From: Sasha Levin <Alexander.Levin@microsoft.com>
To: "stable@vger.kernel.org" <stable@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Cc: Tahsin Erdogan <tahsin@google.com>, Theodore Ts'o <tytso@mit.edu>,
Sasha Levin <Alexander.Levin@microsoft.com>
Subject: [PATCH AUTOSEL for 3.18 061/101] ext4: change fast symlink test to not rely on i_blocks
Date: Mon, 9 Apr 2018 00:41:24 +0000 [thread overview]
Message-ID: <20180409004042.164920-11-alexander.levin@microsoft.com> (raw)
In-Reply-To: <20180409004042.164920-1-alexander.levin@microsoft.com>
From: Tahsin Erdogan <tahsin@google.com>
[ Upstream commit 407cd7fb83c0ebabb490190e673d8c71ee7df97e ]
ext4_inode_info->i_data is the storage area for 4 types of data:
a) Extents data
b) Inline data
c) Block map
d) Fast symlink data (symlink length < 60)
Extents data case is positively identified by EXT4_INODE_EXTENTS flag.
Inline data case is also obvious because of EXT4_INODE_INLINE_DATA
flag.
Distinguishing c) and d) however requires additional logic. This
currently relies on i_blocks count. After subtracting external xattr
block from i_blocks, if it is greater than 0 then we know that some
data blocks exist, so there must be a block map.
This logic got broken after ea_inode feature was added. That feature
charges the data blocks of external xattr inodes to the referencing
inode and so adds them to the i_blocks. To fix this, we could subtract
ea_inode blocks by iterating through all xattr entries and then check
whether remaining i_blocks count is zero. Besides being complicated,
this won't change the fact that the current way of distinguishing
between c) and d) is fragile.
The alternative solution is to test whether i_size is less than 60 to
determine fast symlink case. ext4_symlink() uses the same test to decide
whether to store the symlink in i_data. There is one caveat to address
before this can work though.
If an inode's i_nlink is zero during eviction, its i_size is set to
zero and its data is truncated. If system crashes before inode is removed
from the orphan list, next boot orphan cleanup may find the inode with
zero i_size. So, a symlink that had its data stored in a block may now
appear to be a fast symlink. The solution used in this patch is to treat
i_size = 0 as a non-fast symlink case. A zero sized symlink is not legal
so the only time this can happen is the mentioned scenario. This is also
logically correct because a i_size = 0 symlink has no data stored in
i_data.
Suggested-by: Andreas Dilger <adilger@dilger.ca>
Signed-off-by: Tahsin Erdogan <tahsin@google.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Signed-off-by: Sasha Levin <alexander.levin@microsoft.com>
---
fs/ext4/inode.c | 20 +++++++++++++-------
1 file changed, 13 insertions(+), 7 deletions(-)
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index c2434d72681e..c70d7d3832c4 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -140,16 +140,12 @@ static int ext4_meta_trans_blocks(struct inode *inode, int lblocks,
/*
* Test whether an inode is a fast symlink.
+ * A fast symlink has its symlink data stored in ext4_inode_info->i_data.
*/
static int ext4_inode_is_fast_symlink(struct inode *inode)
{
- int ea_blocks = EXT4_I(inode)->i_file_acl ?
- EXT4_CLUSTER_SIZE(inode->i_sb) >> 9 : 0;
-
- if (ext4_has_inline_data(inode))
- return 0;
-
- return (S_ISLNK(inode->i_mode) && inode->i_blocks - ea_blocks == 0);
+ return S_ISLNK(inode->i_mode) && inode->i_size &&
+ (inode->i_size < EXT4_N_BLOCKS * 4);
}
/*
@@ -253,6 +249,16 @@ void ext4_evict_inode(struct inode *inode)
if (IS_SYNC(inode))
ext4_handle_sync(handle);
+
+ /*
+ * Set inode->i_size to 0 before calling ext4_truncate(). We need
+ * special handling of symlinks here because i_size is used to
+ * determine whether ext4_inode_info->i_data contains symlink data or
+ * block mappings. Setting i_size to 0 will remove its fast symlink
+ * status. Erase i_data so that it becomes a valid empty block map.
+ */
+ if (ext4_inode_is_fast_symlink(inode))
+ memset(EXT4_I(inode)->i_data, 0, sizeof(EXT4_I(inode)->i_data));
inode->i_size = 0;
err = ext4_mark_inode_dirty(handle, inode);
if (err) {
--
2.15.1
next prev parent reply other threads:[~2018-04-09 0:41 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-04-09 0:41 [PATCH AUTOSEL for 3.18 051/101] scsi: sun_esp: fix device reference leaks Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 052/101] powerpc/fadump: avoid duplicates in crash memory ranges Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 053/101] libertas: Fix lbs_prb_rsp_limit_set() Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 055/101] MIPS: CPS: Prevent multi-core with dcache aliasing Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 054/101] PCI: Enable ECRC only if device supports it Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 056/101] MIPS: Handle tlbex-tlbp race condition Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 057/101] arm64: ptrace: Avoid setting compat FP[SC]R to garbage if get_user fails Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 058/101] arm64: ptrace: Fix incorrect get_user() use in compat_vfp_set() Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 059/101] x86/um: thin archives build fix Sasha Levin
2018-04-09 4:29 ` Nicholas Piggin
2018-04-15 15:02 ` Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 060/101] fs: warn in case userspace lied about modprobe return Sasha Levin
2018-04-09 0:41 ` Sasha Levin [this message]
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 062/101] irqchip/gic-v3: Honor forced affinity setting Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 064/101] firewire-ohci: work around oversized DMA reads on JMicron controllers Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 063/101] vmlfb: Fix error handling in cr_pll_init() Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 065/101] ASoC: au1x: Fix timeout tests in au1xac97c_ac97_read() Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 066/101] kvm: x86: fix KVM_XEN_HVM_CONFIG ioctl Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 067/101] tracing/hrtimer: Fix tracing bugs by taking all clock bases and modes into account Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 068/101] PCI: Add function 1 DMA alias quirk for Marvell 9128 Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 070/101] dm thin: fix documentation relative to low water mark threshold Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 069/101] tools lib traceevent: Fix get_field_str() for dynamic strings Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 071/101] nfs: Do not convert nfs_idmap_cache_timeout to jiffies Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 073/101] kconfig: Don't leak main menus during parsing Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 072/101] watchdog: sp5100_tco: Fix watchdog disable bit Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 074/101] kconfig: Fix automatic menu creation mem leak Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 076/101] btrfs: Fix out of bounds access in btrfs_search_slot Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 075/101] kconfig: Fix expr_free() E_NOT leak Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 077/101] scsi: devinfo: fix format of the device list Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 078/101] scsi: fas216: fix sense buffer initialization Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 079/101] HID: roccat: prevent an out of bounds read in kovaplus_profile_activated() Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 080/101] jffs2: Fix use-after-free bug in jffs2_iget()'s error handling path Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 081/101] RDMA/mlx5: Avoid memory leak in case of XRCD dealloc failure Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 082/101] ocfs2: return -EROFS to mount.ocfs2 if inode block is invalid Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 083/101] ocfs2/acl: use 'ip_xattr_sem' to protect getting extended attribute Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 084/101] mm/mempolicy: fix the check of nodemask from user Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 085/101] mm/mempolicy: add nodes_empty check in SYSC_migrate_pages Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 086/101] asm-generic: provide generic_pmdp_establish() Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 087/101] mm: pin address_space before dereferencing it while isolating an LRU page Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 088/101] IB/ipoib: Fix for potential no-carrier state Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 089/101] x86/power: Fix swsusp_arch_resume prototype Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 090/101] firmware: dmi_scan: Fix handling of empty DMI strings Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 091/101] ACPI: processor_perflib: Do not send _PPC change notification if not ready Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 092/101] MIPS: TXx9: use IS_BUILTIN() for CONFIG_LEDS_CLASS Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 093/101] xen/grant-table: Use put_page instead of free_page Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 095/101] cifs: silence compiler warnings showing up with gcc-8.0.0 Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 094/101] proc: fix /proc/*/map_files lookup Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 096/101] bcache: properly set task state in bch_writeback_thread() Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 097/101] bcache: fix for allocator and register thread race Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 098/101] bcache: fix for data collapse after re-attaching an attached device Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 099/101] bcache: return attach error when no cache set exist Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 100/101] nfsd: return RESOURCE not GARBAGE_ARGS on too many ops Sasha Levin
2018-04-09 0:41 ` [PATCH AUTOSEL for 3.18 101/101] irqchip/gic-v3: Change pr_debug message to pr_devel Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180409004042.164920-11-alexander.levin@microsoft.com \
--to=alexander.levin@microsoft.com \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
--cc=tahsin@google.com \
--cc=tytso@mit.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®