFS_DAS_LEAF_REMOVE : - attr->xattri_dela_state =XFS_DAS_DONE ; -return xfs_attr_leaf_removename (args); + error =xfs_attr_leaf_removename (args); + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args)); +break ;case XFS_DAS_NODE_REMOVE :fs/xfs/libxfs/xfs_attr.c +4
-2error =xfs_attr_node_removename_setup (attr);if (error)@@ -678,12 +673,14 @@ xfs_attr_set_iter(case XFS_DAS_LEAF_REMOVE_ATTR : error =xfs_attr_leaf_remove_attr (attr); - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args));break ;case XFS_DAS_NODE_REMOVE_ATTR : error =xfs_attr_node_remove_attr (attr); - attr->xattri_dela_state =XFS_DAS_DONE ; + attr->xattri_dela_state =xfs_attr_complete_op (attr, +xfs_attr_init_add_state (args));break ;fs/xfs/libxfs/xfs_attr.c +2
-7default :ASSERT (0 );@@ -753,14 +750,9 @@ xfs_attr_defer_add(if (error)return error; -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_ADD ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_ADD ; -else - new->xattri_dela_state =XFS_DAS_NODE_ADD ; - + new->xattri_dela_state =xfs_attr_init_add_state (args);xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new- >xattri_list); +trace_xfs_attr_defer_add (new->xattri_dela_state, args->dp);return 0 ;}
fs/xfs/libxfs/xfs_attr.c +3
-8@@ -773,18 +765,13 @@ xfs_attr_defer_replace(struct xfs_attr_item *new;int error =0 ; - error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_SET , &new); + error =xfs_attr_item_init (args,XFS_ATTR_OP_FLAGS_REPLACE , &new);if (error)return error; -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_ADD ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_ADD ; -else - new->xattri_dela_state =XFS_DAS_NODE_ADD ; - + new->xattri_dela_state =xfs_attr_init_replace_state (args);xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new- >xattri_list); +trace_xfs_attr_defer_replace (new->xattri_dela_state, args->dp);fs/xfs/libxfs/xfs_attr.c +2
-7return 0 ; }@@ -802,14 +789,9 @@ xfs_attr_defer_remove(if (error)return error; -if (xfs_attr_is_shortform (args->dp)) - new->xattri_dela_state =XFS_DAS_SF_REMOVE ; -else if (xfs_attr_is_leaf (args->dp)) - new->xattri_dela_state =XFS_DAS_LEAF_REMOVE ; -else - new->xattri_dela_state =XFS_DAS_NODE_REMOVE ; - + new->xattri_dela_state =xfs_attr_init_remove_state (args);xfs_defer_add (args->trans,XFS_DEFER_OPS_TYPE_ATTR , &new- >xattri_list); +trace_xfs_attr_defer_remove (new->xattri_dela_state, args->dp);return 0 ;}
fs/xfs/libxfs/xfs_attr.c +1
-3@@ -1032,6 +1014,7 @@ xfs_attr_shortform_addname(xfs_da_args_t *args)* not being around. */ args->attr_flags &= ~XATTR_REPLACE ; + args->op_flags &= ~XFS_DA_OP_RENAME ; } if (args->namelen >= XFS_ATTR_SF_ENTSIZE_MAX ||@@ -1122,16 +1105,14 @@ xfs_attr_leaf_try_add( goto out_brelse;trace_xfs_attr_leaf_replace (args); - -/* save the attribute state for later removal*/ - args->op_flags |=XFS_DA_OP_RENAME ;/* an atomic rename */
fs/xfs/libxfs/xfs_attr.c +1
-2-xfs_attr_save_rmt_blk (args); +ASSERT (args->op_flags &XFS_DA_OP_RENAME );/* -* clear the remote attr state now that it is saved so that the
fs/xfs/libxfs/xfs_attr.c +0
-1-* values reflect the state of the attribute we are about to
fs/xfs/libxfs/xfs_attr.c +1
-0+* Save the existing remote attr state so that the current
fs/xfs/libxfs/xfs_attr.c +1
-0+* values reflect the state of the new attribute we are about to
* add, not the attribute we just found and will remove
later.fs/xfs/libxfs/xfs_attr.c +1
-3*/ +xfs_attr_save_rmt_blk (args); args->rmtblkno =0 ; args->rmtblkcnt =0 ; args->rmtvaluelen =0 ;@@ -1293,16 +1274,14 @@ xfs_attr_node_addname_find_attr( goto error;trace_xfs_attr_node_replace (args); - -/* save the attribute state for later removal*/ - args->op_flags |=XFS_DA_OP_RENAME ;/* atomic rename op */
fs/xfs/libxfs/xfs_attr.c +1
-2-xfs_attr_save_rmt_blk (args); +ASSERT (args->op_flags &XFS_DA_OP_RENAME );/* -* clear the remote attr state now that it is saved so that the
fs/xfs/libxfs/xfs_attr.c +0
-1-* values reflect the state of the attribute we are about to
fs/xfs/libxfs/xfs_attr.c +1
-0+* Save the existing remote attr state so that the current
fs/xfs/libxfs/xfs_attr.c +1
-0+* values reflect the state of the new attribute we are about to
* add, not the attribute we just found and will remove
later.fs/xfs/libxfs/xfs_attr.c +1
-0*/ +xfs_attr_save_rmt_blk (args); args->rmtblkno =0 ; args->rmtblkcnt =0 ; args->rmtvaluelen =0 ;fs/xfs/libxfs/xfs_attr.h +0
-1@@ -444,18 +444,23 @@ struct xfs_attr_list_context {*/ enum xfs_delattr_state {XFS_DAS_UNINIT =0 ,/* No state has been set yet */ -XFS_DAS_SF_ADD ,/* Initial shortform set iter state */
fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_LEAF_ADD, /* Initial leaf form set iter state*/
fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_NODE_ADD, /* Initial node form set iter state*/
fs/xfs/libxfs/xfs_attr.h +0
-3-XFS_DAS_RMTBLK ,/* Removing remote blks */ -XFS_DAS_RM_NAME ,/* Remove attr name */ - XFS_DAS_RM_SHRINK, /* We are shrinking the tree*/
fs/xfs/libxfs/xfs_attr.h +0
-2- - XFS_DAS_SF_REMOVE, /* Initial shortform set iterstate */
fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_LEAF_REMOVE, /* Initial leaf form set iterstate */
fs/xfs/libxfs/xfs_attr.h +0
-1- XFS_DAS_NODE_REMOVE, /* Initial node form set iterstate */
fs/xfs/libxfs/xfs_attr.h +3
-2- -/* Leaf state set/replace sequence */ + +/* +* Initial sequence states. The replace setup code relies on the
fs/xfs/libxfs/xfs_attr.h +1
-0+* ADD and REMOVE states for a specific format to be sequential so
fs/xfs/libxfs/xfs_attr.h +5
-0+* that we can transform the initial operation to be performed +* according to the xfs_has_larp() state easily. +*/ +XFS_DAS_SF_ADD ,/* Initial sf add state */ + XFS_DAS_SF_REMOVE, /* Initial sf replace/removestate */
fs/xfs/libxfs/xfs_attr.h +3
-0+ +XFS_DAS_LEAF_ADD ,/* Initial leaf add state */ + XFS_DAS_LEAF_REMOVE, /* Initial leafreplace/remove state */
fs/xfs/libxfs/xfs_attr.h +3
-0+ +XFS_DAS_NODE_ADD ,/* Initial node add state */ + XFS_DAS_NODE_REMOVE, /* Initial nodereplace/remove state */
fs/xfs/libxfs/xfs_attr.h +2
-0+ +/* Leaf state set/replace/remove sequence */ XFS_DAS_LEAF_SET_RMT, /* set a remote xattr from aleaf */
XFS_DAS_LEAF_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_LEAF_REPLACE, /* Perform replace ops on a
leaf */fs/xfs/libxfs/xfs_attr.h +1
-1@@ -463,7 +468,7 @@ enum xfs_delattr_state {XFS_DAS_LEAF_REMOVE_RMT, /* A rename is removing remote blocks */ XFS_DAS_LEAF_REMOVE_ATTR ,/* Remove the old attr from a leaf */ -/* Node state set/replace sequence, must match leaf state above */ +/* Node state sequence, must match leaf state above */ XFS_DAS_NODE_SET_RMT, /* set a remote xattr from anode */
XFS_DAS_NODE_ALLOC_RMT, /* We are allocating remote
blocks */
XFS_DAS_NODE_REPLACE, /* Perform replace ops on a
node */fs/xfs/libxfs/xfs_attr.h +6
-5@@ -477,11 +482,11 @@ enum xfs_delattr_state {#define XFS_DAS_STRINGS \ { XFS_DAS_UNINIT, "XFS_DAS_UNINIT" }, \ { XFS_DAS_SF_ADD, "XFS_DAS_SF_ADD" }, \ + { XFS_DAS_SF_REMOVE, "XFS_DAS_SF_REMOVE" }, \ { XFS_DAS_LEAF_ADD, "XFS_DAS_LEAF_ADD" }, \ + { XFS_DAS_LEAF_REMOVE, "XFS_DAS_LEAF_REMOVE" }, \ { XFS_DAS_NODE_ADD, "XFS_DAS_NODE_ADD" }, \ - { XFS_DAS_RMTBLK, "XFS_DAS_RMTBLK" }, \ - { XFS_DAS_RM_NAME, "XFS_DAS_RM_NAME" }, \ - { XFS_DAS_RM_SHRINK, "XFS_DAS_RM_SHRINK" }, \ + { XFS_DAS_NODE_REMOVE, "XFS_DAS_NODE_REMOVE" }, \ { XFS_DAS_LEAF_SET_RMT, "XFS_DAS_LEAF_SET_RMT" }, \ { XFS_DAS_LEAF_ALLOC_RMT, "XFS_DAS_LEAF_ALLOC_RMT" }, \ { XFS_DAS_LEAF_REPLACE, "XFS_DAS_LEAF_REPLACE" }, \@@ -525,8 +530,7 @@ struct xfs_attr_item {enum xfs_delattr_state xattri_dela_state;/* -* Indicates if the attr operation is a set or a remove -* XFS_ATTR_OP_FLAGS_{SET,REMOVE} +* Attr operation being performed - XFS_ATTR_OP_FLAGS_* */ unsigned int xattri_op_flags;@@ -568,4 +572,65 @@ void xfs_attri_destroy_cache(void);int __init xfs_attrd_init_cache (void );void xfs_attrd_destroy_cache (void ); +/* +* Check to see if the attr should be upgraded from non-existent or shortform to
+ * single-leaf-block attribute list.
+ */fs/xfs/libxfs/xfs_attr.h +15
-0+static inline bool +xfs_attr_is_shortform ( +struct xfs_inode *ip) +{ +return ip->i_afp->if_format ==XFS_DINODE_FMT_LOCAL || + (ip->i_afp->if_format ==XFS_DINODE_FMT_EXTENTS && + ip->i_afp->if_nextents ==0 ); +} + +static inline enum xfs_delattr_state +xfs_attr_init_add_state (struct xfs_da_args *args) +{ +/* +* On a pure remove, we can remove the attr fork so when we are +* called to set the add state after the remove to set up the next
fs/xfs/libxfs/xfs_attr.h +1
-0+* replace state, we can hit this case here. However, we will not see
fs/xfs/libxfs/xfs_attr.h +1
-0+* an empty fork on logged replace operation, so if there is no attr
fs/xfs/libxfs/xfs_attr.h +26
-0+* fork then we are done here. +* +* XXX: The replace state changeover needs a bit of rework to +* avoid this quirk. +*/ +if (!args->dp->i_afp) +return XFS_DAS_DONE ; +if (xfs_attr_is_shortform (args->dp)) +return XFS_DAS_SF_ADD ; +if (xfs_attr_is_leaf (args->dp)) +return XFS_DAS_LEAF_ADD ; +return XFS_DAS_NODE_ADD ; +} + +static inline enum xfs_delattr_state +xfs_attr_init_remove_state (struct xfs_da_args *args) +{ +if (xfs_attr_is_shortform (args->dp)) +return XFS_DAS_SF_REMOVE ; +if (xfs_attr_is_leaf (args->dp)) +return XFS_DAS_LEAF_REMOVE ; +return XFS_DAS_NODE_REMOVE ; +} + +/* +* If we are logging the attributes, then we have to start with removal of the
+ * old attribute so that there is always consistent state that we can recover
+ * from if the system goes down part way through. We always log the new attr
+ * value, so even when we remove the attr first we still have the information in
+ * the log to finish the replace operation atomically.
+ */fs/xfs/libxfs/xfs_attr.h +9
-0+static inline enum xfs_delattr_state +xfs_attr_init_replace_state (struct xfs_da_args *args) +{ + args->op_flags |=XFS_DA_OP_RENAME ; +if (xfs_has_larp (args->dp->i_mount)) +return xfs_attr_init_remove_state (args); +return xfs_attr_init_add_state (args); +} +#endif /* __XFS_ATTR_H__ */ fs/xfs/libxfs/xfs_attr_leaf.c +0
-0@@ -446,6 +446,14 @@ xfs_attr3_leaf_read( * Namespace helper routines *========================================================================*/
+/*
+ * If we are in log recovery, then we want the lookup to ignore the INCOMPLETE
+ * flag on disk - if there's an incomplete attr then recovery needs to tear it
+ * down. If there's no incomplete attr, then recovery needs to tear that attr
+ * down to replace it with the attr that has been logged. In this case, the
+ * INCOMPLETE flag will not be set inattr->attr_filter, but rather
+ *XFS_DA_OP_RECOVERYwill be set inargs->op_flags.
+ */
static bool
xfs_attr_match(fs/xfs/libxfs/xfs_attr_leaf.c +1
-2struct xfs_da_args *args,@@ -453,14 +461,18 @@ xfs_attr_match( unsigned char *name, int flags) { +if (args->namelen != namelen)return false;if (memcmp (args->name, name, namelen) !=0 )return false; -/* -* If we are looking for incomplete entries, show only those, else only
fs/xfs/libxfs/xfs_attr_leaf.c +7
-2-* show complete entries. -*/ + +/* Recovery ignores the INCOMPLETE flag. */ +if ((args->op_flags &XFS_DA_OP_RECOVERY ) && + args->attr_filter == (flags &XFS_ATTR_NSP_ONDISK_MASK )) +return true; + +/* All remaining matches need to be filtered by INCOMPLETE state. */
fs/xfs/libxfs/xfs_attr_leaf.c +0
-0if (args->attr_filter != (flags & (XFS_ATTR_NSP_ONDISK_MASK |XFS_ATTR_INCOMPLETE )))return false;@@ -819,7 +831,7 @@ xfs_attr_sf_removename( totsize -= size; if (totsize == sizeof(xfs_attr_sf_hdr_t) && xfs_has_attr2(mp)&&
fs/xfs/libxfs/xfs_attr_leaf.c +1
-1(dp->i_df.if_format !=XFS_DINODE_FMT_BTREE ) && - !(args->op_flags & XFS_DA_OP_ADDNAME)) { + !(args->op_flags & (XFS_DA_OP_ADDNAME | XFS_DA_OP_RENAME))){
fs/xfs/libxfs/xfs_attr_leaf.c +0
-0xfs_attr_fork_remove (dp, args->trans); }else {xfs_idata_realloc (dp, -size,XFS_ATTR_FORK );fs/xfs/libxfs/xfs_da_btree.h +0
-0@@ -90,6 +90,7 @@ typedef struct xfs_da_args {#define XFS_DA_OP_OKNOENT (1u << 3)/* lookup op, ENOENT ok, else die */
#defineXFS_DA_OP_CILOOKUP(1u << 4) /* lookup returns CI name if
found */
#defineXFS_DA_OP_NOTIME(1u << 5) /* don't update inode
timestamps */
+#defineXFS_DA_OP_RECOVERY(1u << 6) /* running a recovery
operation */#define
XFS_DA_OP_FLAGS\fs/xfs/libxfs/xfs_da_btree.h +0
-0{ XFS_DA_OP_JUSTCHECK, "JUSTCHECK" }, \fs/xfs/libxfs/xfs_defer.c +2
-0@@ -23,6 +23,8 @@#include "xfs_bmap.h" #include "xfs_alloc.h" #include "xfs_buf.h" +#include "xfs_da_format.h" +#include "xfs_da_btree.h" #include "xfs_attr.h" static struct kmem_cache *xfs_defer_pending_cache;fs/xfs/libxfs/xfs_log_format.h +1
-3@@ -906,9 +906,10 @@ struct xfs_icreate_log {* Flags for deferred attribute operations. * Upper bits are flags, lower byte is type code */ -#define XFS_ATTR_OP_FLAGS_SET 1/* Set the attribute */ -#define XFS_ATTR_OP_FLAGS_REMOVE 2/* Remove the attribute */ -#define XFS_ATTR_OP_FLAGS_TYPE_MASK 0xFF/* Flags type mask */ +#define XFS_ATTR_OP_FLAGS_SET (1u << 0)/* Set the attribute */
+#defineXFS_ATTR_OP_FLAGS_REMOVE(1u << 1) /* Remove the
attribute */
+#defineXFS_ATTR_OP_FLAGS_REPLACE(1u << 2) /* Replace the
attribute */
+#defineXFS_ATTR_OP_FLAGS_TYPE_MASK0xFF/* Flags type mask
*//*
* This is the structure used to lay out an attr log item in thefs/xfs/xfs_acl.c +1
-1@@ -10,12 +10,12 @@#include "xfs_trans_resv.h" #include "xfs_mount.h" #include "xfs_inode.h" -#include "xfs_attr.h" #include "xfs_trace.h" #include "xfs_error.h" #include "xfs_acl.h" #include "xfs_da_format.h" #include "xfs_da_btree.h" +#include "xfs_attr.h" #include "xfs_trans.h" #include <linux/posix_acl_xattr.h> fs/xfs/xfs_attr_item.c +7
-2@@ -490,9 +490,14 @@ xfs_attri_validate(if (attrp->__pad !=0 )return false; -/* alfi_op_flags should be either a set or remove */ -if (op !=XFS_ATTR_OP_FLAGS_SET && op != XFS_ATTR_OP_FLAGS_REMOVE) +switch (op) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REMOVE : +case XFS_ATTR_OP_FLAGS_REPLACE : +break ; +default :return false; + }if (attrp->alfi_value_len >XATTR_SIZE_MAX )fs/xfs/xfs_attr_item.c +0
-0return false;@@ -553,11 +558,27 @@ xfs_attri_item_recover( args->namelen = attrp->alfi_name_len; args->hashval =xfs_da_hashname (args->name, args->namelen); args->attr_filter = attrp->alfi_attr_flags;
fs/xfs/xfs_attr_item.c xfs_attri_item_recover( · line 569
Quoted 10 lines
fs/xfs/xfs_attr_item.c +5
-1+ args->op_flags =XFS_DA_OP_RECOVERY ; -if (attrp->alfi_op_flags ==XFS_ATTR_OP_FLAGS_SET ) { +switch (attr->xattri_op_flags) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REPLACE : args->value = attrip->attri_value; args->valuelen = attrp->alfi_value_len; args->total =xfs_attr_calc_size (args, &local); +if (xfs_inode_hasattr (args->dp))
I ran into a test failure and tracked it down to the above line. I suspect because xfs_inode_hasattr only checks to see if the inode has an attr fork, it doesnt actually check to see if it has the attr we're replacing. So we fall into the replace code path where it should have been the set code path. If I replace it with the below line, the failure is resolved:
if (attr->xattri_op_flags == XFS_ATTR_OP_FLAGS_REPLACE)
I only encountered this bug after running with parent pointers though which generates a lot more activity, but I figure it's not a bad idea to catch things early. There's one more test failure it's picking up, though I havnt figured out the cause just yet.
The rest looks good though, I see the below lines address the issue of the states needing to be initialized in the replay paths, so that addresses the concerns in patches 4 and 13. Thanks for all the larp improvements!
Allison
Quoted 83 lines
fs/xfs/xfs_attr_item.c +12
-0+ attr->xattri_dela_state =xfs_attr_init_replace_state (args); + else + attr->xattri_dela_state =xfs_attr_init_add_state (args); +break ; +case XFS_ATTR_OP_FLAGS_REMOVE : +if (!xfs_inode_hasattr (args->dp)) + goto out; + attr->xattri_dela_state =xfs_attr_init_remove_state (args); +break ; +default : + error = -EFSCORRUPTED ; + goto out;fs/xfs/xfs_attr_item.c +0
-0}xfs_init_attr_trans (args, &tres, &total);fs/xfs/xfs_ioctl.c +1
-1@@ -15,7 +15,6 @@#include "xfs_iwalk.h" #include "xfs_itable.h" #include "xfs_error.h" -#include "xfs_attr.h" #include "xfs_bmap.h" #include "xfs_bmap_util.h" #include "xfs_fsops.h" @@ -37,6 +36,7 @@#include "xfs_ioctl.h" #include "xfs_da_format.h" #include "xfs_da_btree.h" +#include "xfs_attr.h" #include <linux/mount.h> #include <linux/namei.h> fs/xfs/xfs_trace.h +5
-5@@ -4100,13 +4100,10 @@ DEFINE_ICLOG_EVENT(xlog_iclog_write);TRACE_DEFINE_ENUM (XFS_DAS_UNINIT );TRACE_DEFINE_ENUM (XFS_DAS_SF_ADD ); -TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ADD ); -TRACE_DEFINE_ENUM (XFS_DAS_NODE_ADD ); -TRACE_DEFINE_ENUM (XFS_DAS_RMTBLK ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_NAME ); -TRACE_DEFINE_ENUM (XFS_DAS_RM_SHRINK );TRACE_DEFINE_ENUM (XFS_DAS_SF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ADD );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_REMOVE ); +TRACE_DEFINE_ENUM (XFS_DAS_NODE_ADD );TRACE_DEFINE_ENUM (XFS_DAS_NODE_REMOVE );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_SET_RMT );TRACE_DEFINE_ENUM (XFS_DAS_LEAF_ALLOC_RMT );@@ -4142,6 +4139,9 @@ DECLARE_EVENT_CLASS(xfs_das_state_class,DEFINE_EVENT (xfs_das_state_class, name, \ TP_PROTO(int das, struct xfs_inode *ip), \ TP_ARGS(das, ip)) +DEFINE_DAS_STATE_EVENT (xfs_attr_defer_add); +DEFINE_DAS_STATE_EVENT (xfs_attr_defer_replace); +DEFINE_DAS_STATE_EVENT (xfs_attr_defer_remove);DEFINE_DAS_STATE_EVENT (xfs_attr_sf_addname_return);DEFINE_DAS_STATE_EVENT (xfs_attr_set_iter_return);DEFINE_DAS_STATE_EVENT (xfs_attr_leaf_addname_return);fs/xfs/xfs_xattr.c +1
-1@@ -12,9 +12,9 @@#include "xfs_trans_resv.h" #include "xfs_mount.h" #include "xfs_inode.h" -#include "xfs_attr.h" #include "xfs_acl.h" #include "xfs_da_btree.h" +#include "xfs_attr.h" #include <linux/posix_acl_xattr.h>
Re: [PATCH 16/16] xfs: ATTR_REPLACE algorithm with LARP enabled needs rework
Quoted 14 lines · Alli
Quoted 13 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We can't use the same algorithm for replacing an existing attribute when logging attributes. The existing algorithm is essentially:
1. create new attr w/ INCOMPLETE
2. atomically flip INCOMPLETE flags between old + new attribute
3. remove old attr which is marked w/ INCOMPLETEThis algorithm guarantees that we see either the old or new attribute, and if we fail after the atomic flag flip, we don't have to recover the removal of the old attr because we never see INCOMPLETE attributes in lookups.
....
Quoted 28 lines
Quoted 28 lines
fs/xfs/xfs_attr_item.c +7
-2@@ -490,9 +490,14 @@ xfs_attri_validate(if (attrp->__pad !=0 )return false; -/* alfi_op_flags should be either a set or remove */ -if (op !=XFS_ATTR_OP_FLAGS_SET && op != XFS_ATTR_OP_FLAGS_REMOVE) +switch (op) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REMOVE : +case XFS_ATTR_OP_FLAGS_REPLACE : +break ; +default :return false; + }if (attrp->alfi_value_len >XATTR_SIZE_MAX )fs/xfs/xfs_attr_item.c +1
-0return false;@@ -553,11 +558,27 @@ xfs_attri_item_recover( args->namelen = attrp->alfi_name_len; args->hashval =xfs_da_hashname (args->name, args->namelen); args->attr_filter = attrp->alfi_attr_flags; + args->op_flags =XFS_DA_OP_RECOVERY ;
Quoted 12 lines
fs/xfs/xfs_attr_item.c xfs_attri_item_recover( · line 569
Quoted 8 lines
fs/xfs/xfs_attr_item.c +4
-1-if (attrp->alfi_op_flags ==XFS_ATTR_OP_FLAGS_SET ) { +switch (attr->xattri_op_flags) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REPLACE : args->value = attrip->attri_value; args->valuelen = attrp->alfi_value_len; args->total =xfs_attr_calc_size (args, &local); +if (xfs_inode_hasattr (args->dp))I ran into a test failure and tracked it down to the above line. I suspect because
xfs_inode_hasattronly checks to see if the inode has an attr fork, it doesnt actually check to see if it has the attr we're replacing.
Right, that was intentional. It is based on the fact that if we are recovering a set or a replace operation, we have to remove the INCOMPLETE xattr first. However, if the attr fork is empty, there's no INCOMPLETE xattr to remove, and so we can just go straight to the set operation to create the new value.
Hmmm - what was the failure? Was it a null pointer dereference on ip->i_afp? I wonder if you hit the corner case where attr removal can remove the attr fork, and that's when it crashed and we've tried to recover from?
Oh, I think I might have missed a case there. If you look at xfs_attr_sf_removename() I added a case to avoid removing the attr fork when XFS_DA_OP_RENAME is set because we don't want to remove it when we are about to add to it again. But I didn't add the same logic to xfs_attr3_leaf_to_shortform() which can also trash the attr fork if the last attr we remove from the attr fork is larger than would fit in a sf attr fork. Hence we go straight from leaf form to no attr fork at all.
Ok, that's definitely a bug, I'll need to fix that, and it could be the cause of this issue as removing the attr fork will set forkoff to 0 and so the inode will not have an attr fork instantiated when it is read into memory...
Quoted 10 lines
So we fall into the replace code path where it should have been the set code path. If I replace it with the below line, the failure is resolved:
if (attr->xattri_op_flags == XFS_ATTR_OP_FLAGS_REPLACE)I only encountered this bug after running with parent pointers though which generates a lot more activity, but I figure it's not a bad idea to catch things early. There's one more test failure it's picking up, though I havnt figured out the cause just yet.
Yup, that's a good idea.
Quoted 4 lines
The rest looks good though, I see the below lines address the issue of the states needing to be initialized in the replay paths, so that addresses the concerns in patches 4 and 13. Thanks for all the larp improvements!
I'm going to try to move them up into patches 4 and 13, so that recovery at least tries to function correctly as we move through the patch set.
Cheers,
Dave.
Signature · 2 lines
Dave Chinner [email protected]
Re: [PATCH 16/16] xfs: ATTR_REPLACE algorithm with LARP enabled needs rework
Quoted 67 lines · Dave Chinner
Quoted 16 lines · Alli
Quoted 15 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We can't use the same algorithm for replacing an existing attribute
when logging attributes. The existing algorithm is essentially:1. create new attr w/ INCOMPLETE
2. atomically flip INCOMPLETE flags between old + new attribute
3. remove old attr which is marked w/ INCOMPLETEThis algorithm guarantees that we see either the old or new attribute, and if we fail after the atomic flag flip, we don't have
to recover the removal of the old attr because we never see INCOMPLETE attributes in lookups.....
Quoted 28 lines
Quoted 28 lines
fs/xfs/xfs_attr_item.c +7
-2@@ -490,9 +490,14 @@ xfs_attri_validate(if (attrp->__pad !=0 )return false; -/* alfi_op_flags should be either a set or remove */ -if (op !=XFS_ATTR_OP_FLAGS_SET && op != XFS_ATTR_OP_FLAGS_REMOVE) +switch (op) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REMOVE : +case XFS_ATTR_OP_FLAGS_REPLACE : +break ; +default :return false; + }if (attrp->alfi_value_len >XATTR_SIZE_MAX )fs/xfs/xfs_attr_item.c +1
-0return false;@@ -553,11 +558,27 @@ xfs_attri_item_recover( args->namelen = attrp->alfi_name_len; args->hashval =xfs_da_hashname (args->name, args->namelen); args->attr_filter = attrp->alfi_attr_flags; + args->op_flags =XFS_DA_OP_RECOVERY ;Quoted 14 lines
fs/xfs/xfs_attr_item.c xfs_attri_item_recover( · line 569
Quoted 8 lines
fs/xfs/xfs_attr_item.c +4
-1-if (attrp->alfi_op_flags ==XFS_ATTR_OP_FLAGS_SET ) { +switch (attr->xattri_op_flags) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REPLACE : args->value = attrip->attri_value; args->valuelen = attrp->alfi_value_len; args->total =xfs_attr_calc_size (args, &local); +if (xfs_inode_hasattr (args->dp))I ran into a test failure and tracked it down to the above line. I suspect because
xfs_inode_hasattronly checks to see if the inode has
an attr fork, it doesnt actually check to see if it has the attr we're
replacing.Right, that was intentional. It is based on the fact that if we are recovering a set or a replace operation, we have to remove the INCOMPLETE xattr first. However, if the attr fork is empty, there's no INCOMPLETE xattr to remove, and so we can just go straight to the set operation to create the new value.
Quoted 4 lines
Hmmm - what was the failure? Was it a null pointer dereference on
ip->i_afp? I wonder if you hit the corner case where attr removal can remove the attr fork, and that's when it crashed and we've tried to recover from?
No, the actual shutdown was from the error inject that the test case uses. The unexpected part was a set operation returning -ENODATA because we incorrectly fell down the rename path. I suspect the reason the parent pointers exposed it was because the presence of the parent pointer caused the attr fork to not be empty and so xfs_inode_hasattr succeeds.
Quoted 10 lines
Oh, I think I might have missed a case there. If you look at
xfs_attr_sf_removename()I added a case to avoid removing the attr fork whenXFS_DA_OP_RENAMEis set because we don't want to remove it when we are about to add to it again. But I didn't add the same logic toxfs_attr3_leaf_to_shortform()which can also trash the attr fork if the last attr we remove from the attr fork is larger than would fit in a sf attr fork. Hence we go straight from leaf form to no attr fork at all.
Quoted 6 lines
Ok, that's definitely a bug, I'll need to fix that, and it could be the cause of this issue as removing the attr fork will set forkoff to 0 and so the inode will not have an attr fork instantiated when it is read into memory...
Ah, that could be it then. The last failing test case is: expanding the fork into node form, setting the inject, and attempting a rename.
The correct result should be that it finds the attr correctly renamed,
but instead finds no attr. So that sounds like what you've described.
I will wait for your fix and then retest. Thx for all your help here!
Allison
Quoted 30 lines
Quoted 6 lines
So we fall into the replace code path where it should have been the set code path. If I replace it with the below line, the failure is resolved:
if (attr->xattri_op_flags == XFS_ATTR_OP_FLAGS_REPLACE)Quoted 7 lines
I only encountered this bug after running with parent pointers though
which generates a lot more activity, but I figure it's not a bad idea
to catch things early. There's one more test failure it's picking up,
though I havnt figured out the cause just yet.Yup, that's a good idea.
Quoted 6 lines
The rest looks good though, I see the below lines address the issue of
the states needing to be initialized in the replay paths, so that addresses the concerns in patches 4 and 13. Thanks for all the larp
improvements!I'm going to try to move them up into patches 4 and 13, so that recovery at least tries to function correctly as we move through the patch set.
Cheers,
Dave.
Re: [PATCH 16/16] xfs: ATTR_REPLACE algorithm with LARP enabled needs rework
Quoted 68 lines · Alli
Quoted 67 lines · Dave Chinner
Quoted 16 lines · Alli
Quoted 15 lines · Dave Chinner
From: Dave Chinner <[email protected]>
We can't use the same algorithm for replacing an existing attribute
when logging attributes. The existing algorithm is essentially:1. create new attr w/ INCOMPLETE
2. atomically flip INCOMPLETE flags between old + new attribute
3. remove old attr which is marked w/ INCOMPLETEThis algorithm guarantees that we see either the old or new attribute, and if we fail after the atomic flag flip, we don't have
to recover the removal of the old attr because we never see INCOMPLETE attributes in lookups.....
Quoted 28 lines
Quoted 28 lines
fs/xfs/xfs_attr_item.c +7
-2@@ -490,9 +490,14 @@ xfs_attri_validate(if (attrp->__pad !=0 )return false; -/* alfi_op_flags should be either a set or remove */ -if (op !=XFS_ATTR_OP_FLAGS_SET && op != XFS_ATTR_OP_FLAGS_REMOVE) +switch (op) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REMOVE : +case XFS_ATTR_OP_FLAGS_REPLACE : +break ; +default :return false; + }if (attrp->alfi_value_len >XATTR_SIZE_MAX )fs/xfs/xfs_attr_item.c +1
-0return false;@@ -553,11 +558,27 @@ xfs_attri_item_recover( args->namelen = attrp->alfi_name_len; args->hashval =xfs_da_hashname (args->name, args->namelen); args->attr_filter = attrp->alfi_attr_flags; + args->op_flags =XFS_DA_OP_RECOVERY ;Quoted 14 lines
fs/xfs/xfs_attr_item.c xfs_attri_item_recover( · line 569
Quoted 8 lines
fs/xfs/xfs_attr_item.c +4
-1-if (attrp->alfi_op_flags ==XFS_ATTR_OP_FLAGS_SET ) { +switch (attr->xattri_op_flags) { +case XFS_ATTR_OP_FLAGS_SET : +case XFS_ATTR_OP_FLAGS_REPLACE : args->value = attrip->attri_value; args->valuelen = attrp->alfi_value_len; args->total =xfs_attr_calc_size (args, &local); +if (xfs_inode_hasattr (args->dp))I ran into a test failure and tracked it down to the above line. I suspect because
xfs_inode_hasattronly checks to see if the inode has
an attr fork, it doesnt actually check to see if it has the attr we're
replacing.Right, that was intentional. It is based on the fact that if we are recovering a set or a replace operation, we have to remove the INCOMPLETE xattr first. However, if the attr fork is empty, there's no INCOMPLETE xattr to remove, and so we can just go straight to the set operation to create the new value.
Quoted 7 lines
Quoted 4 lines
Hmmm - what was the failure? Was it a null pointer dereference on
ip->i_afp? I wonder if you hit the corner case where attr removal can remove the attr fork, and that's when it crashed and we've tried to recover from?No, the actual shutdown was from the error inject that the test case uses. The unexpected part was a set operation returning
-ENODATAbecause we incorrectly fell down the rename path.
That's not correct - recovery of a set operation has to remove the INCOMPLETE attr that is was in the process of being built. If we are in recovery and we don't find an existing entry, we should just then fall through the state machine to the set operation as there was nothing to remove.
-ENOATTR/-ENODATA in that case is valid, it just sounds like we didn't find an incomplete attr to remove and didn't handle the case correctly.
Quoted 4 lines
I suspect the reason
the parent pointers exposed it was because the presence of the parent pointer caused the attr fork to not be empty and soxfs_inode_hasattrsucceeds.
I think it's the case where we fail immediately after logging the first SET intent to the journal, and haven't actually logged any other changes yet. That first intent can only be logged after the attr fork has been created and logged, so -ENODATA is a case recovery is supposed to handle for a SET operation.
Quoted 9 lines
Quoted 9 lines
Oh, I think I might have missed a case there. If you look at
xfs_attr_sf_removename()I added a case to avoid removing the attr fork whenXFS_DA_OP_RENAMEis set because we don't want to remove it when we are about to add to it again. But I didn't add the same logic toxfs_attr3_leaf_to_shortform()which can also trash the attr fork if the last attr we remove from the attr fork is larger than would fit in a sf attr fork. Hence we go straight from leaf form to no attr fork at all.
Quoted 8 lines
Quoted 6 lines
Ok, that's definitely a bug, I'll need to fix that, and it could be the cause of this issue as removing the attr fork will set forkoff to 0 and so the inode will not have an attr fork instantiated when it is read into memory...
Ah, that could be it then. The last failing test case is: expanding the fork into node form, setting the inject, and attempting a rename.
Ah, so that's likely how we get the situation I suggested above - we commit the node form expansion without other modifications but include the SET intent, so if we recover that node form transformation, we most definitely have a SET intent without an INCOMPLETE attr to remove...
I'll add it to the list.
Cheers,
Dave.
Signature · 2 lines
Dave Chinner [email protected]