On 1/22/20 9:57 AM, Allison Collins wrote:
On 1/21/20 5:12 PM, Darrick J. Wong wrote:On Sat, Jan 18, 2020 at 03:50:35PM -0700, Allison Collins wrote:This patch modifies the attr set routines to be delay ready. This means they no longer roll or commit transactions, but instead return -EAGAIN to have the calling routine roll and refresh the transaction. In this series, xfs_attr_set_args has become xfs_attr_set_iter, which uses a state machine like switch to keep track of where it was when EAGAIN was returned. Part of xfs_attr_leaf_addname has been factored out into a new helper function xfs_attr_leaf_try_add to allow transaction cycling between the two routines. Two new helper functions have been added: xfs_attr_rmtval_set_init and xfs_attr_rmtval_set_blk. They provide a subset of logic similar to xfs_attr_rmtval_set, but they store the current block in the delay attr context to allow the caller to roll the transaction between allocations. This helps to simplify and consolidate code used by xfs_attr_leaf_addname and xfs_attr_node_addname Finally, xfs_attr_set_args has become a simple loop to refresh the transaction until the operation is completed. Signed-off-by: Allison Collins <allison.henderson@xxxxxxxxxx> ---fs/xfs/libxfs/xfs_attr.c | 347 +++++++++++++++++++++++++---------------fs/xfs/libxfs/xfs_attr.h | 1 + fs/xfs/libxfs/xfs_attr_remote.c | 67 +++++++- fs/xfs/libxfs/xfs_attr_remote.h | 4 + fs/xfs/libxfs/xfs_da_btree.h | 13 ++ 5 files changed, 301 insertions(+), 131 deletions(-) diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c index 92929ad..02a7f55 100644 --- a/fs/xfs/libxfs/xfs_attr.c +++ b/fs/xfs/libxfs/xfs_attr.c @@ -58,6 +58,7 @@ 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); STATIC int xfs_attr_refillstate(xfs_da_state_t *state);+STATIC int xfs_attr_leaf_try_add(struct xfs_da_args *args, struct xfs_buf *bp);STATIC int @@ -259,9 +260,86 @@ int xfs_attr_set_args( struct xfs_da_args *args) { + int error = 0; + int err2 = 0; + struct xfs_buf *leaf_bp = NULL; + + do { + error = xfs_attr_set_iter(args, &leaf_bp); + if (error && error != -EAGAIN) + goto out; + + if (args->dac.flags & XFS_DAC_FINISH_TRANS) { + args->dac.flags &= ~XFS_DAC_FINISH_TRANS; + + err2 = xfs_defer_finish(&args->trans); + if (err2) { + error = err2; + goto out; + } + } + + err2 = xfs_trans_roll_inode(&args->trans, args->dp); + if (err2) { + error = err2; + goto out; + } + + if (leaf_bp) { + xfs_trans_bjoin(args->trans, leaf_bp); + xfs_trans_bhold(args->trans, leaf_bp); + } + + } while (error == -EAGAIN); + +out: + return error; +} + +/* + * 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. + */ +int +xfs_attr_set_iter( + struct xfs_da_args *args, + struct xfs_buf **leaf_bp) +{ struct xfs_inode *dp = args->dp; - struct xfs_buf *leaf_bp = NULL; - int error, error2 = 0; + int error = 0; + int sf_size; + + /* State machine switch */ + switch (args->dac.dela_state) { + case XFS_DAS_SF_TO_LEAF: + goto sf_to_leaf; + case XFS_DAS_ALLOC_LEAF: + case XFS_DAS_FLIP_LFLAG: + case XFS_DAS_FOUND_LBLK: + goto leaf; + case XFS_DAS_FOUND_NBLK: + case XFS_DAS_FLIP_NFLAG: + case XFS_DAS_ALLOC_NODE: + case XFS_DAS_LEAF_TO_NODE: + goto node; + default:Er... do these need to ASSERT on invalid states? e.g. if somehow we end up with XFS_DAS_RM_SHRINK, we ought to blow up loudly, right?+ break; + } + + /*+ * New inodes may not have an attribute fork yet. So set the attribute+ * fork appropriately + */ + if (XFS_IFORK_Q((args->dp)) == 0) { + sf_size = sizeof(struct xfs_attr_sf_hdr) ++ XFS_ATTR_SF_ENTSIZE_BYNAME(args->name.len, args->valuelen);+ xfs_bmap_set_attrforkoff(args->dp, sf_size, NULL); + args->dp->i_afp = kmem_zone_zalloc(xfs_ifork_zone, 0); + args->dp->i_afp->if_flags = XFS_IFEXTENTS; + } /** If the attribute list is non-existent or a shortform list, proceed to@@ -272,22 +350,20 @@ xfs_attr_set_args( (dp->i_d.di_aformat == XFS_DINODE_FMT_EXTENTS && dp->i_d.di_anextents == 0))) goto sf_to_leaf; - /* * Try to add the attr to the attribute list in the inode. */ error = xfs_attr_try_sf_addname(dp, args); - if (error != -ENOSPC) { - error2 = xfs_trans_commit(args->trans); - args->trans = NULL; - return error ? error : error2;Where does this commit go?- } + + /* Should only be 0, -EEXIST or ENOSPC */ + if (error != -ENOSPC) + return error; /* * It won't fit in the shortform, transform to a leaf block. * GROT: another possible req'mt for a double-split btree op. */ - error = xfs_attr_shortform_to_leaf(args, &leaf_bp); + error = xfs_attr_shortform_to_leaf(args, leaf_bp); if (error) return error; @@ -295,42 +371,47 @@ xfs_attr_set_args( * Prevent the leaf buffer from being unlocked so that a * concurrent AIL push cannot grab the half-baked leaf * buffer and run into problems with the write verifier. - * Once we're done rolling the transaction we can release - * the hold and add the attr to the leaf. */ - xfs_trans_bhold(args->trans, leaf_bp); - error = xfs_defer_finish(&args->trans); - xfs_trans_bhold_release(args->trans, leaf_bp); - if (error) { - xfs_trans_brelse(args->trans, leaf_bp); - return error; - } + xfs_trans_bhold(args->trans, *leaf_bp); + args->dac.flags |= XFS_DAC_FINISH_TRANS; + args->dac.dela_state = XFS_DAS_SF_TO_LEAF; + return -EAGAIN; sf_to_leaf: - if (xfs_bmap_one_block(dp, XFS_ATTR_FORK)) { - error = xfs_attr_leaf_addname(args); - if (error != -ENOSPC) - return error; - - /* - * Commit that transaction so that the node_addname() - * call can manage its own transactions. - */ - error = xfs_defer_finish(&args->trans); - if (error) - return error; + /*+ * After a shortform to leaf conversion, we need to hold the leaf and+ * cylce out the transaction. When we get back, we need to release + * the leaf. + */ + if (*leaf_bp != NULL) { + xfs_trans_brelse(args->trans, *leaf_bp); + *leaf_bp = NULL; + } - /* - * Commit the current trans (including the inode) and - * start a new one. - */ - error = xfs_trans_roll_inode(&args->trans, dp); - if (error) + if (xfs_bmap_one_block(dp, XFS_ATTR_FORK)) { + error = xfs_attr_leaf_try_add(args, *leaf_bp); + switch (error) { + case -ENOSPC: + args->dac.flags |= XFS_DAC_FINISH_TRANS; + args->dac.dela_state = XFS_DAS_LEAF_TO_NODE; + return -EAGAIN; + case 0: + args->dac.dela_state = XFS_DAS_FOUND_LBLK; + return -EAGAIN; + default: return error; - + } +leaf: + error = xfs_attr_leaf_addname(args); + if (error == -ENOSPC) { + args->dac.dela_state = XFS_DAS_LEAF_TO_NODE; + return -EAGAIN; + } + return error; } - + args->dac.dela_state = XFS_DAS_LEAF_TO_NODE; +node: error = xfs_attr_node_addname(args); return error; } @@ -759,27 +840,28 @@ xfs_attr_leaf_try_add( ** This leaf block cannot have a "remote" value, we only call this routine * if bmap_one_block() says there is only one block (ie: no remote blks).+ *+ * 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_leaf_addname(struct xfs_da_args *args) { - int error, forkoff; struct xfs_buf *bp = NULL; + int error, forkoff; struct xfs_inode *dp = args->dp; - trace_xfs_attr_leaf_addname(args); - - error = xfs_attr_leaf_try_add(args, bp); - if (error) - return error; - - /* - * Commit the transaction that added the attr name so that - * later routines can manage their own transactions. - */ - error = xfs_trans_roll_inode(&args->trans, dp); - if (error) - return error; + /* State machine switch */ + switch (args->dac.dela_state) { + case XFS_DAS_FLIP_LFLAG: + goto flip_flag; + case XFS_DAS_ALLOC_LEAF: + goto alloc_leaf; + default: + break; + } /* * If there was an out-of-line value, allocate the blocks we @@ -788,7 +870,28 @@ xfs_attr_leaf_addname(struct xfs_da_args *args) * maximum size of a transaction and/or hit a deadlock. */ if (args->rmtblkno > 0) { - error = xfs_attr_rmtval_set(args); + + /* Open coded xfs_attr_rmtval_set without trans handling */ + error = xfs_attr_rmtval_set_init(args); + if (error) + return error; + + /* + * Roll through the "value", allocating blocks on disk as + * required. + */ +alloc_leaf: + while (args->dac.blkcnt > 0) { + error = xfs_attr_rmtval_set_blk(args); + if (error) + return error; + + args->dac.flags |= XFS_DAC_FINISH_TRANS; + args->dac.dela_state = XFS_DAS_ALLOC_LEAF; + return -EAGAIN; + } + + error = xfs_attr_rmtval_set_value(args); if (error) return error; } @@ -807,13 +910,10 @@ xfs_attr_leaf_addname(struct xfs_da_args *args) error = xfs_attr3_leaf_flipflags(args); if (error) return error; - /* - * Commit the flag value change and start the next trans in - * series. - */ - error = xfs_trans_roll_inode(&args->trans, args->dp); - if (error) - return error; + + args->dac.dela_state = XFS_DAS_FLIP_LFLAG; + return -EAGAIN; +flip_flag: /* * Dismantle the "old" attribute/value pair by removing @@ -844,34 +944,17 @@ xfs_attr_leaf_addname(struct xfs_da_args *args) /** If the result is small enough, shrink it all into the inode.*/ - if ((forkoff = xfs_attr_shortform_allfit(bp, dp))) { + 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 */ - if (error) - return error; - error = xfs_defer_finish(&args->trans); - if (error) - return error; - } - /* - * Commit the remove and start the next trans in series. - */ - error = xfs_trans_roll_inode(&args->trans, dp); + args->dac.flags |= XFS_DAC_FINISH_TRANS; } else if (args->rmtblkno > 0) { /* * Added a "remote" value, just clear the incomplete flag. */ error = xfs_attr3_leaf_clearflag(args); - if (error) - return error; - - /* - * Commit the flag value change and start the next trans in - * series. - */ - error = xfs_trans_roll_inode(&args->trans, args->dp); } return error; } @@ -1020,16 +1103,22 @@ xfs_attr_node_hasname( ** "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_addname( struct xfs_da_args *args) { - struct xfs_da_state *state; + struct xfs_da_state *state = NULL; struct xfs_da_state_blk *blk; struct xfs_inode *dp; struct xfs_mount *mp; - int retval, error; + int retval = 0; + int error = 0; trace_xfs_attr_node_addname(args); @@ -1038,7 +1127,19 @@ xfs_attr_node_addname( */ dp = args->dp; mp = dp->i_mount; -restart: + + /* State machine switch */ + switch (args->dac.dela_state) { + case XFS_DAS_FLIP_NFLAG: + goto flip_flag; + case XFS_DAS_FOUND_NBLK: + goto found_nblk; + case XFS_DAS_ALLOC_NODE: + goto alloc_node; + default: + break; + } + /* * Search to see if name already exists, and get back a pointer * to where it should go. @@ -1088,19 +1189,13 @@ xfs_attr_node_addname( error = xfs_attr3_leaf_to_node(args); if (error) goto out; - error = xfs_defer_finish(&args->trans); - if (error) - goto out; /* - * Commit the node conversion and start the next - * trans in the chain. + * Restart routine from the top. No need to set the + * stateI think I missed responding to this comment in my last email:*OH*, dela_state == 0 does not mean "do not use delayed attrs", we're switching /all/ users to delayed attrs. Urk.Not quite. While both inline and delayed operations go through this same dance of EAGAIN ping pong, it's not a delayed operation until we package it into an intent and log it. And then the intent pops back out during defer finish, and the ping pong starts from xfs_defer_finish_noroll instead of xfs_attr_set_iter.
Sorry, that should say xfs_attr_set_args, not xfs_attr_set_iter. :-)
We still need to support in line operations for older versions that do not support the attr log item. Later when we get into delayed attrs, we use a feature bit to control this, and toggle as needed.Hope that helps!*/ - error = xfs_trans_roll_inode(&args->trans, dp); - if (error) - goto out; - - goto restart; + args->dac.flags |= XFS_DAC_FINISH_TRANS; + return -EAGAIN; } /* @@ -1112,9 +1207,7 @@ xfs_attr_node_addname( error = xfs_da3_split(state); if (error) goto out; - error = xfs_defer_finish(&args->trans); - if (error) - goto out; + args->dac.flags |= XFS_DAC_FINISH_TRANS; } else { /* * Addition succeeded, update Btree hashvals. @@ -1129,13 +1222,9 @@ xfs_attr_node_addname( xfs_da_state_free(state); state = NULL; - /* - * Commit the leaf addition or btree split and start the next - * trans in the chain. - */ - error = xfs_trans_roll_inode(&args->trans, dp); - if (error) - goto out; + args->dac.dela_state = XFS_DAS_FOUND_NBLK; + return -EAGAIN; +found_nblk: /* * If there was an out-of-line value, allocate the blocks we @@ -1144,7 +1233,27 @@ xfs_attr_node_addname( * maximum size of a transaction and/or hit a deadlock. */ if (args->rmtblkno > 0) { - error = xfs_attr_rmtval_set(args); + /* Open coded xfs_attr_rmtval_set without trans handling */ + error = xfs_attr_rmtval_set_init(args); + if (error) + return error; + + /* + * Roll through the "value", allocating blocks on disk as + * required. + */ +alloc_node: + while (args->dac.blkcnt > 0) { + error = xfs_attr_rmtval_set_blk(args); + if (error) + return error; + + args->dac.flags |= XFS_DAC_FINISH_TRANS; + args->dac.dela_state = XFS_DAS_ALLOC_NODE; + return -EAGAIN; + } + + error = xfs_attr_rmtval_set_value(args); if (error) return error; } @@ -1167,10 +1276,9 @@ xfs_attr_node_addname( * Commit the flag value change and start the next trans in * series */ - error = xfs_trans_roll_inode(&args->trans, args->dp); - if (error) - goto out; - + args->dac.dela_state = XFS_DAS_FLIP_NFLAG; + return -EAGAIN; +flip_flag: /* * Dismantle the "old" attribute/value pair by removing * a "remote" value (if it exists). @@ -1199,7 +1307,6 @@ xfs_attr_node_addname( error = xfs_da3_node_lookup_int(state, &retval); if (error) goto out; - /* * Remove the name and update the hashvals in the tree. */ @@ -1207,7 +1314,6 @@ xfs_attr_node_addname( ASSERT(blk->magic == XFS_ATTR_LEAF_MAGIC); error = xfs_attr3_leaf_remove(blk->bp, args); xfs_da3_fixhashpath(state, &state->path); - /* * Check to see if the tree needs to be collapsed. */ @@ -1215,18 +1321,9 @@ xfs_attr_node_addname( error = xfs_da3_join(state); if (error) goto out; - error = xfs_defer_finish(&args->trans); - if (error) - goto out; - } - - /* - * Commit and start the next trans in the chain. - */ - error = xfs_trans_roll_inode(&args->trans, dp); - if (error) - goto out; + args->dac.flags |= XFS_DAC_FINISH_TRANS; + } } else if (args->rmtblkno > 0) { /* * Added a "remote" value, just clear the incomplete flag. @@ -1234,14 +1331,6 @@ xfs_attr_node_addname( error = xfs_attr3_leaf_clearflag(args); if (error) goto out; - - /* - * Commit the flag value change and start the next trans in - * series. - */ - error = xfs_trans_roll_inode(&args->trans, args->dp); - if (error) - goto out; } retval = error = 0; diff --git a/fs/xfs/libxfs/xfs_attr.h b/fs/xfs/libxfs/xfs_attr.h index ea873a5..f450d8c 100644 --- a/fs/xfs/libxfs/xfs_attr.h +++ b/fs/xfs/libxfs/xfs_attr.h@@ -152,6 +152,7 @@ int xfs_attr_get(struct xfs_inode *ip, struct xfs_name *name,int xfs_attr_set(struct xfs_inode *dp, struct xfs_name *name, unsigned char *value, int valuelen, int flags); int xfs_attr_set_args(struct xfs_da_args *args);+int xfs_attr_set_iter(struct xfs_da_args *args, struct xfs_buf **leaf_bp); int xfs_attr_remove(struct xfs_inode *dp, struct xfs_name *name, int flags);int xfs_has_attr(struct xfs_da_args *args); int xfs_attr_remove_args(struct xfs_da_args *args);diff --git a/fs/xfs/libxfs/xfs_attr_remote.c b/fs/xfs/libxfs/xfs_attr_remote.cindex f2ee0b8..ad4f14a 100644 --- a/fs/xfs/libxfs/xfs_attr_remote.c +++ b/fs/xfs/libxfs/xfs_attr_remote.c @@ -428,7 +428,7 @@ xfs_attr_rmtval_get(* Find a "hole" in the attribute address space large enough for us to drop the* new attribute's value into */ -STATIC int +int xfs_attr_rmt_find_hole( struct xfs_da_args *args) { @@ -455,7 +455,7 @@ xfs_attr_rmt_find_hole( return 0; } -STATIC int +int xfs_attr_rmtval_set_value( struct xfs_da_args *args) { @@ -588,6 +588,69 @@ xfs_attr_rmtval_set( } /*+ * Find a hole for the attr and store it in the delayed attr context. This + * initializes the context to roll through allocating an attr extent for a+ * delayed attr operation + */ +int +xfs_attr_rmtval_set_init( + struct xfs_da_args *args) +{ + struct xfs_bmbt_irec *map = &args->dac.map; + int error; + + args->dac.lblkno = 0; + args->dac.lfileoff = 0; + args->dac.blkcnt = 0; + args->rmtblkcnt = 0; + args->rmtblkno = 0; + memset(map, 0, sizeof(struct xfs_bmbt_irec)); + + error = xfs_attr_rmt_find_hole(args); + if (error) + return error; + + args->dac.blkcnt = args->rmtblkcnt; + args->dac.lblkno = args->rmtblkno; + + return error; +} + +/* + * Write one block of the value associated with an attribute into the+ * out-of-line buffer that we have defined for it. This is similar to a subset + * of xfs_attr_rmtval_set, but records the current block to the delayed attr+ * context, and leaves transaction handling to the caller. + */ +int +xfs_attr_rmtval_set_blk( + struct xfs_da_args *args) +{ + struct xfs_inode *dp = args->dp; + struct xfs_bmbt_irec *map = &args->dac.map; + int nmap; + int error; + + nmap = 1; + error = xfs_bmapi_write(args->trans, dp, + (xfs_fileoff_t)args->dac.lblkno, + args->dac.blkcnt, XFS_BMAPI_ATTRFORK, + args->total, map, &nmap); + if (error) + return error; + + ASSERT(nmap == 1); + ASSERT((map->br_startblock != DELAYSTARTBLOCK) && + (map->br_startblock != HOLESTARTBLOCK)); + + /* roll attribute extent map forwards */ + args->dac.lblkno += map->br_blockcount; + args->dac.blkcnt -= map->br_blockcount; + + return 0; +} + +/* * Remove the value associated with an attribute by deleting the * out-of-line buffer that it is stored on. */diff --git a/fs/xfs/libxfs/xfs_attr_remote.h b/fs/xfs/libxfs/xfs_attr_remote.hindex 7ab3770..ab03519 100644 --- a/fs/xfs/libxfs/xfs_attr_remote.h +++ b/fs/xfs/libxfs/xfs_attr_remote.h @@ -13,4 +13,8 @@ int xfs_attr_rmtval_set(struct xfs_da_args *args); int xfs_attr_rmtval_remove(struct xfs_da_args *args); int xfs_attr_rmtval_invalidate(struct xfs_da_args *args); int xfs_attr_rmtval_unmap(struct xfs_da_args *args); +int xfs_attr_rmt_find_hole(struct xfs_da_args *args); +int xfs_attr_rmtval_set_value(struct xfs_da_args *args); +int xfs_attr_rmtval_set_blk(struct xfs_da_args *args); +int xfs_attr_rmtval_set_init(struct xfs_da_args *args); #endif /* __XFS_ATTR_REMOTE_H__ */ diff --git a/fs/xfs/libxfs/xfs_da_btree.h b/fs/xfs/libxfs/xfs_da_btree.h index 7fc87da..9943062 100644 --- a/fs/xfs/libxfs/xfs_da_btree.h +++ b/fs/xfs/libxfs/xfs_da_btree.h @@ -55,6 +55,14 @@ enum xfs_dacmp { enum xfs_delattr_state { XFS_DAS_RM_SHRINK = 1, /* We are shrinking the tree */ XFS_DAS_RM_NODE_BLKS = 2, /* We are removing node blocks */ + XFS_DAS_SF_TO_LEAF = 3, /* Converted short form to leaf */ + XFS_DAS_FOUND_LBLK = 4, /* We found leaf blk for attr */ + XFS_DAS_LEAF_TO_NODE = 5, /* Converted leaf to node */ + XFS_DAS_FOUND_NBLK = 6, /* We found node blk for attr */ + XFS_DAS_ALLOC_LEAF = 7, /* We are allocating leaf blocks */ + XFS_DAS_FLIP_LFLAG = 8, /* Flipped leaf INCOMPLETE attr flag */ + XFS_DAS_ALLOC_NODE = 9, /* We are allocating node blocks */ + XFS_DAS_FLIP_NFLAG = 10,/* Flipped node INCOMPLETE attr flag */We've definitely reached the point where a state diagram would be helpful. Can you go from any of the RM_ states to the ones that you've just added? The new state machine code in last two patches would be a lot easier to review if I could look down from above instead of up from the XFS_DAS values and goto labels. I /think/ it looks sane, but I'm only 20% confident of that statement. --D}; /* @@ -66,8 +74,13 @@ enum xfs_delattr_state { * Context used for keeping track of delayed attribute operations */ struct xfs_delattr_context { + struct xfs_bmbt_irec map; + struct xfs_buf *leaf_bp; + xfs_fileoff_t lfileoff; struct xfs_da_state *da_state; struct xfs_da_state_blk *blk; + xfs_dablk_t lblkno; + int blkcnt; int flags; enum xfs_delattr_state dela_state; }; -- 2.7.4
