New ->fiemap infrastructure and ->bmap removal
Subsystems (19)
v4 cover New ->fiemap infrastructure and ->bmap removal
Hi.
This is the 4th version of the complete series with the goal to deprecate and eventually remove ->bmap() interface, in lieu of FIEMAP.
Besides the rebase of the patchset against latest Linus's tree, the only changes in this patchset regarding to the previous version, are concentrated in patches 4/9 and 8/9.
In patch 4 (fibmap: Use bmap instead of ->bmap method in ioctl_fibmap), the difference compared with previous version, is a change in ioclt_fibmap() return value, I spotted while testing this new version with filesystems using data inlined in inodes. It now returns 0 in case of error instead an error value, otherwise it would be an user interface change.
In patch 8 (Use FIEMAP for FIBMAP calls), there are minor changes regarding V3. It just contains a coding style fix pointed by Andreas in the previous version, but now, it also include changes to all filesystems which supports both FIEMAP and FIBMAP, and need some sort of special handling (like inlined data, reflinked inodes, etc).
Again, Patch 9 is xfs-specific removal of ->bmap() interface, without any changes compared to the previous version.
I do apologize for taking so long to rework this patchset, I've got busy with other stuff.
Comments are appreciated, specially regarding if the error values returned by ioctl_fibmap() make sense.
Cheers
Carlos Maiolino (9):
fs: Enable bmap() function to properly return errors
cachefiles: drop direct usage of ->bmap method.
ecryptfs: drop direct calls to ->bmap
fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
fs: Move start and length fiemap fields into fiemap_extent_info
iomap: Remove length and start fields from iomap_fiemap
fiemap: Use a callback to fill fiemap extents
Use FIEMAP for FIBMAP calls
xfs: Get rid of ->bmap
drivers/md/md-bitmap.c | 16 ++++-- fs/bad_inode.c | 3 +- fs/btrfs/inode.c | 5 +- fs/cachefiles/rdwr.c | 27 ++++----- fs/ecryptfs/mmap.c | 16 ++---- fs/ext2/ext2.h | 3 +- fs/ext2/inode.c | 6 +- fs/ext4/ext4.h | 3 +- fs/ext4/extents.c | 15 +++-- fs/f2fs/data.c | 15 ++++- fs/f2fs/f2fs.h | 3 +- fs/gfs2/inode.c | 9 ++- fs/hpfs/file.c | 4 +- fs/inode.c | 105 +++++++++++++++++++++++++++++---- fs/ioctl.c | 128 ++++++++++++++++++++++++++--------------- fs/iomap.c | 40 ++----------- fs/jbd2/journal.c | 22 ++++--- fs/nilfs2/inode.c | 5 +- fs/nilfs2/nilfs.h | 3 +- fs/ocfs2/extent_map.c | 13 ++++- fs/ocfs2/extent_map.h | 3 +- fs/overlayfs/inode.c | 5 +- fs/xfs/xfs_aops.c | 24 -------- fs/xfs/xfs_iops.c | 19 +++--- fs/xfs/xfs_trace.h | 1 - include/linux/fs.h | 33 +++++++---- include/linux/iomap.h | 2 +- mm/page_io.c | 11 ++-- 28 files changed, 320 insertions(+), 219 deletions(-)
Signature · 1 line
2.20.1
PATCH v4 1/9 fs: Enable bmap() function to properly return errors
By now, bmap() will either return the physical block number related to the requested file offset or 0 in case of error or the requested offset maps into a hole.
This patch makes the needed changes to enable bmap() to proper return errors, using the return value as an error return, and now, a pointer must be passed to bmap() to be filled with the mapped physical block.
It will change the behavior of bmap() on return:
- negative value in case of error
- zero on success or map fell into a hole
In case of a hole, the *block will be zero too
Since this is a prep patch, by now, the only error return is -EINVAL if ->bmap doesn't exist.
--- drivers/md/md-bitmap.c | 16 ++++++++++------ fs/inode.c | 30 +++++++++++++++++------------- fs/jbd2/journal.c | 22 +++++++++++++++------- include/linux/fs.h | 2 +- mm/page_io.c | 11 +++++++---- 5 files changed, 50 insertions(+), 31 deletions(-)
drivers/md/md-bitmap.c +10 -6
@@ -363,7 +363,7 @@ static int read_page(struct file *file, unsigned long index,int ret =0 ;struct inode *inode =file_inode (file);struct buffer_head *bh; -sector_t block; +sector_t block, blk_cur;pr_debug ("read bitmap file (%dB @ %llu)\n" , (int )PAGE_SIZE , (unsigned long long )index <<PAGE_SHIFT );
@@ -374,17 +374,21 @@ static int read_page(struct file *file, unsigned long index, goto out; }attach_page_buffers (page, bh); - block = index << (PAGE_SHIFT - inode->i_blkbits); + blk_cur = index << (PAGE_SHIFT - inode->i_blkbits);while (bh) { + block = blk_cur; +if (count ==0 ) bh->b_blocknr =0 ;else { - bh->b_blocknr =bmap (inode, block); -if (bh->b_blocknr ==0 ) { -/* Cannot use this file! */ + ret =bmap (inode, &block); +if (ret || !block) { ret = -EINVAL ; + bh->b_blocknr =0 ; goto out; } + + bh->b_blocknr = block; bh->b_bdev = inode->i_sb->s_bdev;if (count < (1 <<inode->i_blkbits)) count =0 ;
@@ -398,7 +402,7 @@ static int read_page(struct file *file, unsigned long index,set_buffer_mapped (bh);submit_bh (REQ_OP_READ ,0 , bh); } - block++; + blk_cur++; bh = bh->b_this_page; } page->index = index;
fs/inode.c +17 -13
@@ -1577,21 +1577,25 @@ EXPORT_SYMBOL(iput);/** * bmap - find a block number in a file -* @inode: inode of file -* @block: block to find -* -* Returns the block number on the device holding the inode that -* is the disk block number for the block of the file requested. -* That is, asked for block 4 of inode 1 the function will return the -* disk block relative to the disk start that holds that block of the -* file. +* @inode: inode owning the block number being requested +* @*block: pointer containing the block to find +* +* Replaces the value in *block with the block number on the device holding +* corresponding to the requested block number in the file. +* That is, asked for block 4 of inode 1 the function will replace the +* 4 in *block, with disk block relative to the disk start that holds that +* block of the file. +* +* Returns -EINVAL in case of error, 0 otherwise. If mapping falls into a +* hole, returns 0 and *block is also set to 0. */ -sector_t bmap (struct inode *inode,sector_t block) +int bmap (struct inode *inode,sector_t *block) { -sector_t res =0 ; -if (inode->i_mapping->a_ops->bmap) - res = inode->i_mapping->a_ops->bmap (inode->i_mapping, block); -return res; +if (!inode->i_mapping->a_ops->bmap) +return -EINVAL ; + + *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, *block); +return 0 ; }EXPORT_SYMBOL (bmap);
fs/jbd2/journal.c +15 -7
@@ -798,18 +798,23 @@ int jbd2_journal_bmap(journal_t *journal, unsigned long blocknr, {int err =0 ;unsigned long long ret; +sector_t block =0 ;if (journal->j_inode) { - ret =bmap (journal->j_inode, blocknr); -if (ret) - *retp = ret; -else { + block = blocknr; + ret =bmap (journal->j_inode, &block); + +if (ret || !block) {printk (KERN_ALERT "%s: journal block not found " "at offset %lu on %s\n" , __func__, blocknr, journal->j_devname); err = -EIO ;__journal_abort_soft (journal, err); + + }else { + *retp = block; } + }else { *retp = blocknr;/* +journal->j_blk_offset */ }
@@ -1235,11 +1240,14 @@ journal_t *jbd2_journal_init_dev(struct block_device *bdev,journal_t *jbd2_journal_init_inode (struct inode *inode) {journal_t *journal; +sector_t blocknr;char *p; -unsigned long long blocknr; +int err =0 ; + + blocknr =0 ; + err =bmap (inode, &blocknr); - blocknr =bmap (inode,0 ); -if (!blocknr) { +if (err || !blocknr) {pr_err ("%s: Cannot locate journal superblock\n" , __func__);return NULL ;
include/linux/fs.h +1 -1
@@ -2803,7 +2803,7 @@ static inline ssize_t generic_write_sync(struct kiocb *iocb, ssize_t count)extern void emergency_sync (void );extern void emergency_remount (void );#ifdef CONFIG_BLOCK -extern sector_t bmap (struct inode *,sector_t ); +extern int bmap (struct inode *,sector_t *);#endif extern int notify_change (struct dentry *,struct iattr *,struct inode **);extern int inode_permission (struct inode *,int );
mm/page_io.c +7 -4
@@ -177,8 +177,9 @@ int generic_swapfile_activate(struct swap_info_struct *sis,cond_resched (); - first_block =bmap (inode, probe_block); -if (first_block ==0 ) + first_block = probe_block; + ret =bmap (inode, &first_block); +if (ret || !first_block) goto bad_bmap;/*
@@ -193,9 +194,11 @@ int generic_swapfile_activate(struct swap_info_struct *sis, block_in_page++) {sector_t block; - block =bmap (inode, probe_block + block_in_page); -if (block ==0 ) + block = probe_block + block_in_page; + ret =bmap (inode, &block); +if (ret || !block) goto bad_bmap; +if (block != first_block + block_in_page) {/* Discontiguity */ probe_block++;
Signature · 1 line
2.20.1
PATCH v4 2/9 cachefiles: drop direct usage of ->bmap method.
In reply to Carlos MaiolinoReplace the direct usage of ->bmap method by a bmap() call.
--- fs/cachefiles/rdwr.c | 27 ++++++++++++++------------- 1 file changed, 14 insertions(+), 13 deletions(-)
fs/cachefiles/rdwr.c +14 -13
@@ -400,7 +400,7 @@ int cachefiles_read_or_alloc_page(struct fscache_retrieval *op,struct cachefiles_object *object;struct cachefiles_cache *cache;struct inode *inode; -sector_t block0, block; +sector_t block;unsigned shift;int ret;
@@ -416,7 +416,6 @@ int cachefiles_read_or_alloc_page(struct fscache_retrieval *op, inode =d_backing_inode (object->backer);ASSERT (S_ISREG (inode->i_mode)); -ASSERT (inode->i_mapping->a_ops->bmap);ASSERT (inode->i_mapping->a_ops->readpages);/* calculate the shift required to use bmap */
@@ -432,12 +431,14 @@ int cachefiles_read_or_alloc_page(struct fscache_retrieval *op,* enough for this as it doesn't indicate errors, but it's all we've * got for the moment */ - block0 = page->index; - block0 <<= shift; + block = page->index; + block <<= shift; + + ret =bmap (inode, &block); +ASSERT (!ret); - block = inode->i_mapping->a_ops->bmap (inode->i_mapping, block0);_debug ("%llx -> %llx" , - (unsigned long long ) block0, + (unsigned long long ) (page->index << shift), (unsigned long long ) block);if (block) {
@@ -715,7 +716,6 @@ int cachefiles_read_or_alloc_pages(struct fscache_retrieval *op, inode =d_backing_inode (object->backer);ASSERT (S_ISREG (inode->i_mode)); -ASSERT (inode->i_mapping->a_ops->bmap);ASSERT (inode->i_mapping->a_ops->readpages);/* calculate the shift required to use bmap */
@@ -732,7 +732,7 @@ int cachefiles_read_or_alloc_pages(struct fscache_retrieval *op, ret = space ? -ENODATA : -ENOBUFS ;list_for_each_entry_safe (page, _n, pages, lru) { -sector_t block0, block; +sector_t block;/* we assume the absence or presence of the first block is a * good enough indication for the page as a whole
@@ -740,13 +740,14 @@ int cachefiles_read_or_alloc_pages(struct fscache_retrieval *op,* good enough for this as it doesn't indicate errors, but * it's all we've got for the moment */ - block0 = page->index; - block0 <<= shift; + block = page->index; + block <<= shift; + + ret =bmap (inode, &block); +ASSERT (!ret); - block = inode->i_mapping->a_ops->bmap (inode->i_mapping, - block0);_debug ("%llx -> %llx" , - (unsigned long long ) block0, + (unsigned long long ) (page->index << shift), (unsigned long long ) block);if (block) {
Signature · 1 line
2.20.1
PATCH v4 3/9 ecryptfs: drop direct calls to ->bmap
In reply to Carlos MaiolinoReplace direct ->bmap calls by bmap() method.
--- fs/ecryptfs/mmap.c | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-)
fs/ecryptfs/mmap.c +6 -10
@@ -538,16 +538,12 @@ static int ecryptfs_write_end(struct file *file,static sector_t ecryptfs_bmap (struct address_space *mapping,sector_t block) { -int rc =0 ; -struct inode *inode; -struct inode *lower_inode; - - inode = (struct inode *)mapping->host; - lower_inode =ecryptfs_inode_to_lower (inode); -if (lower_inode->i_mapping->a_ops->bmap) - rc = lower_inode->i_mapping->a_ops->bmap (lower_inode->i_mapping, - block); -return rc; +struct inode *lower_inode =ecryptfs_inode_to_lower (mapping->host); +int ret =bmap (lower_inode, &block); + +if (ret) +return 0 ; +return block; }const struct address_space_operations ecryptfs_aops = {
Signature · 1 line
2.20.1
PATCH v4 4/9 fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
In reply to Carlos MaiolinoNow we have the possibility of proper error return in bmap, use bmap() function in ioctl_fibmap() instead of calling ->bmap method directly.
---
Changelog:
V4:
- Ensure ioctl_fibmap() returns 0 in case of error returned from bmap(). Otherwise we'll be changing the user interface (which returns 0 in case of error)
V3:
- Rename usr_blk to ur_block
V2:
- Use a local sector_t variable to asign the block number instead of using direct casting.
fs/ioctl.c | 29 +++++++++++++++++++---------- 1 file changed, 19 insertions(+), 10 deletions(-)
fs/ioctl.c +19 -10
@@ -53,19 +53,28 @@ EXPORT_SYMBOL(vfs_ioctl);static int ioctl_fibmap (struct file *filp,int __user *p) { -struct address_space *mapping = filp->f_mapping; -int res, block; +struct inode *inode =file_inode (filp); +int error, ur_block; +sector_t block; -/* do we support this mess? */ -if (!mapping->a_ops->bmap) -return -EINVAL ;if (!capable (CAP_SYS_RAWIO ))return -EPERM ; - res =get_user (block, p); -if (res) -return res; - res = mapping->a_ops->bmap (mapping, block); -return put_user (res, p); + + error =get_user (ur_block, p); +if (error) +return error; + + block = ur_block; + error =bmap (inode, &block); + +if (error) + ur_block =0 ; +else + ur_block = block; + + error =put_user (ur_block, p); + +return error; }/**
Signature · 1 line
2.20.1
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 48 lines · Carlos Maiolino
Now we have the possibility of proper error return in bmap, use
bmap()function inioctl_fibmap()instead of calling ->bmap method directly.Signed-off-by Carlos Maiolino <[email protected]>---
Changelog:
V4:
- Ensureioctl_fibmap()returns 0 in case of error returned frombmap(). Otherwise we'll be changing the user interface (which returns 0 in case of error)
V3:
- Renameusr_blktour_blockV2:
- Use a localsector_tvariable to asign the block number instead of using direct casting.fs/ioctl.c | 29 +++++++++++++++++++---------- 1 file changed, 19 insertions(+), 10 deletions(-)fs/ioctl.c +5
-10@@ -53,19 +53,28 @@ EXPORT_SYMBOL(vfs_ioctl);static int ioctl_fibmap (struct file *filp,int __user *p) { -struct address_space *mapping = filp->f_mapping; -int res, block; +struct inode *inode =file_inode (filp); +int error, ur_block; +sector_t block; -/* do we support this mess? */ -if (!mapping->a_ops->bmap) -return -EINVAL ;if (!capable (CAP_SYS_RAWIO ))return -EPERM ; - res =get_user (block, p); -if (res) -return res; - res = mapping->a_ops->bmap (mapping, block); -return put_user (res, p); + + error =get_user (ur_block, p);
fs/ioctl.c EXPORT_SYMBOL(vfs_ioctl); · line 73
Quoted 10 lines
fs/ioctl.c +10
-0+if (error) +return error; + + block = ur_block; + error =bmap (inode, &block); + +if (error) + ur_block =0 ; +else + ur_block = block;
What happens if ur_block > INT_MAX? Shouldn't we return zero (i.e. error) instead of truncating the value? Maybe the code does this somewhere else? Here seemed like the obvious place for an overflow check as we go from sector_t to int.
--D
Quoted 10 lines
fs/ioctl.c +4
-0+ + error =put_user (ur_block, p); + +return error; }/** Signature · 1 line
2.20.1
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Hi Darrick.
Quoted 6 lines
Quoted 6 lines
fs/ioctl.c +6
-0+return error; + + block = ur_block; + error =bmap (inode, &block); + +if (error)
Quoted 9 lines
fs/ioctl.c EXPORT_SYMBOL(vfs_ioctl); · line 73
Quoted 3 lines
fs/ioctl.c +3
-0+ ur_block =0 ; + else + ur_block = block;What happens if
ur_block>INT_MAX? Shouldn't we return zero (i.e. error) instead of truncating the value? Maybe the code does this somewhere else? Here seemed like the obvious place for an overflow check as we go fromsector_tto int.
The behavior should still be the same. It will get truncated, unfortunately. I don't think we can actually change this behavior and return zero instead of truncating it.
Quoted 12 lines
--D
Quoted 10 lines
fs/ioctl.c +4
-0+ + error =put_user (ur_block, p); + +return error; }/** Signature · 1 line
2.20.1
Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 18 lines · Carlos Maiolino
Hi Darrick.
Quoted 6 lines
Quoted 6 lines
fs/ioctl.c +6
-0+return error; + + block = ur_block; + error =bmap (inode, &block); + +if (error)Quoted 9 lines
fs/ioctl.c EXPORT_SYMBOL(vfs_ioctl); · line 73
Quoted 3 lines
fs/ioctl.c +3
-0+ ur_block =0 ; + else + ur_block = block;What happens if
ur_block>INT_MAX? Shouldn't we return zero (i.e. error) instead of truncating the value? Maybe the code does this somewhere else? Here seemed like the obvious place for an overflow check as we go fromsector_tto int.
Quoted 3 lines
The behavior should still be the same. It will get truncated, unfortunately. I don't think we can actually change this behavior and return zero instead of truncating it.
But that's even worse, because the programs that rely on FIBMAP will now receive *incorrect* results that may point at a different file and definitely do not point at the correct file block.
Note also that the iomap (and therefore xfs) implementation WARNs on integer overflow and returns 0 (error) to prevent an incorrect access.
--D
Quoted 15 lines
Quoted 12 lines
--D
Quoted 10 lines
fs/ioctl.c +4
-0+ + error =put_user (ur_block, p); + +return error; }/** Signature · 1 line
2.20.1Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 19 lines · Darrick J. Wong
Quoted 18 lines · Carlos Maiolino
Hi Darrick.
Quoted 15 lines
fs/ioctl.c EXPORT_SYMBOL(vfs_ioctl); · line 73
Quoted 9 lines
fs/ioctl.c +9
-0+return error; + + block = ur_block; + error =bmap (inode, &block); + +if (error) + ur_block =0 ; +else + ur_block = block;What happens if
ur_block>INT_MAX? Shouldn't we return zero (i.e. error) instead of truncating the value? Maybe the code does this somewhere else? Here seemed like the obvious place for an overflow check as we go fromsector_tto int.
Quoted 7 lines
Quoted 3 lines
The behavior should still be the same. It will get truncated, unfortunately. I don't think we can actually change this behavior and return zero instead of truncating it.
But that's even worse, because the programs that rely on FIBMAP will now receive *incorrect* results that may point at a different file and definitely do not point at the correct file block.
How is this worse? This is exactly what happens today, on the original FIBMAP implementation.
Maybe I am not seeing something or having a different thinking you have, but this is the behavior we have now, without my patches. And we can't really change it; the user view of this implementation.
That's why I didn't try to change the result, so the truncation still happens.
Quoted 3 lines
Note also that the iomap (and therefore xfs) implementation WARNs on integer overflow and returns 0 (error) to prevent an incorrect access.
It does not really prevent anything. It just issue a warning saying the result will be truncated, in an attempt to notify the FIBMAP interface user that he/she can't trust the result, but it does not prevent a truncated result to be returned. And IIRC, iomap is the only interface now that cares about issuing a warning.
I think the *best* we could do here, is to make the new bmap() to issue the same kind of WARN() iomap does, but we can't really change the end result.
Quoted 18 lines
--D
Quoted 15 lines
Quoted 12 lines
--D
Quoted 10 lines
fs/ioctl.c +4
-0+ + error =put_user (ur_block, p); + +return error; }/** Signature · 1 line
2.20.1Signature · 1 line
Carlos
Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 24 lines · Carlos Maiolino
Quoted 23 lines · Darrick J. Wong
Quoted 21 lines · Carlos Maiolino
Hi Darrick.
Quoted 6 lines
Quoted 6 lines
fs/ioctl.c +6
-0+return error; + + block = ur_block; + error =bmap (inode, &block); + +if (error)Quoted 9 lines
fs/ioctl.c EXPORT_SYMBOL(vfs_ioctl); · line 73
Quoted 3 lines
fs/ioctl.c +3
-0+ ur_block =0 ; + else + ur_block = block;What happens if
ur_block>INT_MAX? Shouldn't we return zero (i.e. error) instead of truncating the value? Maybe the code does this somewhere else? Here seemed like the obvious place for an overflow check as we go fromsector_tto int.The behavior should still be the same. It will get truncated, unfortunately. I don't think we can actually change this behavior and return zero instead of truncating it.
Quoted 6 lines
Quoted 3 lines
But that's even worse, because the programs that rely on FIBMAP will now receive *incorrect* results that may point at a different file and definitely do not point at the correct file block.
How is this worse? This is exactly what happens today, on the original FIBMAP implementation.
Ok, I wasn't being 110% careful with my words. Delete "will now" from the sentence above.
Quoted 4 lines
Maybe I am not seeing something or having a different thinking you have, but this is the behavior we have now, without my patches. And we can't really change it; the user view of this implementation.
That's why I didn't try to change the result, so the truncation still happens.
I understand that we're not generally supposed to change existing userspace interfaces, but the fact remains that allowing truncated responses causes *filesystem corruption*.
We know that the most well known FIBMAP callers are bootloaders, and we know what they do with the information they get -- they use it to record the block map of boot files. So if the IPL/grub/whatever installer queries the boot file and the boot file is at block 12345678901 (a 34-bit number), this interface truncates that to 3755744309 (a 32-bit number) and that's where the bootloader will think its boot files are. The installation succeeds, the user reboots and *kaboom* the system no longer boots because the contents of block 3755744309 is not a bootloader.
Worse yet, grub1 used FIBMAP data to record the location of the grub environment file and installed itself between the MBR and the start of partition 1. If the environment file is at offset 1234578901, grub will write status data to its environment file (which it thinks is at 3755744309) and *KABOOM* we've just destroyed whatever was in that block.
Far better for the bootloader installation script to hit an error and force the admin to deal with the situation than for the system to become unbootable. That's *why* the (newer) iomap bmap implementation does not return truncated mappings, even though the classic implementation does.
The classic code returning truncated results is a broken behavior.
Quoted 6 lines
Quoted 2 lines
Note also that the iomap (and therefore xfs) implementation WARNs on integer overflow and returns 0 (error) to prevent an incorrect access.
It does not really prevent anything. It just issue a warning saying the result will be truncated, in an attempt to notify the FIBMAP interface user that he/she can't trust the result, but it does not prevent a truncated result to be
I disagree; the iomap bmap implementation /does/ prevent truncated responses:
: static loff_t
: iomap_bmap_actor(struct inode *inode, loff_t pos, loff_t length, : void *data, struct iomap *iomap)
: {
: sector_t *bno = data, addr;
:
: if (iomap->type == IOMAP_MAPPED) {
: addr = (pos - iomap->offset + iomap->addr) >> inode->i_blkbits;
: if (addr > INT_MAX)
: WARN(1, "would truncate bmap result\n");
Notice how we don't set *bno here?
: else
: *bno = addr;
And only set it in the case that there isn't an integer overflow?
: }
: return 0;
: }
:
: /* legacy ->bmap interface. 0 is the error return (!) */ : sector_t
: iomap_bmap(struct address_space *mapping, sector_t bno,
: const struct iomap_ops *ops)
: {
: struct inode *inode = mapping->host;
: loff_t pos = bno << inode->i_blkbits;
: unsigned blocksize = i_blocksize(inode);
:
: if (filemap_write_and_wait(mapping))
: return 0;
:
: bno = 0;
We initialize bno to zero here...
: iomap_apply(inode, pos, blocksize, 0, ops, &bno, iomap_bmap_actor);
...then pass bno's address to the apply function to pass to iomap_bmap_actor, so either the _actor function set bno or in the case of overflow it left it set to zero.
: return bno;
: }
Quoted 5 lines
returned. And IIRC, iomap is the only interface now that cares about issuing a warning.
I think the *best* we could do here, is to make the new
bmap()to issue the same kind ofWARN()iomap does, but we can't really change the end result.
I'd rather we break legacy code than corrupt filesystems.
--D
Quoted 22 lines
Quoted 19 lines
--D
Quoted 15 lines
Quoted 12 lines
--D
Quoted 10 lines
fs/ioctl.c +4
-0+ + error =put_user (ur_block, p); + +return error; }/** Signature · 1 line
2.20.1Signature · 1 line
CarlosSignature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 7 lines · Darrick J. Wong
Quoted 5 lines
returned. And IIRC, iomap is the only interface now that cares about issuing a warning.
I think the *best* we could do here, is to make the new
bmap()to issue the same kind ofWARN()iomap does, but we can't really change the end result.I'd rather we break legacy code than corrupt filesystems.
This particular patch should keep existing behavior as is, as the intent is to not change functionality. Throwing in another patch to have saner error behavior now that we have a saner in-kernel interface that cleary documents what it is breaking and why on the other hand sounds like a very good idea.
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 9 lines · Christoph Hellwig
Quoted 7 lines · Darrick J. Wong
Quoted 5 lines
returned. And IIRC, iomap is the only interface now that cares about issuing a warning.
I think the *best* we could do here, is to make the new
bmap()to issue the same kind ofWARN()iomap does, but we can't really change the end result.I'd rather we break legacy code than corrupt filesystems.
Yes, I have the same feeling, but this patchset does not have the goal to fix the broken api.
Quoted 5 lines
This particular patch should keep existing behavior as is, as the intent is to not change functionality. Throwing in another patch to have saner error behavior now that we have a saner in-kernel interface that cleary documents what it is breaking and why on the other hand sounds like a very good idea.
I totally agree here, and to be honest, I think such change should be in a different patchset rather than a new patch in this series. I can do it for sure, but this discussion IMHO should be done not only here in linux-fsdevel, but also in linux-api, which well, I don't think cc'ing this whole patchset there will do any good other than keep the change discussion more complicated than it should be. I'd rather finish the design and implementation of this patchset, and I'll follow-up it, once it's all set, with a new patch to change the truncation behavior, it will make the discussion way easier than mixing up subjects. What you guys think?
Cheers
Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 20 lines · Carlos Maiolino
Quoted 9 lines · Christoph Hellwig
Quoted 7 lines · Darrick J. Wong
Quoted 5 lines
returned. And IIRC, iomap is the only interface now that cares about issuing a warning.
I think the *best* we could do here, is to make the new
bmap()to issue the same kind ofWARN()iomap does, but we can't really change the end result.I'd rather we break legacy code than corrupt filesystems.
Yes, I have the same feeling, but this patchset does not have the goal to fix the broken api.
Quoted 5 lines
This particular patch should keep existing behavior as is, as the intent is to not change functionality. Throwing in another patch to have saner error behavior now that we have a saner in-kernel interface that cleary documents what it is breaking and why on the other hand sounds like a very good idea.
Quoted 9 lines
I totally agree here, and to be honest, I think such change should be in a different patchset rather than a new patch in this series. I can do it for sure, but this discussion IMHO should be done not only here in linux-fsdevel, but also in linux-api, which well, I don't think cc'ing this whole patchset there will do any good other than keep the change discussion more complicated than it should be. I'd rather finish the design and implementation of this patchset, and I'll follow-up it, once it's all set, with a new patch to change the truncation behavior, it will make the discussion way easier than mixing up subjects. What you guys think?
I probably would've fixed the truncation behavior in the old code and based the fiemap-fibmap conversion on that so that anyone who wants to backport the behavior change to an old kernel has an easier time of it.
But afterwards probably works just as well since I don't feel like tying ourselves in more knots over an old interface. ;)
--D
Quoted 5 lines
Cheers
Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 31 lines · Darrick J. Wong
Quoted 20 lines · Carlos Maiolino
Quoted 9 lines · Christoph Hellwig
Quoted 7 lines · Darrick J. Wong
Quoted 5 lines
returned. And IIRC, iomap is the only interface now that cares about issuing a warning.
I think the *best* we could do here, is to make the new
bmap()to issue the same kind ofWARN()iomap does, but we can't really change the end result.I'd rather we break legacy code than corrupt filesystems.
Yes, I have the same feeling, but this patchset does not have the goal to fix the broken api.
Quoted 5 lines
This particular patch should keep existing behavior as is, as the intent is to not change functionality. Throwing in another patch to have saner error behavior now that we have a saner in-kernel interface that cleary documents what it is breaking and why on the other hand sounds like a very good idea.
Quoted 9 lines
I totally agree here, and to be honest, I think such change should be in a different patchset rather than a new patch in this series. I can do it for sure, but this discussion IMHO should be done not only here in linux-fsdevel, but also in linux-api, which well, I don't think cc'ing this whole patchset there will do any good other than keep the change discussion more complicated than it should be. I'd rather finish the design and implementation of this patchset, and I'll follow-up it, once it's all set, with a new patch to change the truncation behavior, it will make the discussion way easier than mixing up subjects. What you guys think?
Quoted 4 lines
I probably would've fixed the truncation behavior in the old code and based the fiemap-fibmap conversion on that so that anyone who wants to backport the behavior change to an old kernel has an easier time of it.
Well, another problem in fixing it in the old code, is that bmap() can't properly return errors :P
After this patchset, bmap() will be able to return errors, so we can easily fix it, once we won't need to 'guess' what a zero return mean from bmap()
Quoted 10 lines
But afterwards probably works just as well since I don't feel like tying ourselves in more knots over an old interface. ;)
--D
Quoted 5 lines
Cheers
Signature · 1 line
Carlos
Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
In reply to Darrick J. WongI'll keep my reply to your email short here, because hch's follow-up email summarizes what I wanted to say, so I'll reply to his email directly.
Quoted 10 lines
: return bno;
: }Quoted 5 lines
returned. And IIRC, iomap is the only interface now that cares about issuing a warning.
I think the *best* we could do here, is to make the new
bmap()to issue the same kind ofWARN()iomap does, but we can't really change the end result.I'd rather we break legacy code than corrupt filesystems.
By now, I wish to apologize Darrick, it was my fault to not pay enough attention to iomap's bmap implementation as you mentioned. Will keep the discussion topic on next e-mail.
Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
In reply to Darrick J. WongQuoted 60 lines · Darrick J. Wong
Quoted 24 lines · Carlos Maiolino
Quoted 23 lines · Darrick J. Wong
Quoted 21 lines · Carlos Maiolino
Hi Darrick.
Quoted 6 lines
Quoted 6 lines
fs/ioctl.c +6
-0+return error; + + block = ur_block; + error =bmap (inode, &block); + +if (error)Quoted 9 lines
fs/ioctl.c EXPORT_SYMBOL(vfs_ioctl); · line 73
Quoted 3 lines
fs/ioctl.c +3
-0+ ur_block =0 ; + else + ur_block = block;What happens if
ur_block>INT_MAX? Shouldn't we return zero (i.e. error) instead of truncating the value? Maybe the code does this somewhere else? Here seemed like the obvious place for an overflow check as we go fromsector_tto int.The behavior should still be the same. It will get truncated, unfortunately. I don't think we can actually change this behavior and return zero instead of truncating it.
Quoted 6 lines
Quoted 3 lines
But that's even worse, because the programs that rely on FIBMAP will now receive *incorrect* results that may point at a different file and definitely do not point at the correct file block.
How is this worse? This is exactly what happens today, on the original FIBMAP implementation.
Ok, I wasn't being 110% careful with my words. Delete "will now" from the sentence above.
Quoted 4 lines
Maybe I am not seeing something or having a different thinking you have, but this is the behavior we have now, without my patches. And we can't really change it; the user view of this implementation.
That's why I didn't try to change the result, so the truncation still happens.I understand that we're not generally supposed to change existing userspace interfaces, but the fact remains that allowing truncated responses causes *filesystem corruption*.
We know that the most well known FIBMAP callers are bootloaders, and we know what they do with the information they get -- they use it to record the block map of boot files. So if the IPL/grub/whatever installer queries the boot file and the boot file is at block 12345678901 (a 34-bit number), this interface truncates that to 3755744309 (a 32-bit number) and that's where the bootloader will think its boot files are. The installation succeeds, the user reboots and *kaboom* the system no longer boots because the contents of block 3755744309 is not a bootloader.
Worse yet, grub1 used FIBMAP data to record the location of the grub environment file and installed itself between the MBR and the start of partition 1. If the environment file is at offset 1234578901, grub will write status data to its environment file (which it thinks is at 3755744309) and *KABOOM* we've just destroyed whatever was in that block.
Quoted 6 lines
Far better for the bootloader installation script to hit an error and force the admin to deal with the situation than for the system to become unbootable. That's *why* the (newer) iomap bmap implementation does not return truncated mappings, even though the classic implementation does.
The classic code returning truncated results is a broken behavior.
How long as it been broken for? And if we do fix it, I'd just like for a nice commit lot describing potential risks of not applying it. *If* the issue exists as-is today, the above contains a lot of information for addressing potential issues, even if theoretical.
Luis
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 61 lines · Luis Chamberlain
Quoted 60 lines · Darrick J. Wong
Quoted 24 lines · Carlos Maiolino
Quoted 23 lines · Darrick J. Wong
Quoted 21 lines · Carlos Maiolino
Hi Darrick.
Quoted 6 lines
Quoted 6 lines
(quoted diff) +6
-0+ return error; + + block = ur_block; + error = bmap(inode, &block); + + if (error)Quoted 9 lines
Review
Quoted 3 lines
(quoted diff) +3
-0+ ur_block = 0; + else + ur_block = block;What happens if
ur_block>INT_MAX? Shouldn't we return zero (i.e. error) instead of truncating the value? Maybe the code does this somewhere else? Here seemed like the obvious place for an overflow check as we go fromsector_tto int.The behavior should still be the same. It will get truncated, unfortunately. I don't think we can actually change this behavior and return zero instead of truncating it.
Quoted 6 lines
Quoted 3 lines
But that's even worse, because the programs that rely on FIBMAP will now receive *incorrect* results that may point at a different file and definitely do not point at the correct file block.
How is this worse? This is exactly what happens today, on the original FIBMAP implementation.
Ok, I wasn't being 110% careful with my words. Delete "will now" from the sentence above.
Quoted 4 lines
Maybe I am not seeing something or having a different thinking you have, but this is the behavior we have now, without my patches. And we can't really change it; the user view of this implementation.
That's why I didn't try to change the result, so the truncation still happens.I understand that we're not generally supposed to change existing userspace interfaces, but the fact remains that allowing truncated responses causes *filesystem corruption*.
We know that the most well known FIBMAP callers are bootloaders, and we know what they do with the information they get -- they use it to record the block map of boot files. So if the IPL/grub/whatever installer queries the boot file and the boot file is at block 12345678901 (a 34-bit number), this interface truncates that to 3755744309 (a 32-bit number) and that's where the bootloader will think its boot files are. The installation succeeds, the user reboots and *kaboom* the system no longer boots because the contents of block 3755744309 is not a bootloader.
Worse yet, grub1 used FIBMAP data to record the location of the grub environment file and installed itself between the MBR and the start of partition 1. If the environment file is at offset 1234578901, grub will write status data to its environment file (which it thinks is at 3755744309) and *KABOOM* we've just destroyed whatever was in that block.
Quoted 8 lines
Quoted 6 lines
Far better for the bootloader installation script to hit an error and force the admin to deal with the situation than for the system to become unbootable. That's *why* the (newer) iomap bmap implementation does not return truncated mappings, even though the classic implementation does.
The classic code returning truncated results is a broken behavior.
How long as it been broken for?
Probably since the beginning (ext2).
Quoted 4 lines
And if we do fix it, I'd just like for
a nice commit lot describing potential risks of not applying it. *If* the issue exists as-is today, the above contains a lot of information for addressing potential issues, even if theoretical.
I think a lot of the filesystems avoid the problem either by not supporting > INT_MAX blocks in the first place or by detecting the truncation in the fs-specific ->bmap method, so that might be why we haven't been deluged by corruption reports.
--D
Quoted 1 line
Luis
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
In reply to Luis ChamberlainQuoted 31 lines
Quoted 31 lines
Quoted 5 lines
Maybe I am not seeing something or having a different thinking you have, but this is the behavior we have now, without my patches. And we can't really change it; the user view of this implementation.
That's why I didn't try to change the result, so the truncation still happens.I understand that we're not generally supposed to change existing userspace interfaces, but the fact remains that allowing truncated responses causes *filesystem corruption*.
We know that the most well known FIBMAP callers are bootloaders, and we know what they do with the information they get -- they use it to record the block map of boot files. So if the IPL/grub/whatever installer queries the boot file and the boot file is at block 12345678901 (a 34-bit number), this interface truncates that to 3755744309 (a 32-bit number) and that's where the bootloader will think its boot files are. The installation succeeds, the user reboots and *kaboom* the system no longer boots because the contents of block 3755744309 is not a bootloader.
Worse yet, grub1 used FIBMAP data to record the location of the grub environment file and installed itself between the MBR and the start of partition 1. If the environment file is at offset 1234578901, grub will write status data to its environment file (which it thinks is at 3755744309) and *KABOOM* we've just destroyed whatever was in that block.
Far better for the bootloader installation script to hit an error and force the admin to deal with the situation than for the system to become unbootable. That's *why* the (newer) iomap bmap implementation does not return truncated mappings, even though the classic implementation does.
Quoted 7 lines
Quoted 1 line
The classic code returning truncated results is a broken behavior.
How long as it been broken for? And if we do fix it, I'd just like for a nice commit lot describing potential risks of not applying it. *If* the issue exists as-is today, the above contains a lot of information for addressing potential issues, even if theoretical.
It's broken since forever. This has always been the FIBMAP behavior.
Quoted 1 line
Luis
Signature · 1 line
Carlos
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Quoted 34 lines · Carlos Maiolino
Quoted 34 lines
Quoted 33 lines
Quoted 6 lines
Maybe I am not seeing something or having a different thinking you have, but this is the behavior we have now, without my patches. And we can't really change it; the user view of this implementation.
That's why I didn't try to change the result, so the truncation still happens.I understand that we're not generally supposed to change existing userspace interfaces, but the fact remains that allowing truncated responses causes *filesystem corruption*.
We know that the most well known FIBMAP callers are bootloaders, and we know what they do with the information they get -- they use it to record the block map of boot files. So if the IPL/grub/whatever installer queries the boot file and the boot file is at block 12345678901 (a 34-bit number), this interface truncates that to 3755744309 (a 32-bit number) and that's where the bootloader will think its boot files are. The installation succeeds, the user reboots and *kaboom* the system no longer boots because the contents of block 3755744309 is not a bootloader.
Worse yet, grub1 used FIBMAP data to record the location of the grub environment file and installed itself between the MBR and the start of partition 1. If the environment file is at offset 1234578901, grub will write status data to its environment file (which it thinks is at 3755744309) and *KABOOM* we've just destroyed whatever was in that block.
Far better for the bootloader installation script to hit an error and force the admin to deal with the situation than for the system to become unbootable. That's *why* the (newer) iomap bmap implementation does not return truncated mappings, even though the classic implementation does.
The classic code returning truncated results is a broken behavior.
Quoted 7 lines
Quoted 5 lines
How long as it been broken for? And if we do fix it, I'd just like for a nice commit lot describing potential risks of not applying it. *If* the issue exists as-is today, the above contains a lot of information for addressing potential issues, even if theoretical.
It's broken since forever. This has always been the FIBMAP behavior.
It's been broken since forever, but only for filesystems larger than 4TB or 16TB (2^32 blocks), which are only becoming commonplace for root disks recently. Also, doesn't LILO have a limit on the location of the kernel image, in the first 1GB or similar?
So maybe this is not an issue that FIBMAP users ever hit in practise anyway, but I agree that it doesn't make sense to return bad data (32-bit wrapped block numbers) and 0 should be returned in such cases.
Cheers, Andreas
Re: [PATCH 4/9] fibmap: Use bmap instead of ->bmap method in ioctl_fibmap
Meh... Sorry andreas, your reply became disconnected from the thread, and I think I didn't reply.
Quoted 43 lines · Andreas Dilger
Quoted 34 lines · Carlos Maiolino
Quoted 34 lines
Quoted 33 lines
Quoted 6 lines
Maybe I am not seeing something or having a different thinking you have, but this is the behavior we have now, without my patches. And we can't really change it; the user view of this implementation.
That's why I didn't try to change the result, so the truncation still happens.I understand that we're not generally supposed to change existing userspace interfaces, but the fact remains that allowing truncated responses causes *filesystem corruption*.
We know that the most well known FIBMAP callers are bootloaders, and we know what they do with the information they get -- they use it to record the block map of boot files. So if the IPL/grub/whatever installer queries the boot file and the boot file is at block 12345678901 (a 34-bit number), this interface truncates that to 3755744309 (a 32-bit number) and that's where the bootloader will think its boot files are. The installation succeeds, the user reboots and *kaboom* the system no longer boots because the contents of block 3755744309 is not a bootloader.
Worse yet, grub1 used FIBMAP data to record the location of the grub environment file and installed itself between the MBR and the start of partition 1. If the environment file is at offset 1234578901, grub will write status data to its environment file (which it thinks is at 3755744309) and *KABOOM* we've just destroyed whatever was in that block.
Far better for the bootloader installation script to hit an error and force the admin to deal with the situation than for the system to become unbootable. That's *why* the (newer) iomap bmap implementation does not return truncated mappings, even though the classic implementation does.
The classic code returning truncated results is a broken behavior.
Quoted 7 lines
Quoted 5 lines
How long as it been broken for? And if we do fix it, I'd just like for a nice commit lot describing potential risks of not applying it. *If* the issue exists as-is today, the above contains a lot of information for addressing potential issues, even if theoretical.
It's broken since forever. This has always been the FIBMAP behavior.
Quoted 9 lines
It's been broken since forever, but only for filesystems larger than 4TB or 16TB (2^32 blocks), which are only becoming commonplace for root disks recently. Also, doesn't LILO have a limit on the location of the kernel image, in the first 1GB or similar?
So maybe this is not an issue that FIBMAP users ever hit in practise anyway, but I agree that it doesn't make sense to return bad data (32-bit wrapped block numbers) and 0 should be returned in such cases.
Thanks for the input, but TBH I don't use LILO for a long time, and I don't remember exactly how it works.
Anyway, I have 2 bugs to fix in this code, after I can get this series in, one is the overflow we'll probably need kernel-api approval, and another one is the acceptance of negative values into FIBMAP, which we have no protection at all. I'll fix both once I can get the main series in.
Cheers
Quoted 7 lines
Cheers, Andreas
Signature · 1 line
Carlos
PATCH v4 5/9 fs: Move start and length fiemap fields into fiemap_extent_info
In reply to Carlos MaiolinoAs the overall goal to deprecate fibmap, Christoph suggested a rework of the ->fiemap API, in a way we could pass to it a callback to fill the fiemap structure (one of these callbacks being fiemap_fill_next_extent).
To avoid the need to add several fields into the ->fiemap method, aggregate everything into a single data structure, and pass it along.
This patch isn't suppose to add any functional change, only to update filesystems providing ->fiemap() method.
--- fs/bad_inode.c | 3 +-- fs/btrfs/inode.c | 5 +++-- fs/ext2/ext2.h | 3 +-- fs/ext2/inode.c | 6 ++---- fs/ext4/ext4.h | 3 +-- fs/ext4/extents.c | 8 ++++---- fs/f2fs/data.c | 5 +++-- fs/f2fs/f2fs.h | 3 +-- fs/gfs2/inode.c | 5 +++-- fs/hpfs/file.c | 4 ++-- fs/ioctl.c | 16 ++++++++++------ fs/nilfs2/inode.c | 5 +++-- fs/nilfs2/nilfs.h | 3 +-- fs/ocfs2/extent_map.c | 5 +++-- fs/ocfs2/extent_map.h | 3 +-- fs/overlayfs/inode.c | 5 ++--- fs/xfs/xfs_iops.c | 10 +++++----- include/linux/fs.h | 21 +++++++++++---------- 18 files changed, 57 insertions(+), 56 deletions(-)
fs/bad_inode.c +1 -2
@@ -120,8 +120,7 @@ static struct posix_acl *bad_inode_get_acl(struct inode *inode, int type) }static int bad_inode_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len) +struct fiemap_extent_info *fieinfo) {return -EIO ; }
fs/btrfs/inode.c +3 -2
@@ -8600,9 +8600,10 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)#define BTRFS_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) -static int btrfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len) +static int btrfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;int ret; ret =fiemap_check_flags (fieinfo,BTRFS_FIEMAP_FLAGS );
fs/ext2/ext2.h +1 -2
@@ -760,8 +760,7 @@ extern int ext2_get_block(struct inode *, sector_t, struct buffer_head *, int);extern int ext2_setattr (struct dentry *,struct iattr *);extern int ext2_getattr (const struct path *,struct kstat *,u32 ,unsigned int );extern void ext2_set_inode_flags (struct inode *inode); -extern int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len); +extern int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);/* ioctl.c */ extern long ext2_ioctl (struct file *,unsigned int ,unsigned long );
fs/ext2/inode.c +2 -4
@@ -855,11 +855,9 @@ const struct iomap_ops ext2_iomap_ops = {const struct iomap_ops ext2_iomap_ops;#endif /* CONFIG_FS_DAX */ -int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { -return generic_block_fiemap (inode, fieinfo, start, len, - ext2_get_block); +return generic_block_fiemap (inode, fieinfo, ext2_get_block); }static int ext2_writepage (struct page *page,struct writeback_control *wbc)
fs/ext4/ext4.h +1 -2
@@ -3154,8 +3154,7 @@ extern struct ext4_ext_path *ext4_find_extent(struct inode *, ext4_lblk_t,extern void ext4_ext_drop_refs (struct ext4_ext_path *);extern int ext4_ext_check_inode (struct inode *inode);extern ext4_lblk_t ext4_ext_next_allocated_block (struct ext4_ext_path *path); -extern int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len); +extern int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);extern int ext4_ext_precache (struct inode *inode);extern int ext4_collapse_range (struct inode *inode,loff_t offset,loff_t len);extern int ext4_insert_range (struct inode *inode,loff_t offset,loff_t len);
fs/ext4/extents.c +4 -4
@@ -5038,9 +5038,10 @@ static int ext4_xattr_fiemap(struct inode *inode,return (error <0 ? error :0 ); } -int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len) +int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;ext4_lblk_t start_blk;int error =0 ;
@@ -5062,8 +5063,7 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,/* fallback to generic here if not in extents fmt */ if (!(ext4_test_inode_flag (inode,EXT4_INODE_EXTENTS ))) -return generic_block_fiemap (inode, fieinfo, start, len, - ext4_get_block); +return generic_block_fiemap (inode, fieinfo, ext4_get_block);if (fiemap_check_flags (fieinfo,EXT4_FIEMAP_FLAGS ))return -EBADR ;
fs/f2fs/data.c +3 -2
@@ -1409,9 +1409,10 @@ static int f2fs_xattr_fiemap(struct inode *inode,return (err <0 ? err :0 ); } -int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;struct buffer_head map_bh;sector_t start_blk, last_blk;pgoff_t next_pgofs;
fs/f2fs/f2fs.h +1 -2
@@ -3155,8 +3155,7 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fio);void __do_map_lock (struct f2fs_sb_info *sbi,int flag,bool lock);int f2fs_map_blocks (struct inode *inode,struct f2fs_map_blocks *map,int create,int flag); -int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len); +int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);bool f2fs_should_update_inplace (struct inode *inode,struct f2fs_io_info *fio);bool f2fs_should_update_outplace (struct inode *inode,struct f2fs_io_info *fio); void f2fs_invalidate_page(struct page *page, unsigned int offset,
fs/gfs2/inode.c +3 -2
@@ -2004,9 +2004,10 @@ static int gfs2_getattr(const struct path *path, struct kstat *stat,return 0 ; } -static int gfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +static int gfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;struct gfs2_inode *ip =GFS2_I (inode);struct gfs2_holder gh;int ret;
fs/hpfs/file.c +2 -2
@@ -190,9 +190,9 @@ static sector_t _hpfs_bmap(struct address_space *mapping, sector_t block)return generic_block_bmap (mapping, block, hpfs_get_block); } -static int hpfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo,u64 start,u64 len) +static int hpfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { -return generic_block_fiemap (inode, fieinfo, start, len, hpfs_get_block); +return generic_block_fiemap (inode, fieinfo, hpfs_get_block); }const struct address_space_operations hpfs_aops = {
fs/ioctl.c +10 -6
@@ -210,6 +210,8 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg) fieinfo.fi_flags = fiemap.fm_flags; fieinfo.fi_extents_max = fiemap.fm_extent_count; fieinfo.fi_extents_start = ufiemap->fm_extents; + fieinfo.fi_start = fiemap.fm_start; + fieinfo.fi_len = len;if (fiemap.fm_extent_count !=0 && !access_ok(fieinfo.fi_extents_start,
@@ -219,7 +221,7 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg)if (fieinfo.fi_flags &FIEMAP_FLAG_SYNC )filemap_write_and_wait (inode->i_mapping); - error = inode->i_op->fiemap (inode, &fieinfo, fiemap.fm_start, len); + error = inode->i_op->fiemap (inode, &fieinfo); fiemap.fm_flags = fieinfo.fi_flags; fiemap.fm_mapped_extents = fieinfo.fi_extents_mapped;if (copy_to_user (ufiemap, &fiemap,sizeof (fiemap)))
@@ -296,9 +298,11 @@ static inline loff_t blk_to_logical(struct inode *inode, sector_t blk)*/ int __generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,loff_t start, -loff_t len,get_block_t *get_block) +struct fiemap_extent_info *fieinfo, +get_block_t *get_block) { +loff_t start = fieinfo->fi_start; +loff_t len = fieinfo->fi_len;struct buffer_head map_bh;sector_t start_blk, last_blk;loff_t isize =i_size_read (inode);
@@ -455,12 +459,12 @@ EXPORT_SYMBOL(__generic_block_fiemap);*/ int generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len,get_block_t *get_block) +struct fiemap_extent_info *fieinfo, +get_block_t *get_block) {int ret;inode_lock (inode); - ret =__generic_block_fiemap (inode, fieinfo, start, len, get_block); + ret =__generic_block_fiemap (inode, fieinfo, get_block);inode_unlock (inode);return ret; }
fs/nilfs2/inode.c +3 -2
@@ -992,9 +992,10 @@ void nilfs_dirty_inode(struct inode *inode, int flags)nilfs_transaction_commit (inode->i_sb);/* never fails */ } -int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len) +int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;struct the_nilfs *nilfs = inode->i_sb->s_fs_info;__u64 logical =0 , phys =0 , size =0 ;__u32 flags =0 ;
fs/nilfs2/nilfs.h +1 -2
@@ -276,8 +276,7 @@ extern int nilfs_inode_dirty(struct inode *);int nilfs_set_file_dirty (struct inode *inode,unsigned int nr_dirty);extern int __nilfs_mark_inode_dirty (struct inode *,int );extern void nilfs_dirty_inode (struct inode *,int flags); -int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len); +int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);static inline int nilfs_mark_inode_dirty (struct inode *inode) {return __nilfs_mark_inode_dirty (inode,I_DIRTY );
fs/ocfs2/extent_map.c +3 -2
@@ -749,8 +749,7 @@ static int ocfs2_fiemap_inline(struct inode *inode, struct buffer_head *di_bh,#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) -int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 map_start,u64 map_len) +int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {int ret, is_last;u32 mapping_end, cpos;
@@ -759,6 +758,8 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,u64 len_bytes, phys_bytes, virt_bytes;struct buffer_head *di_bh =NULL ;struct ocfs2_extent_rec rec; +u64 map_start = fieinfo->fi_start; +u64 map_len = fieinfo->fi_len; ret =fiemap_check_flags (fieinfo,OCFS2_FIEMAP_FLAGS );if (ret)
fs/ocfs2/extent_map.h +1 -2
@@ -50,8 +50,7 @@ int ocfs2_get_clusters(struct inode *inode, u32 v_cluster, u32 *p_cluster,int ocfs2_extent_map_get_blocks (struct inode *inode,u64 v_blkno,u64 *p_blkno,u64 *ret_count,unsigned int *extent_flags); -int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 map_start,u64 map_len); +int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);int ocfs2_overwrite_io (struct inode *inode,struct buffer_head *di_bh,u64 map_start,u64 map_len);
fs/overlayfs/inode.c +2 -3
@@ -456,8 +456,7 @@ int ovl_update_time(struct inode *inode, struct timespec64 *ts, int flags)return 0 ; } -static int ovl_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +static int ovl_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {int err;struct inode *realinode =ovl_inode_real (inode);
@@ -471,7 +470,7 @@ static int ovl_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,if (fieinfo->fi_flags &FIEMAP_FLAG_SYNC )filemap_write_and_wait (realinode->i_mapping); - err = realinode->i_op->fiemap (realinode, fieinfo, start, len); + err = realinode->i_op->fiemap (realinode, fieinfo);revert_creds (old_cred);return err;
fs/xfs/xfs_iops.c +5 -5
@@ -1109,12 +1109,12 @@ xfs_vn_update_time( STATIC intxfs_vn_fiemap ( -struct inode *inode, -struct fiemap_extent_info *fieinfo, -u64 start, -u64 length) +struct inode *inode, +struct fiemap_extent_info *fieinfo) { -int error; +u64 start = fieinfo->fi_start; +u64 length = fieinfo->fi_len; +int error;xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) {
include/linux/fs.h +11 -10
@@ -1705,11 +1705,14 @@ extern bool may_open_dev(const struct path *path);* VFS FS_IOC_FIEMAP helper definitions. */ struct fiemap_extent_info { -unsigned int fi_flags;/* Flags as passed from user */ -unsigned int fi_extents_mapped;/* Number of mapped extents */ -unsigned int fi_extents_max;/* Size of fiemap_extent array */ -struct fiemap_extent __user *fi_extents_start;/* Start of -fiemap_extent array */ +unsigned int fi_flags;/* Flags as passed from user */ +u64 fi_start; +u64 fi_len; +unsigned int fi_extents_mapped;/* Number of mapped extents */ +unsigned int fi_extents_max;/* Size of fiemap_extent array */ +struct fiemap_extent __user *fi_extents_start;/* Start of +fiemap_extent +array */ };int fiemap_fill_next_extent (struct fiemap_extent_info *info,u64 logical,u64 phys,u64 len,u32 flags);
@@ -1841,8 +1844,7 @@ struct inode_operations {int (*setattr) (struct dentry *,struct iattr *);int (*getattr) (const struct path *,struct kstat *,u32 ,unsigned int );ssize_t (*listxattr) (struct dentry *,char *,size_t ); -int (*fiemap)(struct inode *,struct fiemap_extent_info *,u64 start, -u64 len); +int (*fiemap)(struct inode *,struct fiemap_extent_info *);int (*update_time)(struct inode *,struct timespec64 *,int ); int (*atomic_open)(struct inode *, struct dentry *, struct file *, unsigned open_flag,
@@ -3199,11 +3201,10 @@ extern int vfs_readlink(struct dentry *, char __user *, int);extern int __generic_block_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -loff_t start,loff_t len,get_block_t *get_block);extern int generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len,get_block_t *get_block); +struct fiemap_extent_info *fieinfo, +get_block_t *get_block);extern struct file_system_type *get_filesystem (struct file_system_type *fs);extern void put_filesystem (struct file_system_type *fs);
Signature · 1 line
2.20.1
Re: [PATCH 5/9] fs: Move start and length fiemap fields into fiemap_extent_info
Quoted 353 lines · Carlos Maiolino
As the overall goal to deprecate fibmap, Christoph suggested a rework of the ->fiemap API, in a way we could pass to it a callback to fill the fiemap structure (one of these callbacks being
fiemap_fill_next_extent).To avoid the need to add several fields into the ->fiemap method, aggregate everything into a single data structure, and pass it along.
This patch isn't suppose to add any functional change, only to update filesystems providing ->fiemap() method.
Reviewed-by Christoph Hellwig <[email protected]>Signed-off-by Carlos Maiolino <[email protected]>--- fs/bad_inode.c | 3 +-- fs/btrfs/inode.c | 5 +++-- fs/ext2/ext2.h | 3 +-- fs/ext2/inode.c | 6 ++---- fs/ext4/ext4.h | 3 +-- fs/ext4/extents.c | 8 ++++---- fs/f2fs/data.c | 5 +++-- fs/f2fs/f2fs.h | 3 +-- fs/gfs2/inode.c | 5 +++-- fs/hpfs/file.c | 4 ++-- fs/ioctl.c | 16 ++++++++++------ fs/nilfs2/inode.c | 5 +++-- fs/nilfs2/nilfs.h | 3 +-- fs/ocfs2/extent_map.c | 5 +++-- fs/ocfs2/extent_map.h | 3 +-- fs/overlayfs/inode.c | 5 ++--- fs/xfs/xfs_iops.c | 10 +++++----- include/linux/fs.h | 21 +++++++++++---------- 18 files changed, 57 insertions(+), 56 deletions(-)fs/bad_inode.c +1
-2@@ -120,8 +120,7 @@ static struct posix_acl *bad_inode_get_acl(struct inode *inode, int type) }static int bad_inode_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len) +struct fiemap_extent_info *fieinfo) {return -EIO ; }fs/btrfs/inode.c +3
-2@@ -8600,9 +8600,10 @@ static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)#define BTRFS_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) -static int btrfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len) +static int btrfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;int ret; ret =fiemap_check_flags (fieinfo,BTRFS_FIEMAP_FLAGS );fs/ext2/ext2.h +1
-2@@ -760,8 +760,7 @@ extern int ext2_get_block(struct inode *, sector_t, struct buffer_head *, int);extern int ext2_setattr (struct dentry *,struct iattr *);extern int ext2_getattr (const struct path *,struct kstat *,u32 ,unsigned int );extern void ext2_set_inode_flags (struct inode *inode); -extern int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len); +extern int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);/* ioctl.c */ extern long ext2_ioctl (struct file *,unsigned int ,unsigned long );fs/ext2/inode.c +2
-4@@ -855,11 +855,9 @@ const struct iomap_ops ext2_iomap_ops = {const struct iomap_ops ext2_iomap_ops;#endif /* CONFIG_FS_DAX */ -int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +int ext2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { -return generic_block_fiemap (inode, fieinfo, start, len, - ext2_get_block); +return generic_block_fiemap (inode, fieinfo, ext2_get_block); }static int ext2_writepage (struct page *page,struct writeback_control *wbc)fs/ext4/ext4.h +1
-2@@ -3154,8 +3154,7 @@ extern struct ext4_ext_path *ext4_find_extent(struct inode *, ext4_lblk_t,extern void ext4_ext_drop_refs (struct ext4_ext_path *);extern int ext4_ext_check_inode (struct inode *inode);extern ext4_lblk_t ext4_ext_next_allocated_block (struct ext4_ext_path *path); -extern int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len); +extern int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);extern int ext4_ext_precache (struct inode *inode);extern int ext4_collapse_range (struct inode *inode,loff_t offset,loff_t len);extern int ext4_insert_range (struct inode *inode,loff_t offset,loff_t len);fs/ext4/extents.c +4
-4@@ -5038,9 +5038,10 @@ static int ext4_xattr_fiemap(struct inode *inode,return (error <0 ? error :0 ); } -int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len) +int ext4_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;ext4_lblk_t start_blk;int error =0 ;@@ -5062,8 +5063,7 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,/* fallback to generic here if not in extents fmt */ if (!(ext4_test_inode_flag (inode,EXT4_INODE_EXTENTS ))) -return generic_block_fiemap (inode, fieinfo, start, len, - ext4_get_block); +return generic_block_fiemap (inode, fieinfo, ext4_get_block);if (fiemap_check_flags (fieinfo,EXT4_FIEMAP_FLAGS ))return -EBADR ;fs/f2fs/data.c +3
-2@@ -1409,9 +1409,10 @@ static int f2fs_xattr_fiemap(struct inode *inode,return (err <0 ? err :0 ); } -int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;struct buffer_head map_bh;sector_t start_blk, last_blk;pgoff_t next_pgofs;fs/f2fs/f2fs.h +1
-2@@ -3155,8 +3155,7 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fio);void __do_map_lock (struct f2fs_sb_info *sbi,int flag,bool lock);int f2fs_map_blocks (struct inode *inode,struct f2fs_map_blocks *map,int create,int flag); -int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len); +int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);bool f2fs_should_update_inplace (struct inode *inode,struct f2fs_io_info *fio);bool f2fs_should_update_outplace (struct inode *inode,struct f2fs_io_info *fio); void f2fs_invalidate_page(struct page *page, unsigned int offset,fs/gfs2/inode.c +3
-2@@ -2004,9 +2004,10 @@ static int gfs2_getattr(const struct path *path, struct kstat *stat,return 0 ; } -static int gfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +static int gfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;struct gfs2_inode *ip =GFS2_I (inode);struct gfs2_holder gh;int ret;fs/hpfs/file.c +2
-2@@ -190,9 +190,9 @@ static sector_t _hpfs_bmap(struct address_space *mapping, sector_t block)return generic_block_bmap (mapping, block, hpfs_get_block); } -static int hpfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo,u64 start,u64 len) +static int hpfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { -return generic_block_fiemap (inode, fieinfo, start, len, hpfs_get_block); +return generic_block_fiemap (inode, fieinfo, hpfs_get_block); }const struct address_space_operations hpfs_aops = {fs/ioctl.c +10
-6@@ -210,6 +210,8 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg) fieinfo.fi_flags = fiemap.fm_flags; fieinfo.fi_extents_max = fiemap.fm_extent_count; fieinfo.fi_extents_start = ufiemap->fm_extents; + fieinfo.fi_start = fiemap.fm_start; + fieinfo.fi_len = len;if (fiemap.fm_extent_count !=0 && !access_ok(fieinfo.fi_extents_start,@@ -219,7 +221,7 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg)if (fieinfo.fi_flags &FIEMAP_FLAG_SYNC )filemap_write_and_wait (inode->i_mapping); - error = inode->i_op->fiemap (inode, &fieinfo, fiemap.fm_start, len); + error = inode->i_op->fiemap (inode, &fieinfo); fiemap.fm_flags = fieinfo.fi_flags; fiemap.fm_mapped_extents = fieinfo.fi_extents_mapped;if (copy_to_user (ufiemap, &fiemap,sizeof (fiemap)))@@ -296,9 +298,11 @@ static inline loff_t blk_to_logical(struct inode *inode, sector_t blk)*/ int __generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,loff_t start, -loff_t len,get_block_t *get_block) +struct fiemap_extent_info *fieinfo, +get_block_t *get_block) { +loff_t start = fieinfo->fi_start; +loff_t len = fieinfo->fi_len;struct buffer_head map_bh;sector_t start_blk, last_blk;loff_t isize =i_size_read (inode);@@ -455,12 +459,12 @@ EXPORT_SYMBOL(__generic_block_fiemap);*/ int generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len,get_block_t *get_block) +struct fiemap_extent_info *fieinfo, +get_block_t *get_block) {int ret;inode_lock (inode); - ret =__generic_block_fiemap (inode, fieinfo, start, len, get_block); + ret =__generic_block_fiemap (inode, fieinfo, get_block);inode_unlock (inode);return ret; }fs/nilfs2/inode.c +3
-2@@ -992,9 +992,10 @@ void nilfs_dirty_inode(struct inode *inode, int flags)nilfs_transaction_commit (inode->i_sb);/* never fails */ } -int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len) +int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { +u64 start = fieinfo->fi_start; +u64 len = fieinfo->fi_len;struct the_nilfs *nilfs = inode->i_sb->s_fs_info;__u64 logical =0 , phys =0 , size =0 ;__u32 flags =0 ;fs/nilfs2/nilfs.h +1
-2@@ -276,8 +276,7 @@ extern int nilfs_inode_dirty(struct inode *);int nilfs_set_file_dirty (struct inode *inode,unsigned int nr_dirty);extern int __nilfs_mark_inode_dirty (struct inode *,int );extern void nilfs_dirty_inode (struct inode *,int flags); -int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -__u64 start,__u64 len); +int nilfs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);static inline int nilfs_mark_inode_dirty (struct inode *inode) {return __nilfs_mark_inode_dirty (inode,I_DIRTY );fs/ocfs2/extent_map.c +3
-2@@ -749,8 +749,7 @@ static int ocfs2_fiemap_inline(struct inode *inode, struct buffer_head *di_bh,#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) -int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 map_start,u64 map_len) +int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {int ret, is_last;u32 mapping_end, cpos;@@ -759,6 +758,8 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,u64 len_bytes, phys_bytes, virt_bytes;struct buffer_head *di_bh =NULL ;struct ocfs2_extent_rec rec; +u64 map_start = fieinfo->fi_start; +u64 map_len = fieinfo->fi_len; ret =fiemap_check_flags (fieinfo,OCFS2_FIEMAP_FLAGS );if (ret)fs/ocfs2/extent_map.h +1
-2@@ -50,8 +50,7 @@ int ocfs2_get_clusters(struct inode *inode, u32 v_cluster, u32 *p_cluster,int ocfs2_extent_map_get_blocks (struct inode *inode,u64 v_blkno,u64 *p_blkno,u64 *ret_count,unsigned int *extent_flags); -int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 map_start,u64 map_len); +int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo);int ocfs2_overwrite_io (struct inode *inode,struct buffer_head *di_bh,u64 map_start,u64 map_len);fs/overlayfs/inode.c +2
-3@@ -456,8 +456,7 @@ int ovl_update_time(struct inode *inode, struct timespec64 *ts, int flags)return 0 ; } -static int ovl_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -u64 start,u64 len) +static int ovl_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {int err;struct inode *realinode =ovl_inode_real (inode);@@ -471,7 +470,7 @@ static int ovl_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,if (fieinfo->fi_flags &FIEMAP_FLAG_SYNC )filemap_write_and_wait (realinode->i_mapping); - err = realinode->i_op->fiemap (realinode, fieinfo, start, len); + err = realinode->i_op->fiemap (realinode, fieinfo);revert_creds (old_cred);return err;fs/xfs/xfs_iops.c +0
-1@@ -1109,12 +1109,12 @@ xfs_vn_update_time( STATIC int xfs_vn_fiemap( - struct inode *inode,
fs/xfs/xfs_iops.c xfs_vn_update_time( · line 1117
Quoted 10 lines
fs/xfs/xfs_iops.c +5
-4-struct fiemap_extent_info *fieinfo, - u64 start, - u64 length) +struct inode *inode, + struct fiemap_extent_info *fieinfo) { - int error; + u64 start = fieinfo->fi_start; +u64 length = fieinfo->fi_len; +int error;
Would be nice if the variable name indentation was consistent here, but otherwise the xfs part looks ok.
Quoted 9 lines
fs/xfs/xfs_iops.c +0
-0xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) {include/linux/fs.h +0
-0@@ -1705,11 +1705,14 @@ extern bool may_open_dev(const struct path *path);* VFS FS_IOC_FIEMAP helper definitions.
include/linux/fs.h extern bool may_open_dev(const struct path *path); · line 1710
Quoted 10 lines
include/linux/fs.h +3
-5*/ struct fiemap_extent_info { -unsigned int fi_flags;/* Flags as passed from user */ -unsigned int fi_extents_mapped;/* Number of mapped extents */ -unsigned int fi_extents_max;/* Size of fiemap_extent array */ -struct fiemap_extent __user *fi_extents_start;/* Start of -fiemap_extent array */ +unsigned int fi_flags;/* Flags as passed from user */ +u64 fi_start; +u64 fi_len;
Comments for these two new fields?
--D
Quoted 35 lines
include/linux/fs.h +8
-5+unsigned int fi_extents_mapped;/* Number of mapped extents */ +unsigned int fi_extents_max;/* Size of fiemap_extent array */ +struct fiemap_extent __user *fi_extents_start;/* Start of +fiemap_extent +array */ };int fiemap_fill_next_extent (struct fiemap_extent_info *info,u64 logical,u64 phys,u64 len,u32 flags);@@ -1841,8 +1844,7 @@ struct inode_operations {int (*setattr) (struct dentry *,struct iattr *);int (*getattr) (const struct path *,struct kstat *,u32 ,unsigned int );ssize_t (*listxattr) (struct dentry *,char *,size_t ); -int (*fiemap)(struct inode *,struct fiemap_extent_info *,u64 start, -u64 len); +int (*fiemap)(struct inode *,struct fiemap_extent_info *);int (*update_time)(struct inode *,struct timespec64 *,int ); int (*atomic_open)(struct inode *, struct dentry *, struct file *, unsigned open_flag,@@ -3199,11 +3201,10 @@ extern int vfs_readlink(struct dentry *, char __user *, int);extern int __generic_block_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -loff_t start,loff_t len,get_block_t *get_block);extern int generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len,get_block_t *get_block); +struct fiemap_extent_info *fieinfo, +get_block_t *get_block);extern struct file_system_type *get_filesystem (struct file_system_type *fs);extern void put_filesystem (struct file_system_type *fs);Signature · 1 line
2.20.1
Re: [PATCH 5/9] fs: Move start and length fiemap fields into fiemap_extent_info
Quoted 17 lines
Quoted 4 lines
STATIC int
xfs_vn_fiemap(fs/xfs/xfs_iops.c +0
-1- struct inode *inode,fs/xfs/xfs_iops.c xfs_vn_update_time( · line 1117
Quoted 10 lines
fs/xfs/xfs_iops.c +5
-4-struct fiemap_extent_info *fieinfo, - u64 start, - u64 length) +struct inode *inode, + struct fiemap_extent_info *fieinfo) { - int error; + u64 start = fieinfo->fi_start; +u64 length = fieinfo->fi_len; +int error;Would be nice if the variable name indentation was consistent here, but otherwise the xfs part looks ok.
These fields are removed on the next patch, updating it is really required?
Quoted 14 lines
Quoted 14 lines
fs/xfs/xfs_iops.c +0
-0xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) {include/linux/fs.h +0
-2@@ -1705,11 +1705,14 @@ extern bool may_open_dev(const struct path *path);* VFS FS_IOC_FIEMAP helper definitions. */ struct fiemap_extent_info { -unsigned int fi_flags;/* Flags as passed from user */ -unsigned int fi_extents_mapped;/* Number of mapped extents */
Quoted 8 lines
include/linux/fs.h extern bool may_open_dev(const struct path *path); · line 1710
Quoted 6 lines
include/linux/fs.h +3
-3-unsigned int fi_extents_max;/* Size of fiemap_extent array */ -struct fiemap_extent __user *fi_extents_start;/* Start of -fiemap_extent array */ +unsigned int fi_flags;/* Flags as passed from user */ +u64 fi_start; +u64 fi_len;Comments for these two new fields?
Sure, how about this:
u64 fi_start; /* Logical offset at which
start mapping */
u64 fi_len; /* Logical length of mapping
the caller cares about */
btw, Above indentation won't match final result
Christoph, may I keep your reviewed tag by updating the comments as above? Otherwise I'll just remove your tag
Quoted 38 lines
--D
Quoted 35 lines
include/linux/fs.h +8
-5+unsigned int fi_extents_mapped;/* Number of mapped extents */ +unsigned int fi_extents_max;/* Size of fiemap_extent array */ +struct fiemap_extent __user *fi_extents_start;/* Start of +fiemap_extent +array */ };int fiemap_fill_next_extent (struct fiemap_extent_info *info,u64 logical,u64 phys,u64 len,u32 flags);@@ -1841,8 +1844,7 @@ struct inode_operations {int (*setattr) (struct dentry *,struct iattr *);int (*getattr) (const struct path *,struct kstat *,u32 ,unsigned int );ssize_t (*listxattr) (struct dentry *,char *,size_t ); -int (*fiemap)(struct inode *,struct fiemap_extent_info *,u64 start, -u64 len); +int (*fiemap)(struct inode *,struct fiemap_extent_info *);int (*update_time)(struct inode *,struct timespec64 *,int ); int (*atomic_open)(struct inode *, struct dentry *, struct file *, unsigned open_flag,@@ -3199,11 +3201,10 @@ extern int vfs_readlink(struct dentry *, char __user *, int);extern int __generic_block_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -loff_t start,loff_t len,get_block_t *get_block);extern int generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len,get_block_t *get_block); +struct fiemap_extent_info *fieinfo, +get_block_t *get_block);extern struct file_system_type *get_filesystem (struct file_system_type *fs);extern void put_filesystem (struct file_system_type *fs);Signature · 1 line
2.20.1
Signature · 1 line
Carlos
Re: [PATCH 5/9] fs: Move start and length fiemap fields into fiemap_extent_info
Quoted 18 lines · Carlos Maiolino
Quoted 17 lines
Quoted 4 lines
STATIC int
xfs_vn_fiemap(fs/xfs/xfs_iops.c +0
-1- struct inode *inode,fs/xfs/xfs_iops.c xfs_vn_update_time( · line 1117
Quoted 10 lines
fs/xfs/xfs_iops.c +5
-4-struct fiemap_extent_info *fieinfo, - u64 start, - u64 length) +struct inode *inode, + struct fiemap_extent_info *fieinfo) { - int error; + u64 start = fieinfo->fi_start; +u64 length = fieinfo->fi_len; +int error;Would be nice if the variable name indentation was consistent here, but otherwise the xfs part looks ok.
Quoted 1 line
These fields are removed on the next patch, updating it is really required?
Yes, please.
Quoted 25 lines
Quoted 14 lines
Quoted 14 lines
fs/xfs/xfs_iops.c +0
-0xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) {include/linux/fs.h +0
-2@@ -1705,11 +1705,14 @@ extern bool may_open_dev(const struct path *path);* VFS FS_IOC_FIEMAP helper definitions. */ struct fiemap_extent_info { -unsigned int fi_flags;/* Flags as passed from user */ -unsigned int fi_extents_mapped;/* Number of mapped extents */ Quoted 8 lines
include/linux/fs.h extern bool may_open_dev(const struct path *path); · line 1710
Quoted 6 lines
include/linux/fs.h +3
-3-unsigned int fi_extents_max;/* Size of fiemap_extent array */ -struct fiemap_extent __user *fi_extents_start;/* Start of -fiemap_extent array */ +unsigned int fi_flags;/* Flags as passed from user */ +u64 fi_start; +u64 fi_len;Comments for these two new fields?
Sure, how about this:
Quoted 7 lines
u64
fi_start; /* Logical offset at which
start mapping */
u64fi_len; /* Logical length of mapping
the caller cares about */btw, Above indentation won't match final result
Looks good to me.
--D
Quoted 45 lines
Christoph, may I keep your reviewed tag by updating the comments as above? Otherwise I'll just remove your tag
Quoted 38 lines
--D
Quoted 35 lines
include/linux/fs.h +8
-5+unsigned int fi_extents_mapped;/* Number of mapped extents */ +unsigned int fi_extents_max;/* Size of fiemap_extent array */ +struct fiemap_extent __user *fi_extents_start;/* Start of +fiemap_extent +array */ };int fiemap_fill_next_extent (struct fiemap_extent_info *info,u64 logical,u64 phys,u64 len,u32 flags);@@ -1841,8 +1844,7 @@ struct inode_operations {int (*setattr) (struct dentry *,struct iattr *);int (*getattr) (const struct path *,struct kstat *,u32 ,unsigned int );ssize_t (*listxattr) (struct dentry *,char *,size_t ); -int (*fiemap)(struct inode *,struct fiemap_extent_info *,u64 start, -u64 len); +int (*fiemap)(struct inode *,struct fiemap_extent_info *);int (*update_time)(struct inode *,struct timespec64 *,int ); int (*atomic_open)(struct inode *, struct dentry *, struct file *, unsigned open_flag,@@ -3199,11 +3201,10 @@ extern int vfs_readlink(struct dentry *, char __user *, int);extern int __generic_block_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -loff_t start,loff_t len,get_block_t *get_block);extern int generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len,get_block_t *get_block); +struct fiemap_extent_info *fieinfo, +get_block_t *get_block);extern struct file_system_type *get_filesystem (struct file_system_type *fs);extern void put_filesystem (struct file_system_type *fs);Signature · 1 line
2.20.1Signature · 1 line
Carlos
Re: [PATCH 5/9] fs: Move start and length fiemap fields into fiemap_extent_info
Quoted 19 lines · Darrick J. Wong
Quoted 18 lines · Carlos Maiolino
Quoted 17 lines
Quoted 4 lines
STATIC int
xfs_vn_fiemap(fs/xfs/xfs_iops.c +0
-1- struct inode *inode,fs/xfs/xfs_iops.c xfs_vn_update_time( · line 1117
Quoted 10 lines
fs/xfs/xfs_iops.c +5
-4-struct fiemap_extent_info *fieinfo, - u64 start, - u64 length) +struct inode *inode, + struct fiemap_extent_info *fieinfo) { - int error; + u64 start = fieinfo->fi_start; +u64 length = fieinfo->fi_len; +int error;Would be nice if the variable name indentation was consistent here, but otherwise the xfs part looks ok.
Quoted 3 lines
Quoted 1 line
These fields are removed on the next patch, updating it is really required?
Yes, please.
NP then
Quoted 13 lines
Quoted 13 lines
Quoted 4 lines
include/linux/fs.h extern bool may_open_dev(const struct path *path); · line 1710
Quoted 2 lines
include/linux/fs.h +2
-0+u64 fi_start; +u64 fi_len;Comments for these two new fields?
Sure, how about this:
u64
fi_start; /* Logical offset at which
start mapping */
u64fi_len; /* Logical length of mapping
the caller cares about */
Quoted 3 lines
Quoted 1 line
btw, Above indentation won't match final result
Looks good to me.
Will update the fields, thanks for the review
Quoted 46 lines
Quoted 46 lines
Christoph, may I keep your reviewed tag by updating the comments as above? Otherwise I'll just remove your tag
Quoted 38 lines
--D
Quoted 35 lines
include/linux/fs.h +8
-5+unsigned int fi_extents_mapped;/* Number of mapped extents */ +unsigned int fi_extents_max;/* Size of fiemap_extent array */ +struct fiemap_extent __user *fi_extents_start;/* Start of +fiemap_extent +array */ };int fiemap_fill_next_extent (struct fiemap_extent_info *info,u64 logical,u64 phys,u64 len,u32 flags);@@ -1841,8 +1844,7 @@ struct inode_operations {int (*setattr) (struct dentry *,struct iattr *);int (*getattr) (const struct path *,struct kstat *,u32 ,unsigned int );ssize_t (*listxattr) (struct dentry *,char *,size_t ); -int (*fiemap)(struct inode *,struct fiemap_extent_info *,u64 start, -u64 len); +int (*fiemap)(struct inode *,struct fiemap_extent_info *);int (*update_time)(struct inode *,struct timespec64 *,int ); int (*atomic_open)(struct inode *, struct dentry *, struct file *, unsigned open_flag,@@ -3199,11 +3201,10 @@ extern int vfs_readlink(struct dentry *, char __user *, int);extern int __generic_block_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -loff_t start,loff_t len,get_block_t *get_block);extern int generic_block_fiemap (struct inode *inode, -struct fiemap_extent_info *fieinfo,u64 start, -u64 len,get_block_t *get_block); +struct fiemap_extent_info *fieinfo, +get_block_t *get_block);extern struct file_system_type *get_filesystem (struct file_system_type *fs);extern void put_filesystem (struct file_system_type *fs);Signature · 1 line
2.20.1Signature · 1 line
Carlos
Signature · 1 line
Carlos
Re: [PATCH 5/9] fs: Move start and length fiemap fields into fiemap_extent_info
In reply to Carlos MaiolinoQuoted 2 lines · Carlos Maiolino
Christoph, may I keep your reviewed tag by updating the comments as above? Otherwise I'll just remove your tag
Feel free to keep them for any trivial changes. Indentation changes and adding comments always qualify as trivial.
PATCH v4 6/9 iomap: Remove length and start fields from iomap_fiemap
In reply to Carlos Maiolinofiemap_extent_info now embeds start and length parameters, users of iomap_fiemap() doesn't need to pass it individually anymore.
--- fs/gfs2/inode.c | 4 +--- fs/iomap.c | 4 +++- fs/xfs/xfs_iops.c | 8 ++------ include/linux/iomap.h | 2 +- 4 files changed, 7 insertions(+), 11 deletions(-)
fs/gfs2/inode.c +1 -3
@@ -2006,8 +2006,6 @@ static int gfs2_getattr(const struct path *path, struct kstat *stat,static int gfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { -u64 start = fieinfo->fi_start; -u64 len = fieinfo->fi_len;struct gfs2_inode *ip =GFS2_I (inode);struct gfs2_holder gh;int ret;
@@ -2018,7 +2016,7 @@ static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret) goto out; - ret =iomap_fiemap (inode, fieinfo, start, len, &gfs2_iomap_ops); + ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops);gfs2_glock_dq_uninit (&gh);
fs/iomap.c +3 -1
@@ -1184,9 +1184,11 @@ iomap_fiemap_actor(struct inode *inode, loff_t pos, loff_t length, void *data, }int iomap_fiemap (struct inode *inode,struct fiemap_extent_info *fi, -loff_t start,loff_t len,const struct iomap_ops *ops) +const struct iomap_ops *ops) {struct fiemap_ctx ctx; +loff_t start = fi->fi_start; +loff_t len = fi->fi_len;loff_t ret;memset (&ctx,0 ,sizeof (ctx));
fs/xfs/xfs_iops.c +2 -6
@@ -1112,18 +1112,14 @@ xfs_vn_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) { - u64 start = fieinfo->fi_start; -u64 length = fieinfo->fi_len; int error;xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) { fieinfo->fi_flags &= ~FIEMAP_FLAG_XATTR ; - error =iomap_fiemap (inode, fieinfo, start, length, - &xfs_xattr_iomap_ops); + error =iomap_fiemap (inode, fieinfo, &xfs_xattr_iomap_ops); }else { - error =iomap_fiemap (inode, fieinfo, start, length, - &xfs_iomap_ops); + error =iomap_fiemap (inode, fieinfo, &xfs_iomap_ops); }xfs_iunlock (XFS_I (inode),XFS_IOLOCK_SHARED );
include/linux/iomap.h +1 -1
@@ -145,7 +145,7 @@ int iomap_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,vm_fault_t iomap_page_mkwrite (struct vm_fault *vmf,const struct iomap_ops *ops);int iomap_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -loff_t start,loff_t len,const struct iomap_ops *ops); +const struct iomap_ops *ops);loff_t iomap_seek_hole (struct inode *inode,loff_t offset,const struct iomap_ops *ops); loff_t iomap_seek_data(struct inode *inode, loff_t offset,
Signature · 1 line
2.20.1
Re: [PATCH 6/9] iomap: Remove length and start fields from iomap_fiemap
Quoted 5 lines · Carlos Maiolino
fiemap_extent_infonow embeds start and length parameters, users ofiomap_fiemap()doesn't need to pass it individually anymore.Reviewed-by Christoph Hellwig <[email protected]>Signed-off-by Carlos Maiolino <[email protected]>
Looks ok,
--D
Quoted 87 lines
--- fs/gfs2/inode.c | 4 +--- fs/iomap.c | 4 +++- fs/xfs/xfs_iops.c | 8 ++------ include/linux/iomap.h | 2 +- 4 files changed, 7 insertions(+), 11 deletions(-)fs/gfs2/inode.c +1
-3@@ -2006,8 +2006,6 @@ static int gfs2_getattr(const struct path *path, struct kstat *stat,static int gfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) { -u64 start = fieinfo->fi_start; -u64 len = fieinfo->fi_len;struct gfs2_inode *ip =GFS2_I (inode);struct gfs2_holder gh;int ret;@@ -2018,7 +2016,7 @@ static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret) goto out; - ret =iomap_fiemap (inode, fieinfo, start, len, &gfs2_iomap_ops); + ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops);gfs2_glock_dq_uninit (&gh);fs/iomap.c +3
-1@@ -1184,9 +1184,11 @@ iomap_fiemap_actor(struct inode *inode, loff_t pos, loff_t length, void *data, }int iomap_fiemap (struct inode *inode,struct fiemap_extent_info *fi, -loff_t start,loff_t len,const struct iomap_ops *ops) +const struct iomap_ops *ops) {struct fiemap_ctx ctx; +loff_t start = fi->fi_start; +loff_t len = fi->fi_len;loff_t ret;memset (&ctx,0 ,sizeof (ctx));fs/xfs/xfs_iops.c +2
-6@@ -1112,18 +1112,14 @@ xfs_vn_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) { - u64 start = fieinfo->fi_start; -u64 length = fieinfo->fi_len; int error;xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) { fieinfo->fi_flags &= ~FIEMAP_FLAG_XATTR ; - error =iomap_fiemap (inode, fieinfo, start, length, - &xfs_xattr_iomap_ops); + error =iomap_fiemap (inode, fieinfo, &xfs_xattr_iomap_ops); }else { - error =iomap_fiemap (inode, fieinfo, start, length, - &xfs_iomap_ops); + error =iomap_fiemap (inode, fieinfo, &xfs_iomap_ops); }xfs_iunlock (XFS_I (inode),XFS_IOLOCK_SHARED );include/linux/iomap.h +1
-1@@ -145,7 +145,7 @@ int iomap_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,vm_fault_t iomap_page_mkwrite (struct vm_fault *vmf,const struct iomap_ops *ops);int iomap_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo, -loff_t start,loff_t len,const struct iomap_ops *ops); +const struct iomap_ops *ops);loff_t iomap_seek_hole (struct inode *inode,loff_t offset,const struct iomap_ops *ops); loff_t iomap_seek_data(struct inode *inode, loff_t offset,Signature · 1 line
2.20.1
PATCH v4 7/9 fiemap: Use a callback to fill fiemap extents
In reply to Carlos MaiolinoAs a goal to enable fiemap infrastructure to be used by fibmap too, we need a way to use different helpers to fill extent data, depending on its usage. One helper to fill extent data stored in user address space (used in fiemap), and another fo fill extent data stored in kernel address space (will be used in fibmap).
This patch sets up the usage of a callback to be used to fill in the extents. It transforms the current fiemap_fill_next_extent, into a simple helper to call the callback, avoiding unneeded changes on any filesystem, and reutilizes the original function as the callback used by FIEMAP.
---
Changelog:
V3:
- Rebase to current linux-next
- Fix conflict on rebase
- Merge this patch into patch 07 from V2
- Rename fi_extents_start to fi_cb_data
V2:
- Now based on the rework on fiemap_extent_info (previous was based on fiemap_ctx)
fs/ioctl.c | 45 ++++++++++++++++++++++++++------------------- include/linux/fs.h | 12 +++++++++--- 2 files changed, 35 insertions(+), 22 deletions(-)
fs/ioctl.c +26 -19
@@ -77,29 +77,14 @@ static int ioctl_fibmap(struct file *filp, int __user *p)return error; } -/** -* fiemap_fill_next_extent - Fiemap helper function -* @fieinfo: Fiemap context passed into ->fiemap -* @logical: Extent logical start offset, in bytes -* @phys: Extent physical start offset, in bytes -* @len: Extent length, in bytes -* @flags: FIEMAP_EXTENT flags that describe this extent -* -* Called from file system ->fiemap callback. Will populate extent -* info as passed in via arguments and copy to user memory. On -* success, extent count on fieinfo is incremented. -* -* Returns 0 on success, -errno on error, 1 if this was the last -* extent that will fit in user array. -*/ #define SET_UNKNOWN_FLAGS (FIEMAP_EXTENT_DELALLOC)#define SET_NO_UNMOUNTED_IO_FLAGS (FIEMAP_EXTENT_DATA_ENCRYPTED)#define SET_NOT_ALIGNED_FLAGS (FIEMAP_EXTENT_DATA_TAIL|FIEMAP_EXTENT_DATA_INLINE) -int fiemap_fill_next_extent (struct fiemap_extent_info *fieinfo,u64 logical, +int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo,u64 logical,u64 phys,u64 len,u32 flags) {struct fiemap_extent extent; -struct fiemap_extent __user *dest = fieinfo->fi_extents_start; +struct fiemap_extent __user *dest = fieinfo->fi_cb_data;/* only count the extents */ if (fieinfo->fi_extents_max ==0 ) {
@@ -132,6 +117,27 @@ int fiemap_fill_next_extent(struct fiemap_extent_info *fieinfo, u64 logical,return 1 ;return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; } + +/** +* fiemap_fill_next_extent - Fiemap helper function +* @fieinfo: Fiemap context passed into ->fiemap +* @logical: Extent logical start offset, in bytes +* @phys: Extent physical start offset, in bytes +* @len: Extent length, in bytes +* @flags: FIEMAP_EXTENT flags that describe this extent +* +* Called from file system ->fiemap callback. Will populate extent +* info as passed in via arguments and copy to user memory. On +* success, extent count on fieinfo is incremented. +* +* Returns 0 on success, -errno on error, 1 if this was the last +* extent that will fit in user array. +*/ +int fiemap_fill_next_extent (struct fiemap_extent_info *fieinfo,u64 logical, +u64 phys,u64 len,u32 flags) +{ +return fieinfo->fi_cb (fieinfo, logical, phys, len, flags); +}EXPORT_SYMBOL (fiemap_fill_next_extent);/**
@@ -209,12 +215,13 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg) fieinfo.fi_flags = fiemap.fm_flags; fieinfo.fi_extents_max = fiemap.fm_extent_count; - fieinfo.fi_extents_start = ufiemap->fm_extents; + fieinfo.fi_cb_data = ufiemap->fm_extents; fieinfo.fi_start = fiemap.fm_start; fieinfo.fi_len = len; + fieinfo.fi_cb = fiemap_fill_user_extent;if (fiemap.fm_extent_count !=0 && - !access_ok (fieinfo.fi_extents_start, + !access_ok (fieinfo.fi_cb_data, fieinfo.fi_extents_max *sizeof (struct fiemap_extent )))return -EFAULT ;
include/linux/fs.h +9 -3
@@ -66,6 +66,7 @@ struct fscrypt_info;struct fscrypt_operations ;struct fs_context ;struct fs_parameter_description ; +struct fiemap_extent_info ;extern void __init inode_init (void );extern void __init inode_init_early (void );
@@ -1704,16 +1705,21 @@ extern bool may_open_dev(const struct path *path);/* * VFS FS_IOC_FIEMAP helper definitions. */ + +typedef int (*fiemap_fill_cb )(struct fiemap_extent_info *fieinfo,u64 logical, +u64 phys,u64 len,u32 flags); +struct fiemap_extent_info {unsigned int fi_flags;/* Flags as passed from user */ u64 fi_start;u64 fi_len;unsigned int fi_extents_mapped;/* Number of mapped extents */ unsigned int fi_extents_max;/* Size of fiemap_extent array */ -struct fiemap_extent __user *fi_extents_start;/* Start of -fiemap_extent -array */ +void *fi_cb_data;/* Start of fiemap_extent +array */ +fiemap_fill_cb fi_cb; }; +int fiemap_fill_next_extent (struct fiemap_extent_info *info,u64 logical,u64 phys,u64 len,u32 flags);int fiemap_check_flags (struct fiemap_extent_info *fieinfo,u32 fs_flags);
Signature · 1 line
2.20.1
Re: [PATCH 7/9] fiemap: Use a callback to fill fiemap extents
Quoted 6 lines · Carlos Maiolino
As a goal to enable fiemap infrastructure to be used by fibmap too, we need a way to use different helpers to fill extent data, depending on its usage. One helper to fill extent data stored in user address space (used in fiemap), and another fo fill extent data stored in kernel address space (will be used in fibmap).
Quoted 6 lines
This patch sets up the usage of a callback to be used to fill in the extents. It transforms the current
fiemap_fill_next_extent, into a simple helper to call the callback, avoiding unneeded changes on any filesystem, and reutilizes the original function as the callback used by FIEMAP.Signed-off-by Carlos Maiolino <[email protected]>
Looks ok,
--D
Quoted 138 lines
---
Changelog:
V3:
- Rebase to current linux-next
- Fix conflict on rebase
- Merge this patch into patch 07 from V2
- Renamefi_extents_starttofi_cb_dataV2:
- Now based on the rework onfiemap_extent_info(previous was based onfiemap_ctx)fs/ioctl.c | 45 ++++++++++++++++++++++++++------------------- include/linux/fs.h | 12 +++++++++--- 2 files changed, 35 insertions(+), 22 deletions(-)fs/ioctl.c +26
-19@@ -77,29 +77,14 @@ static int ioctl_fibmap(struct file *filp, int __user *p)return error; } -/** -* fiemap_fill_next_extent - Fiemap helper function -* @fieinfo: Fiemap context passed into ->fiemap -* @logical: Extent logical start offset, in bytes -* @phys: Extent physical start offset, in bytes -* @len: Extent length, in bytes -* @flags: FIEMAP_EXTENT flags that describe this extent -* -* Called from file system ->fiemap callback. Will populate extent -* info as passed in via arguments and copy to user memory. On -* success, extent count on fieinfo is incremented. -* -* Returns 0 on success, -errno on error, 1 if this was the last -* extent that will fit in user array. -*/ #define SET_UNKNOWN_FLAGS (FIEMAP_EXTENT_DELALLOC)#define SET_NO_UNMOUNTED_IO_FLAGS (FIEMAP_EXTENT_DATA_ENCRYPTED)#define SET_NOT_ALIGNED_FLAGS (FIEMAP_EXTENT_DATA_TAIL|FIEMAP_EXTENT_DATA_INLINE) -int fiemap_fill_next_extent (struct fiemap_extent_info *fieinfo,u64 logical, +int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo,u64 logical,u64 phys,u64 len,u32 flags) {struct fiemap_extent extent; -struct fiemap_extent __user *dest = fieinfo->fi_extents_start; +struct fiemap_extent __user *dest = fieinfo->fi_cb_data;/* only count the extents */ if (fieinfo->fi_extents_max ==0 ) {@@ -132,6 +117,27 @@ int fiemap_fill_next_extent(struct fiemap_extent_info *fieinfo, u64 logical,return 1 ;return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; } + +/** +* fiemap_fill_next_extent - Fiemap helper function +* @fieinfo: Fiemap context passed into ->fiemap +* @logical: Extent logical start offset, in bytes +* @phys: Extent physical start offset, in bytes +* @len: Extent length, in bytes +* @flags: FIEMAP_EXTENT flags that describe this extent +* +* Called from file system ->fiemap callback. Will populate extent +* info as passed in via arguments and copy to user memory. On +* success, extent count on fieinfo is incremented. +* +* Returns 0 on success, -errno on error, 1 if this was the last +* extent that will fit in user array. +*/ +int fiemap_fill_next_extent (struct fiemap_extent_info *fieinfo,u64 logical, +u64 phys,u64 len,u32 flags) +{ +return fieinfo->fi_cb (fieinfo, logical, phys, len, flags); +}EXPORT_SYMBOL (fiemap_fill_next_extent);/** @@ -209,12 +215,13 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg) fieinfo.fi_flags = fiemap.fm_flags; fieinfo.fi_extents_max = fiemap.fm_extent_count; - fieinfo.fi_extents_start = ufiemap->fm_extents; + fieinfo.fi_cb_data = ufiemap->fm_extents; fieinfo.fi_start = fiemap.fm_start; fieinfo.fi_len = len; + fieinfo.fi_cb = fiemap_fill_user_extent;if (fiemap.fm_extent_count !=0 && - !access_ok (fieinfo.fi_extents_start, + !access_ok (fieinfo.fi_cb_data, fieinfo.fi_extents_max *sizeof (struct fiemap_extent )))return -EFAULT ;include/linux/fs.h +9
-3@@ -66,6 +66,7 @@ struct fscrypt_info;struct fscrypt_operations ;struct fs_context ;struct fs_parameter_description ; +struct fiemap_extent_info ;extern void __init inode_init (void );extern void __init inode_init_early (void );@@ -1704,16 +1705,21 @@ extern bool may_open_dev(const struct path *path);/* * VFS FS_IOC_FIEMAP helper definitions. */ + +typedef int (*fiemap_fill_cb )(struct fiemap_extent_info *fieinfo,u64 logical, +u64 phys,u64 len,u32 flags); +struct fiemap_extent_info {unsigned int fi_flags;/* Flags as passed from user */ u64 fi_start;u64 fi_len;unsigned int fi_extents_mapped;/* Number of mapped extents */ unsigned int fi_extents_max;/* Size of fiemap_extent array */ -struct fiemap_extent __user *fi_extents_start;/* Start of -fiemap_extent -array */ +void *fi_cb_data;/* Start of fiemap_extent +array */ +fiemap_fill_cb fi_cb; }; +int fiemap_fill_next_extent (struct fiemap_extent_info *info,u64 logical,u64 phys,u64 len,u32 flags);int fiemap_check_flags (struct fiemap_extent_info *fieinfo,u32 fs_flags);Signature · 1 line
2.20.1
PATCH v4 8/9 Use FIEMAP for FIBMAP calls
In reply to Carlos MaiolinoEnables the usage of FIEMAP ioctl infrastructure to handle FIBMAP calls.
Quoted 1 line
From now on, ->bmap() methods can start to be removed from filesystems
which already provides ->fiemap().
This adds a new helper - bmap_fiemap() - which is used to fill in the fiemap request, call into the underlying filesystem and check the flags set in the extent requested.
Add a new fiemap fill extent callback to handle the in-kernel only fiemap_extent structure used for FIBMAP.
The new FIEMAP_KERNEL_FIBMAP flag, is used to tell the filesystem ->fiemap interface, that the call is coming from ioctl_fibmap. The addition of this new flag, requires an update to fiemap_check_flags(), so it doesn't treat FIBMAP requests as invalid.
---
Changelog:
V4:
- Fix if conditional in bmap()
- Add filesystem-specific modifications
V3:
- Add FIEMAP_EXTENT_SHARED to the list of invalid extents in bmap_fiemap()
- Rename fi_extents_start to fi_cb_data
- Use if conditional instead of ternary operator
- Make fiemap_fill_* callbacks static (which required the removal of some macros
- Set FIEMAP_FLAG_SYNC when calling in ->fiemap method from fibmap
- Add FIEMAP_KERNEL_FIBMAP flag, to identify the usage of fiemap infrastructure for fibmap calls, defined in fs.h so it's not exported to userspace.
- Update fiemap_check_flags() to understand FIEMAP_KERNEL_FIBMAP
- Update filesystems supporting both FIBMAP and FIEMAP, which need extra checks on FIBMAP calls
V2:
- Now based on the updated fiemap_extent_info,
- move the fiemap call itself to a new helper
fs/ext4/extents.c | 7 +++- fs/f2fs/data.c | 10 +++++- fs/gfs2/inode.c | 6 +++- fs/inode.c | 81 +++++++++++++++++++++++++++++++++++++++++-- fs/ioctl.c | 40 ++++++++++++++------- fs/iomap.c | 2 +- fs/ocfs2/extent_map.c | 8 ++++- fs/xfs/xfs_iops.c | 5 +++ include/linux/fs.h | 4 +++ 9 files changed, 144 insertions(+), 19 deletions(-)
fs/ext4/extents.c +6 -1
@@ -5001,7 +5001,9 @@ static int ext4_find_delayed_extent(struct inode *inode,return next_del; }/* fiemap flags we can handle specified here */ -#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) +#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)static int ext4_xattr_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo)
@@ -5048,6 +5050,9 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ext4_has_inline_data (inode)) {int has_inline =1 ; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +return -EINVAL ; + error =ext4_inline_data_fiemap (inode, fieinfo, &has_inline, start, len);
fs/f2fs/data.c +9 -1
@@ -1409,6 +1409,9 @@ static int f2fs_xattr_fiemap(struct inode *inode,return (err <0 ? err :0 ); } +#define F2FS_FIEMAP_COMPAT (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {u64 start = fieinfo->fi_start;
@@ -1426,7 +1429,7 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)return ret; } - ret =fiemap_check_flags (fieinfo,FIEMAP_FLAG_SYNC |FIEMAP_FLAG_XATTR ); + ret =fiemap_check_flags (fieinfo,F2FS_FIEMAP_COMPAT );if (ret)return ret;
@@ -1438,6 +1441,11 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) }if (f2fs_has_inline_data (inode)) { + + ret = -EINVAL ; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) + goto out; + ret =f2fs_inline_data_fiemap (inode, fieinfo, start, len);if (ret != -EAGAIN ) goto out;
fs/gfs2/inode.c +5 -1
@@ -2016,7 +2016,11 @@ static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret) goto out; - ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops); +if (gfs2_is_stuffed (ip) && + (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP )) + ret = -EINVAL ; +else + ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops);gfs2_glock_dq_uninit (&gh);
fs/inode.c +79 -2
95 lines not on this page · raw message
fs/ioctl.c +28 -12
@@ -77,11 +77,8 @@ static int ioctl_fibmap(struct file *filp, int __user *p)return error; } -#define SET_UNKNOWN_FLAGS (FIEMAP_EXTENT_DELALLOC) -#define SET_NO_UNMOUNTED_IO_FLAGS (FIEMAP_EXTENT_DATA_ENCRYPTED) -#define SET_NOT_ALIGNED_FLAGS (FIEMAP_EXTENT_DATA_TAIL|FIEMAP_EXTENT_DATA_INLINE) -int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo,u64 logical, -u64 phys,u64 len,u32 flags) +static int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo, +u64 logical,u64 phys,u64 len,u32 flags) {struct fiemap_extent extent;struct fiemap_extent __user *dest = fieinfo->fi_cb_data;
@@ -89,17 +86,17 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical,/* only count the extents */ if (fieinfo->fi_extents_max ==0 ) { fieinfo->fi_extents_mapped++; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + goto out; }if (fieinfo->fi_extents_mapped >= fieinfo->fi_extents_max)return 1 ; -if (flags &SET_UNKNOWN_FLAGS ) +if (flags &FIEMAP_EXTENT_DELALLOC ) flags |=FIEMAP_EXTENT_UNKNOWN ; -if (flags &SET_NO_UNMOUNTED_IO_FLAGS ) +if (flags &FIEMAP_EXTENT_DATA_ENCRYPTED ) flags |=FIEMAP_EXTENT_ENCODED ; -if (flags &SET_NOT_ALIGNED_FLAGS ) +if (flags & (FIEMAP_EXTENT_DATA_TAIL |FIEMAP_EXTENT_DATA_INLINE )) flags |=FIEMAP_EXTENT_NOT_ALIGNED ;memset (&extent,0 ,sizeof (extent));
@@ -115,7 +112,11 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, fieinfo->fi_extents_mapped++;if (fieinfo->fi_extents_mapped == fieinfo->fi_extents_max)return 1 ; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + +out: +if (flags &FIEMAP_EXTENT_LAST ) +return 1 ; +return 0 ; }/**
@@ -151,13 +152,23 @@ EXPORT_SYMBOL(fiemap_fill_next_extent);* flags, the invalid values will be written into the fieinfo structure, and * -EBADR is returned, which tells ioctl_fiemap() to return those values to * userspace. For this reason, a return code of -EBADR should be preserved. +* In case ->fiemap is being used for FIBMAP calls, and the filesystem does not +* support it, return -EINVAL. * -* Returns 0 on success, -EBADR on bad flags. +* Returns 0 on success, -EBADR on bad flags, -EINVAL for an unsupported FIBMAP +* request. */ int fiemap_check_flags (struct fiemap_extent_info *fieinfo,u32 fs_flags) {u32 incompat_flags; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) { +if (fs_flags &FIEMAP_KERNEL_FIBMAP ) +return 0 ; + +return -EINVAL ; + } + incompat_flags = fieinfo->fi_flags & ~(FIEMAP_FLAGS_COMPAT & fs_flags);if (incompat_flags) { fieinfo->fi_flags = incompat_flags;
@@ -208,6 +219,10 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg)if (fiemap.fm_extent_count >FIEMAP_MAX_EXTENTS )return -EINVAL ; +/* Userspace has no access to this flag */ +if (fiemap.fm_flags &FIEMAP_KERNEL_FIBMAP ) +return -EINVAL ; + error =fiemap_check_ranges (sb, fiemap.fm_start, fiemap.fm_length, &len);if (error)
@@ -318,7 +333,8 @@ int __generic_block_fiemap(struct inode *inode,bool past_eof = false, whole_file = false;int ret =0 ; - ret =fiemap_check_flags (fieinfo,FIEMAP_FLAG_SYNC ); + ret =fiemap_check_flags (fieinfo, +FIEMAP_FLAG_SYNC |FIEMAP_KERNEL_FIBMAP );if (ret)return ret;
fs/iomap.c +1 -1
@@ -1195,7 +1195,7 @@ int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fi, ctx.fi = fi; ctx.prev.type =IOMAP_HOLE ; - ret =fiemap_check_flags (fi,FIEMAP_FLAG_SYNC ); + ret =fiemap_check_flags (fi,FIEMAP_FLAG_SYNC |FIEMAP_KERNEL_FIBMAP );if (ret)return ret;
fs/ocfs2/extent_map.c +7 -1
@@ -747,7 +747,7 @@ static int ocfs2_fiemap_inline(struct inode *inode, struct buffer_head *di_bh,return 0 ; } -#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) +#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | FIEMAP_KERNEL_FIBMAP)int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {
@@ -756,6 +756,7 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)unsigned int hole_size;struct ocfs2_super *osb =OCFS2_SB (inode->i_sb);u64 len_bytes, phys_bytes, virt_bytes; +struct buffer_head *di_bh =NULL ;struct ocfs2_extent_rec rec;u64 map_start = fieinfo->fi_start;
@@ -765,6 +766,11 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret)return ret; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) { +if (ocfs2_is_refcount_inode (inode)) +return -EINVAL ; + } + ret =ocfs2_inode_lock (inode, &di_bh,0 );if (ret) {mlog_errno (ret);
fs/xfs/xfs_iops.c +5 -0
@@ -1113,6 +1113,11 @@ xfs_vn_fiemap( struct fiemap_extent_info *fieinfo) {int error; +struct xfs_inode *ip =XFS_I (inode); + +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +if (xfs_is_reflink_inode (ip) ||XFS_IS_REALTIME_INODE (ip)) +return -EINVAL ;xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) {
include/linux/fs.h +4 -0
@@ -1709,6 +1709,10 @@ extern bool may_open_dev(const struct path *path);typedef int (*fiemap_fill_cb )(struct fiemap_extent_info *fieinfo,u64 logical,u64 phys,u64 len,u32 flags); +#define FIEMAP_KERNEL_FIBMAP 0x10000000/* FIBMAP call through FIEMAP +interface. This is a kernel +only flag */ +struct fiemap_extent_info {unsigned int fi_flags;/* Flags as passed from user */ u64 fi_start;
Signature · 1 line
2.20.1
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
Quoted 71 lines · Carlos Maiolino
Enables the usage of FIEMAP ioctl infrastructure to handle FIBMAP calls. From now on, ->bmap() methods can start to be removed from filesystems which already provides ->fiemap().
This adds a new helper -
bmap_fiemap()- which is used to fill in the fiemap request, call into the underlying filesystem and check the flags set in the extent requested.Add a new fiemap fill extent callback to handle the in-kernel only
fiemap_extentstructure used for FIBMAP.The new
FIEMAP_KERNEL_FIBMAPflag, is used to tell the filesystem ->fiemap interface, that the call is coming fromioctl_fibmap. The addition of this new flag, requires an update tofiemap_check_flags(), so it doesn't treat FIBMAP requests as invalid.Signed-off-by Carlos Maiolino <[email protected]>---
Changelog:
V4:
- Fix if conditional inbmap()
- Add filesystem-specific modifications
V3:
- AddFIEMAP_EXTENT_SHAREDto the list of invalid extents inbmap_fiemap()
- Renamefi_extents_starttofi_cb_data
- Use if conditional instead of ternary operator
- Make fiemap_fill_* callbacks static (which required the removal of some macros
- SetFIEMAP_FLAG_SYNCwhen calling in ->fiemap method from fibmap
- AddFIEMAP_KERNEL_FIBMAPflag, to identify the usage of fiemap infrastructure for fibmap calls, defined infs.hso it's not exported to userspace.
- Updatefiemap_check_flags()to understandFIEMAP_KERNEL_FIBMAP
- Update filesystems supporting both FIBMAP and FIEMAP, which need extra checks on FIBMAP callsV2:
- Now based on the updatedfiemap_extent_info,
- move the fiemap call itself to a new helperfs/ext4/extents.c | 7 +++- fs/f2fs/data.c | 10 +++++- fs/gfs2/inode.c | 6 +++- fs/inode.c | 81 +++++++++++++++++++++++++++++++++++++++++-- fs/ioctl.c | 40 ++++++++++++++------- fs/iomap.c | 2 +- fs/ocfs2/extent_map.c | 8 ++++- fs/xfs/xfs_iops.c | 5 +++ include/linux/fs.h | 4 +++ 9 files changed, 144 insertions(+), 19 deletions(-)fs/ext4/extents.c +3
-1@@ -5001,7 +5001,9 @@ static int ext4_find_delayed_extent(struct inode *inode,return next_del; }/* fiemap flags we can handle specified here */ -#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) +#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)static int ext4_xattr_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo)
fs/ext4/extents.c int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) · line 5054
Quoted 5 lines
fs/ext4/extents.c +2
-0@@ -5048,6 +5050,9 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ext4_has_inline_data (inode)) {int has_inline =1 ; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +return -EINVAL ;
Wouldn't the inline data case be caught by fiemap_bmap and turned into -EINVAL?
Quoted 183 lines
fs/ext4/extents.c +1
-0+ error =ext4_inline_data_fiemap (inode, fieinfo, &has_inline, start, len);fs/f2fs/data.c +9
-1@@ -1409,6 +1409,9 @@ static int f2fs_xattr_fiemap(struct inode *inode,return (err <0 ? err :0 ); } +#define F2FS_FIEMAP_COMPAT (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {u64 start = fieinfo->fi_start;@@ -1426,7 +1429,7 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)return ret; } - ret =fiemap_check_flags (fieinfo,FIEMAP_FLAG_SYNC |FIEMAP_FLAG_XATTR ); + ret =fiemap_check_flags (fieinfo,F2FS_FIEMAP_COMPAT );if (ret)return ret;@@ -1438,6 +1441,11 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) }if (f2fs_has_inline_data (inode)) { + + ret = -EINVAL ; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) + goto out; + ret =f2fs_inline_data_fiemap (inode, fieinfo, start, len);if (ret != -EAGAIN ) goto out;fs/gfs2/inode.c +5
-1@@ -2016,7 +2016,11 @@ static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret) goto out; - ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops); +if (gfs2_is_stuffed (ip) && + (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP )) + ret = -EINVAL ; +else + ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops);gfs2_glock_dq_uninit (&gh);fs/inode.c +79
-2@@ -1575,6 +1575,78 @@ void iput(struct inode *inode) }EXPORT_SYMBOL (iput); +static int fiemap_fill_kernel_extent (struct fiemap_extent_info *fieinfo, +u64 logical,u64 phys,u64 len,u32 flags) +{ +struct fiemap_extent *extent = fieinfo->fi_cb_data; + +/* only count the extents */ +if (fieinfo->fi_cb_data ==0 ) { + fieinfo->fi_extents_mapped++; + goto out; + } + +if (fieinfo->fi_extents_mapped >= fieinfo->fi_extents_max) +return 1 ; + +if (flags &FIEMAP_EXTENT_DELALLOC ) + flags |=FIEMAP_EXTENT_UNKNOWN ; +if (flags &FIEMAP_EXTENT_DATA_ENCRYPTED ) + flags |=FIEMAP_EXTENT_ENCODED ; +if (flags & (FIEMAP_EXTENT_DATA_TAIL |FIEMAP_EXTENT_DATA_INLINE )) + flags |=FIEMAP_EXTENT_NOT_ALIGNED ; + + extent->fe_logical = logical; + extent->fe_physical = phys; + extent->fe_length = len; + extent->fe_flags = flags; + + fieinfo->fi_extents_mapped++; + +if (fieinfo->fi_extents_mapped == fieinfo->fi_extents_max) +return 1 ; + +out: +if (flags &FIEMAP_EXTENT_LAST ) +return 1 ; +return 0 ; +} + +static int bmap_fiemap (struct inode *inode,sector_t *block) +{ +struct fiemap_extent_info fieinfo = {0 , }; +struct fiemap_extent fextent; +u64 start = *block << inode->i_blkbits; +int error = -EINVAL ; + + fextent.fe_logical =0 ; + fextent.fe_physical =0 ; + fieinfo.fi_extents_max =1 ; + fieinfo.fi_extents_mapped =0 ; + fieinfo.fi_cb_data = &fextent; + fieinfo.fi_start = start; + fieinfo.fi_len =1 << inode->i_blkbits; + fieinfo.fi_cb = fiemap_fill_kernel_extent; + fieinfo.fi_flags = (FIEMAP_KERNEL_FIBMAP |FIEMAP_FLAG_SYNC ); + + error = inode->i_op->fiemap (inode, &fieinfo); + +if (error) +return error; + +if (fieinfo.fi_flags & (FIEMAP_EXTENT_UNKNOWN | +FIEMAP_EXTENT_ENCODED | +FIEMAP_EXTENT_DATA_INLINE | +FIEMAP_EXTENT_UNWRITTEN | +FIEMAP_EXTENT_SHARED )) +return -EINVAL ; + + *block = (fextent.fe_physical + + (start - fextent.fe_logical)) >> inode->i_blkbits; + +return error; +} +/** * bmap - find a block number in a file * @inode: inode owning the block number being requested @@ -1591,10 +1663,15 @@ EXPORT_SYMBOL(iput);*/ int bmap (struct inode *inode,sector_t *block) { -if (!inode->i_mapping->a_ops->bmap) +if (inode->i_op->fiemap) +return bmap_fiemap (inode, block); + +if (inode->i_mapping->a_ops->bmap) + *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, + *block); +else return -EINVAL ; - *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, *block);return 0 ; }EXPORT_SYMBOL (bmap);fs/ioctl.c +3
-6@@ -77,11 +77,8 @@ static int ioctl_fibmap(struct file *filp, int __user *p)return error; } -#define SET_UNKNOWN_FLAGS (FIEMAP_EXTENT_DELALLOC) -#define SET_NO_UNMOUNTED_IO_FLAGS (FIEMAP_EXTENT_DATA_ENCRYPTED) -#define SET_NOT_ALIGNED_FLAGS (FIEMAP_EXTENT_DATA_TAIL|FIEMAP_EXTENT_DATA_INLINE) -int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo,u64 logical, -u64 phys,u64 len,u32 flags) +static int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo, +u64 logical,u64 phys,u64 len,u32 flags) {struct fiemap_extent extent;struct fiemap_extent __user *dest = fieinfo->fi_cb_data;@@ -89,17 +86,17 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical,/* only count the extents */ if (fieinfo->fi_extents_max ==0 ) { fieinfo->fi_extents_mapped++; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + goto out; }
fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 102
Quoted 10 lines
fs/ioctl.c +2
-3if (fieinfo->fi_extents_mapped >= fieinfo->fi_extents_max)return 1 ; -if (flags &SET_UNKNOWN_FLAGS ) +if (flags &FIEMAP_EXTENT_DELALLOC ) flags |=FIEMAP_EXTENT_UNKNOWN ; -if (flags &SET_NO_UNMOUNTED_IO_FLAGS ) +if (flags &FIEMAP_EXTENT_DATA_ENCRYPTED ) flags |=FIEMAP_EXTENT_ENCODED ; -if (flags &SET_NOT_ALIGNED_FLAGS )
It's too bad that we lose the "not aligned" semantic meaning here.
fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 100
Quoted 2 lines
fs/ioctl.c +1
-0+if (flags & (FIEMAP_EXTENT_DATA_TAIL |FIEMAP_EXTENT_DATA_INLINE )) flags |=FIEMAP_EXTENT_NOT_ALIGNED ;
Why doesn't this function just call fiemap_fill_kernel_extent to fill out the onstack @extent structure? We've now implemented "fill out out a struct fiemap_extent" twice.
Quoted 112 lines
fs/ioctl.c +22
-3memset (&extent,0 ,sizeof (extent));@@ -115,7 +112,11 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, fieinfo->fi_extents_mapped++;if (fieinfo->fi_extents_mapped == fieinfo->fi_extents_max)return 1 ; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + +out: +if (flags &FIEMAP_EXTENT_LAST ) +return 1 ; +return 0 ; }/** @@ -151,13 +152,23 @@ EXPORT_SYMBOL(fiemap_fill_next_extent);* flags, the invalid values will be written into the fieinfo structure, and * -EBADR is returned, which tells ioctl_fiemap() to return those values to * userspace. For this reason, a return code of -EBADR should be preserved. +* In case ->fiemap is being used for FIBMAP calls, and the filesystem does not +* support it, return -EINVAL. * -* Returns 0 on success, -EBADR on bad flags. +* Returns 0 on success, -EBADR on bad flags, -EINVAL for an unsupported FIBMAP +* request. */ int fiemap_check_flags (struct fiemap_extent_info *fieinfo,u32 fs_flags) {u32 incompat_flags; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) { +if (fs_flags &FIEMAP_KERNEL_FIBMAP ) +return 0 ; + +return -EINVAL ; + } + incompat_flags = fieinfo->fi_flags & ~(FIEMAP_FLAGS_COMPAT & fs_flags);if (incompat_flags) { fieinfo->fi_flags = incompat_flags;@@ -208,6 +219,10 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg)if (fiemap.fm_extent_count >FIEMAP_MAX_EXTENTS )return -EINVAL ; +/* Userspace has no access to this flag */ +if (fiemap.fm_flags &FIEMAP_KERNEL_FIBMAP ) +return -EINVAL ; + error =fiemap_check_ranges (sb, fiemap.fm_start, fiemap.fm_length, &len);if (error)@@ -318,7 +333,8 @@ int __generic_block_fiemap(struct inode *inode,bool past_eof = false, whole_file = false;int ret =0 ; - ret =fiemap_check_flags (fieinfo,FIEMAP_FLAG_SYNC ); + ret =fiemap_check_flags (fieinfo, +FIEMAP_FLAG_SYNC |FIEMAP_KERNEL_FIBMAP );if (ret)return ret;fs/iomap.c +1
-1@@ -1195,7 +1195,7 @@ int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fi, ctx.fi = fi; ctx.prev.type =IOMAP_HOLE ; - ret =fiemap_check_flags (fi,FIEMAP_FLAG_SYNC ); + ret =fiemap_check_flags (fi,FIEMAP_FLAG_SYNC |FIEMAP_KERNEL_FIBMAP );if (ret)return ret;fs/ocfs2/extent_map.c +7
-1@@ -747,7 +747,7 @@ static int ocfs2_fiemap_inline(struct inode *inode, struct buffer_head *di_bh,return 0 ; } -#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) +#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | FIEMAP_KERNEL_FIBMAP)int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {@@ -756,6 +756,7 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)unsigned int hole_size;struct ocfs2_super *osb =OCFS2_SB (inode->i_sb);u64 len_bytes, phys_bytes, virt_bytes; +struct buffer_head *di_bh =NULL ;struct ocfs2_extent_rec rec;u64 map_start = fieinfo->fi_start;@@ -765,6 +766,11 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret)return ret; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) { +if (ocfs2_is_refcount_inode (inode)) +return -EINVAL ; + } + ret =ocfs2_inode_lock (inode, &di_bh,0 );if (ret) {mlog_errno (ret);fs/xfs/xfs_iops.c +0
-0
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1116
Quoted 4 lines
fs/xfs/xfs_iops.c +1
-0@@ -1113,6 +1113,11 @@ xfs_vn_fiemap( struct fiemap_extent_info *fieinfo) {int error; +struct xfs_inode *ip =XFS_I (inode);
Would you mind fixing the indentation to match usual xfs style?
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1120
Quoted 4 lines
fs/xfs/xfs_iops.c +4
-0+ +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +if (xfs_is_reflink_inode (ip) ||XFS_IS_REALTIME_INODE (ip)) +return -EINVAL ;
The xfs part looks ok to me.
--D
Quoted 21 lines
fs/xfs/xfs_iops.c +0
-0xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) {include/linux/fs.h +4
-0@@ -1709,6 +1709,10 @@ extern bool may_open_dev(const struct path *path);typedef int (*fiemap_fill_cb )(struct fiemap_extent_info *fieinfo,u64 logical,u64 phys,u64 len,u32 flags); +#define FIEMAP_KERNEL_FIBMAP 0x10000000/* FIBMAP call through FIEMAP +interface. This is a kernel +only flag */ +struct fiemap_extent_info {unsigned int fi_flags;/* Flags as passed from user */ u64 fi_start;Signature · 1 line
2.20.1
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
Quoted 403 lines · Darrick J. Wong
Quoted 71 lines · Carlos Maiolino
Enables the usage of FIEMAP ioctl infrastructure to handle FIBMAP calls. From now on, ->bmap() methods can start to be removed from filesystems which already provides ->fiemap().
This adds a new helper -
bmap_fiemap()- which is used to fill in the fiemap request, call into the underlying filesystem and check the flags set in the extent requested.Add a new fiemap fill extent callback to handle the in-kernel only
fiemap_extentstructure used for FIBMAP.The new
FIEMAP_KERNEL_FIBMAPflag, is used to tell the filesystem ->fiemap interface, that the call is coming fromioctl_fibmap. The addition of this new flag, requires an update tofiemap_check_flags(), so it doesn't treat FIBMAP requests as invalid.Signed-off-by Carlos Maiolino <[email protected]>---
Changelog:
V4:
- Fix if conditional inbmap()
- Add filesystem-specific modifications
V3:
- AddFIEMAP_EXTENT_SHAREDto the list of invalid extents inbmap_fiemap()
- Renamefi_extents_starttofi_cb_data
- Use if conditional instead of ternary operator
- Make fiemap_fill_* callbacks static (which required the removal of some macros
- SetFIEMAP_FLAG_SYNCwhen calling in ->fiemap method from fibmap
- AddFIEMAP_KERNEL_FIBMAPflag, to identify the usage of fiemap infrastructure for fibmap calls, defined infs.hso it's not exported to userspace.
- Updatefiemap_check_flags()to understandFIEMAP_KERNEL_FIBMAP
- Update filesystems supporting both FIBMAP and FIEMAP, which need extra checks on FIBMAP callsV2:
- Now based on the updatedfiemap_extent_info,
- move the fiemap call itself to a new helperfs/ext4/extents.c | 7 +++- fs/f2fs/data.c | 10 +++++- fs/gfs2/inode.c | 6 +++- fs/inode.c | 81 +++++++++++++++++++++++++++++++++++++++++-- fs/ioctl.c | 40 ++++++++++++++------- fs/iomap.c | 2 +- fs/ocfs2/extent_map.c | 8 ++++- fs/xfs/xfs_iops.c | 5 +++ include/linux/fs.h | 4 +++ 9 files changed, 144 insertions(+), 19 deletions(-)fs/ext4/extents.c +3
-1@@ -5001,7 +5001,9 @@ static int ext4_find_delayed_extent(struct inode *inode,return next_del; }/* fiemap flags we can handle specified here */ -#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) +#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)static int ext4_xattr_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo)fs/ext4/extents.c int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) · line 5054
Quoted 5 lines
fs/ext4/extents.c +2
-0@@ -5048,6 +5050,9 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ext4_has_inline_data (inode)) {int has_inline =1 ; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +return -EINVAL ;Wouldn't the inline data case be caught by
fiemap_bmapand turned into-EINVAL?Quoted 183 lines
fs/ext4/extents.c +1
-0+ error =ext4_inline_data_fiemap (inode, fieinfo, &has_inline, start, len);fs/f2fs/data.c +9
-1@@ -1409,6 +1409,9 @@ static int f2fs_xattr_fiemap(struct inode *inode,return (err <0 ? err :0 ); } +#define F2FS_FIEMAP_COMPAT (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)int f2fs_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {u64 start = fieinfo->fi_start;@@ -1426,7 +1429,7 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)return ret; } - ret =fiemap_check_flags (fieinfo,FIEMAP_FLAG_SYNC |FIEMAP_FLAG_XATTR ); + ret =fiemap_check_flags (fieinfo,F2FS_FIEMAP_COMPAT );if (ret)return ret;@@ -1438,6 +1441,11 @@ int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) }if (f2fs_has_inline_data (inode)) { + + ret = -EINVAL ; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) + goto out; + ret =f2fs_inline_data_fiemap (inode, fieinfo, start, len);if (ret != -EAGAIN ) goto out;fs/gfs2/inode.c +5
-1@@ -2016,7 +2016,11 @@ static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret) goto out; - ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops); +if (gfs2_is_stuffed (ip) && + (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP )) + ret = -EINVAL ; +else + ret =iomap_fiemap (inode, fieinfo, &gfs2_iomap_ops);gfs2_glock_dq_uninit (&gh);fs/inode.c +79
-2@@ -1575,6 +1575,78 @@ void iput(struct inode *inode) }EXPORT_SYMBOL (iput); +static int fiemap_fill_kernel_extent (struct fiemap_extent_info *fieinfo, +u64 logical,u64 phys,u64 len,u32 flags) +{ +struct fiemap_extent *extent = fieinfo->fi_cb_data; + +/* only count the extents */ +if (fieinfo->fi_cb_data ==0 ) { + fieinfo->fi_extents_mapped++; + goto out; + } + +if (fieinfo->fi_extents_mapped >= fieinfo->fi_extents_max) +return 1 ; + +if (flags &FIEMAP_EXTENT_DELALLOC ) + flags |=FIEMAP_EXTENT_UNKNOWN ; +if (flags &FIEMAP_EXTENT_DATA_ENCRYPTED ) + flags |=FIEMAP_EXTENT_ENCODED ; +if (flags & (FIEMAP_EXTENT_DATA_TAIL |FIEMAP_EXTENT_DATA_INLINE )) + flags |=FIEMAP_EXTENT_NOT_ALIGNED ; + + extent->fe_logical = logical; + extent->fe_physical = phys; + extent->fe_length = len; + extent->fe_flags = flags; + + fieinfo->fi_extents_mapped++; + +if (fieinfo->fi_extents_mapped == fieinfo->fi_extents_max) +return 1 ; + +out: +if (flags &FIEMAP_EXTENT_LAST ) +return 1 ; +return 0 ; +} + +static int bmap_fiemap (struct inode *inode,sector_t *block) +{ +struct fiemap_extent_info fieinfo = {0 , }; +struct fiemap_extent fextent; +u64 start = *block << inode->i_blkbits; +int error = -EINVAL ; + + fextent.fe_logical =0 ; + fextent.fe_physical =0 ; + fieinfo.fi_extents_max =1 ; + fieinfo.fi_extents_mapped =0 ; + fieinfo.fi_cb_data = &fextent; + fieinfo.fi_start = start; + fieinfo.fi_len =1 << inode->i_blkbits; + fieinfo.fi_cb = fiemap_fill_kernel_extent; + fieinfo.fi_flags = (FIEMAP_KERNEL_FIBMAP |FIEMAP_FLAG_SYNC ); + + error = inode->i_op->fiemap (inode, &fieinfo); + +if (error) +return error; + +if (fieinfo.fi_flags & (FIEMAP_EXTENT_UNKNOWN | +FIEMAP_EXTENT_ENCODED | +FIEMAP_EXTENT_DATA_INLINE | +FIEMAP_EXTENT_UNWRITTEN | +FIEMAP_EXTENT_SHARED )) +return -EINVAL ; + + *block = (fextent.fe_physical + + (start - fextent.fe_logical)) >> inode->i_blkbits; + +return error; +} +/** * bmap - find a block number in a file * @inode: inode owning the block number being requested @@ -1591,10 +1663,15 @@ EXPORT_SYMBOL(iput);*/ int bmap (struct inode *inode,sector_t *block) { -if (!inode->i_mapping->a_ops->bmap) +if (inode->i_op->fiemap) +return bmap_fiemap (inode, block); + +if (inode->i_mapping->a_ops->bmap) + *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, + *block); +else return -EINVAL ; - *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, *block);return 0 ; }EXPORT_SYMBOL (bmap);fs/ioctl.c +3
-6@@ -77,11 +77,8 @@ static int ioctl_fibmap(struct file *filp, int __user *p)return error; } -#define SET_UNKNOWN_FLAGS (FIEMAP_EXTENT_DELALLOC) -#define SET_NO_UNMOUNTED_IO_FLAGS (FIEMAP_EXTENT_DATA_ENCRYPTED) -#define SET_NOT_ALIGNED_FLAGS (FIEMAP_EXTENT_DATA_TAIL|FIEMAP_EXTENT_DATA_INLINE) -int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo,u64 logical, -u64 phys,u64 len,u32 flags) +static int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo, +u64 logical,u64 phys,u64 len,u32 flags) {struct fiemap_extent extent;struct fiemap_extent __user *dest = fieinfo->fi_cb_data;@@ -89,17 +86,17 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical,/* only count the extents */ if (fieinfo->fi_extents_max ==0 ) { fieinfo->fi_extents_mapped++; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + goto out; }fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 102
Quoted 10 lines
fs/ioctl.c +2
-3if (fieinfo->fi_extents_mapped >= fieinfo->fi_extents_max)return 1 ; -if (flags &SET_UNKNOWN_FLAGS ) +if (flags &FIEMAP_EXTENT_DELALLOC ) flags |=FIEMAP_EXTENT_UNKNOWN ; -if (flags &SET_NO_UNMOUNTED_IO_FLAGS ) +if (flags &FIEMAP_EXTENT_DATA_ENCRYPTED ) flags |=FIEMAP_EXTENT_ENCODED ; -if (flags &SET_NOT_ALIGNED_FLAGS )It's too bad that we lose the "not aligned" semantic meaning here.
fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 100
Quoted 2 lines
fs/ioctl.c +1
-0+if (flags & (FIEMAP_EXTENT_DATA_TAIL |FIEMAP_EXTENT_DATA_INLINE )) flags |=FIEMAP_EXTENT_NOT_ALIGNED ;Why doesn't this function just call
fiemap_fill_kernel_extentto fill out the onstack @extent structure? We've now implemented "fill out out a structfiemap_extent" twice.Quoted 112 lines
fs/ioctl.c +22
-3memset (&extent,0 ,sizeof (extent));@@ -115,7 +112,11 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, fieinfo->fi_extents_mapped++;if (fieinfo->fi_extents_mapped == fieinfo->fi_extents_max)return 1 ; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + +out: +if (flags &FIEMAP_EXTENT_LAST ) +return 1 ; +return 0 ; }/** @@ -151,13 +152,23 @@ EXPORT_SYMBOL(fiemap_fill_next_extent);* flags, the invalid values will be written into the fieinfo structure, and * -EBADR is returned, which tells ioctl_fiemap() to return those values to * userspace. For this reason, a return code of -EBADR should be preserved. +* In case ->fiemap is being used for FIBMAP calls, and the filesystem does not +* support it, return -EINVAL. * -* Returns 0 on success, -EBADR on bad flags. +* Returns 0 on success, -EBADR on bad flags, -EINVAL for an unsupported FIBMAP +* request. */ int fiemap_check_flags (struct fiemap_extent_info *fieinfo,u32 fs_flags) {u32 incompat_flags; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) { +if (fs_flags &FIEMAP_KERNEL_FIBMAP ) +return 0 ; + +return -EINVAL ; + } + incompat_flags = fieinfo->fi_flags & ~(FIEMAP_FLAGS_COMPAT & fs_flags);if (incompat_flags) { fieinfo->fi_flags = incompat_flags;@@ -208,6 +219,10 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg)if (fiemap.fm_extent_count >FIEMAP_MAX_EXTENTS )return -EINVAL ; +/* Userspace has no access to this flag */ +if (fiemap.fm_flags &FIEMAP_KERNEL_FIBMAP ) +return -EINVAL ; + error =fiemap_check_ranges (sb, fiemap.fm_start, fiemap.fm_length, &len);if (error)@@ -318,7 +333,8 @@ int __generic_block_fiemap(struct inode *inode,bool past_eof = false, whole_file = false;int ret =0 ; - ret =fiemap_check_flags (fieinfo,FIEMAP_FLAG_SYNC ); + ret =fiemap_check_flags (fieinfo, +FIEMAP_FLAG_SYNC |FIEMAP_KERNEL_FIBMAP );if (ret)return ret;fs/iomap.c +1
-1@@ -1195,7 +1195,7 @@ int iomap_fiemap(struct inode *inode, struct fiemap_extent_info *fi, ctx.fi = fi; ctx.prev.type =IOMAP_HOLE ; - ret =fiemap_check_flags (fi,FIEMAP_FLAG_SYNC ); + ret =fiemap_check_flags (fi,FIEMAP_FLAG_SYNC |FIEMAP_KERNEL_FIBMAP );if (ret)return ret;fs/ocfs2/extent_map.c +7
-1@@ -747,7 +747,7 @@ static int ocfs2_fiemap_inline(struct inode *inode, struct buffer_head *di_bh,return 0 ; } -#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC) +#define OCFS2_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | FIEMAP_KERNEL_FIBMAP)int ocfs2_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo) {@@ -756,6 +756,7 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)unsigned int hole_size;struct ocfs2_super *osb =OCFS2_SB (inode->i_sb);u64 len_bytes, phys_bytes, virt_bytes; +struct buffer_head *di_bh =NULL ;struct ocfs2_extent_rec rec;u64 map_start = fieinfo->fi_start;@@ -765,6 +766,11 @@ int ocfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ret)return ret; +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) { +if (ocfs2_is_refcount_inode (inode)) +return -EINVAL ; + } + ret =ocfs2_inode_lock (inode, &di_bh,0 );if (ret) {mlog_errno (ret);fs/xfs/xfs_iops.c +0
-0fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1116
Quoted 4 lines
fs/xfs/xfs_iops.c +1
-0@@ -1113,6 +1113,11 @@ xfs_vn_fiemap( struct fiemap_extent_info *fieinfo) {int error; +struct xfs_inode *ip =XFS_I (inode);Would you mind fixing the indentation to match usual xfs style?
Quoted 6 lines
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1120
Quoted 4 lines
fs/xfs/xfs_iops.c +4
-0+ +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +if (xfs_is_reflink_inode (ip) ||XFS_IS_REALTIME_INODE (ip)) +return -EINVAL ;The xfs part looks ok to me.
No it doesn't. This check ought to be xfs_is_cow_inode() to match the code being removed in the last patch:
if ((fieinfo->fi_flags & FIEMAP_KERNEL_FIBMAP) &&
(xfs_is_cow_inode(ip) || XFS_IS_REALTIME_INODE(ip))) return -EINVAL;
Quoted 23 lines
--D
Quoted 21 lines
fs/xfs/xfs_iops.c +0
-0xfs_ilock (XFS_I (inode),XFS_IOLOCK_SHARED );if (fieinfo->fi_flags &FIEMAP_FLAG_XATTR ) {include/linux/fs.h +4
-0@@ -1709,6 +1709,10 @@ extern bool may_open_dev(const struct path *path);typedef int (*fiemap_fill_cb )(struct fiemap_extent_info *fieinfo,u64 logical,u64 phys,u64 len,u32 flags); +#define FIEMAP_KERNEL_FIBMAP 0x10000000/* FIBMAP call through FIEMAP +interface. This is a kernel +only flag */ +struct fiemap_extent_info {unsigned int fi_flags;/* Flags as passed from user */ u64 fi_start;Signature · 1 line
2.20.1
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
Quoted 19 lines
Quoted 12 lines
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1116
Quoted 9 lines
fs/xfs/xfs_iops.c +1
-0@@ -1113,6 +1113,11 @@ xfs_vn_fiemap( struct fiemap_extent_info *fieinfo) {int error; +struct xfs_inode *ip =XFS_I (inode);Would you mind fixing the indentation to match usual xfs style?
Quoted 6 lines
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1120
Quoted 4 lines
fs/xfs/xfs_iops.c +4
-0+ +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +if (xfs_is_reflink_inode (ip) ||XFS_IS_REALTIME_INODE (ip)) +return -EINVAL ;The xfs part looks ok to me.
Quoted 7 lines
No it doesn't. This check ought to be
xfs_is_cow_inode()to match the code being removed in the last patch:if ((
fieinfo->fi_flags&FIEMAP_KERNEL_FIBMAP) &&(xfs_is_cow_inode(ip) || XFS_IS_REALTIME_INODE(ip))) return -EINVAL;
Agreed, will fix
Quoted 2 lines
Quoted 2 lines
--D
Signature · 1 line
Carlos
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
In reply to Darrick J. WongDarrick, and chance to not send out a double full quote? I just can't not find the actual information in this mail even if I try.
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
In reply to Darrick J. WongQuoted 11 lines
Quoted 11 lines
fs/ext4/extents.c +3
-1-#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) +#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)static int ext4_xattr_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo)@@ -5048,6 +5050,9 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ext4_has_inline_data (inode)) {int has_inline =1 ;
Quoted 5 lines
Review int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) · line 5054
Quoted 2 lines
(quoted diff) +2
-0+ if (fieinfo->fi_flags & FIEMAP_KERNEL_FIBMAP) + return -EINVAL;Wouldn't the inline data case be caught by
fiemap_bmapand turned into-EINVAL?
Yes, it does, but until ext4_fiemap() returns the extent with the INLINE flag, it does need to go through the whole fiemap mapping mechanism when we already know the result... So, instead of letting the ext4_fiemap() map the extent, just take the shortcut and return -EINVAL directly.
The check in fiemap_bmap() is a 'safe measure' (if it does have other name I don't know :), but if the filesystem already knows it's gonna fall into an inline inode, taking the shortcut is better, isn't it?
Quoted 89 lines
Quoted 89 lines
fs/inode.c +46
-2+return 1 ; +return 0 ; +} + +static int bmap_fiemap (struct inode *inode,sector_t *block) +{ +struct fiemap_extent_info fieinfo = {0 , }; +struct fiemap_extent fextent; +u64 start = *block << inode->i_blkbits; +int error = -EINVAL ; + + fextent.fe_logical =0 ; + fextent.fe_physical =0 ; + fieinfo.fi_extents_max =1 ; + fieinfo.fi_extents_mapped =0 ; + fieinfo.fi_cb_data = &fextent; + fieinfo.fi_start = start; + fieinfo.fi_len =1 << inode->i_blkbits; + fieinfo.fi_cb = fiemap_fill_kernel_extent; + fieinfo.fi_flags = (FIEMAP_KERNEL_FIBMAP |FIEMAP_FLAG_SYNC ); + + error = inode->i_op->fiemap (inode, &fieinfo); + +if (error) +return error; + +if (fieinfo.fi_flags & (FIEMAP_EXTENT_UNKNOWN | +FIEMAP_EXTENT_ENCODED | +FIEMAP_EXTENT_DATA_INLINE | +FIEMAP_EXTENT_UNWRITTEN | +FIEMAP_EXTENT_SHARED )) +return -EINVAL ; + + *block = (fextent.fe_physical + + (start - fextent.fe_logical)) >> inode->i_blkbits; + +return error; +} +/** * bmap - find a block number in a file * @inode: inode owning the block number being requested @@ -1591,10 +1663,15 @@ EXPORT_SYMBOL(iput);*/ int bmap (struct inode *inode,sector_t *block) { -if (!inode->i_mapping->a_ops->bmap) +if (inode->i_op->fiemap) +return bmap_fiemap (inode, block); + +if (inode->i_mapping->a_ops->bmap) + *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, + *block); +else return -EINVAL ; - *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, *block);return 0 ; }EXPORT_SYMBOL (bmap);fs/ioctl.c +3
-6@@ -77,11 +77,8 @@ static int ioctl_fibmap(struct file *filp, int __user *p)return error; } -#define SET_UNKNOWN_FLAGS (FIEMAP_EXTENT_DELALLOC) -#define SET_NO_UNMOUNTED_IO_FLAGS (FIEMAP_EXTENT_DATA_ENCRYPTED) -#define SET_NOT_ALIGNED_FLAGS (FIEMAP_EXTENT_DATA_TAIL|FIEMAP_EXTENT_DATA_INLINE) -int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo,u64 logical, -u64 phys,u64 len,u32 flags) +static int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo, +u64 logical,u64 phys,u64 len,u32 flags) {struct fiemap_extent extent;struct fiemap_extent __user *dest = fieinfo->fi_cb_data;@@ -89,17 +86,17 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical,/* only count the extents */ if (fieinfo->fi_extents_max ==0 ) { fieinfo->fi_extents_mapped++; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + goto out; }if (fieinfo->fi_extents_mapped >= fieinfo->fi_extents_max)return 1 ;
Quoted 9 lines
fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 102
Quoted 7 lines
fs/ioctl.c +2
-3-if (flags &SET_UNKNOWN_FLAGS ) +if (flags &FIEMAP_EXTENT_DELALLOC ) flags |=FIEMAP_EXTENT_UNKNOWN ; -if (flags &SET_NO_UNMOUNTED_IO_FLAGS ) +if (flags &FIEMAP_EXTENT_DATA_ENCRYPTED ) flags |=FIEMAP_EXTENT_ENCODED ; -if (flags &SET_NOT_ALIGNED_FLAGS )It's too bad that we lose the "not aligned" semantic meaning here.
May you explain a bit better what you mean? We don't lose it, just the define goes away, the reason I dropped these defines is because the same flags are used in both functions, fiemap_fill_{user,kernel}_extent(), and I didn't think defining them on both places (or in fs.h) has any benefit here, so I opted to remove them.
Quoted 7 lines
fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 100
Quoted 3 lines
fs/ioctl.c +1
-0+if (flags & (FIEMAP_EXTENT_DATA_TAIL |FIEMAP_EXTENT_DATA_INLINE )) flags |=FIEMAP_EXTENT_NOT_ALIGNED ;Why doesn't this function just call
fiemap_fill_kernel_extentto fill out the onstack @extent structure? We've now implemented "fill out out a structfiemap_extent" twice.
fiemap_fill_{user, kernel}_extent() have different purposes, and the big difference is one handles a userspace pointer memory and the other don't. IIRC the original proposal was some sort of sharing a single function, but then Christoph suggested a new design, using different functions as callbacks.
Quoted 11 lines
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1116
Quoted 9 lines
fs/xfs/xfs_iops.c +1
-0@@ -1113,6 +1113,11 @@ xfs_vn_fiemap( struct fiemap_extent_info *fieinfo) {int error; +struct xfs_inode *ip =XFS_I (inode);Would you mind fixing the indentation to match usual xfs style?
Sure, will fix it
Quoted 10 lines
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1120
Quoted 5 lines
fs/xfs/xfs_iops.c +4
-0+ +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +if (xfs_is_reflink_inode (ip) ||XFS_IS_REALTIME_INODE (ip)) +return -EINVAL ;The xfs part looks ok to me.
--D
Signature · 1 line
Carlos
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
Quoted 17 lines · Carlos Maiolino
Quoted 11 lines
Quoted 11 lines
fs/ext4/extents.c +3
-1-#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) +#define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC | \ + FIEMAP_FLAG_XATTR| \ + FIEMAP_KERNEL_FIBMAP)static int ext4_xattr_fiemap (struct inode *inode,struct fiemap_extent_info *fieinfo)@@ -5048,6 +5050,9 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo)if (ext4_has_inline_data (inode)) {int has_inline =1 ;Quoted 5 lines
Review int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo) · line 5054
Quoted 2 lines
(quoted diff) +2
-0+ if (fieinfo->fi_flags & FIEMAP_KERNEL_FIBMAP) + return -EINVAL;Wouldn't the inline data case be caught by
fiemap_bmapand turned into-EINVAL?
Quoted 8 lines
Yes, it does, but until
ext4_fiemap()returns the extent with the INLINE flag, it does need to go through the whole fiemap mapping mechanism when we already know the result... So, instead of letting theext4_fiemap()map the extent, just take the shortcut and return-EINVALdirectly.The check in
fiemap_bmap()is a 'safe measure' (if it does have other name I don't know :), but if the filesystem already knows it's gonna fall into an inline inode, taking the shortcut is better, isn't it?
I suppose so. Just wondering, that was all... :)
Quoted 100 lines
Quoted 90 lines
Quoted 90 lines
fs/inode.c +46
-2+return 1 ; +return 0 ; +} + +static int bmap_fiemap (struct inode *inode,sector_t *block) +{ +struct fiemap_extent_info fieinfo = {0 , }; +struct fiemap_extent fextent; +u64 start = *block << inode->i_blkbits; +int error = -EINVAL ; + + fextent.fe_logical =0 ; + fextent.fe_physical =0 ; + fieinfo.fi_extents_max =1 ; + fieinfo.fi_extents_mapped =0 ; + fieinfo.fi_cb_data = &fextent; + fieinfo.fi_start = start; + fieinfo.fi_len =1 << inode->i_blkbits; + fieinfo.fi_cb = fiemap_fill_kernel_extent; + fieinfo.fi_flags = (FIEMAP_KERNEL_FIBMAP |FIEMAP_FLAG_SYNC ); + + error = inode->i_op->fiemap (inode, &fieinfo); + +if (error) +return error; + +if (fieinfo.fi_flags & (FIEMAP_EXTENT_UNKNOWN | +FIEMAP_EXTENT_ENCODED | +FIEMAP_EXTENT_DATA_INLINE | +FIEMAP_EXTENT_UNWRITTEN | +FIEMAP_EXTENT_SHARED )) +return -EINVAL ; + + *block = (fextent.fe_physical + + (start - fextent.fe_logical)) >> inode->i_blkbits; + +return error; +} +/** * bmap - find a block number in a file * @inode: inode owning the block number being requested @@ -1591,10 +1663,15 @@ EXPORT_SYMBOL(iput);*/ int bmap (struct inode *inode,sector_t *block) { -if (!inode->i_mapping->a_ops->bmap) +if (inode->i_op->fiemap) +return bmap_fiemap (inode, block); + +if (inode->i_mapping->a_ops->bmap) + *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, + *block); +else return -EINVAL ; - *block = inode->i_mapping->a_ops->bmap (inode->i_mapping, *block);return 0 ; }EXPORT_SYMBOL (bmap);fs/ioctl.c +3
-6@@ -77,11 +77,8 @@ static int ioctl_fibmap(struct file *filp, int __user *p)return error; } -#define SET_UNKNOWN_FLAGS (FIEMAP_EXTENT_DELALLOC) -#define SET_NO_UNMOUNTED_IO_FLAGS (FIEMAP_EXTENT_DATA_ENCRYPTED) -#define SET_NOT_ALIGNED_FLAGS (FIEMAP_EXTENT_DATA_TAIL|FIEMAP_EXTENT_DATA_INLINE) -int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo,u64 logical, -u64 phys,u64 len,u32 flags) +static int fiemap_fill_user_extent (struct fiemap_extent_info *fieinfo, +u64 logical,u64 phys,u64 len,u32 flags) {struct fiemap_extent extent;struct fiemap_extent __user *dest = fieinfo->fi_cb_data;@@ -89,17 +86,17 @@ int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical,/* only count the extents */ if (fieinfo->fi_extents_max ==0 ) { fieinfo->fi_extents_mapped++; -return (flags &FIEMAP_EXTENT_LAST ) ?1 :0 ; + goto out; }if (fieinfo->fi_extents_mapped >= fieinfo->fi_extents_max)return 1 ;Quoted 9 lines
fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 102
Quoted 7 lines
fs/ioctl.c +2
-3-if (flags &SET_UNKNOWN_FLAGS ) +if (flags &FIEMAP_EXTENT_DELALLOC ) flags |=FIEMAP_EXTENT_UNKNOWN ; -if (flags &SET_NO_UNMOUNTED_IO_FLAGS ) +if (flags &FIEMAP_EXTENT_DATA_ENCRYPTED ) flags |=FIEMAP_EXTENT_ENCODED ; -if (flags &SET_NOT_ALIGNED_FLAGS )It's too bad that we lose the "not aligned" semantic meaning here.
Quoted 5 lines
May you explain a bit better what you mean? We don't lose it, just the define goes away, the reason I dropped these defines is because the same flags are used in both functions, fiemap_fill_{user,kernel}_extent(), and I didn't think defining them on both places (or in
fs.h) has any benefit here, so I opted to remove them.
Eh, I changed my mind. It's easy enough to tell which flags map to "No umounted IO" from the code even if the #defines go away.
Quoted 8 lines
Quoted 7 lines
fs/ioctl.c int fiemap_fill_user_extent(struct fiemap_extent_info *fieinfo, u64 logical, · line 100
Quoted 3 lines
fs/ioctl.c +1
-0+if (flags & (FIEMAP_EXTENT_DATA_TAIL |FIEMAP_EXTENT_DATA_INLINE )) flags |=FIEMAP_EXTENT_NOT_ALIGNED ;Why doesn't this function just call
fiemap_fill_kernel_extentto fill out the onstack @extent structure? We've now implemented "fill out out a structfiemap_extent" twice.
Quoted 4 lines
fiemap_fill_{user, kernel}_extent() have different purposes, and the big difference is one handles a userspace pointer memory and the other don't. IIRC the original proposal was some sort of sharing a single function, but then Christoph suggested a new design, using different functions as callbacks.
It's harder for me to tell when I don't have a branch containing the final product to look at, but I'd have thought that _fill_kernel fills out an in-kernel fiemap extent; and then _fill_user would declare one on the stack, call _fill_kernel to set the fields, and then copy_to_user?
(But maybe the code already does this and I can't tell...)
Quoted 29 lines
Quoted 12 lines
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1116
Quoted 10 lines
fs/xfs/xfs_iops.c +1
-0@@ -1113,6 +1113,11 @@ xfs_vn_fiemap( struct fiemap_extent_info *fieinfo) {int error; +struct xfs_inode *ip =XFS_I (inode);Would you mind fixing the indentation to match usual xfs style?
Sure, will fix it
Quoted 10 lines
fs/xfs/xfs_iops.c xfs_vn_fiemap( · line 1120
Quoted 5 lines
fs/xfs/xfs_iops.c +4
-0+ +if (fieinfo->fi_flags &FIEMAP_KERNEL_FIBMAP ) +if (xfs_is_reflink_inode (ip) ||XFS_IS_REALTIME_INODE (ip)) +return -EINVAL ;The xfs part looks ok to me.
--D
Signature · 1 line
Carlos
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
Quoted 5 lines
Quoted 5 lines
Quoted 4 lines
Why doesn't this function just call
fiemap_fill_kernel_extentto fill out the onstack @extent structure? We've now implemented "fill out out a structfiemap_extent" twice.
Quoted 7 lines
Quoted 4 lines
fiemap_fill_{user, kernel}_extent() have different purposes, and the big difference is one handles a userspace pointer memory and the other don't. IIRC the original proposal was some sort of sharing a single function, but then Christoph suggested a new design, using different functions as callbacks.
It's harder for me to tell when I don't have a branch containing the final product to look at,
Good, I though I was the only one having issues with it :)
You can see the work here:
https://github.com/cmaiolino/linux/commits/FIEMAP_V5
^ This already includes changes addressing your concerns as we discussed int this thread btw.
Quoted 3 lines
but I'd have thought that
_fill_kernelfills
out an in-kernel fiemap extent; and then_fill_userwould declare one on the stack, call_fill_kernelto set the fields, and thencopy_to_user?
None of those functions will declare a fiemap_extent, the fiemap extent will be declared before (in ioctl_fiemap() or bmap_fiemap()) calling either function and then passed to the proper callback via fieinfo.fi_cb_data. This will be a user or a kernel memory address, and the callbacks will handle the memory accordingly.
Cheers
Signature · 1 line
Carlos
Re: [PATCH 8/9] Use FIEMAP for FIBMAP calls
In reply to Darrick J. WongQuoted 4 lines · Darrick J. Wong
It's harder for me to tell when I don't have a branch containing the final product to look at, but I'd have thought that
_fill_kernelfills out an in-kernel fiemap extent; and then_fill_userwould declare one on the stack, call_fill_kernelto set the fields, and thencopy_to_user?
That works fine for small, fixed-sized structures. But for large variable sized structures it is very inefficient, as we need to do a possibly large kernel allocation and just copy it on. Thus I told Carlos to follow the readdir model with the ->actor (used to be ->filldir) callback that can fill out the actual kernel or user data directly. Another example of this high-level model is our struct iov_iter used in the I/O path.
PATCH v4 9/9 xfs: Get rid of ->bmap
In reply to Carlos MaiolinoWe don't need ->bmap anymore, only usage for it was FIBMAP, which is now gone.
Also kill iomap_bmap() and iomap_bmap_actor once it has no users anymore.
---
Changelog:
V2:
- Kill iomap_bmap() and iomap_bmap_actor()
fs/iomap.c | 34 ---------------------------------- fs/xfs/xfs_aops.c | 24 ------------------------ fs/xfs/xfs_trace.h | 1 - 3 files changed, 59 deletions(-)
fs/iomap.c +0 -34
@@ -2153,37 +2153,3 @@ int iomap_swapfile_activate(struct swap_info_struct *sis, }EXPORT_SYMBOL_GPL (iomap_swapfile_activate);#endif /* CONFIG_SWAP */ - -static loff_t -iomap_bmap_actor (struct inode *inode,loff_t pos,loff_t length, -void *data,struct iomap *iomap) -{ -sector_t *bno = data, addr; - -if (iomap->type ==IOMAP_MAPPED ) { - addr = (pos - iomap->offset + iomap->addr) >> inode->i_blkbits; -if (addr >INT_MAX ) -WARN (1 ,"would truncate bmap result\n" ); -else - *bno = addr; - } -return 0 ; -} - -/* legacy ->bmap interface. 0 is the error return (!) */ -sector_t -iomap_bmap (struct address_space *mapping,sector_t bno, -const struct iomap_ops *ops) -{ -struct inode *inode = mapping->host; -loff_t pos = bno << inode->i_blkbits; -unsigned blocksize =i_blocksize (inode); - -if (filemap_write_and_wait (mapping)) -return 0 ; - - bno =0 ; -iomap_apply (inode, pos, blocksize,0 , ops, &bno, iomap_bmap_actor); -return bno; -} -EXPORT_SYMBOL_GPL (iomap_bmap);
fs/xfs/xfs_aops.c +0 -24
@@ -1006,29 +1006,6 @@ xfs_vm_releasepage(return iomap_releasepage (page, gfp_mask); } -STATIC sector_t -xfs_vm_bmap ( -struct address_space *mapping, -sector_t block) -{ -struct xfs_inode *ip =XFS_I (mapping->host); - -trace_xfs_vm_bmap (ip); - -/* -* The swap code (ab-)uses ->bmap to get a block mapping and then -* bypasses the file system for actual I/O. We really can't allow -* that on reflinks inodes, so we have to skip out here. And yes, -* 0 is the magic code for a bmap error. -* -* Since we don't pass back blockdev info, we can't return bmap -* information for rt files either. -*/ -if (xfs_is_cow_inode (ip) ||XFS_IS_REALTIME_INODE (ip)) -return 0 ; -return iomap_bmap (mapping, block, &xfs_iomap_ops); -} - STATIC int xfs_vm_readpage( struct file *unused,
@@ -1067,7 +1044,6 @@ const struct address_space_operations xfs_address_space_operations = { .set_page_dirty = iomap_set_page_dirty, .releasepage = xfs_vm_releasepage, .invalidatepage = xfs_vm_invalidatepage, - .bmap = xfs_vm_bmap, .direct_IO = noop_direct_IO, .migratepage = iomap_migrate_page, .is_partially_uptodate = iomap_is_partially_uptodate,
fs/xfs/xfs_trace.h +0 -1
@@ -621,7 +621,6 @@ DEFINE_INODE_EVENT(xfs_readdir);#ifdef CONFIG_XFS_POSIX_ACL DEFINE_INODE_EVENT (xfs_get_acl);#endif -DEFINE_INODE_EVENT (xfs_vm_bmap);DEFINE_INODE_EVENT (xfs_file_ioctl);DEFINE_INODE_EVENT (xfs_file_compat_ioctl);DEFINE_INODE_EVENT (xfs_ioctl_setattr);
Signature · 1 line
2.20.1
Re: [PATCH 9/9] xfs: Get rid of ->bmap
fs/iomap.c
Quoted 22 lines · Carlos Maiolino
We don't need ->bmap anymore, only usage for it was FIBMAP, which is now gone.
Also kill
iomap_bmap()andiomap_bmap_actoronce it has no users anymore.Signed-off-by Carlos Maiolino <[email protected]>---
Changelog:
V2:
- Killiomap_bmap()andiomap_bmap_actor()fs/iomap.c | 34 ---------------------------------- fs/xfs/xfs_aops.c | 24 ------------------------ fs/xfs/xfs_trace.h | 1 - 3 files changed, 59 deletions(-)fs/iomap.c +0
-0
Needs rebasing against 5.3-rc1.
Quoted 55 lines
fs/iomap.c +0
-34@@ -2153,37 +2153,3 @@ int iomap_swapfile_activate(struct swap_info_struct *sis, }EXPORT_SYMBOL_GPL (iomap_swapfile_activate);#endif /* CONFIG_SWAP */ - -static loff_t -iomap_bmap_actor (struct inode *inode,loff_t pos,loff_t length, -void *data,struct iomap *iomap) -{ -sector_t *bno = data, addr; - -if (iomap->type ==IOMAP_MAPPED ) { - addr = (pos - iomap->offset + iomap->addr) >> inode->i_blkbits; -if (addr >INT_MAX ) -WARN (1 ,"would truncate bmap result\n" ); -else - *bno = addr; - } -return 0 ; -} - -/* legacy ->bmap interface. 0 is the error return (!) */ -sector_t -iomap_bmap (struct address_space *mapping,sector_t bno, -const struct iomap_ops *ops) -{ -struct inode *inode = mapping->host; -loff_t pos = bno << inode->i_blkbits; -unsigned blocksize =i_blocksize (inode); - -if (filemap_write_and_wait (mapping)) -return 0 ; - - bno =0 ; -iomap_apply (inode, pos, blocksize,0 , ops, &bno, iomap_bmap_actor); -return bno; -} -EXPORT_SYMBOL_GPL (iomap_bmap);fs/xfs/xfs_aops.c +0
-9@@ -1006,29 +1006,6 @@ xfs_vm_releasepage(return iomap_releasepage (page, gfp_mask); } -STATIC sector_t -xfs_vm_bmap ( -struct address_space *mapping, -sector_t block) -{ -struct xfs_inode *ip =XFS_I (mapping->host); - -trace_xfs_vm_bmap (ip); -
fs/xfs/xfs_aops.c xfs_vm_releasepage( · line 1027
Quoted 10 lines
fs/xfs/xfs_aops.c +0
-10-/* -* The swap code (ab-)uses ->bmap to get a block mapping and then -* bypasses the file system for actual I/O. We really can't allow -* that on reflinks inodes, so we have to skip out here. And yes, -* 0 is the magic code for a bmap error. -* -* Since we don't pass back blockdev info, we can't return bmap -* information for rt files either. -*/ -if (xfs_is_cow_inode (ip) ||XFS_IS_REALTIME_INODE (ip))
Wait, this isn't the same check that was added to the xfs fiemap implementation.
--D
Quoted 30 lines
fs/xfs/xfs_aops.c +0
-5-return 0 ; -return iomap_bmap (mapping, block, &xfs_iomap_ops); -} - STATIC int xfs_vm_readpage( struct file *unused,@@ -1067,7 +1044,6 @@ const struct address_space_operations xfs_address_space_operations = { .set_page_dirty = iomap_set_page_dirty, .releasepage = xfs_vm_releasepage, .invalidatepage = xfs_vm_invalidatepage, - .bmap = xfs_vm_bmap, .direct_IO = noop_direct_IO, .migratepage = iomap_migrate_page, .is_partially_uptodate = iomap_is_partially_uptodate,fs/xfs/xfs_trace.h +0
-1@@ -621,7 +621,6 @@ DEFINE_INODE_EVENT(xfs_readdir);#ifdef CONFIG_XFS_POSIX_ACL DEFINE_INODE_EVENT (xfs_get_acl);#endif -DEFINE_INODE_EVENT (xfs_vm_bmap);DEFINE_INODE_EVENT (xfs_file_ioctl);DEFINE_INODE_EVENT (xfs_file_compat_ioctl);DEFINE_INODE_EVENT (xfs_ioctl_setattr);Signature · 1 line
2.20.1
Quoted 1 line · Carlos Maiolino
Hi.
Please, disconsider this patchset, I wrongly rebased it and there are many missing changes that should be done to properly rebase it agains 5.3
Quoted 79 lines
This is the 4th version of the complete series with the goal to deprecate and eventually remove ->bmap() interface, in lieu of FIEMAP.
Besides the rebase of the patchset against latest Linus's tree, the only changes in this patchset regarding to the previous version, are concentrated in patches 4/9 and 8/9.
In patch 4 (fibmap: Use bmap instead of ->bmap method in
ioctl_fibmap), the difference compared with previous version, is a change inioclt_fibmap()return value, I spotted while testing this new version with filesystems using data inlined in inodes. It now returns 0 in case of error instead an error value, otherwise it would be an user interface change.In patch 8 (Use FIEMAP for FIBMAP calls), there are minor changes regarding V3. It just contains a coding style fix pointed by Andreas in the previous version, but now, it also include changes to all filesystems which supports both FIEMAP and FIBMAP, and need some sort of special handling (like inlined data, reflinked inodes, etc).
Again, Patch 9 is xfs-specific removal of ->bmap() interface, without any changes compared to the previous version.
I do apologize for taking so long to rework this patchset, I've got busy with other stuff.
Comments are appreciated, specially regarding if the error values returned by
ioctl_fibmap()make sense.Cheers
Carlos Maiolino (9):
fs: Enablebmap()function to properly return errors
cachefiles: drop direct usage of ->bmap method.
ecryptfs: drop direct calls to ->bmap
fibmap: Use bmap instead of ->bmap method inioctl_fibmap
fs: Move start and length fiemap fields intofiemap_extent_info
iomap: Remove length and start fields fromiomap_fiemap
fiemap: Use a callback to fill fiemap extents
Use FIEMAP for FIBMAP calls
xfs: Get rid of ->bmapdrivers/md/md-bitmap.c | 16 ++++-- fs/bad_inode.c | 3 +- fs/btrfs/inode.c | 5 +- fs/cachefiles/rdwr.c | 27 ++++----- fs/ecryptfs/mmap.c | 16 ++---- fs/ext2/ext2.h | 3 +- fs/ext2/inode.c | 6 +- fs/ext4/ext4.h | 3 +- fs/ext4/extents.c | 15 +++-- fs/f2fs/data.c | 15 ++++- fs/f2fs/f2fs.h | 3 +- fs/gfs2/inode.c | 9 ++- fs/hpfs/file.c | 4 +- fs/inode.c | 105 +++++++++++++++++++++++++++++---- fs/ioctl.c | 128 ++++++++++++++++++++++++++--------------- fs/iomap.c | 40 ++----------- fs/jbd2/journal.c | 22 ++++--- fs/nilfs2/inode.c | 5 +- fs/nilfs2/nilfs.h | 3 +- fs/ocfs2/extent_map.c | 13 ++++- fs/ocfs2/extent_map.h | 3 +- fs/overlayfs/inode.c | 5 +- fs/xfs/xfs_aops.c | 24 -------- fs/xfs/xfs_iops.c | 19 +++--- fs/xfs/xfs_trace.h | 1 - include/linux/fs.h | 33 +++++++---- include/linux/iomap.h | 2 +- mm/page_io.c | 11 ++-- 28 files changed, 320 insertions(+), 219 deletions(-)Signature · 1 line
2.20.1
Signature · 1 line
Carlos