License cleanup: add SPDX GPL-2.0 license identifier to files with no license
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 07:07:57 -07:00
|
|
|
// SPDX-License-Identifier: GPL-2.0
|
2010-04-06 15:14:15 -07:00
|
|
|
#include <linux/ceph/ceph_debug.h>
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/fs.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/uaccess.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/writeback.h>
|
|
|
|
#include <linux/vmalloc.h>
|
2016-04-13 15:30:17 -07:00
|
|
|
#include <linux/xattr.h>
|
ceph: Fix up after semantic merge conflict
The previous ceph-client merge resulted in ceph not even building,
because there was a merge conflict that wasn't visible as an actual data
conflict: commit 7221fe4c2ed7 ("ceph: add acl for cephfs") added support
for POSIX ACL's into Ceph, but unluckily we also had the VFS tree change
a lot of the POSIX ACL helper functions to be much more helpful to
filesystems (see for example commits 2aeccbe957d0 "fs: add generic
xattr_acl handlers", 5bf3258fd2ac "fs: make posix_acl_chmod more useful"
and 37bc15392a23 "fs: make posix_acl_create more useful")
The reason this conflict wasn't obvious was many-fold: because it was a
semantic conflict rather than a data conflict, it wasn't visible in the
git merge as a conflict. And because the VFS tree hadn't been in
linux-next, people hadn't become aware of it that way. And because I
was at jury duty this morning, I was using my laptop and as a result not
doing constant "allmodconfig" builds.
Anyway, this fixes the build and generally removes a fair chunk of the
Ceph POSIX ACL support code, since the improved helpers seem to match
really well for Ceph too. But I don't actually have any way to *test*
the end result, and I was really hoping for some ACK's for this. Oh,
well.
Not compiling certainly doesn't make things easier to test, so I'm
committing this without the acks after having waited for four hours...
Plus it's what I would have done for the merge had I noticed the
semantic conflict..
Reported-by: Dave Jones <davej@redhat.com>
Cc: Sage Weil <sage@inktank.com>
Cc: Guangliang Zhao <lucienchao@gmail.com>
Cc: Li Wang <li.wang@ubuntykylin.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-01-28 19:06:18 -07:00
|
|
|
#include <linux/posix_acl.h>
|
2014-04-18 01:00:30 -07:00
|
|
|
#include <linux/random.h>
|
2016-05-03 05:55:50 -07:00
|
|
|
#include <linux/sort.h>
|
2019-06-06 04:29:23 -07:00
|
|
|
#include <linux/iversion.h>
|
2020-07-27 07:16:09 -07:00
|
|
|
#include <linux/fscrypt.h>
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
#include "super.h"
|
2010-04-06 15:14:15 -07:00
|
|
|
#include "mds_client.h"
|
2013-08-21 14:29:54 -07:00
|
|
|
#include "cache.h"
|
2020-07-27 07:16:09 -07:00
|
|
|
#include "crypto.h"
|
2010-04-06 15:14:15 -07:00
|
|
|
#include <linux/ceph/decode.h>
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Ceph inode operations
|
|
|
|
*
|
|
|
|
* Implement basic inode helpers (get, alloc) and inode ops (getattr,
|
|
|
|
* setattr, etc.), xattr helpers, and helpers for assimilating
|
|
|
|
* metadata returned by the MDS into our cache.
|
|
|
|
*
|
|
|
|
* Also define helpers for doing asynchronous writeback, invalidation,
|
|
|
|
* and truncation for the benefit of those who can't afford to block
|
|
|
|
* (typically because they are in the message handler path).
|
|
|
|
*/
|
|
|
|
|
|
|
|
static const struct inode_operations ceph_symlink_iops;
|
2020-09-03 10:31:10 -07:00
|
|
|
static const struct inode_operations ceph_encrypted_symlink_iops;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2019-05-17 20:18:44 -07:00
|
|
|
static void ceph_inode_work(struct work_struct *work);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* find or create an inode, given the ceph ino number
|
|
|
|
*/
|
2011-01-21 17:44:03 -07:00
|
|
|
static int ceph_set_ino_cb(struct inode *inode, void *data)
|
|
|
|
{
|
ceph: fix inode number handling on arches with 32-bit ino_t
Tuan and Ulrich mentioned that they were hitting a problem on s390x,
which has a 32-bit ino_t value, even though it's a 64-bit arch (for
historical reasons).
I think the current handling of inode numbers in the ceph driver is
wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's
actually not a problem. 32-bit arches can deal with 64-bit inode numbers
just fine when userland code is compiled with LFS support (the common
case these days).
What we really want to do is just use 64-bit numbers everywhere, unless
someone has mounted with the ino32 mount option. In that case, we want
to ensure that we hash the inode number down to something that will fit
in 32 bits before presenting the value to userland.
Add new helper functions that do this, and only do the conversion before
presenting these values to userland in getattr and readdir.
The inode table hashvalue is changed to just cast the inode number to
unsigned long, as low-order bits are the most likely to vary anyway.
While it's not strictly required, we do want to put something in
inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on
the size of the ino_t type.
NOTE: This is a user-visible change on 32-bit arches:
1/ inode numbers will be seen to have changed between kernel versions.
32-bit arches will see large inode numbers now instead of the hashed
ones they saw before.
2/ any really old software not built with LFS support may start failing
stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we
can do about these, but hopefully the intersection of people running
such code on ceph will be very small.
The workaround for both problems is to mount with "-o ino32".
[ idryomov: changelog tweak ]
URL: https://tracker.ceph.com/issues/46828
Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 05:03:48 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2020-09-03 06:01:40 -07:00
|
|
|
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(inode->i_sb);
|
ceph: fix inode number handling on arches with 32-bit ino_t
Tuan and Ulrich mentioned that they were hitting a problem on s390x,
which has a 32-bit ino_t value, even though it's a 64-bit arch (for
historical reasons).
I think the current handling of inode numbers in the ceph driver is
wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's
actually not a problem. 32-bit arches can deal with 64-bit inode numbers
just fine when userland code is compiled with LFS support (the common
case these days).
What we really want to do is just use 64-bit numbers everywhere, unless
someone has mounted with the ino32 mount option. In that case, we want
to ensure that we hash the inode number down to something that will fit
in 32 bits before presenting the value to userland.
Add new helper functions that do this, and only do the conversion before
presenting these values to userland in getattr and readdir.
The inode table hashvalue is changed to just cast the inode number to
unsigned long, as low-order bits are the most likely to vary anyway.
While it's not strictly required, we do want to put something in
inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on
the size of the ino_t type.
NOTE: This is a user-visible change on 32-bit arches:
1/ inode numbers will be seen to have changed between kernel versions.
32-bit arches will see large inode numbers now instead of the hashed
ones they saw before.
2/ any really old software not built with LFS support may start failing
stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we
can do about these, but hopefully the intersection of people running
such code on ceph will be very small.
The workaround for both problems is to mount with "-o ino32".
[ idryomov: changelog tweak ]
URL: https://tracker.ceph.com/issues/46828
Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 05:03:48 -07:00
|
|
|
|
|
|
|
ci->i_vino = *(struct ceph_vino *)data;
|
|
|
|
inode->i_ino = ceph_vino_to_ino_t(ci->i_vino);
|
2019-06-06 04:29:23 -07:00
|
|
|
inode_set_iversion_raw(inode, 0);
|
2020-09-03 06:01:40 -07:00
|
|
|
percpu_counter_inc(&mdsc->metric.total_inodes);
|
|
|
|
|
2011-01-21 17:44:03 -07:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
/**
|
|
|
|
* ceph_new_inode - allocate a new inode in advance of an expected create
|
|
|
|
* @dir: parent directory for new inode
|
|
|
|
* @dentry: dentry that may eventually point to new inode
|
|
|
|
* @mode: mode of new inode
|
|
|
|
* @as_ctx: pointer to inherited security context
|
|
|
|
*
|
|
|
|
* Allocate a new inode in advance of an operation to create a new inode.
|
|
|
|
* This allocates the inode and sets up the acl_sec_ctx with appropriate
|
|
|
|
* info for the new inode.
|
|
|
|
*
|
|
|
|
* Returns a pointer to the new inode or an ERR_PTR.
|
|
|
|
*/
|
|
|
|
struct inode *ceph_new_inode(struct inode *dir, struct dentry *dentry,
|
|
|
|
umode_t *mode, struct ceph_acl_sec_ctx *as_ctx)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
struct inode *inode;
|
|
|
|
|
|
|
|
inode = new_inode(dir->i_sb);
|
|
|
|
if (!inode)
|
|
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
|
|
|
|
if (!S_ISLNK(*mode)) {
|
|
|
|
err = ceph_pre_init_acls(dir, mode, as_ctx);
|
|
|
|
if (err < 0)
|
|
|
|
goto out_err;
|
|
|
|
}
|
|
|
|
|
2020-09-08 06:47:40 -07:00
|
|
|
inode->i_state = 0;
|
|
|
|
inode->i_mode = *mode;
|
|
|
|
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
err = ceph_security_init_secctx(dentry, *mode, as_ctx);
|
|
|
|
if (err < 0)
|
|
|
|
goto out_err;
|
|
|
|
|
2022-08-25 06:31:28 -07:00
|
|
|
/*
|
|
|
|
* We'll skip setting fscrypt context for snapshots, leaving that for
|
|
|
|
* the handle_reply().
|
|
|
|
*/
|
|
|
|
if (ceph_snap(dir) != CEPH_SNAPDIR) {
|
|
|
|
err = ceph_fscrypt_prepare_context(dir, inode, as_ctx);
|
|
|
|
if (err)
|
|
|
|
goto out_err;
|
|
|
|
}
|
2020-09-08 06:47:40 -07:00
|
|
|
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
return inode;
|
|
|
|
out_err:
|
|
|
|
iput(inode);
|
|
|
|
return ERR_PTR(err);
|
|
|
|
}
|
|
|
|
|
|
|
|
void ceph_as_ctx_to_req(struct ceph_mds_request *req,
|
|
|
|
struct ceph_acl_sec_ctx *as_ctx)
|
|
|
|
{
|
|
|
|
if (as_ctx->pagelist) {
|
|
|
|
req->r_pagelist = as_ctx->pagelist;
|
|
|
|
as_ctx->pagelist = NULL;
|
|
|
|
}
|
2020-09-08 06:47:40 -07:00
|
|
|
ceph_fscrypt_as_ctx_to_req(req, as_ctx);
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* ceph_get_inode - find or create/hash a new inode
|
|
|
|
* @sb: superblock to search and allocate in
|
|
|
|
* @vino: vino to search for
|
|
|
|
* @newino: optional new inode to insert if one isn't found (may be NULL)
|
|
|
|
*
|
|
|
|
* Search for or insert a new inode into the hash for the given vino, and
|
|
|
|
* return a reference to it. If new is non-NULL, its reference is consumed.
|
|
|
|
*/
|
|
|
|
struct inode *ceph_get_inode(struct super_block *sb, struct ceph_vino vino,
|
|
|
|
struct inode *newino)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(sb);
|
|
|
|
struct ceph_client *cl = mdsc->fsc->client;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct inode *inode;
|
|
|
|
|
2021-04-01 10:55:11 -07:00
|
|
|
if (ceph_vino_is_reserved(vino))
|
|
|
|
return ERR_PTR(-EREMOTEIO);
|
|
|
|
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
if (newino) {
|
|
|
|
inode = inode_insert5(newino, (unsigned long)vino.ino,
|
|
|
|
ceph_ino_compare, ceph_set_ino_cb, &vino);
|
|
|
|
if (inode != newino)
|
|
|
|
iput(newino);
|
|
|
|
} else {
|
|
|
|
inode = iget5_locked(sb, (unsigned long)vino.ino,
|
|
|
|
ceph_ino_compare, ceph_set_ino_cb, &vino);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!inode) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "no inode found for %llx.%llx\n", vino.ino, vino.snap);
|
2009-10-06 11:31:08 -07:00
|
|
|
return ERR_PTR(-ENOMEM);
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "on %llx=%llx.%llx got %p new %d\n",
|
|
|
|
ceph_present_inode(inode), ceph_vinop(inode), inode,
|
|
|
|
!!(inode->i_state & I_NEW));
|
2009-10-06 11:31:08 -07:00
|
|
|
return inode;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* get/constuct snapdir inode for a given directory
|
|
|
|
*/
|
|
|
|
struct inode *ceph_get_snapdir(struct inode *parent)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(parent);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_vino vino = {
|
|
|
|
.ino = ceph_ino(parent),
|
|
|
|
.snap = CEPH_SNAPDIR,
|
|
|
|
};
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
struct inode *inode = ceph_get_inode(parent->i_sb, vino, NULL);
|
2009-11-11 16:22:37 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2022-08-25 06:31:28 -07:00
|
|
|
int ret = -ENOTDIR;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (IS_ERR(inode))
|
2010-05-22 03:01:14 -07:00
|
|
|
return inode;
|
2021-02-25 13:04:15 -07:00
|
|
|
|
|
|
|
if (!S_ISDIR(parent->i_mode)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_warn_once_client(cl, "bad snapdir parent type (mode=0%o)\n",
|
|
|
|
parent->i_mode);
|
2022-03-01 23:51:53 -07:00
|
|
|
goto err;
|
2021-02-25 13:04:15 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!(inode->i_state & I_NEW) && !S_ISDIR(inode->i_mode)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_warn_once_client(cl, "bad snapdir inode type (mode=0%o)\n",
|
|
|
|
inode->i_mode);
|
2022-03-01 23:51:53 -07:00
|
|
|
goto err;
|
2021-02-25 13:04:15 -07:00
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
inode->i_mode = parent->i_mode;
|
|
|
|
inode->i_uid = parent->i_uid;
|
|
|
|
inode->i_gid = parent->i_gid;
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_mtime_to_ts(inode, inode_get_mtime(parent));
|
2023-07-05 12:00:55 -07:00
|
|
|
inode_set_ctime_to_ts(inode, inode_get_ctime(parent));
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_atime_to_ts(inode, inode_get_atime(parent));
|
2009-11-11 16:22:37 -07:00
|
|
|
ci->i_rbytes = 0;
|
2020-03-11 04:05:38 -07:00
|
|
|
ci->i_btime = ceph_inode(parent)->i_btime;
|
2019-12-11 13:21:24 -07:00
|
|
|
|
2022-08-25 06:31:28 -07:00
|
|
|
#ifdef CONFIG_FS_ENCRYPTION
|
|
|
|
/* if encrypted, just borrow fscrypt_auth from parent */
|
|
|
|
if (IS_ENCRYPTED(parent)) {
|
|
|
|
struct ceph_inode_info *pci = ceph_inode(parent);
|
|
|
|
|
|
|
|
ci->fscrypt_auth = kmemdup(pci->fscrypt_auth,
|
|
|
|
pci->fscrypt_auth_len,
|
|
|
|
GFP_KERNEL);
|
|
|
|
if (ci->fscrypt_auth) {
|
|
|
|
inode->i_flags |= S_ENCRYPTED;
|
|
|
|
ci->fscrypt_auth_len = pci->fscrypt_auth_len;
|
|
|
|
} else {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "Failed to alloc snapdir fscrypt_auth\n");
|
2022-08-25 06:31:28 -07:00
|
|
|
ret = -ENOMEM;
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif
|
2021-03-01 05:38:01 -07:00
|
|
|
if (inode->i_state & I_NEW) {
|
|
|
|
inode->i_op = &ceph_snapdir_iops;
|
|
|
|
inode->i_fop = &ceph_snapdir_fops;
|
|
|
|
ci->i_snap_caps = CEPH_CAP_PIN; /* so we can open */
|
2019-12-11 13:21:24 -07:00
|
|
|
unlock_new_inode(inode);
|
2021-03-01 05:38:01 -07:00
|
|
|
}
|
2019-12-11 13:21:24 -07:00
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
return inode;
|
2022-03-01 23:51:53 -07:00
|
|
|
err:
|
|
|
|
if ((inode->i_state & I_NEW))
|
|
|
|
discard_new_inode(inode);
|
|
|
|
else
|
|
|
|
iput(inode);
|
2022-08-25 06:31:28 -07:00
|
|
|
return ERR_PTR(ret);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
const struct inode_operations ceph_file_iops = {
|
|
|
|
.permission = ceph_permission,
|
|
|
|
.setattr = ceph_setattr,
|
|
|
|
.getattr = ceph_getattr,
|
|
|
|
.listxattr = ceph_listxattr,
|
2022-09-22 08:17:00 -07:00
|
|
|
.get_inode_acl = ceph_get_acl,
|
2014-01-29 07:22:25 -07:00
|
|
|
.set_acl = ceph_set_acl,
|
2009-10-06 11:31:08 -07:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We use a 'frag tree' to keep track of the MDS's directory fragments
|
|
|
|
* for a given inode (usually there is just a single fragment). We
|
|
|
|
* need to know when a child frag is delegated to a new MDS, or when
|
|
|
|
* it is flagged as replicated, so we can direct our requests
|
|
|
|
* accordingly.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
|
|
|
* find/create a frag in the tree
|
|
|
|
*/
|
|
|
|
static struct ceph_inode_frag *__get_or_create_frag(struct ceph_inode_info *ci,
|
|
|
|
u32 f)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct inode *inode = &ci->netfs.inode;
|
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct rb_node **p;
|
|
|
|
struct rb_node *parent = NULL;
|
|
|
|
struct ceph_inode_frag *frag;
|
|
|
|
int c;
|
|
|
|
|
|
|
|
p = &ci->i_fragtree.rb_node;
|
|
|
|
while (*p) {
|
|
|
|
parent = *p;
|
|
|
|
frag = rb_entry(parent, struct ceph_inode_frag, node);
|
|
|
|
c = ceph_frag_compare(f, frag->frag);
|
|
|
|
if (c < 0)
|
|
|
|
p = &(*p)->rb_left;
|
|
|
|
else if (c > 0)
|
|
|
|
p = &(*p)->rb_right;
|
|
|
|
else
|
|
|
|
return frag;
|
|
|
|
}
|
|
|
|
|
|
|
|
frag = kmalloc(sizeof(*frag), GFP_NOFS);
|
2017-08-20 11:00:09 -07:00
|
|
|
if (!frag)
|
2009-10-06 11:31:08 -07:00
|
|
|
return ERR_PTR(-ENOMEM);
|
2017-08-20 11:00:09 -07:00
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
frag->frag = f;
|
|
|
|
frag->split_by = 0;
|
|
|
|
frag->mds = -1;
|
|
|
|
frag->ndist = 0;
|
|
|
|
|
|
|
|
rb_link_node(&frag->node, parent, p);
|
|
|
|
rb_insert_color(&frag->node, &ci->i_fragtree);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "added %p %llx.%llx frag %x\n", inode, ceph_vinop(inode), f);
|
2009-10-06 11:31:08 -07:00
|
|
|
return frag;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* find a specific frag @f
|
|
|
|
*/
|
|
|
|
struct ceph_inode_frag *__ceph_find_frag(struct ceph_inode_info *ci, u32 f)
|
|
|
|
{
|
|
|
|
struct rb_node *n = ci->i_fragtree.rb_node;
|
|
|
|
|
|
|
|
while (n) {
|
|
|
|
struct ceph_inode_frag *frag =
|
|
|
|
rb_entry(n, struct ceph_inode_frag, node);
|
|
|
|
int c = ceph_frag_compare(f, frag->frag);
|
|
|
|
if (c < 0)
|
|
|
|
n = n->rb_left;
|
|
|
|
else if (c > 0)
|
|
|
|
n = n->rb_right;
|
|
|
|
else
|
|
|
|
return frag;
|
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Choose frag containing the given value @v. If @pfrag is
|
|
|
|
* specified, copy the frag delegation info to the caller if
|
|
|
|
* it is present.
|
|
|
|
*/
|
2014-04-18 01:00:30 -07:00
|
|
|
static u32 __ceph_choose_frag(struct ceph_inode_info *ci, u32 v,
|
|
|
|
struct ceph_inode_frag *pfrag, int *found)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(&ci->netfs.inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
u32 t = ceph_frag_make(0, 0);
|
|
|
|
struct ceph_inode_frag *frag;
|
|
|
|
unsigned nway, i;
|
|
|
|
u32 n;
|
|
|
|
|
|
|
|
if (found)
|
|
|
|
*found = 0;
|
|
|
|
|
|
|
|
while (1) {
|
|
|
|
WARN_ON(!ceph_frag_contains_value(t, v));
|
|
|
|
frag = __ceph_find_frag(ci, t);
|
|
|
|
if (!frag)
|
|
|
|
break; /* t is a leaf */
|
|
|
|
if (frag->split_by == 0) {
|
|
|
|
if (pfrag)
|
|
|
|
memcpy(pfrag, frag, sizeof(*pfrag));
|
|
|
|
if (found)
|
|
|
|
*found = 1;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* choose child */
|
|
|
|
nway = 1 << frag->split_by;
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "frag(%x) %x splits by %d (%d ways)\n", v, t,
|
|
|
|
frag->split_by, nway);
|
2009-10-06 11:31:08 -07:00
|
|
|
for (i = 0; i < nway; i++) {
|
|
|
|
n = ceph_frag_make_child(t, frag->split_by, i);
|
|
|
|
if (ceph_frag_contains_value(n, v)) {
|
|
|
|
t = n;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
BUG_ON(i == nway);
|
|
|
|
}
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "frag(%x) = %x\n", v, t);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
return t;
|
|
|
|
}
|
|
|
|
|
2014-04-18 01:00:30 -07:00
|
|
|
u32 ceph_choose_frag(struct ceph_inode_info *ci, u32 v,
|
|
|
|
struct ceph_inode_frag *pfrag, int *found)
|
|
|
|
{
|
|
|
|
u32 ret;
|
|
|
|
mutex_lock(&ci->i_fragtree_mutex);
|
|
|
|
ret = __ceph_choose_frag(ci, v, pfrag, found);
|
|
|
|
mutex_unlock(&ci->i_fragtree_mutex);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* Process dirfrag (delegation) info from the mds. Include leaf
|
|
|
|
* fragment in tree ONLY if ndist > 0. Otherwise, only
|
|
|
|
* branches/splits are included in i_fragtree)
|
|
|
|
*/
|
|
|
|
static int ceph_fill_dirfrag(struct inode *inode,
|
|
|
|
struct ceph_mds_reply_dirfrag *dirinfo)
|
|
|
|
{
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_inode_frag *frag;
|
|
|
|
u32 id = le32_to_cpu(dirinfo->frag);
|
|
|
|
int mds = le32_to_cpu(dirinfo->auth);
|
|
|
|
int ndist = le32_to_cpu(dirinfo->ndist);
|
2014-04-18 07:01:38 -07:00
|
|
|
int diri_auth = -1;
|
2009-10-06 11:31:08 -07:00
|
|
|
int i;
|
|
|
|
int err = 0;
|
|
|
|
|
2014-04-18 07:01:38 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
|
|
|
if (ci->i_auth_cap)
|
|
|
|
diri_auth = ci->i_auth_cap->mds;
|
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
|
|
|
|
2016-05-03 07:33:20 -07:00
|
|
|
if (mds == -1) /* CDIR_AUTH_PARENT */
|
|
|
|
mds = diri_auth;
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
mutex_lock(&ci->i_fragtree_mutex);
|
2014-04-18 07:01:38 -07:00
|
|
|
if (ndist == 0 && mds == diri_auth) {
|
2009-10-06 11:31:08 -07:00
|
|
|
/* no delegation info needed. */
|
|
|
|
frag = __ceph_find_frag(ci, id);
|
|
|
|
if (!frag)
|
|
|
|
goto out;
|
|
|
|
if (frag->split_by == 0) {
|
|
|
|
/* tree leaf, remove */
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "removed %p %llx.%llx frag %x (no ref)\n",
|
|
|
|
inode, ceph_vinop(inode), id);
|
2009-10-06 11:31:08 -07:00
|
|
|
rb_erase(&frag->node, &ci->i_fragtree);
|
|
|
|
kfree(frag);
|
|
|
|
} else {
|
|
|
|
/* tree branch, keep and clear */
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "cleared %p %llx.%llx frag %x referral\n",
|
|
|
|
inode, ceph_vinop(inode), id);
|
2009-10-06 11:31:08 -07:00
|
|
|
frag->mds = -1;
|
|
|
|
frag->ndist = 0;
|
|
|
|
}
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* find/add this frag to store mds delegation info */
|
|
|
|
frag = __get_or_create_frag(ci, id);
|
|
|
|
if (IS_ERR(frag)) {
|
|
|
|
/* this is not the end of the world; we can continue
|
|
|
|
with bad/inaccurate delegation info */
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "ENOMEM on mds ref %p %llx.%llx fg %x\n",
|
|
|
|
inode, ceph_vinop(inode),
|
|
|
|
le32_to_cpu(dirinfo->frag));
|
2009-10-06 11:31:08 -07:00
|
|
|
err = -ENOMEM;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
frag->mds = mds;
|
|
|
|
frag->ndist = min_t(u32, ndist, CEPH_MAX_DIRFRAG_REP);
|
|
|
|
for (i = 0; i < frag->ndist; i++)
|
|
|
|
frag->dist[i] = le32_to_cpu(dirinfo->dist[i]);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx frag %x ndist=%d\n", inode,
|
|
|
|
ceph_vinop(inode), frag->frag, frag->ndist);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
out:
|
|
|
|
mutex_unlock(&ci->i_fragtree_mutex);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2016-05-03 05:55:50 -07:00
|
|
|
static int frag_tree_split_cmp(const void *l, const void *r)
|
|
|
|
{
|
|
|
|
struct ceph_frag_tree_split *ls = (struct ceph_frag_tree_split*)l;
|
|
|
|
struct ceph_frag_tree_split *rs = (struct ceph_frag_tree_split*)r;
|
2017-01-12 12:42:39 -07:00
|
|
|
return ceph_frag_compare(le32_to_cpu(ls->frag),
|
|
|
|
le32_to_cpu(rs->frag));
|
2016-05-03 05:55:50 -07:00
|
|
|
}
|
|
|
|
|
2016-05-03 20:05:10 -07:00
|
|
|
static bool is_frag_child(u32 f, struct ceph_inode_frag *frag)
|
|
|
|
{
|
|
|
|
if (!frag)
|
|
|
|
return f == ceph_frag_make(0, 0);
|
|
|
|
if (ceph_frag_bits(f) != ceph_frag_bits(frag->frag) + frag->split_by)
|
|
|
|
return false;
|
|
|
|
return ceph_frag_contains_value(frag->frag, ceph_frag_value(f));
|
|
|
|
}
|
|
|
|
|
2014-04-18 01:00:30 -07:00
|
|
|
static int ceph_fill_fragtree(struct inode *inode,
|
|
|
|
struct ceph_frag_tree_head *fragtree,
|
|
|
|
struct ceph_mds_reply_dirfrag *dirinfo)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2014-04-18 01:00:30 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2016-05-03 20:05:10 -07:00
|
|
|
struct ceph_inode_frag *frag, *prev_frag = NULL;
|
2014-04-18 01:00:30 -07:00
|
|
|
struct rb_node *rb_node;
|
2016-05-03 20:40:30 -07:00
|
|
|
unsigned i, split_by, nsplits;
|
|
|
|
u32 id;
|
2014-04-18 01:00:30 -07:00
|
|
|
bool update = false;
|
|
|
|
|
|
|
|
mutex_lock(&ci->i_fragtree_mutex);
|
|
|
|
nsplits = le32_to_cpu(fragtree->nsplits);
|
2016-05-03 20:40:30 -07:00
|
|
|
if (nsplits != ci->i_fragtree_nsplits) {
|
|
|
|
update = true;
|
|
|
|
} else if (nsplits) {
|
2022-10-09 19:44:02 -07:00
|
|
|
i = get_random_u32_below(nsplits);
|
2014-04-18 01:00:30 -07:00
|
|
|
id = le32_to_cpu(fragtree->splits[i].frag);
|
|
|
|
if (!__ceph_find_frag(ci, id))
|
|
|
|
update = true;
|
|
|
|
} else if (!RB_EMPTY_ROOT(&ci->i_fragtree)) {
|
|
|
|
rb_node = rb_first(&ci->i_fragtree);
|
|
|
|
frag = rb_entry(rb_node, struct ceph_inode_frag, node);
|
|
|
|
if (frag->frag != ceph_frag_make(0, 0) || rb_next(rb_node))
|
|
|
|
update = true;
|
|
|
|
}
|
|
|
|
if (!update && dirinfo) {
|
|
|
|
id = le32_to_cpu(dirinfo->frag);
|
|
|
|
if (id != __ceph_choose_frag(ci, id, NULL, NULL))
|
|
|
|
update = true;
|
|
|
|
}
|
|
|
|
if (!update)
|
|
|
|
goto out_unlock;
|
|
|
|
|
2016-05-03 05:55:50 -07:00
|
|
|
if (nsplits > 1) {
|
|
|
|
sort(fragtree->splits, nsplits, sizeof(fragtree->splits[0]),
|
|
|
|
frag_tree_split_cmp, NULL);
|
|
|
|
}
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx\n", inode, ceph_vinop(inode));
|
2014-04-18 01:00:30 -07:00
|
|
|
rb_node = rb_first(&ci->i_fragtree);
|
|
|
|
for (i = 0; i < nsplits; i++) {
|
|
|
|
id = le32_to_cpu(fragtree->splits[i].frag);
|
2016-05-03 20:40:30 -07:00
|
|
|
split_by = le32_to_cpu(fragtree->splits[i].by);
|
|
|
|
if (split_by == 0 || ceph_frag_bits(id) + split_by > 24) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "%p %llx.%llx invalid split %d/%u, "
|
|
|
|
"frag %x split by %d\n", inode,
|
|
|
|
ceph_vinop(inode), i, nsplits, id, split_by);
|
2016-05-03 20:40:30 -07:00
|
|
|
continue;
|
|
|
|
}
|
2014-04-18 01:00:30 -07:00
|
|
|
frag = NULL;
|
|
|
|
while (rb_node) {
|
|
|
|
frag = rb_entry(rb_node, struct ceph_inode_frag, node);
|
|
|
|
if (ceph_frag_compare(frag->frag, id) >= 0) {
|
|
|
|
if (frag->frag != id)
|
|
|
|
frag = NULL;
|
|
|
|
else
|
|
|
|
rb_node = rb_next(rb_node);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
rb_node = rb_next(rb_node);
|
2016-05-03 20:05:10 -07:00
|
|
|
/* delete stale split/leaf node */
|
|
|
|
if (frag->split_by > 0 ||
|
|
|
|
!is_frag_child(frag->frag, prev_frag)) {
|
|
|
|
rb_erase(&frag->node, &ci->i_fragtree);
|
2016-05-03 20:40:30 -07:00
|
|
|
if (frag->split_by > 0)
|
|
|
|
ci->i_fragtree_nsplits--;
|
2016-05-03 20:05:10 -07:00
|
|
|
kfree(frag);
|
|
|
|
}
|
2014-04-18 01:00:30 -07:00
|
|
|
frag = NULL;
|
|
|
|
}
|
|
|
|
if (!frag) {
|
|
|
|
frag = __get_or_create_frag(ci, id);
|
|
|
|
if (IS_ERR(frag))
|
|
|
|
continue;
|
|
|
|
}
|
2016-05-03 20:40:30 -07:00
|
|
|
if (frag->split_by == 0)
|
|
|
|
ci->i_fragtree_nsplits++;
|
|
|
|
frag->split_by = split_by;
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " frag %x split by %d\n", frag->frag, frag->split_by);
|
2016-05-03 20:05:10 -07:00
|
|
|
prev_frag = frag;
|
2014-04-18 01:00:30 -07:00
|
|
|
}
|
|
|
|
while (rb_node) {
|
|
|
|
frag = rb_entry(rb_node, struct ceph_inode_frag, node);
|
|
|
|
rb_node = rb_next(rb_node);
|
2016-05-03 20:05:10 -07:00
|
|
|
/* delete stale split/leaf node */
|
|
|
|
if (frag->split_by > 0 ||
|
|
|
|
!is_frag_child(frag->frag, prev_frag)) {
|
|
|
|
rb_erase(&frag->node, &ci->i_fragtree);
|
2016-05-03 20:40:30 -07:00
|
|
|
if (frag->split_by > 0)
|
|
|
|
ci->i_fragtree_nsplits--;
|
2016-05-03 20:05:10 -07:00
|
|
|
kfree(frag);
|
|
|
|
}
|
2014-04-18 01:00:30 -07:00
|
|
|
}
|
|
|
|
out_unlock:
|
|
|
|
mutex_unlock(&ci->i_fragtree_mutex);
|
|
|
|
return 0;
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* initialize a newly allocated inode.
|
|
|
|
*/
|
|
|
|
struct inode *ceph_alloc_inode(struct super_block *sb)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_inode_info *ci;
|
|
|
|
int i;
|
|
|
|
|
2022-03-22 14:41:03 -07:00
|
|
|
ci = alloc_inode_sb(sb, ceph_inode_cachep, GFP_NOFS);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (!ci)
|
|
|
|
return NULL;
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(fsc->client, "%p\n", &ci->netfs.inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2021-06-29 14:37:05 -07:00
|
|
|
/* Set parameters for the netfs library */
|
netfs: Optimise away reads above the point at which there can be no data
Track the file position above which the server is not expected to have any
data (the "zero point") and preemptively assume that we can satisfy
requests by filling them with zeroes locally rather than attempting to
download them if they're over that line - even if we've written data back
to the server. Assume that any data that was written back above that
position is held in the local cache. Note that we have to split requests
that straddle the line.
Make use of this to optimise away some reads from the server. We need to
set the zero point in the following circumstances:
(1) When we see an extant remote inode and have no cache for it, we set
the zero_point to i_size.
(2) On local inode creation, we set zero_point to 0.
(3) On local truncation down, we reduce zero_point to the new i_size if
the new i_size is lower.
(4) On local truncation up, we don't change zero_point.
(5) On local modification, we don't change zero_point.
(6) On remote invalidation, we set zero_point to the new i_size.
(7) If stored data is discarded from the pagecache or culled from fscache,
we must set zero_point above that if the data also got written to the
server.
(8) If dirty data is written back to the server, but not fscache, we must
set zero_point above that.
(9) If a direct I/O write is made, set zero_point above that.
Assuming the above, any read from the server at or above the zero_point
position will return all zeroes.
The zero_point value can be stored in the cache, provided the above rules
are applied to it by any code that culls part of the local cache.
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Jeff Layton <jlayton@kernel.org>
cc: linux-cachefs@redhat.com
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
2023-11-24 06:39:02 -07:00
|
|
|
netfs_inode_init(&ci->netfs, &ceph_netfs_ops, false);
|
2021-06-29 14:37:05 -07:00
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock_init(&ci->i_ceph_lock);
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_version = 0;
|
2014-11-14 06:41:55 -07:00
|
|
|
ci->i_inline_version = 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_time_warp_seq = 0;
|
|
|
|
ci->i_ceph_flags = 0;
|
2015-06-16 05:48:56 -07:00
|
|
|
atomic64_set(&ci->i_ordered_count, 1);
|
|
|
|
atomic64_set(&ci->i_release_count, 1);
|
|
|
|
atomic64_set(&ci->i_complete_seq[0], 0);
|
|
|
|
atomic64_set(&ci->i_complete_seq[1], 0);
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_symlink = NULL;
|
|
|
|
|
2018-01-05 03:47:18 -07:00
|
|
|
ci->i_max_bytes = 0;
|
|
|
|
ci->i_max_files = 0;
|
|
|
|
|
2010-11-16 12:14:34 -07:00
|
|
|
memset(&ci->i_dir_layout, 0, sizeof(ci->i_dir_layout));
|
2020-01-02 05:11:38 -07:00
|
|
|
memset(&ci->i_cached_layout, 0, sizeof(ci->i_cached_layout));
|
2016-02-13 20:24:31 -07:00
|
|
|
RCU_INIT_POINTER(ci->i_layout.pool_ns, NULL);
|
2010-11-16 12:14:34 -07:00
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_fragtree = RB_ROOT;
|
|
|
|
mutex_init(&ci->i_fragtree_mutex);
|
|
|
|
|
|
|
|
ci->i_xattrs.blob = NULL;
|
|
|
|
ci->i_xattrs.prealloc_blob = NULL;
|
|
|
|
ci->i_xattrs.dirty = false;
|
|
|
|
ci->i_xattrs.index = RB_ROOT;
|
|
|
|
ci->i_xattrs.count = 0;
|
|
|
|
ci->i_xattrs.names_size = 0;
|
|
|
|
ci->i_xattrs.vals_size = 0;
|
|
|
|
ci->i_xattrs.version = 0;
|
|
|
|
ci->i_xattrs.index_version = 0;
|
|
|
|
|
|
|
|
ci->i_caps = RB_ROOT;
|
|
|
|
ci->i_auth_cap = NULL;
|
|
|
|
ci->i_dirty_caps = 0;
|
|
|
|
ci->i_flushing_caps = 0;
|
|
|
|
INIT_LIST_HEAD(&ci->i_dirty_item);
|
|
|
|
INIT_LIST_HEAD(&ci->i_flushing_item);
|
2015-06-10 02:26:13 -07:00
|
|
|
ci->i_prealloc_cap_flush = NULL;
|
2016-07-05 20:12:56 -07:00
|
|
|
INIT_LIST_HEAD(&ci->i_cap_flush_list);
|
2009-10-06 11:31:08 -07:00
|
|
|
init_waitqueue_head(&ci->i_cap_wq);
|
|
|
|
ci->i_hold_caps_max = 0;
|
|
|
|
INIT_LIST_HEAD(&ci->i_cap_delay_list);
|
|
|
|
INIT_LIST_HEAD(&ci->i_cap_snaps);
|
|
|
|
ci->i_head_snapc = NULL;
|
|
|
|
ci->i_snap_caps = 0;
|
|
|
|
|
2020-03-05 05:21:00 -07:00
|
|
|
ci->i_last_rd = ci->i_last_wr = jiffies - 3600 * HZ;
|
2016-06-06 01:01:39 -07:00
|
|
|
for (i = 0; i < CEPH_FILE_MODE_BITS; i++)
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_nr_by_mode[i] = 0;
|
|
|
|
|
2013-08-12 21:42:15 -07:00
|
|
|
mutex_init(&ci->i_truncate_mutex);
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_truncate_seq = 0;
|
|
|
|
ci->i_truncate_size = 0;
|
|
|
|
ci->i_truncate_pending = 0;
|
2022-08-25 06:31:11 -07:00
|
|
|
ci->i_truncate_pagecache_size = 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
ci->i_max_size = 0;
|
|
|
|
ci->i_reported_size = 0;
|
|
|
|
ci->i_wanted_max_size = 0;
|
|
|
|
ci->i_requested_max_size = 0;
|
|
|
|
|
|
|
|
ci->i_pin_ref = 0;
|
|
|
|
ci->i_rd_ref = 0;
|
|
|
|
ci->i_rdcache_ref = 0;
|
|
|
|
ci->i_wr_ref = 0;
|
2011-05-11 03:29:54 -07:00
|
|
|
ci->i_wb_ref = 0;
|
2019-04-02 05:04:30 -07:00
|
|
|
ci->i_fx_ref = 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_wrbuffer_ref = 0;
|
|
|
|
ci->i_wrbuffer_ref_head = 0;
|
2017-09-08 00:23:18 -07:00
|
|
|
atomic_set(&ci->i_filelock_ref, 0);
|
2019-01-28 05:43:55 -07:00
|
|
|
atomic_set(&ci->i_shared_gen, 1);
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_rdcache_gen = 0;
|
|
|
|
ci->i_rdcache_revoking = 0;
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&ci->i_unsafe_dirops);
|
2015-10-27 03:36:06 -07:00
|
|
|
INIT_LIST_HEAD(&ci->i_unsafe_iops);
|
2009-10-06 11:31:08 -07:00
|
|
|
spin_lock_init(&ci->i_unsafe_lock);
|
|
|
|
|
|
|
|
ci->i_snap_realm = NULL;
|
|
|
|
INIT_LIST_HEAD(&ci->i_snap_realm_item);
|
|
|
|
INIT_LIST_HEAD(&ci->i_snap_flush_item);
|
|
|
|
|
2019-05-17 20:18:44 -07:00
|
|
|
INIT_WORK(&ci->i_work, ceph_inode_work);
|
|
|
|
ci->i_work_mask = 0;
|
2019-05-29 08:19:42 -07:00
|
|
|
memset(&ci->i_btime, '\0', sizeof(ci->i_btime));
|
2020-07-27 07:16:09 -07:00
|
|
|
#ifdef CONFIG_FS_ENCRYPTION
|
|
|
|
ci->fscrypt_auth = NULL;
|
|
|
|
ci->fscrypt_auth_len = 0;
|
|
|
|
#endif
|
netfs: Fix gcc-12 warning by embedding vfs inode in netfs_i_context
While randstruct was satisfied with using an open-coded "void *" offset
cast for the netfs_i_context <-> inode casting, __builtin_object_size() as
used by FORTIFY_SOURCE was not as easily fooled. This was causing the
following complaint[1] from gcc v12:
In file included from include/linux/string.h:253,
from include/linux/ceph/ceph_debug.h:7,
from fs/ceph/inode.c:2:
In function 'fortify_memset_chk',
inlined from 'netfs_i_context_init' at include/linux/netfs.h:326:2,
inlined from 'ceph_alloc_inode' at fs/ceph/inode.c:463:2:
include/linux/fortify-string.h:242:25: warning: call to '__write_overflow_field' declared with attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Wattribute-warning]
242 | __write_overflow_field(p_size_field, size);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fix this by embedding a struct inode into struct netfs_i_context (which
should perhaps be renamed to struct netfs_inode). The struct inode
vfs_inode fields are then removed from the 9p, afs, ceph and cifs inode
structs and vfs_inode is then simply changed to "netfs.inode" in those
filesystems.
Further, rename netfs_i_context to netfs_inode, get rid of the
netfs_inode() function that converted a netfs_i_context pointer to an
inode pointer (that can now be done with &ctx->inode) and rename the
netfs_i_context() function to netfs_inode() (which is now a wrapper
around container_of()).
Most of the changes were done with:
perl -p -i -e 's/vfs_inode/netfs.inode/'g \
`git grep -l 'vfs_inode' -- fs/{9p,afs,ceph,cifs}/*.[ch]`
Kees suggested doing it with a pair structure[2] and a special
declarator to insert that into the network filesystem's inode
wrapper[3], but I think it's cleaner to embed it - and then it doesn't
matter if struct randomisation reorders things.
Dave Chinner suggested using a filesystem-specific VFS_I() function in
each filesystem to convert that filesystem's own inode wrapper struct
into the VFS inode struct[4].
Version #2:
- Fix a couple of missed name changes due to a disabled cifs option.
- Rename nfs_i_context to nfs_inode
- Use "netfs" instead of "nic" as the member name in per-fs inode wrapper
structs.
[ This also undoes commit 507160f46c55 ("netfs: gcc-12: temporarily
disable '-Wattribute-warning' for now") that is no longer needed ]
Fixes: bc899ee1c898 ("netfs: Add a netfs inode context")
Reported-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
cc: Jonathan Corbet <corbet@lwn.net>
cc: Eric Van Hensbergen <ericvh@gmail.com>
cc: Latchesar Ionkov <lucho@ionkov.net>
cc: Dominique Martinet <asmadeus@codewreck.org>
cc: Christian Schoenebeck <linux_oss@crudebyte.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Ilya Dryomov <idryomov@gmail.com>
cc: Steve French <smfrench@gmail.com>
cc: William Kucharski <william.kucharski@oracle.com>
cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
cc: Dave Chinner <david@fromorbit.com>
cc: linux-doc@vger.kernel.org
cc: v9fs-developer@lists.sourceforge.net
cc: linux-afs@lists.infradead.org
cc: ceph-devel@vger.kernel.org
cc: linux-cifs@vger.kernel.org
cc: samba-technical@lists.samba.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-hardening@vger.kernel.org
Link: https://lore.kernel.org/r/d2ad3a3d7bdd794c6efb562d2f2b655fb67756b9.camel@kernel.org/ [1]
Link: https://lore.kernel.org/r/20220517210230.864239-1-keescook@chromium.org/ [2]
Link: https://lore.kernel.org/r/20220518202212.2322058-1-keescook@chromium.org/ [3]
Link: https://lore.kernel.org/r/20220524101205.GI2306852@dread.disaster.area/ [4]
Link: https://lore.kernel.org/r/165296786831.3591209.12111293034669289733.stgit@warthog.procyon.org.uk/ # v1
Link: https://lore.kernel.org/r/165305805651.4094995.7763502506786714216.stgit@warthog.procyon.org.uk # v2
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2022-06-09 13:46:04 -07:00
|
|
|
return &ci->netfs.inode;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
2019-04-10 12:18:50 -07:00
|
|
|
void ceph_free_inode(struct inode *inode)
|
2011-01-06 23:49:49 -07:00
|
|
|
{
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
|
|
|
|
2019-03-25 18:38:58 -07:00
|
|
|
kfree(ci->i_symlink);
|
2020-07-27 07:16:09 -07:00
|
|
|
#ifdef CONFIG_FS_ENCRYPTION
|
|
|
|
kfree(ci->fscrypt_auth);
|
|
|
|
#endif
|
2020-09-03 10:31:10 -07:00
|
|
|
fscrypt_free_inode(inode);
|
2011-01-06 23:49:49 -07:00
|
|
|
kmem_cache_free(ceph_inode_cachep, ci);
|
|
|
|
}
|
|
|
|
|
2019-06-01 18:45:38 -07:00
|
|
|
void ceph_evict_inode(struct inode *inode)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2020-09-03 06:01:40 -07:00
|
|
|
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(inode->i_sb);
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_inode_frag *frag;
|
|
|
|
struct rb_node *n;
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p ino %llx.%llx\n", inode, ceph_vinop(inode));
|
2019-06-01 18:45:38 -07:00
|
|
|
|
2020-09-03 06:01:40 -07:00
|
|
|
percpu_counter_dec(&mdsc->metric.total_inodes);
|
|
|
|
|
2019-06-01 18:45:38 -07:00
|
|
|
truncate_inode_pages_final(&inode->i_data);
|
2023-11-27 06:58:07 -07:00
|
|
|
if (inode->i_state & I_PINNING_NETFS_WB)
|
2021-12-07 06:44:50 -07:00
|
|
|
ceph_fscache_unuse_cookie(inode, true);
|
2019-06-01 18:45:38 -07:00
|
|
|
clear_inode(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2013-08-21 14:29:54 -07:00
|
|
|
ceph_fscache_unregister_inode_cookie(ci);
|
2020-07-27 07:16:09 -07:00
|
|
|
fscrypt_put_encryption_info(inode);
|
2013-08-21 14:29:54 -07:00
|
|
|
|
2019-05-22 20:01:37 -07:00
|
|
|
__ceph_remove_caps(ci);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2022-04-27 08:57:04 -07:00
|
|
|
if (__ceph_has_quota(ci, QUOTA_GET_ANY))
|
2018-01-12 10:19:29 -07:00
|
|
|
ceph_adjust_quota_realms_count(inode, false);
|
|
|
|
|
2010-03-09 13:59:08 -07:00
|
|
|
/*
|
|
|
|
* we may still have a snap_realm reference if there are stray
|
2014-04-17 18:57:11 -07:00
|
|
|
* caps in i_snap_caps.
|
2010-03-09 13:59:08 -07:00
|
|
|
*/
|
|
|
|
if (ci->i_snap_realm) {
|
2017-12-14 00:11:09 -07:00
|
|
|
if (ceph_snap(inode) == CEPH_NOSNAP) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " dropping residual ref to snap realm %p\n",
|
|
|
|
ci->i_snap_realm);
|
2021-08-02 08:01:26 -07:00
|
|
|
ceph_change_snap_realm(inode, NULL);
|
2017-12-14 00:11:09 -07:00
|
|
|
} else {
|
|
|
|
ceph_put_snapid_map(mdsc, ci->i_snapid_map);
|
|
|
|
ci->i_snap_realm = NULL;
|
|
|
|
}
|
2010-03-09 13:59:08 -07:00
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
while ((n = rb_first(&ci->i_fragtree)) != NULL) {
|
|
|
|
frag = rb_entry(n, struct ceph_inode_frag, node);
|
|
|
|
rb_erase(n, &ci->i_fragtree);
|
|
|
|
kfree(frag);
|
|
|
|
}
|
2016-05-03 20:40:30 -07:00
|
|
|
ci->i_fragtree_nsplits = 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
__ceph_destroy_xattrs(ci);
|
2009-12-07 13:17:17 -07:00
|
|
|
if (ci->i_xattrs.blob)
|
|
|
|
ceph_buffer_put(ci->i_xattrs.blob);
|
|
|
|
if (ci->i_xattrs.prealloc_blob)
|
|
|
|
ceph_buffer_put(ci->i_xattrs.prealloc_blob);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2016-03-06 18:35:06 -07:00
|
|
|
ceph_put_string(rcu_dereference_raw(ci->i_layout.pool_ns));
|
2020-01-02 05:11:38 -07:00
|
|
|
ceph_put_string(rcu_dereference_raw(ci->i_cached_layout.pool_ns));
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
2016-05-05 01:40:17 -07:00
|
|
|
static inline blkcnt_t calc_inode_blocks(u64 size)
|
|
|
|
{
|
|
|
|
return (size + (1<<9) - 1) >> 9;
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* Helpers to fill in size, ctime, mtime, and atime. We have to be
|
|
|
|
* careful because either the client or MDS may have more up to date
|
|
|
|
* info, depending on which capabilities are held, and whether
|
|
|
|
* time_warp_seq or truncate_seq have increased. (Ordinarily, mtime
|
|
|
|
* and size are monotonically increasing, except when utimes() or
|
|
|
|
* truncate() increments the corresponding _seq values.)
|
|
|
|
*/
|
|
|
|
int ceph_fill_file_size(struct inode *inode, int issued,
|
|
|
|
u32 truncate_seq, u64 truncate_size, u64 size)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
|
|
|
int queue_trunc = 0;
|
2021-04-09 12:58:35 -07:00
|
|
|
loff_t isize = i_size_read(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (ceph_seq_cmp(truncate_seq, ci->i_truncate_seq) > 0 ||
|
2021-04-09 12:58:35 -07:00
|
|
|
(truncate_seq == ci->i_truncate_seq && size > isize)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "size %lld -> %llu\n", isize, size);
|
2016-02-26 02:16:32 -07:00
|
|
|
if (size > 0 && S_ISDIR(inode->i_mode)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "non-zero size for directory\n");
|
2016-02-26 02:16:32 -07:00
|
|
|
size = 0;
|
|
|
|
}
|
2015-12-29 20:32:46 -07:00
|
|
|
i_size_write(inode, size);
|
2016-05-05 01:40:17 -07:00
|
|
|
inode->i_blocks = calc_inode_blocks(size);
|
2021-12-07 06:44:50 -07:00
|
|
|
/*
|
|
|
|
* If we're expanding, then we should be able to just update
|
|
|
|
* the existing cookie.
|
|
|
|
*/
|
|
|
|
if (size > isize)
|
|
|
|
ceph_fscache_update(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_reported_size = size;
|
|
|
|
if (truncate_seq != ci->i_truncate_seq) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "truncate_seq %u -> %u\n",
|
|
|
|
ci->i_truncate_seq, truncate_seq);
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_truncate_seq = truncate_seq;
|
2013-08-12 21:42:15 -07:00
|
|
|
|
|
|
|
/* the MDS should have revoked these caps */
|
2023-09-05 23:22:07 -07:00
|
|
|
WARN_ON_ONCE(issued & (CEPH_CAP_FILE_RD |
|
2013-08-12 21:42:15 -07:00
|
|
|
CEPH_CAP_FILE_LAZYIO));
|
2010-02-09 12:08:40 -07:00
|
|
|
/*
|
|
|
|
* If we hold relevant caps, or in the case where we're
|
|
|
|
* not the only client referencing this file and we
|
|
|
|
* don't hold those caps, then we need to check whether
|
|
|
|
* the file is either opened or mmaped
|
|
|
|
*/
|
2013-08-12 21:42:15 -07:00
|
|
|
if ((issued & (CEPH_CAP_FILE_CACHE|
|
|
|
|
CEPH_CAP_FILE_BUFFER)) ||
|
2010-02-09 12:08:40 -07:00
|
|
|
mapping_mapped(inode->i_mapping) ||
|
2020-03-05 05:21:00 -07:00
|
|
|
__ceph_is_file_opened(ci)) {
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_truncate_pending++;
|
|
|
|
queue_trunc = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2023-03-06 00:01:15 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* It's possible that the new sizes of the two consecutive
|
|
|
|
* size truncations will be in the same fscrypt last block,
|
|
|
|
* and we need to truncate the corresponding page caches
|
|
|
|
* anyway.
|
|
|
|
*/
|
|
|
|
if (ceph_seq_cmp(truncate_seq, ci->i_truncate_seq) >= 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "truncate_size %lld -> %llu, encrypted %d\n",
|
|
|
|
ci->i_truncate_size, truncate_size,
|
|
|
|
!!IS_ENCRYPTED(inode));
|
2023-03-06 00:01:15 -07:00
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_truncate_size = truncate_size;
|
2023-03-06 00:01:15 -07:00
|
|
|
|
|
|
|
if (IS_ENCRYPTED(inode)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "truncate_pagecache_size %lld -> %llu\n",
|
|
|
|
ci->i_truncate_pagecache_size, size);
|
2022-08-25 06:31:11 -07:00
|
|
|
ci->i_truncate_pagecache_size = size;
|
2023-03-06 00:01:15 -07:00
|
|
|
} else {
|
2022-08-25 06:31:11 -07:00
|
|
|
ci->i_truncate_pagecache_size = truncate_size;
|
2023-03-06 00:01:15 -07:00
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
return queue_trunc;
|
|
|
|
}
|
|
|
|
|
|
|
|
void ceph_fill_file_time(struct inode *inode, int issued,
|
2018-07-13 13:18:36 -07:00
|
|
|
u64 time_warp_seq, struct timespec64 *ctime,
|
|
|
|
struct timespec64 *mtime, struct timespec64 *atime)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2023-07-05 12:00:55 -07:00
|
|
|
struct timespec64 ictime = inode_get_ctime(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
int warn = 0;
|
|
|
|
|
|
|
|
if (issued & (CEPH_CAP_FILE_EXCL|
|
|
|
|
CEPH_CAP_FILE_WR|
|
2010-11-08 10:24:34 -07:00
|
|
|
CEPH_CAP_FILE_BUFFER|
|
|
|
|
CEPH_CAP_AUTH_EXCL|
|
|
|
|
CEPH_CAP_XATTR_EXCL)) {
|
2018-03-26 01:46:39 -07:00
|
|
|
if (ci->i_version == 0 ||
|
2023-07-05 12:00:55 -07:00
|
|
|
timespec64_compare(ctime, &ictime) > 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "ctime %lld.%09ld -> %lld.%09ld inc w/ cap\n",
|
2023-07-05 12:00:55 -07:00
|
|
|
ictime.tv_sec, ictime.tv_nsec,
|
2018-07-13 13:18:36 -07:00
|
|
|
ctime->tv_sec, ctime->tv_nsec);
|
2023-07-05 12:00:55 -07:00
|
|
|
inode_set_ctime_to_ts(inode, *ctime);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
2018-03-26 01:46:39 -07:00
|
|
|
if (ci->i_version == 0 ||
|
|
|
|
ceph_seq_cmp(time_warp_seq, ci->i_time_warp_seq) > 0) {
|
2009-10-06 11:31:08 -07:00
|
|
|
/* the MDS did a utimes() */
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "mtime %lld.%09ld -> %lld.%09ld tw %d -> %d\n",
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_get_mtime_sec(inode),
|
|
|
|
inode_get_mtime_nsec(inode),
|
2018-07-13 13:18:36 -07:00
|
|
|
mtime->tv_sec, mtime->tv_nsec,
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_time_warp_seq, (int)time_warp_seq);
|
|
|
|
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_mtime_to_ts(inode, *mtime);
|
|
|
|
inode_set_atime_to_ts(inode, *atime);
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_time_warp_seq = time_warp_seq;
|
|
|
|
} else if (time_warp_seq == ci->i_time_warp_seq) {
|
2023-10-04 11:52:09 -07:00
|
|
|
struct timespec64 ts;
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/* nobody did utimes(); take the max */
|
2023-10-04 11:52:09 -07:00
|
|
|
ts = inode_get_mtime(inode);
|
|
|
|
if (timespec64_compare(mtime, &ts) > 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "mtime %lld.%09ld -> %lld.%09ld inc\n",
|
2023-10-04 11:52:09 -07:00
|
|
|
ts.tv_sec, ts.tv_nsec,
|
2018-07-13 13:18:36 -07:00
|
|
|
mtime->tv_sec, mtime->tv_nsec);
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_mtime_to_ts(inode, *mtime);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
2023-10-04 11:52:09 -07:00
|
|
|
ts = inode_get_atime(inode);
|
|
|
|
if (timespec64_compare(atime, &ts) > 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "atime %lld.%09ld -> %lld.%09ld inc\n",
|
2023-10-04 11:52:09 -07:00
|
|
|
ts.tv_sec, ts.tv_nsec,
|
2018-07-13 13:18:36 -07:00
|
|
|
atime->tv_sec, atime->tv_nsec);
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_atime_to_ts(inode, *atime);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
} else if (issued & CEPH_CAP_FILE_EXCL) {
|
|
|
|
/* we did a utimes(); ignore mds values */
|
|
|
|
} else {
|
|
|
|
warn = 1;
|
|
|
|
}
|
|
|
|
} else {
|
2010-11-08 10:24:34 -07:00
|
|
|
/* we have no write|excl caps; whatever the MDS says is true */
|
2009-10-06 11:31:08 -07:00
|
|
|
if (ceph_seq_cmp(time_warp_seq, ci->i_time_warp_seq) >= 0) {
|
2023-07-05 12:00:55 -07:00
|
|
|
inode_set_ctime_to_ts(inode, *ctime);
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_mtime_to_ts(inode, *mtime);
|
|
|
|
inode_set_atime_to_ts(inode, *atime);
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_time_warp_seq = time_warp_seq;
|
|
|
|
} else {
|
|
|
|
warn = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (warn) /* time_warp_seq shouldn't go backwards */
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p mds time_warp_seq %llu < %u\n", inode,
|
|
|
|
time_warp_seq, ci->i_time_warp_seq);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
2020-09-03 10:31:10 -07:00
|
|
|
#if IS_ENABLED(CONFIG_FS_ENCRYPTION)
|
2023-06-11 18:04:07 -07:00
|
|
|
static int decode_encrypted_symlink(struct ceph_mds_client *mdsc,
|
|
|
|
const char *encsym,
|
|
|
|
int enclen, u8 **decsym)
|
2020-09-03 10:31:10 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = mdsc->fsc->client;
|
2020-09-03 10:31:10 -07:00
|
|
|
int declen;
|
|
|
|
u8 *sym;
|
|
|
|
|
|
|
|
sym = kmalloc(enclen + 1, GFP_NOFS);
|
|
|
|
if (!sym)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
declen = ceph_base64_decode(encsym, enclen, sym);
|
|
|
|
if (declen < 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl,
|
|
|
|
"can't decode symlink (%d). Content: %.*s\n",
|
|
|
|
declen, enclen, encsym);
|
2020-09-03 10:31:10 -07:00
|
|
|
kfree(sym);
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
sym[declen + 1] = '\0';
|
|
|
|
*decsym = sym;
|
|
|
|
return declen;
|
|
|
|
}
|
|
|
|
#else
|
2023-06-11 18:04:07 -07:00
|
|
|
static int decode_encrypted_symlink(struct ceph_mds_client *mdsc,
|
|
|
|
const char *encsym,
|
|
|
|
int symlen, u8 **decsym)
|
2020-09-03 10:31:10 -07:00
|
|
|
{
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* Populate an inode based on info from mds. May be called on new or
|
|
|
|
* existing inodes.
|
|
|
|
*/
|
2019-12-05 07:09:25 -07:00
|
|
|
int ceph_fill_inode(struct inode *inode, struct page *locked_page,
|
|
|
|
struct ceph_mds_reply_info_in *iinfo,
|
|
|
|
struct ceph_mds_reply_dirfrag *dirinfo,
|
|
|
|
struct ceph_mds_session *session, int cap_fmode,
|
|
|
|
struct ceph_cap_reservation *caps_reservation)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2020-09-03 06:01:39 -07:00
|
|
|
struct ceph_mds_client *mdsc = ceph_sb_to_mdsc(inode->i_sb);
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = mdsc->fsc->client;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_mds_reply_inode *info = iinfo->in;
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2018-04-26 20:14:39 -07:00
|
|
|
int issued, new_issued, info_caps;
|
2018-07-13 13:18:36 -07:00
|
|
|
struct timespec64 mtime, atime, ctime;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_buffer *xattr_blob = NULL;
|
ceph: fix buffer free while holding i_ceph_lock in fill_inode()
Calling ceph_buffer_put() in fill_inode() may result in freeing the
i_xattrs.blob buffer while holding the i_ceph_lock. This can be fixed by
postponing the call until later, when the lock is released.
The following backtrace was triggered by fstests generic/070.
BUG: sleeping function called from invalid context at mm/vmalloc.c:2283
in_atomic(): 1, irqs_disabled(): 0, pid: 3852, name: kworker/0:4
6 locks held by kworker/0:4/3852:
#0: 000000004270f6bb ((wq_completion)ceph-msgr){+.+.}, at: process_one_work+0x1b8/0x5f0
#1: 00000000eb420803 ((work_completion)(&(&con->work)->work)){+.+.}, at: process_one_work+0x1b8/0x5f0
#2: 00000000be1c53a4 (&s->s_mutex){+.+.}, at: dispatch+0x288/0x1476
#3: 00000000559cb958 (&mdsc->snap_rwsem){++++}, at: dispatch+0x2eb/0x1476
#4: 000000000d5ebbae (&req->r_fill_mutex){+.+.}, at: dispatch+0x2fc/0x1476
#5: 00000000a83d0514 (&(&ci->i_ceph_lock)->rlock){+.+.}, at: fill_inode.isra.0+0xf8/0xf70
CPU: 0 PID: 3852 Comm: kworker/0:4 Not tainted 5.2.0+ #441
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.1-0-ga5cab58-prebuilt.qemu.org 04/01/2014
Workqueue: ceph-msgr ceph_con_workfn
Call Trace:
dump_stack+0x67/0x90
___might_sleep.cold+0x9f/0xb1
vfree+0x4b/0x60
ceph_buffer_release+0x1b/0x60
fill_inode.isra.0+0xa9b/0xf70
ceph_fill_trace+0x13b/0xc70
? dispatch+0x2eb/0x1476
dispatch+0x320/0x1476
? __mutex_unlock_slowpath+0x4d/0x2a0
ceph_con_workfn+0xc97/0x2ec0
? process_one_work+0x1b8/0x5f0
process_one_work+0x244/0x5f0
worker_thread+0x4d/0x3e0
kthread+0x105/0x140
? process_one_work+0x5f0/0x5f0
? kthread_park+0x90/0x90
ret_from_fork+0x3a/0x50
Signed-off-by: Luis Henriques <lhenriques@suse.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-07-19 07:32:22 -07:00
|
|
|
struct ceph_buffer *old_blob = NULL;
|
2016-03-06 18:35:06 -07:00
|
|
|
struct ceph_string *pool_ns = NULL;
|
2014-04-17 18:57:11 -07:00
|
|
|
struct ceph_cap *new_cap = NULL;
|
2009-10-06 11:31:08 -07:00
|
|
|
int err = 0;
|
2014-04-17 18:57:11 -07:00
|
|
|
bool wake = false;
|
2014-04-16 17:55:50 -07:00
|
|
|
bool queue_trunc = false;
|
|
|
|
bool new_version = false;
|
2014-11-14 06:41:55 -07:00
|
|
|
bool fill_inline = false;
|
2021-02-25 13:04:16 -07:00
|
|
|
umode_t mode = le32_to_cpu(info->mode);
|
|
|
|
dev_t rdev = le32_to_cpu(info->rdev);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2021-06-01 06:40:25 -07:00
|
|
|
lockdep_assert_held(&mdsc->snap_rwsem);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p ino %llx.%llx v %llu had %llu\n", inode, ceph_vinop(inode),
|
|
|
|
le64_to_cpu(info->version), ci->i_version);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2021-02-25 13:04:16 -07:00
|
|
|
/* Once I_NEW is cleared, we can't change type or dev numbers */
|
|
|
|
if (inode->i_state & I_NEW) {
|
|
|
|
inode->i_mode = mode;
|
|
|
|
} else {
|
|
|
|
if (inode_wrong_type(inode, mode)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_warn_once_client(cl,
|
|
|
|
"inode type changed! (ino %llx.%llx is 0%o, mds says 0%o)\n",
|
|
|
|
ceph_vinop(inode), inode->i_mode, mode);
|
2021-02-25 13:04:16 -07:00
|
|
|
return -ESTALE;
|
|
|
|
}
|
|
|
|
|
|
|
|
if ((S_ISCHR(mode) || S_ISBLK(mode)) && inode->i_rdev != rdev) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_warn_once_client(cl,
|
|
|
|
"dev inode rdev changed! (ino %llx.%llx is %u:%u, mds says %u:%u)\n",
|
|
|
|
ceph_vinop(inode), MAJOR(inode->i_rdev),
|
|
|
|
MINOR(inode->i_rdev), MAJOR(rdev),
|
|
|
|
MINOR(rdev));
|
2021-02-25 13:04:16 -07:00
|
|
|
return -ESTALE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-04-26 20:14:39 -07:00
|
|
|
info_caps = le32_to_cpu(info->cap.caps);
|
|
|
|
|
2014-04-17 18:57:11 -07:00
|
|
|
/* prealloc new cap struct */
|
2019-12-05 06:41:25 -07:00
|
|
|
if (info_caps && ceph_snap(inode) == CEPH_NOSNAP) {
|
2014-04-17 18:57:11 -07:00
|
|
|
new_cap = ceph_get_cap(mdsc, caps_reservation);
|
2019-12-05 06:41:25 -07:00
|
|
|
if (!new_cap)
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
2014-04-17 18:57:11 -07:00
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* prealloc xattr data, if it looks like we'll need it. only
|
|
|
|
* if len > 4 (meaning there are actually xattrs; the first 4
|
|
|
|
* bytes are the xattr count).
|
|
|
|
*/
|
|
|
|
if (iinfo->xattr_len > 4) {
|
2009-12-07 13:17:17 -07:00
|
|
|
xattr_blob = ceph_buffer_new(iinfo->xattr_len, GFP_NOFS);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (!xattr_blob)
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "ENOMEM xattr blob %d bytes\n",
|
|
|
|
iinfo->xattr_len);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
2016-03-06 18:35:06 -07:00
|
|
|
if (iinfo->pool_ns_len > 0)
|
|
|
|
pool_ns = ceph_find_or_create_string(iinfo->pool_ns_data,
|
|
|
|
iinfo->pool_ns_len);
|
|
|
|
|
2017-12-14 00:11:09 -07:00
|
|
|
if (ceph_snap(inode) != CEPH_NOSNAP && !ci->i_snapid_map)
|
|
|
|
ci->i_snapid_map = ceph_get_snapid_map(mdsc, ceph_snap(inode));
|
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* provided version will be odd if inode value is projected,
|
2010-11-08 10:23:12 -07:00
|
|
|
* even if stable. skip the update if we have newer stable
|
|
|
|
* info (ours>=theirs, e.g. due to racing mds replies), unless
|
|
|
|
* we are getting projected (unstable) info (in which case the
|
|
|
|
* version is odd, and we want ours>theirs).
|
|
|
|
* us them
|
|
|
|
* 2 2 skip
|
|
|
|
* 3 2 skip
|
|
|
|
* 3 3 update
|
2009-10-06 11:31:08 -07:00
|
|
|
*/
|
2014-04-16 17:55:50 -07:00
|
|
|
if (ci->i_version == 0 ||
|
|
|
|
((info->cap.flags & CEPH_CAP_FLAG_AUTH) &&
|
|
|
|
le64_to_cpu(info->version) > (ci->i_version & ~1)))
|
|
|
|
new_version = true;
|
|
|
|
|
2019-06-06 04:29:23 -07:00
|
|
|
/* Update change_attribute */
|
|
|
|
inode_set_max_iversion_raw(inode, iinfo->change_attr);
|
|
|
|
|
2018-04-26 20:14:39 -07:00
|
|
|
__ceph_caps_issued(ci, &issued);
|
|
|
|
issued |= __ceph_caps_dirty(ci);
|
|
|
|
new_issued = ~issued & info_caps;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2018-01-12 10:19:29 -07:00
|
|
|
__ceph_update_quota(ci, iinfo->max_bytes, iinfo->max_files);
|
2018-01-05 03:47:18 -07:00
|
|
|
|
2020-07-27 07:16:09 -07:00
|
|
|
#ifdef CONFIG_FS_ENCRYPTION
|
2022-11-03 07:06:40 -07:00
|
|
|
if (iinfo->fscrypt_auth_len &&
|
|
|
|
((inode->i_state & I_NEW) || (ci->fscrypt_auth_len == 0))) {
|
2020-07-27 07:16:09 -07:00
|
|
|
kfree(ci->fscrypt_auth);
|
|
|
|
ci->fscrypt_auth_len = iinfo->fscrypt_auth_len;
|
|
|
|
ci->fscrypt_auth = iinfo->fscrypt_auth;
|
|
|
|
iinfo->fscrypt_auth = NULL;
|
|
|
|
iinfo->fscrypt_auth_len = 0;
|
|
|
|
inode_set_flags(inode, S_ENCRYPTED, S_ENCRYPTED);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2014-04-16 17:55:50 -07:00
|
|
|
if ((new_version || (new_issued & CEPH_CAP_AUTH_SHARED)) &&
|
|
|
|
(issued & CEPH_CAP_AUTH_EXCL) == 0) {
|
2021-02-25 13:04:16 -07:00
|
|
|
inode->i_mode = mode;
|
2013-01-31 04:40:12 -07:00
|
|
|
inode->i_uid = make_kuid(&init_user_ns, le32_to_cpu(info->uid));
|
|
|
|
inode->i_gid = make_kgid(&init_user_ns, le32_to_cpu(info->gid));
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx mode 0%o uid.gid %d.%d\n", inode,
|
|
|
|
ceph_vinop(inode), inode->i_mode,
|
|
|
|
from_kuid(&init_user_ns, inode->i_uid),
|
|
|
|
from_kgid(&init_user_ns, inode->i_gid));
|
2019-05-29 08:19:42 -07:00
|
|
|
ceph_decode_timespec64(&ci->i_btime, &iinfo->btime);
|
2019-04-18 05:15:46 -07:00
|
|
|
ceph_decode_timespec64(&ci->i_snap_btime, &iinfo->snap_btime);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
2021-07-01 11:40:51 -07:00
|
|
|
/* directories have fl_stripe_unit set to zero */
|
|
|
|
if (IS_ENCRYPTED(inode))
|
|
|
|
inode->i_blkbits = CEPH_FSCRYPT_BLOCK_SHIFT;
|
|
|
|
else if (le32_to_cpu(info->layout.fl_stripe_unit))
|
|
|
|
inode->i_blkbits =
|
|
|
|
fls(le32_to_cpu(info->layout.fl_stripe_unit)) - 1;
|
|
|
|
else
|
|
|
|
inode->i_blkbits = CEPH_BLOCK_SHIFT;
|
|
|
|
|
2014-04-16 17:55:50 -07:00
|
|
|
if ((new_version || (new_issued & CEPH_CAP_LINK_SHARED)) &&
|
|
|
|
(issued & CEPH_CAP_LINK_EXCL) == 0)
|
2011-10-28 05:13:29 -07:00
|
|
|
set_nlink(inode, le32_to_cpu(info->nlink));
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2014-04-16 17:55:50 -07:00
|
|
|
if (new_version || (new_issued & CEPH_CAP_ANY_RD)) {
|
|
|
|
/* be careful with mtime, atime, size */
|
2018-07-13 13:18:36 -07:00
|
|
|
ceph_decode_timespec64(&atime, &info->atime);
|
|
|
|
ceph_decode_timespec64(&mtime, &info->mtime);
|
|
|
|
ceph_decode_timespec64(&ctime, &info->ctime);
|
2014-04-16 17:55:50 -07:00
|
|
|
ceph_fill_file_time(inode, issued,
|
|
|
|
le32_to_cpu(info->time_warp_seq),
|
|
|
|
&ctime, &mtime, &atime);
|
|
|
|
}
|
|
|
|
|
2018-04-26 20:14:39 -07:00
|
|
|
if (new_version || (info_caps & CEPH_CAP_FILE_SHARED)) {
|
|
|
|
ci->i_files = le64_to_cpu(info->files);
|
|
|
|
ci->i_subdirs = le64_to_cpu(info->subdirs);
|
|
|
|
}
|
|
|
|
|
2014-04-16 17:55:50 -07:00
|
|
|
if (new_version ||
|
|
|
|
(new_issued & (CEPH_CAP_ANY_FILE_RD | CEPH_CAP_ANY_FILE_WR))) {
|
2022-08-25 06:31:06 -07:00
|
|
|
u64 size = le64_to_cpu(info->size);
|
2016-02-03 06:24:49 -07:00
|
|
|
s64 old_pool = ci->i_layout.pool_id;
|
2016-03-06 18:35:06 -07:00
|
|
|
struct ceph_string *old_ns;
|
|
|
|
|
2016-02-03 06:24:49 -07:00
|
|
|
ceph_file_layout_from_legacy(&ci->i_layout, &info->layout);
|
2016-03-06 18:35:06 -07:00
|
|
|
old_ns = rcu_dereference_protected(ci->i_layout.pool_ns,
|
|
|
|
lockdep_is_held(&ci->i_ceph_lock));
|
|
|
|
rcu_assign_pointer(ci->i_layout.pool_ns, pool_ns);
|
|
|
|
|
|
|
|
if (ci->i_layout.pool_id != old_pool || pool_ns != old_ns)
|
2015-04-27 00:33:28 -07:00
|
|
|
ci->i_ceph_flags &= ~CEPH_I_POOL_PERM;
|
|
|
|
|
2016-03-06 18:35:06 -07:00
|
|
|
pool_ns = old_ns;
|
2015-04-27 00:33:28 -07:00
|
|
|
|
2022-08-25 06:31:06 -07:00
|
|
|
if (IS_ENCRYPTED(inode) && size &&
|
|
|
|
iinfo->fscrypt_file_len == sizeof(__le64)) {
|
|
|
|
u64 fsize = __le64_to_cpu(*(__le64 *)iinfo->fscrypt_file);
|
|
|
|
|
|
|
|
if (size == round_up(fsize, CEPH_FSCRYPT_BLOCK_SIZE)) {
|
|
|
|
size = fsize;
|
|
|
|
} else {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_warn_client(cl,
|
|
|
|
"fscrypt size mismatch: size=%llu fscrypt_file=%llu, discarding fscrypt_file size.\n",
|
2022-08-25 06:31:06 -07:00
|
|
|
info->size, size);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-04-16 17:55:50 -07:00
|
|
|
queue_trunc = ceph_fill_file_size(inode, issued,
|
|
|
|
le32_to_cpu(info->truncate_seq),
|
|
|
|
le64_to_cpu(info->truncate_size),
|
2022-08-25 06:31:06 -07:00
|
|
|
size);
|
2014-04-16 17:55:50 -07:00
|
|
|
/* only update max_size on auth cap */
|
|
|
|
if ((info->cap.flags & CEPH_CAP_FLAG_AUTH) &&
|
|
|
|
ci->i_max_size != le64_to_cpu(info->max_size)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "max_size %lld -> %llu\n",
|
|
|
|
ci->i_max_size, le64_to_cpu(info->max_size));
|
2014-04-16 17:55:50 -07:00
|
|
|
ci->i_max_size = le64_to_cpu(info->max_size);
|
|
|
|
}
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2018-04-25 02:30:23 -07:00
|
|
|
/* layout and rstat are not tracked by capability, update them if
|
|
|
|
* the inode info is from auth mds */
|
|
|
|
if (new_version || (info->cap.flags & CEPH_CAP_FLAG_AUTH)) {
|
|
|
|
if (S_ISDIR(inode->i_mode)) {
|
|
|
|
ci->i_dir_layout = iinfo->dir_layout;
|
|
|
|
ci->i_rbytes = le64_to_cpu(info->rbytes);
|
|
|
|
ci->i_rfiles = le64_to_cpu(info->rfiles);
|
|
|
|
ci->i_rsubdirs = le64_to_cpu(info->rsubdirs);
|
2019-01-08 20:07:02 -07:00
|
|
|
ci->i_dir_pin = iinfo->dir_pin;
|
2020-08-27 18:28:44 -07:00
|
|
|
ci->i_rsnaps = iinfo->rsnaps;
|
2018-07-13 13:18:36 -07:00
|
|
|
ceph_decode_timespec64(&ci->i_rctime, &info->rctime);
|
2018-04-25 02:30:23 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/* xattrs */
|
|
|
|
/* note that if i_xattrs.len <= 4, i_xattrs.data will still be NULL. */
|
2014-09-16 06:46:17 -07:00
|
|
|
if ((ci->i_xattrs.version == 0 || !(issued & CEPH_CAP_XATTR_EXCL)) &&
|
2009-10-06 11:31:08 -07:00
|
|
|
le64_to_cpu(info->xattr_version) > ci->i_xattrs.version) {
|
|
|
|
if (ci->i_xattrs.blob)
|
ceph: fix buffer free while holding i_ceph_lock in fill_inode()
Calling ceph_buffer_put() in fill_inode() may result in freeing the
i_xattrs.blob buffer while holding the i_ceph_lock. This can be fixed by
postponing the call until later, when the lock is released.
The following backtrace was triggered by fstests generic/070.
BUG: sleeping function called from invalid context at mm/vmalloc.c:2283
in_atomic(): 1, irqs_disabled(): 0, pid: 3852, name: kworker/0:4
6 locks held by kworker/0:4/3852:
#0: 000000004270f6bb ((wq_completion)ceph-msgr){+.+.}, at: process_one_work+0x1b8/0x5f0
#1: 00000000eb420803 ((work_completion)(&(&con->work)->work)){+.+.}, at: process_one_work+0x1b8/0x5f0
#2: 00000000be1c53a4 (&s->s_mutex){+.+.}, at: dispatch+0x288/0x1476
#3: 00000000559cb958 (&mdsc->snap_rwsem){++++}, at: dispatch+0x2eb/0x1476
#4: 000000000d5ebbae (&req->r_fill_mutex){+.+.}, at: dispatch+0x2fc/0x1476
#5: 00000000a83d0514 (&(&ci->i_ceph_lock)->rlock){+.+.}, at: fill_inode.isra.0+0xf8/0xf70
CPU: 0 PID: 3852 Comm: kworker/0:4 Not tainted 5.2.0+ #441
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.1-0-ga5cab58-prebuilt.qemu.org 04/01/2014
Workqueue: ceph-msgr ceph_con_workfn
Call Trace:
dump_stack+0x67/0x90
___might_sleep.cold+0x9f/0xb1
vfree+0x4b/0x60
ceph_buffer_release+0x1b/0x60
fill_inode.isra.0+0xa9b/0xf70
ceph_fill_trace+0x13b/0xc70
? dispatch+0x2eb/0x1476
dispatch+0x320/0x1476
? __mutex_unlock_slowpath+0x4d/0x2a0
ceph_con_workfn+0xc97/0x2ec0
? process_one_work+0x1b8/0x5f0
process_one_work+0x244/0x5f0
worker_thread+0x4d/0x3e0
kthread+0x105/0x140
? process_one_work+0x5f0/0x5f0
? kthread_park+0x90/0x90
ret_from_fork+0x3a/0x50
Signed-off-by: Luis Henriques <lhenriques@suse.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-07-19 07:32:22 -07:00
|
|
|
old_blob = ci->i_xattrs.blob;
|
2009-10-06 11:31:08 -07:00
|
|
|
ci->i_xattrs.blob = xattr_blob;
|
|
|
|
if (xattr_blob)
|
|
|
|
memcpy(ci->i_xattrs.blob->vec.iov_base,
|
|
|
|
iinfo->xattr_data, iinfo->xattr_len);
|
|
|
|
ci->i_xattrs.version = le64_to_cpu(info->xattr_version);
|
2013-11-11 00:18:03 -07:00
|
|
|
ceph_forget_all_cached_acls(inode);
|
2019-05-26 01:27:56 -07:00
|
|
|
ceph_security_invalidate_secctx(inode);
|
2010-04-29 09:28:11 -07:00
|
|
|
xattr_blob = NULL;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
2018-03-26 01:46:39 -07:00
|
|
|
/* finally update i_version */
|
2018-05-26 01:54:39 -07:00
|
|
|
if (le64_to_cpu(info->version) > ci->i_version)
|
|
|
|
ci->i_version = le64_to_cpu(info->version);
|
2018-03-26 01:46:39 -07:00
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
inode->i_mapping->a_ops = &ceph_aops;
|
|
|
|
|
|
|
|
switch (inode->i_mode & S_IFMT) {
|
|
|
|
case S_IFIFO:
|
|
|
|
case S_IFBLK:
|
|
|
|
case S_IFCHR:
|
|
|
|
case S_IFSOCK:
|
2018-12-17 20:03:38 -07:00
|
|
|
inode->i_blkbits = PAGE_SHIFT;
|
2021-02-25 13:04:16 -07:00
|
|
|
init_special_inode(inode, inode->i_mode, rdev);
|
2009-10-06 11:31:08 -07:00
|
|
|
inode->i_op = &ceph_file_iops;
|
|
|
|
break;
|
|
|
|
case S_IFREG:
|
|
|
|
inode->i_op = &ceph_file_iops;
|
|
|
|
inode->i_fop = &ceph_file_fops;
|
|
|
|
break;
|
|
|
|
case S_IFLNK:
|
|
|
|
if (!ci->i_symlink) {
|
2012-02-03 07:55:36 -07:00
|
|
|
u32 symlen = iinfo->symlink_len;
|
2009-10-06 11:31:08 -07:00
|
|
|
char *sym;
|
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2020-09-03 10:31:10 -07:00
|
|
|
if (IS_ENCRYPTED(inode)) {
|
|
|
|
if (symlen != i_size_read(inode))
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl,
|
|
|
|
"%p %llx.%llx BAD symlink size %lld\n",
|
|
|
|
inode, ceph_vinop(inode),
|
2020-09-03 10:31:10 -07:00
|
|
|
i_size_read(inode));
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
err = decode_encrypted_symlink(mdsc, iinfo->symlink,
|
2020-09-03 10:31:10 -07:00
|
|
|
symlen, (u8 **)&sym);
|
|
|
|
if (err < 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl,
|
|
|
|
"decoding encrypted symlink failed: %d\n",
|
|
|
|
err);
|
2020-09-03 10:31:10 -07:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
symlen = err;
|
2016-05-05 01:40:17 -07:00
|
|
|
i_size_write(inode, symlen);
|
|
|
|
inode->i_blocks = calc_inode_blocks(symlen);
|
2020-09-03 10:31:10 -07:00
|
|
|
} else {
|
|
|
|
if (symlen != i_size_read(inode)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl,
|
|
|
|
"%p %llx.%llx BAD symlink size %lld\n",
|
|
|
|
inode, ceph_vinop(inode),
|
2020-09-03 10:31:10 -07:00
|
|
|
i_size_read(inode));
|
|
|
|
i_size_write(inode, symlen);
|
|
|
|
inode->i_blocks = calc_inode_blocks(symlen);
|
|
|
|
}
|
2012-02-03 07:55:36 -07:00
|
|
|
|
2020-09-03 10:31:10 -07:00
|
|
|
err = -ENOMEM;
|
|
|
|
sym = kstrndup(iinfo->symlink, symlen, GFP_NOFS);
|
|
|
|
if (!sym)
|
|
|
|
goto out;
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (!ci->i_symlink)
|
|
|
|
ci->i_symlink = sym;
|
|
|
|
else
|
|
|
|
kfree(sym); /* lost a race */
|
|
|
|
}
|
2020-09-03 10:31:10 -07:00
|
|
|
|
|
|
|
if (IS_ENCRYPTED(inode)) {
|
|
|
|
/*
|
|
|
|
* Encrypted symlinks need to be decrypted before we can
|
|
|
|
* cache their targets in i_link. Don't touch it here.
|
|
|
|
*/
|
|
|
|
inode->i_op = &ceph_encrypted_symlink_iops;
|
|
|
|
} else {
|
|
|
|
inode->i_link = ci->i_symlink;
|
|
|
|
inode->i_op = &ceph_symlink_iops;
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
break;
|
|
|
|
case S_IFDIR:
|
|
|
|
inode->i_op = &ceph_dir_iops;
|
|
|
|
inode->i_fop = &ceph_dir_fops;
|
|
|
|
break;
|
|
|
|
default:
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "%p %llx.%llx BAD mode 0%o\n", inode,
|
|
|
|
ceph_vinop(inode), inode->i_mode);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/* were we issued a capability? */
|
2018-04-26 20:14:39 -07:00
|
|
|
if (info_caps) {
|
2009-10-06 11:31:08 -07:00
|
|
|
if (ceph_snap(inode) == CEPH_NOSNAP) {
|
|
|
|
ceph_add_cap(inode, session,
|
|
|
|
le64_to_cpu(info->cap.cap_id),
|
2020-03-05 05:21:02 -07:00
|
|
|
info_caps,
|
2009-10-06 11:31:08 -07:00
|
|
|
le32_to_cpu(info->cap.wanted),
|
|
|
|
le32_to_cpu(info->cap.seq),
|
|
|
|
le32_to_cpu(info->cap.mseq),
|
|
|
|
le64_to_cpu(info->cap.realm),
|
2014-04-17 18:57:11 -07:00
|
|
|
info->cap.flags, &new_cap);
|
2015-01-18 22:12:24 -07:00
|
|
|
|
|
|
|
/* set dir completion flag? */
|
|
|
|
if (S_ISDIR(inode->i_mode) &&
|
|
|
|
ci->i_files == 0 && ci->i_subdirs == 0 &&
|
2018-04-26 20:14:39 -07:00
|
|
|
(info_caps & CEPH_CAP_FILE_SHARED) &&
|
2015-01-18 22:12:24 -07:00
|
|
|
(issued & CEPH_CAP_FILE_EXCL) == 0 &&
|
|
|
|
!__ceph_dir_is_complete(ci)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " marking %p complete (empty)\n",
|
|
|
|
inode);
|
2015-06-16 05:48:56 -07:00
|
|
|
i_size_write(inode, 0);
|
2015-01-18 22:12:24 -07:00
|
|
|
__ceph_dir_set_complete(ci,
|
2015-06-16 05:48:56 -07:00
|
|
|
atomic64_read(&ci->i_release_count),
|
|
|
|
atomic64_read(&ci->i_ordered_count));
|
2015-01-18 22:12:24 -07:00
|
|
|
}
|
|
|
|
|
2014-04-17 18:57:11 -07:00
|
|
|
wake = true;
|
2009-10-06 11:31:08 -07:00
|
|
|
} else {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " %p got snap_caps %s\n", inode,
|
|
|
|
ceph_cap_string(info_caps));
|
2018-04-26 20:14:39 -07:00
|
|
|
ci->i_snap_caps |= info_caps;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
}
|
2014-11-14 06:41:55 -07:00
|
|
|
|
|
|
|
if (iinfo->inline_version > 0 &&
|
|
|
|
iinfo->inline_version >= ci->i_inline_version) {
|
|
|
|
int cache_caps = CEPH_CAP_FILE_CACHE | CEPH_CAP_FILE_LAZYIO;
|
|
|
|
ci->i_inline_version = iinfo->inline_version;
|
2022-06-06 19:13:53 -07:00
|
|
|
if (ceph_has_inline_data(ci) &&
|
2018-04-26 20:14:39 -07:00
|
|
|
(locked_page || (info_caps & cache_caps)))
|
2014-11-14 06:41:55 -07:00
|
|
|
fill_inline = true;
|
|
|
|
}
|
|
|
|
|
2020-03-05 05:21:00 -07:00
|
|
|
if (cap_fmode >= 0) {
|
|
|
|
if (!info_caps)
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_warn_client(cl, "mds issued no caps on %llx.%llx\n",
|
|
|
|
ceph_vinop(inode));
|
2020-03-05 05:21:00 -07:00
|
|
|
__ceph_touch_fmode(ci, mdsc, cap_fmode);
|
|
|
|
}
|
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2021-12-07 06:44:50 -07:00
|
|
|
ceph_fscache_register_inode_cookie(inode);
|
|
|
|
|
2014-11-14 06:41:55 -07:00
|
|
|
if (fill_inline)
|
2014-11-14 06:56:29 -07:00
|
|
|
ceph_fill_inline_data(inode, locked_page,
|
2014-11-14 06:41:55 -07:00
|
|
|
iinfo->inline_data, iinfo->inline_len);
|
|
|
|
|
2014-04-17 18:57:11 -07:00
|
|
|
if (wake)
|
|
|
|
wake_up_all(&ci->i_cap_wq);
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/* queue truncate if we saw i_size decrease */
|
|
|
|
if (queue_trunc)
|
2010-02-09 16:24:44 -07:00
|
|
|
ceph_queue_vmtruncate(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/* populate frag tree */
|
2014-04-18 01:00:30 -07:00
|
|
|
if (S_ISDIR(inode->i_mode))
|
|
|
|
ceph_fill_fragtree(inode, &info->fragtree, dirinfo);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/* update delegation info? */
|
|
|
|
if (dirinfo)
|
|
|
|
ceph_fill_dirfrag(inode, dirinfo);
|
|
|
|
|
|
|
|
err = 0;
|
|
|
|
out:
|
2014-04-17 18:57:11 -07:00
|
|
|
if (new_cap)
|
|
|
|
ceph_put_cap(mdsc, new_cap);
|
ceph: fix buffer free while holding i_ceph_lock in fill_inode()
Calling ceph_buffer_put() in fill_inode() may result in freeing the
i_xattrs.blob buffer while holding the i_ceph_lock. This can be fixed by
postponing the call until later, when the lock is released.
The following backtrace was triggered by fstests generic/070.
BUG: sleeping function called from invalid context at mm/vmalloc.c:2283
in_atomic(): 1, irqs_disabled(): 0, pid: 3852, name: kworker/0:4
6 locks held by kworker/0:4/3852:
#0: 000000004270f6bb ((wq_completion)ceph-msgr){+.+.}, at: process_one_work+0x1b8/0x5f0
#1: 00000000eb420803 ((work_completion)(&(&con->work)->work)){+.+.}, at: process_one_work+0x1b8/0x5f0
#2: 00000000be1c53a4 (&s->s_mutex){+.+.}, at: dispatch+0x288/0x1476
#3: 00000000559cb958 (&mdsc->snap_rwsem){++++}, at: dispatch+0x2eb/0x1476
#4: 000000000d5ebbae (&req->r_fill_mutex){+.+.}, at: dispatch+0x2fc/0x1476
#5: 00000000a83d0514 (&(&ci->i_ceph_lock)->rlock){+.+.}, at: fill_inode.isra.0+0xf8/0xf70
CPU: 0 PID: 3852 Comm: kworker/0:4 Not tainted 5.2.0+ #441
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.1-0-ga5cab58-prebuilt.qemu.org 04/01/2014
Workqueue: ceph-msgr ceph_con_workfn
Call Trace:
dump_stack+0x67/0x90
___might_sleep.cold+0x9f/0xb1
vfree+0x4b/0x60
ceph_buffer_release+0x1b/0x60
fill_inode.isra.0+0xa9b/0xf70
ceph_fill_trace+0x13b/0xc70
? dispatch+0x2eb/0x1476
dispatch+0x320/0x1476
? __mutex_unlock_slowpath+0x4d/0x2a0
ceph_con_workfn+0xc97/0x2ec0
? process_one_work+0x1b8/0x5f0
process_one_work+0x244/0x5f0
worker_thread+0x4d/0x3e0
kthread+0x105/0x140
? process_one_work+0x5f0/0x5f0
? kthread_park+0x90/0x90
ret_from_fork+0x3a/0x50
Signed-off-by: Luis Henriques <lhenriques@suse.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2019-07-19 07:32:22 -07:00
|
|
|
ceph_buffer_put(old_blob);
|
|
|
|
ceph_buffer_put(xattr_blob);
|
2016-03-06 18:35:06 -07:00
|
|
|
ceph_put_string(pool_ns);
|
2009-10-06 11:31:08 -07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2019-05-22 06:49:44 -07:00
|
|
|
* caller should hold session s_mutex and dentry->d_lock.
|
2009-10-06 11:31:08 -07:00
|
|
|
*/
|
2019-05-22 06:49:44 -07:00
|
|
|
static void __update_dentry_lease(struct inode *dir, struct dentry *dentry,
|
|
|
|
struct ceph_mds_reply_lease *lease,
|
|
|
|
struct ceph_mds_session *session,
|
|
|
|
unsigned long from_time,
|
|
|
|
struct ceph_mds_session **old_lease_session)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(dir);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_dentry_info *di = ceph_dentry(dentry);
|
2020-02-18 12:12:32 -07:00
|
|
|
unsigned mask = le16_to_cpu(lease->mask);
|
2009-10-06 11:31:08 -07:00
|
|
|
long unsigned duration = le32_to_cpu(lease->duration_ms);
|
|
|
|
long unsigned ttl = from_time + (duration * HZ) / 1000;
|
|
|
|
long unsigned half_ttl = from_time + (duration * HZ / 2) / 1000;
|
2017-01-27 07:13:57 -07:00
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p duration %lu ms ttl %lu\n", dentry, duration, ttl);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2016-10-28 18:52:50 -07:00
|
|
|
/* only track leases on regular dentries */
|
|
|
|
if (ceph_snap(dir) != CEPH_NOSNAP)
|
2019-05-22 06:49:44 -07:00
|
|
|
return;
|
2016-10-28 18:52:50 -07:00
|
|
|
|
2020-02-18 12:12:32 -07:00
|
|
|
if (mask & CEPH_LEASE_PRIMARY_LINK)
|
|
|
|
di->flags |= CEPH_DENTRY_PRIMARY_LINK;
|
|
|
|
else
|
|
|
|
di->flags &= ~CEPH_DENTRY_PRIMARY_LINK;
|
|
|
|
|
2017-11-26 19:47:46 -07:00
|
|
|
di->lease_shared_gen = atomic_read(&ceph_inode(dir)->i_shared_gen);
|
2020-02-18 12:12:32 -07:00
|
|
|
if (!(mask & CEPH_LEASE_VALID)) {
|
2019-01-31 01:55:51 -07:00
|
|
|
__ceph_dentry_dir_lease_touch(di);
|
2019-05-22 06:49:44 -07:00
|
|
|
return;
|
2019-01-31 01:55:51 -07:00
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2021-06-04 09:03:09 -07:00
|
|
|
if (di->lease_gen == atomic_read(&session->s_cap_gen) &&
|
2016-06-22 07:35:04 -07:00
|
|
|
time_before(ttl, di->time))
|
2019-05-22 06:49:44 -07:00
|
|
|
return; /* we already have a newer lease. */
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2017-07-02 18:09:10 -07:00
|
|
|
if (di->lease_session && di->lease_session != session) {
|
2019-05-22 06:49:44 -07:00
|
|
|
*old_lease_session = di->lease_session;
|
2017-07-02 18:09:10 -07:00
|
|
|
di->lease_session = NULL;
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (!di->lease_session)
|
|
|
|
di->lease_session = ceph_get_mds_session(session);
|
2021-06-04 09:03:09 -07:00
|
|
|
di->lease_gen = atomic_read(&session->s_cap_gen);
|
2009-10-06 11:31:08 -07:00
|
|
|
di->lease_seq = le32_to_cpu(lease->seq);
|
|
|
|
di->lease_renew_after = half_ttl;
|
|
|
|
di->lease_renew_from = 0;
|
2016-06-22 07:35:04 -07:00
|
|
|
di->time = ttl;
|
2019-01-31 01:55:51 -07:00
|
|
|
|
|
|
|
__ceph_dentry_lease_touch(di);
|
2019-05-22 06:49:44 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void update_dentry_lease(struct inode *dir, struct dentry *dentry,
|
|
|
|
struct ceph_mds_reply_lease *lease,
|
|
|
|
struct ceph_mds_session *session,
|
|
|
|
unsigned long from_time)
|
|
|
|
{
|
|
|
|
struct ceph_mds_session *old_lease_session = NULL;
|
|
|
|
spin_lock(&dentry->d_lock);
|
|
|
|
__update_dentry_lease(dir, dentry, lease, session, from_time,
|
|
|
|
&old_lease_session);
|
|
|
|
spin_unlock(&dentry->d_lock);
|
2021-06-09 11:09:52 -07:00
|
|
|
ceph_put_mds_session(old_lease_session);
|
2019-05-22 06:49:44 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* update dentry lease without having parent inode locked
|
|
|
|
*/
|
|
|
|
static void update_dentry_lease_careful(struct dentry *dentry,
|
|
|
|
struct ceph_mds_reply_lease *lease,
|
|
|
|
struct ceph_mds_session *session,
|
|
|
|
unsigned long from_time,
|
|
|
|
char *dname, u32 dname_len,
|
|
|
|
struct ceph_vino *pdvino,
|
|
|
|
struct ceph_vino *ptvino)
|
|
|
|
|
|
|
|
{
|
|
|
|
struct inode *dir;
|
|
|
|
struct ceph_mds_session *old_lease_session = NULL;
|
|
|
|
|
|
|
|
spin_lock(&dentry->d_lock);
|
|
|
|
/* make sure dentry's name matches target */
|
|
|
|
if (dentry->d_name.len != dname_len ||
|
|
|
|
memcmp(dentry->d_name.name, dname, dname_len))
|
|
|
|
goto out_unlock;
|
|
|
|
|
|
|
|
dir = d_inode(dentry->d_parent);
|
|
|
|
/* make sure parent matches dvino */
|
|
|
|
if (!ceph_ino_compare(dir, pdvino))
|
|
|
|
goto out_unlock;
|
|
|
|
|
|
|
|
/* make sure dentry's inode matches target. NULL ptvino means that
|
|
|
|
* we expect a negative dentry */
|
|
|
|
if (ptvino) {
|
|
|
|
if (d_really_is_negative(dentry))
|
|
|
|
goto out_unlock;
|
|
|
|
if (!ceph_ino_compare(d_inode(dentry), ptvino))
|
|
|
|
goto out_unlock;
|
|
|
|
} else {
|
|
|
|
if (d_really_is_positive(dentry))
|
|
|
|
goto out_unlock;
|
|
|
|
}
|
|
|
|
|
|
|
|
__update_dentry_lease(dir, dentry, lease, session,
|
|
|
|
from_time, &old_lease_session);
|
2009-10-06 11:31:08 -07:00
|
|
|
out_unlock:
|
|
|
|
spin_unlock(&dentry->d_lock);
|
2021-06-09 11:09:52 -07:00
|
|
|
ceph_put_mds_session(old_lease_session);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* splice a dentry to an inode.
|
2022-02-10 20:33:25 -07:00
|
|
|
* caller must hold directory i_rwsem for this to be safe.
|
2009-10-06 11:31:08 -07:00
|
|
|
*/
|
2018-10-25 02:30:30 -07:00
|
|
|
static int splice_dentry(struct dentry **pdn, struct inode *in)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(in);
|
2018-10-25 02:30:30 -07:00
|
|
|
struct dentry *dn = *pdn;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct dentry *realdn;
|
|
|
|
|
2015-03-17 15:25:59 -07:00
|
|
|
BUG_ON(d_inode(dn));
|
2010-05-03 22:08:02 -07:00
|
|
|
|
2017-11-26 20:23:48 -07:00
|
|
|
if (S_ISDIR(in->i_mode)) {
|
|
|
|
/* If inode is directory, d_splice_alias() below will remove
|
|
|
|
* 'realdn' from its origin parent. We need to ensure that
|
|
|
|
* origin parent's readdir cache will not reference 'realdn'
|
|
|
|
*/
|
|
|
|
realdn = d_find_any_alias(in);
|
|
|
|
if (realdn) {
|
|
|
|
struct ceph_dentry_info *di = ceph_dentry(realdn);
|
|
|
|
spin_lock(&realdn->d_lock);
|
|
|
|
|
|
|
|
realdn->d_op->d_prune(realdn);
|
|
|
|
|
|
|
|
di->time = jiffies;
|
|
|
|
di->lease_shared_gen = 0;
|
|
|
|
di->offset = 0;
|
|
|
|
|
|
|
|
spin_unlock(&realdn->d_lock);
|
|
|
|
dput(realdn);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/* dn must be unhashed */
|
|
|
|
if (!d_unhashed(dn))
|
|
|
|
d_drop(dn);
|
2014-10-12 19:24:21 -07:00
|
|
|
realdn = d_splice_alias(in, dn);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (IS_ERR(realdn)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "error %ld %p inode %p ino %llx.%llx\n",
|
|
|
|
PTR_ERR(realdn), dn, in, ceph_vinop(in));
|
2018-10-25 02:30:30 -07:00
|
|
|
return PTR_ERR(realdn);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (realdn) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "dn %p (%d) spliced with %p (%d) inode %p ino %llx.%llx\n",
|
|
|
|
dn, d_count(dn), realdn, d_count(realdn),
|
|
|
|
d_inode(realdn), ceph_vinop(d_inode(realdn)));
|
2009-10-06 11:31:08 -07:00
|
|
|
dput(dn);
|
2018-10-25 02:30:30 -07:00
|
|
|
*pdn = realdn;
|
2009-10-06 11:31:08 -07:00
|
|
|
} else {
|
|
|
|
BUG_ON(!ceph_dentry(dn));
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "dn %p attached to %p ino %llx.%llx\n", dn,
|
|
|
|
d_inode(dn), ceph_vinop(d_inode(dn)));
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
2018-10-25 02:30:30 -07:00
|
|
|
return 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Incorporate results into the local cache. This is either just
|
|
|
|
* one inode, or a directory, dentry, and possibly linked-to inode (e.g.,
|
|
|
|
* after a lookup).
|
|
|
|
*
|
|
|
|
* A reply may contain
|
|
|
|
* a directory inode along with a dentry.
|
|
|
|
* and/or a target inode
|
|
|
|
*
|
|
|
|
* Called with snap_rwsem (read).
|
|
|
|
*/
|
2017-01-31 09:06:13 -07:00
|
|
|
int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2017-01-31 09:06:13 -07:00
|
|
|
struct ceph_mds_session *session = req->r_session;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
|
|
|
|
struct inode *in = NULL;
|
2017-01-27 07:13:57 -07:00
|
|
|
struct ceph_vino tvino, dvino;
|
2023-06-11 19:50:38 -07:00
|
|
|
struct ceph_fs_client *fsc = ceph_sb_to_fs_client(sb);
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = fsc->client;
|
2009-10-06 11:31:08 -07:00
|
|
|
int err = 0;
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p is_dentry %d is_target %d\n", req,
|
|
|
|
rinfo->head->is_dentry, rinfo->head->is_target);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (!rinfo->head->is_target && !rinfo->head->is_dentry) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "reply is empty!\n");
|
2017-01-31 08:28:26 -07:00
|
|
|
if (rinfo->head->result == 0 && req->r_parent)
|
2010-05-14 10:02:57 -07:00
|
|
|
ceph_invalidate_dir_request(req);
|
2009-10-06 11:31:08 -07:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (rinfo->head->is_dentry) {
|
2017-01-31 08:28:26 -07:00
|
|
|
struct inode *dir = req->r_parent;
|
2010-01-25 12:33:08 -07:00
|
|
|
|
2012-08-21 15:55:25 -07:00
|
|
|
if (dir) {
|
2019-12-05 07:09:25 -07:00
|
|
|
err = ceph_fill_inode(dir, NULL, &rinfo->diri,
|
|
|
|
rinfo->dirfrag, session, -1,
|
|
|
|
&req->r_caps_reservation);
|
2012-08-21 15:55:25 -07:00
|
|
|
if (err < 0)
|
2014-03-06 01:40:32 -07:00
|
|
|
goto done;
|
2012-08-21 15:55:25 -07:00
|
|
|
} else {
|
|
|
|
WARN_ON_ONCE(1);
|
|
|
|
}
|
2014-03-06 01:40:32 -07:00
|
|
|
|
2018-09-27 20:34:42 -07:00
|
|
|
if (dir && req->r_op == CEPH_MDS_OP_LOOKUPNAME &&
|
|
|
|
test_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags) &&
|
|
|
|
!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
|
2020-08-10 12:13:29 -07:00
|
|
|
bool is_nokey = false;
|
2014-03-06 01:40:32 -07:00
|
|
|
struct qstr dname;
|
|
|
|
struct dentry *dn, *parent;
|
2020-08-10 12:13:29 -07:00
|
|
|
struct fscrypt_str oname = FSTR_INIT(NULL, 0);
|
|
|
|
struct ceph_fname fname = { .dir = dir,
|
|
|
|
.name = rinfo->dname,
|
|
|
|
.ctext = rinfo->altname,
|
|
|
|
.name_len = rinfo->dname_len,
|
|
|
|
.ctext_len = rinfo->altname_len };
|
2014-03-06 01:40:32 -07:00
|
|
|
|
|
|
|
BUG_ON(!rinfo->head->is_target);
|
|
|
|
BUG_ON(req->r_dentry);
|
|
|
|
|
|
|
|
parent = d_find_any_alias(dir);
|
|
|
|
BUG_ON(!parent);
|
|
|
|
|
2020-08-10 12:13:29 -07:00
|
|
|
err = ceph_fname_alloc_buffer(dir, &oname);
|
|
|
|
if (err < 0) {
|
|
|
|
dput(parent);
|
|
|
|
goto done;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = ceph_fname_to_usr(&fname, NULL, &oname, &is_nokey);
|
|
|
|
if (err < 0) {
|
|
|
|
dput(parent);
|
|
|
|
ceph_fname_free_buffer(dir, &oname);
|
|
|
|
goto done;
|
|
|
|
}
|
|
|
|
dname.name = oname.name;
|
|
|
|
dname.len = oname.len;
|
2016-06-10 07:51:30 -07:00
|
|
|
dname.hash = full_name_hash(parent, dname.name, dname.len);
|
2017-01-27 07:13:57 -07:00
|
|
|
tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
|
|
|
|
tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
|
2014-03-06 01:40:32 -07:00
|
|
|
retry_lookup:
|
|
|
|
dn = d_lookup(parent, &dname);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "d_lookup on parent=%p name=%.*s got %p\n",
|
|
|
|
parent, dname.len, dname.name, dn);
|
2014-03-06 01:40:32 -07:00
|
|
|
|
|
|
|
if (!dn) {
|
|
|
|
dn = d_alloc(parent, &dname);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "d_alloc %p '%.*s' = %p\n", parent,
|
|
|
|
dname.len, dname.name, dn);
|
2017-08-20 11:22:02 -07:00
|
|
|
if (!dn) {
|
2014-03-06 01:40:32 -07:00
|
|
|
dput(parent);
|
2020-08-10 12:13:29 -07:00
|
|
|
ceph_fname_free_buffer(dir, &oname);
|
2014-03-06 01:40:32 -07:00
|
|
|
err = -ENOMEM;
|
|
|
|
goto done;
|
|
|
|
}
|
2020-08-10 12:13:29 -07:00
|
|
|
if (is_nokey) {
|
|
|
|
spin_lock(&dn->d_lock);
|
|
|
|
dn->d_flags |= DCACHE_NOKEY_NAME;
|
|
|
|
spin_unlock(&dn->d_lock);
|
|
|
|
}
|
2016-10-28 19:05:13 -07:00
|
|
|
err = 0;
|
2015-03-17 15:25:59 -07:00
|
|
|
} else if (d_really_is_positive(dn) &&
|
2017-01-27 07:13:57 -07:00
|
|
|
(ceph_ino(d_inode(dn)) != tvino.ino ||
|
|
|
|
ceph_snap(d_inode(dn)) != tvino.snap)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " dn %p points to wrong inode %p\n",
|
|
|
|
dn, d_inode(dn));
|
2017-10-10 02:06:25 -07:00
|
|
|
ceph_dir_clear_ordered(dir);
|
2014-03-06 01:40:32 -07:00
|
|
|
d_delete(dn);
|
|
|
|
dput(dn);
|
|
|
|
goto retry_lookup;
|
|
|
|
}
|
2020-08-10 12:13:29 -07:00
|
|
|
ceph_fname_free_buffer(dir, &oname);
|
2014-03-06 01:40:32 -07:00
|
|
|
|
|
|
|
req->r_dentry = dn;
|
|
|
|
dput(parent);
|
|
|
|
}
|
2010-01-25 12:33:08 -07:00
|
|
|
}
|
|
|
|
|
2013-12-04 21:38:59 -07:00
|
|
|
if (rinfo->head->is_target) {
|
2020-11-12 08:03:38 -07:00
|
|
|
/* Should be filled in by handle_reply */
|
|
|
|
BUG_ON(!req->r_target_inode);
|
2013-12-04 21:38:59 -07:00
|
|
|
|
2020-11-12 08:03:38 -07:00
|
|
|
in = req->r_target_inode;
|
2019-12-05 07:09:25 -07:00
|
|
|
err = ceph_fill_inode(in, req->r_locked_page, &rinfo->targeti,
|
|
|
|
NULL, session,
|
2017-02-01 11:49:09 -07:00
|
|
|
(!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags) &&
|
2019-12-02 11:47:57 -07:00
|
|
|
!test_bit(CEPH_MDS_R_ASYNC, &req->r_req_flags) &&
|
2019-12-04 13:28:17 -07:00
|
|
|
rinfo->head->result == 0) ? req->r_fmode : -1,
|
2013-12-04 21:38:59 -07:00
|
|
|
&req->r_caps_reservation);
|
|
|
|
if (err < 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "badness %p %llx.%llx\n", in,
|
|
|
|
ceph_vinop(in));
|
2020-11-12 08:03:38 -07:00
|
|
|
req->r_target_inode = NULL;
|
2019-12-11 13:21:24 -07:00
|
|
|
if (in->i_state & I_NEW)
|
|
|
|
discard_new_inode(in);
|
2020-11-12 07:37:59 -07:00
|
|
|
else
|
|
|
|
iput(in);
|
2013-12-04 21:38:59 -07:00
|
|
|
goto done;
|
|
|
|
}
|
2019-12-11 13:21:24 -07:00
|
|
|
if (in->i_state & I_NEW)
|
|
|
|
unlock_new_inode(in);
|
2013-12-04 21:38:59 -07:00
|
|
|
}
|
|
|
|
|
2010-03-30 13:54:41 -07:00
|
|
|
/*
|
|
|
|
* ignore null lease/binding on snapdir ENOENT, or else we
|
|
|
|
* will have trouble splicing in the virtual snapdir later
|
|
|
|
*/
|
2017-01-31 08:28:26 -07:00
|
|
|
if (rinfo->head->is_dentry &&
|
|
|
|
!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags) &&
|
|
|
|
test_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags) &&
|
2010-03-30 13:54:41 -07:00
|
|
|
(rinfo->head->is_target || strncmp(req->r_dentry->d_name.name,
|
2010-04-06 15:14:15 -07:00
|
|
|
fsc->mount_options->snapdir_name,
|
2010-03-30 13:54:41 -07:00
|
|
|
req->r_dentry->d_name.len))) {
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* lookup link rename : null -> possibly existing inode
|
|
|
|
* mknod symlink mkdir : null -> new inode
|
|
|
|
* unlink : linked -> null
|
|
|
|
*/
|
2017-01-31 08:28:26 -07:00
|
|
|
struct inode *dir = req->r_parent;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct dentry *dn = req->r_dentry;
|
|
|
|
bool have_dir_cap, have_lease;
|
|
|
|
|
|
|
|
BUG_ON(!dn);
|
|
|
|
BUG_ON(!dir);
|
2015-03-17 15:25:59 -07:00
|
|
|
BUG_ON(d_inode(dn->d_parent) != dir);
|
2017-01-27 07:13:57 -07:00
|
|
|
|
|
|
|
dvino.ino = le64_to_cpu(rinfo->diri.in->ino);
|
|
|
|
dvino.snap = le64_to_cpu(rinfo->diri.in->snapid);
|
|
|
|
|
|
|
|
BUG_ON(ceph_ino(dir) != dvino.ino);
|
|
|
|
BUG_ON(ceph_snap(dir) != dvino.snap);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/* do we have a lease on the whole dir? */
|
|
|
|
have_dir_cap =
|
|
|
|
(le32_to_cpu(rinfo->diri.in->cap.caps) &
|
|
|
|
CEPH_CAP_FILE_SHARED);
|
|
|
|
|
|
|
|
/* do we have a dn lease? */
|
|
|
|
have_lease = have_dir_cap ||
|
2011-07-26 11:28:25 -07:00
|
|
|
le32_to_cpu(rinfo->dlease->duration_ms);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (!have_lease)
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "no dentry lease or dir cap\n");
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/* rename? */
|
|
|
|
if (req->r_old_dentry && req->r_op == CEPH_MDS_OP_RENAME) {
|
2014-04-13 22:13:02 -07:00
|
|
|
struct inode *olddir = req->r_old_dentry_dir;
|
|
|
|
BUG_ON(!olddir);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " src %p '%pd' dst %p '%pd'\n",
|
|
|
|
req->r_old_dentry, req->r_old_dentry, dn, dn);
|
|
|
|
doutc(cl, "doing d_move %p -> %p\n", req->r_old_dentry, dn);
|
2010-04-16 12:56:11 -07:00
|
|
|
|
2015-06-16 05:48:56 -07:00
|
|
|
/* d_move screws up sibling dentries' offsets */
|
|
|
|
ceph_dir_clear_ordered(dir);
|
|
|
|
ceph_dir_clear_ordered(olddir);
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
d_move(req->r_old_dentry, dn);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " src %p '%pd' dst %p '%pd'\n",
|
|
|
|
req->r_old_dentry, req->r_old_dentry, dn, dn);
|
2010-05-14 09:35:38 -07:00
|
|
|
|
2009-12-21 12:42:18 -07:00
|
|
|
/* ensure target dentry is invalidated, despite
|
|
|
|
rehashing bug in vfs_rename_dir */
|
2010-05-14 09:35:38 -07:00
|
|
|
ceph_invalidate_dentry_lease(dn);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "dn %p gets new offset %lld\n",
|
|
|
|
req->r_old_dentry,
|
|
|
|
ceph_dentry(req->r_old_dentry)->offset);
|
2010-05-14 09:35:38 -07:00
|
|
|
|
2018-10-25 02:30:30 -07:00
|
|
|
/* swap r_dentry and r_old_dentry in case that
|
|
|
|
* splice_dentry() gets called later. This is safe
|
|
|
|
* because no other place will use them */
|
|
|
|
req->r_dentry = req->r_old_dentry;
|
|
|
|
req->r_old_dentry = dn;
|
|
|
|
dn = req->r_dentry;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/* null dentry? */
|
|
|
|
if (!rinfo->head->is_target) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "null dentry\n");
|
2015-03-17 15:25:59 -07:00
|
|
|
if (d_really_is_positive(dn)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "d_delete %p\n", dn);
|
2017-11-26 20:23:48 -07:00
|
|
|
ceph_dir_clear_ordered(dir);
|
2009-10-06 11:31:08 -07:00
|
|
|
d_delete(dn);
|
2017-01-26 14:14:18 -07:00
|
|
|
} else if (have_lease) {
|
|
|
|
if (d_unhashed(dn))
|
2016-03-07 13:11:43 -07:00
|
|
|
d_add(dn, NULL);
|
2022-03-29 21:21:12 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!d_unhashed(dn) && have_lease)
|
2019-05-22 06:49:44 -07:00
|
|
|
update_dentry_lease(dir, dn,
|
|
|
|
rinfo->dlease, session,
|
|
|
|
req->r_request_started);
|
2009-10-06 11:31:08 -07:00
|
|
|
goto done;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* attach proper inode */
|
2015-03-17 15:25:59 -07:00
|
|
|
if (d_really_is_negative(dn)) {
|
2014-10-21 18:09:56 -07:00
|
|
|
ceph_dir_clear_ordered(dir);
|
2013-12-04 21:38:59 -07:00
|
|
|
ihold(in);
|
2018-10-25 02:30:30 -07:00
|
|
|
err = splice_dentry(&req->r_dentry, in);
|
|
|
|
if (err < 0)
|
2009-10-06 11:31:08 -07:00
|
|
|
goto done;
|
2018-10-25 02:30:30 -07:00
|
|
|
dn = req->r_dentry; /* may have spliced */
|
2015-03-17 15:25:59 -07:00
|
|
|
} else if (d_really_is_positive(dn) && d_inode(dn) != in) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " %p links to %p %llx.%llx, not %llx.%llx\n",
|
|
|
|
dn, d_inode(dn), ceph_vinop(d_inode(dn)),
|
|
|
|
ceph_vinop(in));
|
2016-03-16 23:41:59 -07:00
|
|
|
d_invalidate(dn);
|
2009-10-06 11:31:08 -07:00
|
|
|
have_lease = false;
|
|
|
|
}
|
|
|
|
|
2017-01-27 07:13:57 -07:00
|
|
|
if (have_lease) {
|
2019-05-22 06:49:44 -07:00
|
|
|
update_dentry_lease(dir, dn,
|
|
|
|
rinfo->dlease, session,
|
|
|
|
req->r_request_started);
|
2017-01-27 07:13:57 -07:00
|
|
|
}
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " final dn %p\n", dn);
|
2017-02-01 11:49:09 -07:00
|
|
|
} else if ((req->r_op == CEPH_MDS_OP_LOOKUPSNAP ||
|
|
|
|
req->r_op == CEPH_MDS_OP_MKSNAP) &&
|
2019-10-29 06:53:29 -07:00
|
|
|
test_bit(CEPH_MDS_R_PARENT_LOCKED, &req->r_req_flags) &&
|
2017-02-01 11:49:09 -07:00
|
|
|
!test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags)) {
|
2017-01-31 08:28:26 -07:00
|
|
|
struct inode *dir = req->r_parent;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/* fill out a snapdir LOOKUPSNAP dentry */
|
2014-04-13 22:13:02 -07:00
|
|
|
BUG_ON(!dir);
|
|
|
|
BUG_ON(ceph_snap(dir) != CEPH_SNAPDIR);
|
2018-10-25 02:30:30 -07:00
|
|
|
BUG_ON(!req->r_dentry);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " linking snapped dir %p to dn %p\n", in,
|
|
|
|
req->r_dentry);
|
2014-10-21 18:09:56 -07:00
|
|
|
ceph_dir_clear_ordered(dir);
|
2013-12-04 21:38:59 -07:00
|
|
|
ihold(in);
|
2018-10-25 02:30:30 -07:00
|
|
|
err = splice_dentry(&req->r_dentry, in);
|
|
|
|
if (err < 0)
|
2009-10-06 11:31:08 -07:00
|
|
|
goto done;
|
2019-05-22 06:49:44 -07:00
|
|
|
} else if (rinfo->head->is_dentry && req->r_dentry) {
|
|
|
|
/* parent inode is not locked, be carefull */
|
2017-01-27 11:07:10 -07:00
|
|
|
struct ceph_vino *ptvino = NULL;
|
2019-05-22 06:49:44 -07:00
|
|
|
dvino.ino = le64_to_cpu(rinfo->diri.in->ino);
|
|
|
|
dvino.snap = le64_to_cpu(rinfo->diri.in->snapid);
|
|
|
|
if (rinfo->head->is_target) {
|
|
|
|
tvino.ino = le64_to_cpu(rinfo->targeti.in->ino);
|
|
|
|
tvino.snap = le64_to_cpu(rinfo->targeti.in->snapid);
|
|
|
|
ptvino = &tvino;
|
2017-01-27 11:07:10 -07:00
|
|
|
}
|
2019-05-22 06:49:44 -07:00
|
|
|
update_dentry_lease_careful(req->r_dentry, rinfo->dlease,
|
|
|
|
session, req->r_request_started,
|
|
|
|
rinfo->dname, rinfo->dname_len,
|
|
|
|
&dvino, ptvino);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
done:
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "done err=%d\n", err);
|
2009-10-06 11:31:08 -07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Prepopulate our cache with readdir results, leases, etc.
|
|
|
|
*/
|
2013-01-29 00:55:31 -07:00
|
|
|
static int readdir_prepopulate_inodes_only(struct ceph_mds_request *req,
|
|
|
|
struct ceph_mds_session *session)
|
|
|
|
{
|
|
|
|
struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = session->s_mdsc->fsc->client;
|
2013-01-29 00:55:31 -07:00
|
|
|
int i, err = 0;
|
|
|
|
|
|
|
|
for (i = 0; i < rinfo->dir_nr; i++) {
|
2016-04-27 18:37:39 -07:00
|
|
|
struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
|
2013-01-29 00:55:31 -07:00
|
|
|
struct ceph_vino vino;
|
|
|
|
struct inode *in;
|
|
|
|
int rc;
|
|
|
|
|
2016-04-27 18:37:39 -07:00
|
|
|
vino.ino = le64_to_cpu(rde->inode.in->ino);
|
|
|
|
vino.snap = le64_to_cpu(rde->inode.in->snapid);
|
2013-01-29 00:55:31 -07:00
|
|
|
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
in = ceph_get_inode(req->r_dentry->d_sb, vino, NULL);
|
2013-01-29 00:55:31 -07:00
|
|
|
if (IS_ERR(in)) {
|
|
|
|
err = PTR_ERR(in);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "badness got %d\n", err);
|
2013-01-29 00:55:31 -07:00
|
|
|
continue;
|
|
|
|
}
|
2019-12-05 07:09:25 -07:00
|
|
|
rc = ceph_fill_inode(in, NULL, &rde->inode, NULL, session,
|
|
|
|
-1, &req->r_caps_reservation);
|
2013-01-29 00:55:31 -07:00
|
|
|
if (rc < 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "inode badness on %p got %d\n", in,
|
|
|
|
rc);
|
2013-01-29 00:55:31 -07:00
|
|
|
err = rc;
|
2019-12-11 13:21:24 -07:00
|
|
|
if (in->i_state & I_NEW) {
|
|
|
|
ihold(in);
|
|
|
|
discard_new_inode(in);
|
|
|
|
}
|
|
|
|
} else if (in->i_state & I_NEW) {
|
|
|
|
unlock_new_inode(in);
|
2013-01-29 00:55:31 -07:00
|
|
|
}
|
2019-12-11 13:21:24 -07:00
|
|
|
|
2021-06-04 09:03:09 -07:00
|
|
|
iput(in);
|
2013-01-29 00:55:31 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2015-06-16 05:48:56 -07:00
|
|
|
void ceph_readdir_cache_release(struct ceph_readdir_cache_control *ctl)
|
|
|
|
{
|
|
|
|
if (ctl->page) {
|
|
|
|
kunmap(ctl->page);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 05:29:47 -07:00
|
|
|
put_page(ctl->page);
|
2015-06-16 05:48:56 -07:00
|
|
|
ctl->page = NULL;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int fill_readdir_cache(struct inode *dir, struct dentry *dn,
|
|
|
|
struct ceph_readdir_cache_control *ctl,
|
|
|
|
struct ceph_mds_request *req)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(dir);
|
2015-06-16 05:48:56 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(dir);
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 05:29:47 -07:00
|
|
|
unsigned nsize = PAGE_SIZE / sizeof(struct dentry*);
|
2015-06-16 05:48:56 -07:00
|
|
|
unsigned idx = ctl->index % nsize;
|
|
|
|
pgoff_t pgoff = ctl->index / nsize;
|
|
|
|
|
|
|
|
if (!ctl->page || pgoff != page_index(ctl->page)) {
|
|
|
|
ceph_readdir_cache_release(ctl);
|
2016-02-26 01:27:13 -07:00
|
|
|
if (idx == 0)
|
|
|
|
ctl->page = grab_cache_page(&dir->i_data, pgoff);
|
|
|
|
else
|
|
|
|
ctl->page = find_lock_page(&dir->i_data, pgoff);
|
2015-06-16 05:48:56 -07:00
|
|
|
if (!ctl->page) {
|
|
|
|
ctl->index = -1;
|
2016-02-26 01:27:13 -07:00
|
|
|
return idx == 0 ? -ENOMEM : 0;
|
2015-06-16 05:48:56 -07:00
|
|
|
}
|
|
|
|
/* reading/filling the cache are serialized by
|
2022-02-10 20:33:25 -07:00
|
|
|
* i_rwsem, no need to use page lock */
|
2015-06-16 05:48:56 -07:00
|
|
|
unlock_page(ctl->page);
|
|
|
|
ctl->dentries = kmap(ctl->page);
|
2016-02-26 01:27:13 -07:00
|
|
|
if (idx == 0)
|
mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macros
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.
This promise never materialized. And unlikely will.
We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE. And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.
Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.
Let's stop pretending that pages in page cache are special. They are
not.
The changes are pretty straight-forward:
- <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;
- PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};
- page_cache_get() -> get_page();
- page_cache_release() -> put_page();
This patch contains automated changes generated with coccinelle using
script below. For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.
The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.
There are few places in the code where coccinelle didn't reach. I'll
fix them manually in a separate patch. Comments and documentation also
will be addressed with the separate patch.
virtual patch
@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E
@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT
@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE
@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK
@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)
@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)
@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-04-01 05:29:47 -07:00
|
|
|
memset(ctl->dentries, 0, PAGE_SIZE);
|
2015-06-16 05:48:56 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
if (req->r_dir_release_cnt == atomic64_read(&ci->i_release_count) &&
|
|
|
|
req->r_dir_ordered_cnt == atomic64_read(&ci->i_ordered_count)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "dn %p idx %d\n", dn, ctl->index);
|
2015-06-16 05:48:56 -07:00
|
|
|
ctl->dentries[idx] = dn;
|
|
|
|
ctl->index++;
|
|
|
|
} else {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "disable readdir cache\n");
|
2015-06-16 05:48:56 -07:00
|
|
|
ctl->index = -1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
int ceph_readdir_prepopulate(struct ceph_mds_request *req,
|
|
|
|
struct ceph_mds_session *session)
|
|
|
|
{
|
|
|
|
struct dentry *parent = req->r_dentry;
|
2022-03-13 19:28:35 -07:00
|
|
|
struct inode *inode = d_inode(parent);
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_mds_reply_info_parsed *rinfo = &req->r_reply_info;
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = session->s_mdsc->fsc->client;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct qstr dname;
|
|
|
|
struct dentry *dn;
|
|
|
|
struct inode *in;
|
2016-03-06 19:34:50 -07:00
|
|
|
int err = 0, skipped = 0, ret, i;
|
2020-12-09 08:27:01 -07:00
|
|
|
u32 frag = le32_to_cpu(req->r_args.readdir.frag);
|
2016-04-28 20:27:30 -07:00
|
|
|
u32 last_hash = 0;
|
|
|
|
u32 fpos_offset;
|
2015-06-16 05:48:56 -07:00
|
|
|
struct ceph_readdir_cache_control cache_ctl = {};
|
|
|
|
|
2017-02-01 11:49:09 -07:00
|
|
|
if (test_bit(CEPH_MDS_R_ABORTED, &req->r_req_flags))
|
2015-06-16 05:48:56 -07:00
|
|
|
return readdir_prepopulate_inodes_only(req, session);
|
2013-09-17 18:44:13 -07:00
|
|
|
|
2017-04-05 09:54:05 -07:00
|
|
|
if (rinfo->hash_order) {
|
|
|
|
if (req->r_path2) {
|
|
|
|
last_hash = ceph_str_hash(ci->i_dir_layout.dl_dir_hash,
|
|
|
|
req->r_path2,
|
|
|
|
strlen(req->r_path2));
|
|
|
|
last_hash = ceph_frag_value(last_hash);
|
|
|
|
} else if (rinfo->offset_hash) {
|
|
|
|
/* mds understands offset_hash */
|
|
|
|
WARN_ON_ONCE(req->r_readdir_offset != 2);
|
2020-12-09 08:27:01 -07:00
|
|
|
last_hash = le32_to_cpu(req->r_args.readdir.offset_hash);
|
2017-04-05 09:54:05 -07:00
|
|
|
}
|
2016-04-28 20:27:30 -07:00
|
|
|
}
|
|
|
|
|
2013-09-17 18:44:13 -07:00
|
|
|
if (rinfo->dir_dir &&
|
|
|
|
le32_to_cpu(rinfo->dir_dir->frag) != frag) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "got new frag %x -> %x\n", frag,
|
|
|
|
le32_to_cpu(rinfo->dir_dir->frag));
|
2013-09-17 18:44:13 -07:00
|
|
|
frag = le32_to_cpu(rinfo->dir_dir->frag);
|
2016-04-28 20:27:30 -07:00
|
|
|
if (!rinfo->hash_order)
|
2015-06-16 05:48:56 -07:00
|
|
|
req->r_readdir_offset = 2;
|
2013-09-17 18:44:13 -07:00
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (le32_to_cpu(rinfo->head->op) == CEPH_MDS_OP_LSSNAP) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%d items under SNAPDIR dn %p\n",
|
|
|
|
rinfo->dir_nr, parent);
|
2009-10-06 11:31:08 -07:00
|
|
|
} else {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%d items under dn %p\n", rinfo->dir_nr, parent);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (rinfo->dir_dir)
|
2015-03-17 15:25:59 -07:00
|
|
|
ceph_fill_dirfrag(d_inode(parent), rinfo->dir_dir);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2017-07-25 21:07:51 -07:00
|
|
|
if (ceph_frag_is_leftmost(frag) &&
|
|
|
|
req->r_readdir_offset == 2 &&
|
|
|
|
!(rinfo->hash_order && last_hash)) {
|
|
|
|
/* note dir version at start of readdir so we can
|
|
|
|
* tell if any dentries get dropped */
|
|
|
|
req->r_dir_release_cnt =
|
|
|
|
atomic64_read(&ci->i_release_count);
|
|
|
|
req->r_dir_ordered_cnt =
|
|
|
|
atomic64_read(&ci->i_ordered_count);
|
|
|
|
req->r_readdir_cache_idx = 0;
|
|
|
|
}
|
2015-06-16 05:48:56 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
cache_ctl.index = req->r_readdir_cache_idx;
|
2016-04-28 20:27:30 -07:00
|
|
|
fpos_offset = req->r_readdir_offset;
|
2015-06-16 05:48:56 -07:00
|
|
|
|
2013-12-04 21:38:59 -07:00
|
|
|
/* FIXME: release caps/leases if error occurs */
|
2009-10-06 11:31:08 -07:00
|
|
|
for (i = 0; i < rinfo->dir_nr; i++) {
|
2016-04-27 18:37:39 -07:00
|
|
|
struct ceph_mds_reply_dir_entry *rde = rinfo->dir_entries + i;
|
2019-05-22 06:49:44 -07:00
|
|
|
struct ceph_vino tvino;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2016-04-27 18:37:39 -07:00
|
|
|
dname.name = rde->name;
|
|
|
|
dname.len = rde->name_len;
|
2016-06-10 07:51:30 -07:00
|
|
|
dname.hash = full_name_hash(parent, dname.name, dname.len);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2017-01-27 07:13:57 -07:00
|
|
|
tvino.ino = le64_to_cpu(rde->inode.in->ino);
|
|
|
|
tvino.snap = le64_to_cpu(rde->inode.in->snapid);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2016-04-28 20:27:30 -07:00
|
|
|
if (rinfo->hash_order) {
|
2022-03-13 19:28:35 -07:00
|
|
|
u32 hash = ceph_frag_value(rde->raw_hash);
|
2016-04-28 20:27:30 -07:00
|
|
|
if (hash != last_hash)
|
|
|
|
fpos_offset = 2;
|
|
|
|
last_hash = hash;
|
|
|
|
rde->offset = ceph_make_fpos(hash, fpos_offset++, true);
|
|
|
|
} else {
|
|
|
|
rde->offset = ceph_make_fpos(frag, fpos_offset++, false);
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
retry_lookup:
|
|
|
|
dn = d_lookup(parent, &dname);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "d_lookup on parent=%p name=%.*s got %p\n",
|
|
|
|
parent, dname.len, dname.name, dn);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (!dn) {
|
|
|
|
dn = d_alloc(parent, &dname);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "d_alloc %p '%.*s' = %p\n", parent,
|
|
|
|
dname.len, dname.name, dn);
|
2017-08-20 11:22:02 -07:00
|
|
|
if (!dn) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "d_alloc badness\n");
|
2009-10-06 11:31:08 -07:00
|
|
|
err = -ENOMEM;
|
|
|
|
goto out;
|
|
|
|
}
|
2022-03-13 19:28:35 -07:00
|
|
|
if (rde->is_nokey) {
|
|
|
|
spin_lock(&dn->d_lock);
|
|
|
|
dn->d_flags |= DCACHE_NOKEY_NAME;
|
|
|
|
spin_unlock(&dn->d_lock);
|
|
|
|
}
|
2015-03-17 15:25:59 -07:00
|
|
|
} else if (d_really_is_positive(dn) &&
|
2017-01-27 07:13:57 -07:00
|
|
|
(ceph_ino(d_inode(dn)) != tvino.ino ||
|
|
|
|
ceph_snap(d_inode(dn)) != tvino.snap)) {
|
2017-11-26 20:23:48 -07:00
|
|
|
struct ceph_dentry_info *di = ceph_dentry(dn);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " dn %p points to wrong inode %p\n",
|
|
|
|
dn, d_inode(dn));
|
2017-11-26 20:23:48 -07:00
|
|
|
|
|
|
|
spin_lock(&dn->d_lock);
|
|
|
|
if (di->offset > 0 &&
|
|
|
|
di->lease_shared_gen ==
|
|
|
|
atomic_read(&ci->i_shared_gen)) {
|
|
|
|
__ceph_dir_clear_ordered(ci);
|
|
|
|
di->offset = 0;
|
|
|
|
}
|
|
|
|
spin_unlock(&dn->d_lock);
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
d_delete(dn);
|
|
|
|
dput(dn);
|
|
|
|
goto retry_lookup;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* inode */
|
2015-03-17 15:25:59 -07:00
|
|
|
if (d_really_is_positive(dn)) {
|
|
|
|
in = d_inode(dn);
|
2009-10-06 11:31:08 -07:00
|
|
|
} else {
|
ceph: preallocate inode for ops that may create one
When creating a new inode, we need to determine the crypto context
before we can transmit the RPC. The fscrypt API has a routine for getting
a crypto context before a create occurs, but it requires an inode.
Change the ceph code to preallocate an inode in advance of a create of
any sort (open(), mknod(), symlink(), etc). Move the existing code that
generates the ACL and SELinux blobs into this routine since that's
mostly common across all the different codepaths.
In most cases, we just want to allow ceph_fill_trace to use that inode
after the reply comes in, so add a new field to the MDS request for it
(r_new_inode).
The async create codepath is a bit different though. In that case, we
want to hash the inode in advance of the RPC so that it can be used
before the reply comes in. If the call subsequently fails with
-EJUKEBOX, then just put the references and clean up the as_ctx. Note
that with this change, we now need to regenerate the as_ctx when this
occurs, but it's quite rare for it to happen.
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-and-tested-by: Luís Henriques <lhenriques@suse.de>
Reviewed-by: Milind Changire <mchangir@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-26 10:11:00 -07:00
|
|
|
in = ceph_get_inode(parent->d_sb, tvino, NULL);
|
2010-08-25 00:11:35 -07:00
|
|
|
if (IS_ERR(in)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "new_inode badness\n");
|
2012-09-26 18:41:05 -07:00
|
|
|
d_drop(dn);
|
2009-10-06 11:31:08 -07:00
|
|
|
dput(dn);
|
2010-08-25 00:11:35 -07:00
|
|
|
err = PTR_ERR(in);
|
2009-10-06 11:31:08 -07:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2019-12-05 07:09:25 -07:00
|
|
|
ret = ceph_fill_inode(in, NULL, &rde->inode, NULL, session,
|
|
|
|
-1, &req->r_caps_reservation);
|
2015-06-16 05:48:56 -07:00
|
|
|
if (ret < 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "badness on %p %llx.%llx\n", in,
|
|
|
|
ceph_vinop(in));
|
2019-05-18 05:39:55 -07:00
|
|
|
if (d_really_is_negative(dn)) {
|
2019-12-11 13:21:24 -07:00
|
|
|
if (in->i_state & I_NEW) {
|
|
|
|
ihold(in);
|
|
|
|
discard_new_inode(in);
|
|
|
|
}
|
2021-06-04 09:03:09 -07:00
|
|
|
iput(in);
|
2019-05-18 05:39:55 -07:00
|
|
|
}
|
2013-12-04 21:38:59 -07:00
|
|
|
d_drop(dn);
|
2015-06-16 05:48:56 -07:00
|
|
|
err = ret;
|
2010-06-21 10:38:14 -07:00
|
|
|
goto next_item;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
2019-12-11 13:21:24 -07:00
|
|
|
if (in->i_state & I_NEW)
|
|
|
|
unlock_new_inode(in);
|
2013-12-04 21:38:59 -07:00
|
|
|
|
2015-03-17 15:25:59 -07:00
|
|
|
if (d_really_is_negative(dn)) {
|
2016-03-06 19:34:50 -07:00
|
|
|
if (ceph_security_xattr_deadlock(in)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, " skip splicing dn %p to inode %p"
|
|
|
|
" (security xattr deadlock)\n", dn, in);
|
2021-06-04 09:03:09 -07:00
|
|
|
iput(in);
|
2016-03-06 19:34:50 -07:00
|
|
|
skipped++;
|
|
|
|
goto next_item;
|
|
|
|
}
|
|
|
|
|
2018-10-25 02:30:30 -07:00
|
|
|
err = splice_dentry(&dn, in);
|
|
|
|
if (err < 0)
|
2013-12-04 21:38:59 -07:00
|
|
|
goto next_item;
|
|
|
|
}
|
|
|
|
|
2016-04-28 20:27:30 -07:00
|
|
|
ceph_dentry(dn)->offset = rde->offset;
|
2013-12-04 21:38:59 -07:00
|
|
|
|
2019-05-22 06:49:44 -07:00
|
|
|
update_dentry_lease(d_inode(parent), dn,
|
|
|
|
rde->lease, req->r_session,
|
|
|
|
req->r_request_started);
|
2015-06-16 05:48:56 -07:00
|
|
|
|
2016-03-06 19:34:50 -07:00
|
|
|
if (err == 0 && skipped == 0 && cache_ctl.index >= 0) {
|
2015-06-16 05:48:56 -07:00
|
|
|
ret = fill_readdir_cache(d_inode(parent), dn,
|
|
|
|
&cache_ctl, req);
|
|
|
|
if (ret < 0)
|
|
|
|
err = ret;
|
|
|
|
}
|
2010-06-21 10:38:14 -07:00
|
|
|
next_item:
|
2018-10-25 02:30:30 -07:00
|
|
|
dput(dn);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
out:
|
2016-03-06 19:34:50 -07:00
|
|
|
if (err == 0 && skipped == 0) {
|
2017-02-01 11:49:09 -07:00
|
|
|
set_bit(CEPH_MDS_R_DID_PREPOPULATE, &req->r_req_flags);
|
2015-06-16 05:48:56 -07:00
|
|
|
req->r_readdir_cache_idx = cache_ctl.index;
|
|
|
|
}
|
|
|
|
ceph_readdir_cache_release(&cache_ctl);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "done\n");
|
2009-10-06 11:31:08 -07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2017-05-21 21:03:32 -07:00
|
|
|
bool ceph_inode_set_size(struct inode *inode, loff_t size)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2017-05-21 21:03:32 -07:00
|
|
|
bool ret;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "set_size %p %llu -> %llu\n", inode, i_size_read(inode), size);
|
2015-12-29 20:32:46 -07:00
|
|
|
i_size_write(inode, size);
|
2021-12-07 06:44:50 -07:00
|
|
|
ceph_fscache_update(inode);
|
2016-05-05 01:40:17 -07:00
|
|
|
inode->i_blocks = calc_inode_blocks(size);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2017-05-21 21:03:32 -07:00
|
|
|
ret = __ceph_should_report_size(ci);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2021-12-07 06:44:50 -07:00
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2020-10-09 11:24:34 -07:00
|
|
|
void ceph_queue_inode_work(struct inode *inode, int work_bit)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 19:50:38 -07:00
|
|
|
struct ceph_fs_client *fsc = ceph_inode_to_fs_client(inode);
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = fsc->client;
|
2019-05-17 20:18:44 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2020-10-09 11:24:34 -07:00
|
|
|
set_bit(work_bit, &ci->i_work_mask);
|
2019-05-17 20:18:44 -07:00
|
|
|
|
|
|
|
ihold(inode);
|
2020-10-09 11:24:34 -07:00
|
|
|
if (queue_work(fsc->inode_wq, &ci->i_work)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx mask=%lx\n", inode,
|
|
|
|
ceph_vinop(inode), ci->i_work_mask);
|
2019-05-17 20:18:44 -07:00
|
|
|
} else {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx already queued, mask=%lx\n",
|
|
|
|
inode, ceph_vinop(inode), ci->i_work_mask);
|
2019-05-17 20:18:44 -07:00
|
|
|
iput(inode);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ceph_do_invalidate_pages(struct inode *inode)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2019-05-17 20:18:44 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
u32 orig_gen;
|
|
|
|
int check = 0;
|
|
|
|
|
2021-12-07 06:44:50 -07:00
|
|
|
ceph_fscache_invalidate(inode, false);
|
|
|
|
|
2013-08-12 21:42:15 -07:00
|
|
|
mutex_lock(&ci->i_truncate_mutex);
|
2016-04-14 22:56:12 -07:00
|
|
|
|
2021-08-31 10:39:13 -07:00
|
|
|
if (ceph_inode_is_shutdown(inode)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_warn_ratelimited_client(cl,
|
|
|
|
"%p %llx.%llx is shut down\n", inode,
|
|
|
|
ceph_vinop(inode));
|
2016-04-14 22:56:12 -07:00
|
|
|
mapping_set_error(inode->i_mapping, -EIO);
|
|
|
|
truncate_pagecache(inode, 0);
|
|
|
|
mutex_unlock(&ci->i_truncate_mutex);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx gen %d revoking %d\n", inode,
|
|
|
|
ceph_vinop(inode), ci->i_rdcache_gen, ci->i_rdcache_revoking);
|
2010-11-04 11:05:05 -07:00
|
|
|
if (ci->i_rdcache_revoking != ci->i_rdcache_gen) {
|
2013-11-21 22:50:45 -07:00
|
|
|
if (__ceph_caps_revoking_other(ci, NULL, CEPH_CAP_FILE_CACHE))
|
|
|
|
check = 1;
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2013-08-12 21:42:15 -07:00
|
|
|
mutex_unlock(&ci->i_truncate_mutex);
|
2009-10-06 11:31:08 -07:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
orig_gen = ci->i_rdcache_gen;
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2016-05-18 05:58:26 -07:00
|
|
|
if (invalidate_inode_pages2(inode->i_mapping) < 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
pr_err_client(cl, "invalidate_inode_pages2 %llx.%llx failed\n",
|
|
|
|
ceph_vinop(inode));
|
2016-05-18 05:58:26 -07:00
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2010-11-04 11:05:05 -07:00
|
|
|
if (orig_gen == ci->i_rdcache_gen &&
|
|
|
|
orig_gen == ci->i_rdcache_revoking) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx gen %d successful\n", inode,
|
|
|
|
ceph_vinop(inode), ci->i_rdcache_gen);
|
2010-11-04 11:05:05 -07:00
|
|
|
ci->i_rdcache_revoking--;
|
2009-10-06 11:31:08 -07:00
|
|
|
check = 1;
|
|
|
|
} else {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx gen %d raced, now %d revoking %d\n",
|
|
|
|
inode, ceph_vinop(inode), orig_gen, ci->i_rdcache_gen,
|
|
|
|
ci->i_rdcache_revoking);
|
2013-11-21 22:50:45 -07:00
|
|
|
if (__ceph_caps_revoking_other(ci, NULL, CEPH_CAP_FILE_CACHE))
|
|
|
|
check = 1;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2013-08-12 21:42:15 -07:00
|
|
|
mutex_unlock(&ci->i_truncate_mutex);
|
2013-11-21 22:50:45 -07:00
|
|
|
out:
|
2009-10-06 11:31:08 -07:00
|
|
|
if (check)
|
2022-10-17 18:03:29 -07:00
|
|
|
ceph_check_caps(ci, 0);
|
2010-02-09 16:24:44 -07:00
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* Make sure any pending truncation is applied before doing anything
|
|
|
|
* that may depend on it.
|
|
|
|
*/
|
2013-07-01 21:40:19 -07:00
|
|
|
void __ceph_do_pending_vmtruncate(struct inode *inode)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
|
|
|
u64 to;
|
2012-11-18 19:49:08 -07:00
|
|
|
int wrbuffer_refs, finish = 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2013-08-12 21:42:15 -07:00
|
|
|
mutex_lock(&ci->i_truncate_mutex);
|
2009-10-06 11:31:08 -07:00
|
|
|
retry:
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (ci->i_truncate_pending == 0) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx none pending\n", inode,
|
|
|
|
ceph_vinop(inode));
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2013-08-12 21:42:15 -07:00
|
|
|
mutex_unlock(&ci->i_truncate_mutex);
|
2009-10-06 11:31:08 -07:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* make sure any dirty snapped pages are flushed before we
|
|
|
|
* possibly truncate them.. so write AND block!
|
|
|
|
*/
|
|
|
|
if (ci->i_wrbuffer_ref_head < ci->i_wrbuffer_ref) {
|
2017-08-28 00:41:28 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx flushing snaps first\n", inode,
|
|
|
|
ceph_vinop(inode));
|
2009-10-06 11:31:08 -07:00
|
|
|
filemap_write_and_wait_range(&inode->i_data, 0,
|
|
|
|
inode->i_sb->s_maxbytes);
|
|
|
|
goto retry;
|
|
|
|
}
|
|
|
|
|
2013-08-12 21:42:15 -07:00
|
|
|
/* there should be no reader or writer */
|
|
|
|
WARN_ON_ONCE(ci->i_rd_ref || ci->i_wr_ref);
|
|
|
|
|
2022-08-25 06:31:11 -07:00
|
|
|
to = ci->i_truncate_pagecache_size;
|
2009-10-06 11:31:08 -07:00
|
|
|
wrbuffer_refs = ci->i_wrbuffer_ref;
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx (%d) to %lld\n", inode, ceph_vinop(inode),
|
|
|
|
ci->i_truncate_pending, to);
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2021-12-07 06:44:50 -07:00
|
|
|
ceph_fscache_resize(inode, to);
|
2014-06-07 14:08:56 -07:00
|
|
|
truncate_pagecache(inode, to);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2022-08-25 06:31:11 -07:00
|
|
|
if (to == ci->i_truncate_pagecache_size) {
|
2012-11-18 19:49:08 -07:00
|
|
|
ci->i_truncate_pending = 0;
|
|
|
|
finish = 1;
|
|
|
|
}
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2012-11-18 19:49:08 -07:00
|
|
|
if (!finish)
|
|
|
|
goto retry;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2013-08-12 21:42:15 -07:00
|
|
|
mutex_unlock(&ci->i_truncate_mutex);
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
if (wrbuffer_refs == 0)
|
2022-10-17 18:03:29 -07:00
|
|
|
ceph_check_caps(ci, 0);
|
2012-11-18 19:49:08 -07:00
|
|
|
|
|
|
|
wake_up_all(&ci->i_cap_wq);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
2019-05-17 20:18:44 -07:00
|
|
|
static void ceph_inode_work(struct work_struct *work)
|
|
|
|
{
|
|
|
|
struct ceph_inode_info *ci = container_of(work, struct ceph_inode_info,
|
|
|
|
i_work);
|
netfs: Fix gcc-12 warning by embedding vfs inode in netfs_i_context
While randstruct was satisfied with using an open-coded "void *" offset
cast for the netfs_i_context <-> inode casting, __builtin_object_size() as
used by FORTIFY_SOURCE was not as easily fooled. This was causing the
following complaint[1] from gcc v12:
In file included from include/linux/string.h:253,
from include/linux/ceph/ceph_debug.h:7,
from fs/ceph/inode.c:2:
In function 'fortify_memset_chk',
inlined from 'netfs_i_context_init' at include/linux/netfs.h:326:2,
inlined from 'ceph_alloc_inode' at fs/ceph/inode.c:463:2:
include/linux/fortify-string.h:242:25: warning: call to '__write_overflow_field' declared with attribute warning: detected write beyond size of field (1st parameter); maybe use struct_group()? [-Wattribute-warning]
242 | __write_overflow_field(p_size_field, size);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fix this by embedding a struct inode into struct netfs_i_context (which
should perhaps be renamed to struct netfs_inode). The struct inode
vfs_inode fields are then removed from the 9p, afs, ceph and cifs inode
structs and vfs_inode is then simply changed to "netfs.inode" in those
filesystems.
Further, rename netfs_i_context to netfs_inode, get rid of the
netfs_inode() function that converted a netfs_i_context pointer to an
inode pointer (that can now be done with &ctx->inode) and rename the
netfs_i_context() function to netfs_inode() (which is now a wrapper
around container_of()).
Most of the changes were done with:
perl -p -i -e 's/vfs_inode/netfs.inode/'g \
`git grep -l 'vfs_inode' -- fs/{9p,afs,ceph,cifs}/*.[ch]`
Kees suggested doing it with a pair structure[2] and a special
declarator to insert that into the network filesystem's inode
wrapper[3], but I think it's cleaner to embed it - and then it doesn't
matter if struct randomisation reorders things.
Dave Chinner suggested using a filesystem-specific VFS_I() function in
each filesystem to convert that filesystem's own inode wrapper struct
into the VFS inode struct[4].
Version #2:
- Fix a couple of missed name changes due to a disabled cifs option.
- Rename nfs_i_context to nfs_inode
- Use "netfs" instead of "nic" as the member name in per-fs inode wrapper
structs.
[ This also undoes commit 507160f46c55 ("netfs: gcc-12: temporarily
disable '-Wattribute-warning' for now") that is no longer needed ]
Fixes: bc899ee1c898 ("netfs: Add a netfs inode context")
Reported-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
cc: Jonathan Corbet <corbet@lwn.net>
cc: Eric Van Hensbergen <ericvh@gmail.com>
cc: Latchesar Ionkov <lucho@ionkov.net>
cc: Dominique Martinet <asmadeus@codewreck.org>
cc: Christian Schoenebeck <linux_oss@crudebyte.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Ilya Dryomov <idryomov@gmail.com>
cc: Steve French <smfrench@gmail.com>
cc: William Kucharski <william.kucharski@oracle.com>
cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
cc: Dave Chinner <david@fromorbit.com>
cc: linux-doc@vger.kernel.org
cc: v9fs-developer@lists.sourceforge.net
cc: linux-afs@lists.infradead.org
cc: ceph-devel@vger.kernel.org
cc: linux-cifs@vger.kernel.org
cc: samba-technical@lists.samba.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-hardening@vger.kernel.org
Link: https://lore.kernel.org/r/d2ad3a3d7bdd794c6efb562d2f2b655fb67756b9.camel@kernel.org/ [1]
Link: https://lore.kernel.org/r/20220517210230.864239-1-keescook@chromium.org/ [2]
Link: https://lore.kernel.org/r/20220518202212.2322058-1-keescook@chromium.org/ [3]
Link: https://lore.kernel.org/r/20220524101205.GI2306852@dread.disaster.area/ [4]
Link: https://lore.kernel.org/r/165296786831.3591209.12111293034669289733.stgit@warthog.procyon.org.uk/ # v1
Link: https://lore.kernel.org/r/165305805651.4094995.7763502506786714216.stgit@warthog.procyon.org.uk # v2
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2022-06-09 13:46:04 -07:00
|
|
|
struct inode *inode = &ci->netfs.inode;
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2019-05-17 20:18:44 -07:00
|
|
|
|
|
|
|
if (test_and_clear_bit(CEPH_I_WORK_WRITEBACK, &ci->i_work_mask)) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "writeback %p %llx.%llx\n", inode, ceph_vinop(inode));
|
2019-05-17 20:18:44 -07:00
|
|
|
filemap_fdatawrite(&inode->i_data);
|
|
|
|
}
|
|
|
|
if (test_and_clear_bit(CEPH_I_WORK_INVALIDATE_PAGES, &ci->i_work_mask))
|
|
|
|
ceph_do_invalidate_pages(inode);
|
|
|
|
|
|
|
|
if (test_and_clear_bit(CEPH_I_WORK_VMTRUNCATE, &ci->i_work_mask))
|
|
|
|
__ceph_do_pending_vmtruncate(inode);
|
|
|
|
|
2020-12-10 12:39:26 -07:00
|
|
|
if (test_and_clear_bit(CEPH_I_WORK_CHECK_CAPS, &ci->i_work_mask))
|
2022-10-17 18:03:29 -07:00
|
|
|
ceph_check_caps(ci, 0);
|
2020-12-10 12:39:26 -07:00
|
|
|
|
|
|
|
if (test_and_clear_bit(CEPH_I_WORK_FLUSH_SNAPS, &ci->i_work_mask))
|
|
|
|
ceph_flush_snaps(ci, NULL);
|
|
|
|
|
2019-05-17 20:18:44 -07:00
|
|
|
iput(inode);
|
|
|
|
}
|
|
|
|
|
2020-09-03 10:31:10 -07:00
|
|
|
static const char *ceph_encrypted_get_link(struct dentry *dentry,
|
|
|
|
struct inode *inode,
|
|
|
|
struct delayed_call *done)
|
|
|
|
{
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
|
|
|
|
|
|
|
if (!dentry)
|
|
|
|
return ERR_PTR(-ECHILD);
|
|
|
|
|
|
|
|
return fscrypt_get_symlink(inode, ci->i_symlink, i_size_read(inode),
|
|
|
|
done);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ceph_encrypted_symlink_getattr(struct mnt_idmap *idmap,
|
|
|
|
const struct path *path,
|
|
|
|
struct kstat *stat, u32 request_mask,
|
|
|
|
unsigned int query_flags)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = ceph_getattr(idmap, path, stat, request_mask, query_flags);
|
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
return fscrypt_symlink_getattr(path, stat);
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* symlinks
|
|
|
|
*/
|
|
|
|
static const struct inode_operations ceph_symlink_iops = {
|
2015-11-17 08:20:54 -07:00
|
|
|
.get_link = simple_get_link,
|
2013-04-07 01:28:49 -07:00
|
|
|
.setattr = ceph_setattr,
|
|
|
|
.getattr = ceph_getattr,
|
|
|
|
.listxattr = ceph_listxattr,
|
2009-10-06 11:31:08 -07:00
|
|
|
};
|
|
|
|
|
2020-09-03 10:31:10 -07:00
|
|
|
static const struct inode_operations ceph_encrypted_symlink_iops = {
|
|
|
|
.get_link = ceph_encrypted_get_link,
|
|
|
|
.setattr = ceph_setattr,
|
|
|
|
.getattr = ceph_encrypted_symlink_getattr,
|
|
|
|
.listxattr = ceph_listxattr,
|
|
|
|
};
|
|
|
|
|
2022-08-25 06:31:11 -07:00
|
|
|
/*
|
|
|
|
* Transfer the encrypted last block to the MDS and the MDS
|
|
|
|
* will help update it when truncating a smaller size.
|
|
|
|
*
|
|
|
|
* We don't support a PAGE_SIZE that is smaller than the
|
|
|
|
* CEPH_FSCRYPT_BLOCK_SIZE.
|
|
|
|
*/
|
|
|
|
static int fill_fscrypt_truncate(struct inode *inode,
|
|
|
|
struct ceph_mds_request *req,
|
|
|
|
struct iattr *attr)
|
|
|
|
{
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2022-08-25 06:31:11 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
|
|
|
int boff = attr->ia_size % CEPH_FSCRYPT_BLOCK_SIZE;
|
|
|
|
loff_t pos, orig_pos = round_down(attr->ia_size,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE);
|
|
|
|
u64 block = orig_pos >> CEPH_FSCRYPT_BLOCK_SHIFT;
|
|
|
|
struct ceph_pagelist *pagelist = NULL;
|
|
|
|
struct kvec iov = {0};
|
|
|
|
struct iov_iter iter;
|
|
|
|
struct page *page = NULL;
|
|
|
|
struct ceph_fscrypt_truncate_size_header header;
|
|
|
|
int retry_op = 0;
|
|
|
|
int len = CEPH_FSCRYPT_BLOCK_SIZE;
|
|
|
|
loff_t i_size = i_size_read(inode);
|
|
|
|
int got, ret, issued;
|
|
|
|
u64 objver;
|
|
|
|
|
|
|
|
ret = __ceph_get_caps(inode, NULL, CEPH_CAP_FILE_RD, 0, -1, &got);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
issued = __ceph_caps_issued(ci, NULL);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "size %lld -> %lld got cap refs on %s, issued %s\n",
|
|
|
|
i_size, attr->ia_size, ceph_cap_string(got),
|
|
|
|
ceph_cap_string(issued));
|
2022-08-25 06:31:11 -07:00
|
|
|
|
|
|
|
/* Try to writeback the dirty pagecaches */
|
|
|
|
if (issued & (CEPH_CAP_FILE_BUFFER)) {
|
|
|
|
loff_t lend = orig_pos + CEPH_FSCRYPT_BLOCK_SHIFT - 1;
|
|
|
|
|
|
|
|
ret = filemap_write_and_wait_range(inode->i_mapping,
|
|
|
|
orig_pos, lend);
|
|
|
|
if (ret < 0)
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
page = __page_cache_alloc(GFP_KERNEL);
|
|
|
|
if (page == NULL) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
pagelist = ceph_pagelist_alloc(GFP_KERNEL);
|
|
|
|
if (!pagelist) {
|
|
|
|
ret = -ENOMEM;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
iov.iov_base = kmap_local_page(page);
|
|
|
|
iov.iov_len = len;
|
|
|
|
iov_iter_kvec(&iter, READ, &iov, 1, len);
|
|
|
|
|
|
|
|
pos = orig_pos;
|
|
|
|
ret = __ceph_sync_read(inode, &pos, &iter, &retry_op, &objver);
|
|
|
|
if (ret < 0)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
/* Insert the header first */
|
|
|
|
header.ver = 1;
|
|
|
|
header.compat = 1;
|
|
|
|
header.change_attr = cpu_to_le64(inode_peek_iversion_raw(inode));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Always set the block_size to CEPH_FSCRYPT_BLOCK_SIZE,
|
|
|
|
* because in MDS it may need this to do the truncate.
|
|
|
|
*/
|
|
|
|
header.block_size = cpu_to_le32(CEPH_FSCRYPT_BLOCK_SIZE);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we hit a hole here, we should just skip filling
|
|
|
|
* the fscrypt for the request, because once the fscrypt
|
|
|
|
* is enabled, the file will be split into many blocks
|
|
|
|
* with the size of CEPH_FSCRYPT_BLOCK_SIZE, if there
|
|
|
|
* has a hole, the hole size should be multiple of block
|
|
|
|
* size.
|
|
|
|
*
|
|
|
|
* If the Rados object doesn't exist, it will be set to 0.
|
|
|
|
*/
|
|
|
|
if (!objver) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "hit hole, ppos %lld < size %lld\n", pos, i_size);
|
2022-08-25 06:31:11 -07:00
|
|
|
|
|
|
|
header.data_len = cpu_to_le32(8 + 8 + 4);
|
|
|
|
header.file_offset = 0;
|
|
|
|
ret = 0;
|
|
|
|
} else {
|
|
|
|
header.data_len = cpu_to_le32(8 + 8 + 4 + CEPH_FSCRYPT_BLOCK_SIZE);
|
|
|
|
header.file_offset = cpu_to_le64(orig_pos);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "encrypt block boff/bsize %d/%lu\n", boff,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE);
|
2023-03-06 00:01:15 -07:00
|
|
|
|
2022-08-25 06:31:11 -07:00
|
|
|
/* truncate and zero out the extra contents for the last block */
|
|
|
|
memset(iov.iov_base + boff, 0, PAGE_SIZE - boff);
|
|
|
|
|
|
|
|
/* encrypt the last block */
|
|
|
|
ret = ceph_fscrypt_encrypt_block_inplace(inode, page,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE,
|
|
|
|
0, block,
|
|
|
|
GFP_KERNEL);
|
|
|
|
if (ret)
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Insert the header */
|
|
|
|
ret = ceph_pagelist_append(pagelist, &header, sizeof(header));
|
|
|
|
if (ret)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
if (header.block_size) {
|
|
|
|
/* Append the last block contents to pagelist */
|
|
|
|
ret = ceph_pagelist_append(pagelist, iov.iov_base,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE);
|
|
|
|
if (ret)
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
req->r_pagelist = pagelist;
|
|
|
|
out:
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx size dropping cap refs on %s\n", inode,
|
|
|
|
ceph_vinop(inode), ceph_cap_string(got));
|
2022-08-25 06:31:11 -07:00
|
|
|
ceph_put_cap_refs(ci, got);
|
|
|
|
if (iov.iov_base)
|
|
|
|
kunmap_local(iov.iov_base);
|
|
|
|
if (page)
|
|
|
|
__free_pages(page, 0);
|
|
|
|
if (ret && pagelist)
|
|
|
|
ceph_pagelist_release(pagelist);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2023-08-07 06:26:22 -07:00
|
|
|
int __ceph_setattr(struct mnt_idmap *idmap, struct inode *inode,
|
|
|
|
struct iattr *attr, struct ceph_iattr *cia)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2019-07-25 10:03:32 -07:00
|
|
|
unsigned int ia_valid = attr->ia_valid;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_mds_request *req;
|
2023-06-11 19:50:38 -07:00
|
|
|
struct ceph_mds_client *mdsc = ceph_sb_to_fs_client(inode->i_sb)->mdsc;
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = ceph_inode_to_client(inode);
|
2015-06-10 02:26:13 -07:00
|
|
|
struct ceph_cap_flush *prealloc_cf;
|
2022-08-25 06:31:11 -07:00
|
|
|
loff_t isize = i_size_read(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
int issued;
|
|
|
|
int release = 0, dirtied = 0;
|
|
|
|
int mask = 0;
|
|
|
|
int err = 0;
|
2011-05-04 11:33:47 -07:00
|
|
|
int inode_dirty_flags = 0;
|
2015-05-01 02:49:16 -07:00
|
|
|
bool lock_snap_rwsem = false;
|
2022-08-25 06:31:11 -07:00
|
|
|
bool fill_fscrypt;
|
|
|
|
int truncate_retry = 20; /* The RMW will take around 50ms */
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2022-08-25 06:31:11 -07:00
|
|
|
retry:
|
2015-06-10 02:26:13 -07:00
|
|
|
prealloc_cf = ceph_alloc_cap_flush();
|
|
|
|
if (!prealloc_cf)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_SETATTR,
|
|
|
|
USE_AUTH_MDS);
|
2015-06-10 02:26:13 -07:00
|
|
|
if (IS_ERR(req)) {
|
|
|
|
ceph_free_cap_flush(prealloc_cf);
|
2009-10-06 11:31:08 -07:00
|
|
|
return PTR_ERR(req);
|
2015-06-10 02:26:13 -07:00
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2022-08-25 06:31:11 -07:00
|
|
|
fill_fscrypt = false;
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_lock(&ci->i_ceph_lock);
|
2009-10-06 11:31:08 -07:00
|
|
|
issued = __ceph_caps_issued(ci, NULL);
|
2015-05-01 02:49:16 -07:00
|
|
|
|
|
|
|
if (!ci->i_head_snapc &&
|
|
|
|
(issued & (CEPH_CAP_ANY_EXCL | CEPH_CAP_FILE_WR))) {
|
|
|
|
lock_snap_rwsem = true;
|
|
|
|
if (!down_read_trylock(&mdsc->snap_rwsem)) {
|
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
|
|
|
down_read(&mdsc->snap_rwsem);
|
|
|
|
spin_lock(&ci->i_ceph_lock);
|
|
|
|
issued = __ceph_caps_issued(ci, NULL);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx issued %s\n", inode, ceph_vinop(inode),
|
|
|
|
ceph_cap_string(issued));
|
2020-07-27 07:16:09 -07:00
|
|
|
#if IS_ENABLED(CONFIG_FS_ENCRYPTION)
|
|
|
|
if (cia && cia->fscrypt_auth) {
|
|
|
|
u32 len = ceph_fscrypt_auth_len(cia->fscrypt_auth);
|
|
|
|
|
|
|
|
if (len > sizeof(*cia->fscrypt_auth)) {
|
|
|
|
err = -EINVAL;
|
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx fscrypt_auth len %u to %u)\n", inode,
|
|
|
|
ceph_vinop(inode), ci->fscrypt_auth_len, len);
|
2020-07-27 07:16:09 -07:00
|
|
|
|
|
|
|
/* It should never be re-set once set */
|
|
|
|
WARN_ON_ONCE(ci->fscrypt_auth);
|
|
|
|
|
|
|
|
if (issued & CEPH_CAP_AUTH_EXCL) {
|
|
|
|
dirtied |= CEPH_CAP_AUTH_EXCL;
|
|
|
|
kfree(ci->fscrypt_auth);
|
|
|
|
ci->fscrypt_auth = (u8 *)cia->fscrypt_auth;
|
|
|
|
ci->fscrypt_auth_len = len;
|
|
|
|
} else if ((issued & CEPH_CAP_AUTH_SHARED) == 0 ||
|
|
|
|
ci->fscrypt_auth_len != len ||
|
|
|
|
memcmp(ci->fscrypt_auth, cia->fscrypt_auth, len)) {
|
|
|
|
req->r_fscrypt_auth = cia->fscrypt_auth;
|
|
|
|
mask |= CEPH_SETATTR_FSCRYPT_AUTH;
|
|
|
|
release |= CEPH_CAP_AUTH_SHARED;
|
|
|
|
}
|
|
|
|
cia->fscrypt_auth = NULL;
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
if (cia && cia->fscrypt_auth) {
|
|
|
|
err = -EINVAL;
|
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_FS_ENCRYPTION */
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (ia_valid & ATTR_UID) {
|
2023-08-07 06:26:23 -07:00
|
|
|
kuid_t fsuid = from_vfsuid(idmap, i_user_ns(inode), attr->ia_vfsuid);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx uid %d -> %d\n", inode,
|
|
|
|
ceph_vinop(inode),
|
|
|
|
from_kuid(&init_user_ns, inode->i_uid),
|
|
|
|
from_kuid(&init_user_ns, attr->ia_uid));
|
2009-10-06 11:31:08 -07:00
|
|
|
if (issued & CEPH_CAP_AUTH_EXCL) {
|
2023-08-07 06:26:23 -07:00
|
|
|
inode->i_uid = fsuid;
|
2009-10-06 11:31:08 -07:00
|
|
|
dirtied |= CEPH_CAP_AUTH_EXCL;
|
|
|
|
} else if ((issued & CEPH_CAP_AUTH_SHARED) == 0 ||
|
2023-08-07 06:26:23 -07:00
|
|
|
!uid_eq(fsuid, inode->i_uid)) {
|
2013-01-31 04:40:12 -07:00
|
|
|
req->r_args.setattr.uid = cpu_to_le32(
|
2023-08-07 06:26:23 -07:00
|
|
|
from_kuid(&init_user_ns, fsuid));
|
2009-10-06 11:31:08 -07:00
|
|
|
mask |= CEPH_SETATTR_UID;
|
|
|
|
release |= CEPH_CAP_AUTH_SHARED;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (ia_valid & ATTR_GID) {
|
2023-08-07 06:26:23 -07:00
|
|
|
kgid_t fsgid = from_vfsgid(idmap, i_user_ns(inode), attr->ia_vfsgid);
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx gid %d -> %d\n", inode,
|
|
|
|
ceph_vinop(inode),
|
|
|
|
from_kgid(&init_user_ns, inode->i_gid),
|
|
|
|
from_kgid(&init_user_ns, attr->ia_gid));
|
2009-10-06 11:31:08 -07:00
|
|
|
if (issued & CEPH_CAP_AUTH_EXCL) {
|
2023-08-07 06:26:23 -07:00
|
|
|
inode->i_gid = fsgid;
|
2009-10-06 11:31:08 -07:00
|
|
|
dirtied |= CEPH_CAP_AUTH_EXCL;
|
|
|
|
} else if ((issued & CEPH_CAP_AUTH_SHARED) == 0 ||
|
2023-08-07 06:26:23 -07:00
|
|
|
!gid_eq(fsgid, inode->i_gid)) {
|
2013-01-31 04:40:12 -07:00
|
|
|
req->r_args.setattr.gid = cpu_to_le32(
|
2023-08-07 06:26:23 -07:00
|
|
|
from_kgid(&init_user_ns, fsgid));
|
2009-10-06 11:31:08 -07:00
|
|
|
mask |= CEPH_SETATTR_GID;
|
|
|
|
release |= CEPH_CAP_AUTH_SHARED;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (ia_valid & ATTR_MODE) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx mode 0%o -> 0%o\n", inode,
|
|
|
|
ceph_vinop(inode), inode->i_mode, attr->ia_mode);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (issued & CEPH_CAP_AUTH_EXCL) {
|
|
|
|
inode->i_mode = attr->ia_mode;
|
|
|
|
dirtied |= CEPH_CAP_AUTH_EXCL;
|
|
|
|
} else if ((issued & CEPH_CAP_AUTH_SHARED) == 0 ||
|
|
|
|
attr->ia_mode != inode->i_mode) {
|
2013-11-11 00:18:03 -07:00
|
|
|
inode->i_mode = attr->ia_mode;
|
2009-10-06 11:31:08 -07:00
|
|
|
req->r_args.setattr.mode = cpu_to_le32(attr->ia_mode);
|
|
|
|
mask |= CEPH_SETATTR_MODE;
|
|
|
|
release |= CEPH_CAP_AUTH_SHARED;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ia_valid & ATTR_ATIME) {
|
2023-10-04 11:52:09 -07:00
|
|
|
struct timespec64 atime = inode_get_atime(inode);
|
|
|
|
|
Two items:
- support for idmapped mounts in CephFS (Christian Brauner, Alexander
Mikhalitsyn). The series was originally developed by Christian and
later picked up and brought over the finish line by Alexander, who
also contributed an enabler on the MDS side (separate owner_{u,g}id
fields on the wire). The required exports for mnt_idmap_{get,put}()
in VFS have been acked by Christian and received no objection from
Christoph.
- a churny change in CephFS logging to include cluster and client
identifiers in log and debug messages (Xiubo Li). This would help
in scenarios with dozens of CephFS mounts on the same node which are
getting increasingly common, especially in the Kubernetes world.
-----BEGIN PGP SIGNATURE-----
iQFHBAABCAAxFiEEydHwtzie9C7TfviiSn/eOAIR84sFAmVNDyMTHGlkcnlvbW92
QGdtYWlsLmNvbQAKCRBKf944AhHziy6xCACmUikhgo+pZDO7oQS7HChdZFz5Q2Jb
5K9K7sJr7Zb2gFKLAV93cyFrz0JRljmgZuA3DmDTH1omrkrVAJfHJ6Md1UFG3W7o
LaowP3kECsykOiz+YtVU2957sfoFqds/q6KCXdp1Pc8WfNnL1vysCim6EGYtCqUm
c6vv0zkvEDQp+3kjTN01LHzHFZdHg8+STNM4BMiB0sO/NqbADnaQBEtDpYrgzwh7
YPwVIKcXutUmAKlb7vjUF9ptSICYkGV7B+loPS4BDva1I6dT42xqLQOu89tTKU0D
DiQAfE2oRgU+fl+mMooFJNSqD25Q6IkXrPs0HuiSHS4wtLGqYZAwwfHB
=F9o1
-----END PGP SIGNATURE-----
Merge tag 'ceph-for-6.7-rc1' of https://github.com/ceph/ceph-client
Pull ceph updates from Ilya Dryomov:
- support for idmapped mounts in CephFS (Christian Brauner, Alexander
Mikhalitsyn).
The series was originally developed by Christian and later picked up
and brought over the finish line by Alexander, who also contributed
an enabler on the MDS side (separate owner_{u,g}id fields on the
wire).
The required exports for mnt_idmap_{get,put}() in VFS have been acked
by Christian and received no objection from Christoph.
- a churny change in CephFS logging to include cluster and client
identifiers in log and debug messages (Xiubo Li).
This would help in scenarios with dozens of CephFS mounts on the same
node which are getting increasingly common, especially in the
Kubernetes world.
* tag 'ceph-for-6.7-rc1' of https://github.com/ceph/ceph-client:
ceph: allow idmapped mounts
ceph: allow idmapped atomic_open inode op
ceph: allow idmapped set_acl inode op
ceph: allow idmapped setattr inode op
ceph: pass idmap to __ceph_setattr
ceph: allow idmapped permission inode op
ceph: allow idmapped getattr inode op
ceph: pass an idmapping to mknod/symlink/mkdir
ceph: add enable_unsafe_idmap module parameter
ceph: handle idmapped mounts in create_request_message()
ceph: stash idmapping in mdsc request
fs: export mnt_idmap_get/mnt_idmap_put
libceph, ceph: move mdsmap.h to fs/ceph
ceph: print cluster fsid and client global_id in all debug logs
ceph: rename _to_client() to _to_fs_client()
ceph: pass the mdsc to several helpers
libceph: add doutc and *_client debug macros support
2023-11-10 10:52:56 -07:00
|
|
|
doutc(cl, "%p %llx.%llx atime %lld.%09ld -> %lld.%09ld\n",
|
|
|
|
inode, ceph_vinop(inode),
|
|
|
|
atime.tv_sec, atime.tv_nsec,
|
|
|
|
attr->ia_atime.tv_sec, attr->ia_atime.tv_nsec);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (issued & CEPH_CAP_FILE_EXCL) {
|
|
|
|
ci->i_time_warp_seq++;
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_atime_to_ts(inode, attr->ia_atime);
|
2009-10-06 11:31:08 -07:00
|
|
|
dirtied |= CEPH_CAP_FILE_EXCL;
|
|
|
|
} else if ((issued & CEPH_CAP_FILE_WR) &&
|
2023-10-04 11:52:09 -07:00
|
|
|
timespec64_compare(&atime,
|
|
|
|
&attr->ia_atime) < 0) {
|
|
|
|
inode_set_atime_to_ts(inode, attr->ia_atime);
|
2009-10-06 11:31:08 -07:00
|
|
|
dirtied |= CEPH_CAP_FILE_WR;
|
|
|
|
} else if ((issued & CEPH_CAP_FILE_SHARED) == 0 ||
|
2023-10-04 11:52:09 -07:00
|
|
|
!timespec64_equal(&atime, &attr->ia_atime)) {
|
2018-07-13 13:18:36 -07:00
|
|
|
ceph_encode_timespec64(&req->r_args.setattr.atime,
|
|
|
|
&attr->ia_atime);
|
2009-10-06 11:31:08 -07:00
|
|
|
mask |= CEPH_SETATTR_ATIME;
|
2017-11-23 03:22:06 -07:00
|
|
|
release |= CEPH_CAP_FILE_SHARED |
|
|
|
|
CEPH_CAP_FILE_RD | CEPH_CAP_FILE_WR;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
}
|
2019-07-25 10:03:32 -07:00
|
|
|
if (ia_valid & ATTR_SIZE) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx size %lld -> %lld\n", inode,
|
|
|
|
ceph_vinop(inode), isize, attr->ia_size);
|
2022-08-25 06:31:11 -07:00
|
|
|
/*
|
|
|
|
* Only when the new size is smaller and not aligned to
|
|
|
|
* CEPH_FSCRYPT_BLOCK_SIZE will the RMW is needed.
|
|
|
|
*/
|
|
|
|
if (IS_ENCRYPTED(inode) && attr->ia_size < isize &&
|
|
|
|
(attr->ia_size % CEPH_FSCRYPT_BLOCK_SIZE)) {
|
|
|
|
mask |= CEPH_SETATTR_SIZE;
|
|
|
|
release |= CEPH_CAP_FILE_SHARED | CEPH_CAP_FILE_EXCL |
|
|
|
|
CEPH_CAP_FILE_RD | CEPH_CAP_FILE_WR;
|
|
|
|
set_bit(CEPH_MDS_R_FSCRYPT_FILE, &req->r_req_flags);
|
|
|
|
mask |= CEPH_SETATTR_FSCRYPT_FILE;
|
|
|
|
req->r_args.setattr.size =
|
|
|
|
cpu_to_le64(round_up(attr->ia_size,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE));
|
|
|
|
req->r_args.setattr.old_size =
|
|
|
|
cpu_to_le64(round_up(isize,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE));
|
|
|
|
req->r_fscrypt_file = attr->ia_size;
|
|
|
|
fill_fscrypt = true;
|
|
|
|
} else if ((issued & CEPH_CAP_FILE_EXCL) && attr->ia_size >= isize) {
|
2021-10-08 01:23:58 -07:00
|
|
|
if (attr->ia_size > isize) {
|
|
|
|
i_size_write(inode, attr->ia_size);
|
|
|
|
inode->i_blocks = calc_inode_blocks(attr->ia_size);
|
|
|
|
ci->i_reported_size = attr->ia_size;
|
|
|
|
dirtied |= CEPH_CAP_FILE_EXCL;
|
|
|
|
ia_valid |= ATTR_MTIME;
|
|
|
|
}
|
2019-07-25 10:03:32 -07:00
|
|
|
} else if ((issued & CEPH_CAP_FILE_SHARED) == 0 ||
|
2021-04-09 12:58:35 -07:00
|
|
|
attr->ia_size != isize) {
|
2019-07-25 10:03:32 -07:00
|
|
|
mask |= CEPH_SETATTR_SIZE;
|
|
|
|
release |= CEPH_CAP_FILE_SHARED | CEPH_CAP_FILE_EXCL |
|
|
|
|
CEPH_CAP_FILE_RD | CEPH_CAP_FILE_WR;
|
2022-08-25 06:31:06 -07:00
|
|
|
if (IS_ENCRYPTED(inode) && attr->ia_size) {
|
|
|
|
set_bit(CEPH_MDS_R_FSCRYPT_FILE, &req->r_req_flags);
|
|
|
|
mask |= CEPH_SETATTR_FSCRYPT_FILE;
|
|
|
|
req->r_args.setattr.size =
|
|
|
|
cpu_to_le64(round_up(attr->ia_size,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE));
|
|
|
|
req->r_args.setattr.old_size =
|
|
|
|
cpu_to_le64(round_up(isize,
|
|
|
|
CEPH_FSCRYPT_BLOCK_SIZE));
|
|
|
|
req->r_fscrypt_file = attr->ia_size;
|
|
|
|
} else {
|
|
|
|
req->r_args.setattr.size = cpu_to_le64(attr->ia_size);
|
|
|
|
req->r_args.setattr.old_size = cpu_to_le64(isize);
|
|
|
|
req->r_fscrypt_file = 0;
|
|
|
|
}
|
2019-07-25 10:03:32 -07:00
|
|
|
}
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
if (ia_valid & ATTR_MTIME) {
|
2023-10-04 11:52:09 -07:00
|
|
|
struct timespec64 mtime = inode_get_mtime(inode);
|
|
|
|
|
Two items:
- support for idmapped mounts in CephFS (Christian Brauner, Alexander
Mikhalitsyn). The series was originally developed by Christian and
later picked up and brought over the finish line by Alexander, who
also contributed an enabler on the MDS side (separate owner_{u,g}id
fields on the wire). The required exports for mnt_idmap_{get,put}()
in VFS have been acked by Christian and received no objection from
Christoph.
- a churny change in CephFS logging to include cluster and client
identifiers in log and debug messages (Xiubo Li). This would help
in scenarios with dozens of CephFS mounts on the same node which are
getting increasingly common, especially in the Kubernetes world.
-----BEGIN PGP SIGNATURE-----
iQFHBAABCAAxFiEEydHwtzie9C7TfviiSn/eOAIR84sFAmVNDyMTHGlkcnlvbW92
QGdtYWlsLmNvbQAKCRBKf944AhHziy6xCACmUikhgo+pZDO7oQS7HChdZFz5Q2Jb
5K9K7sJr7Zb2gFKLAV93cyFrz0JRljmgZuA3DmDTH1omrkrVAJfHJ6Md1UFG3W7o
LaowP3kECsykOiz+YtVU2957sfoFqds/q6KCXdp1Pc8WfNnL1vysCim6EGYtCqUm
c6vv0zkvEDQp+3kjTN01LHzHFZdHg8+STNM4BMiB0sO/NqbADnaQBEtDpYrgzwh7
YPwVIKcXutUmAKlb7vjUF9ptSICYkGV7B+loPS4BDva1I6dT42xqLQOu89tTKU0D
DiQAfE2oRgU+fl+mMooFJNSqD25Q6IkXrPs0HuiSHS4wtLGqYZAwwfHB
=F9o1
-----END PGP SIGNATURE-----
Merge tag 'ceph-for-6.7-rc1' of https://github.com/ceph/ceph-client
Pull ceph updates from Ilya Dryomov:
- support for idmapped mounts in CephFS (Christian Brauner, Alexander
Mikhalitsyn).
The series was originally developed by Christian and later picked up
and brought over the finish line by Alexander, who also contributed
an enabler on the MDS side (separate owner_{u,g}id fields on the
wire).
The required exports for mnt_idmap_{get,put}() in VFS have been acked
by Christian and received no objection from Christoph.
- a churny change in CephFS logging to include cluster and client
identifiers in log and debug messages (Xiubo Li).
This would help in scenarios with dozens of CephFS mounts on the same
node which are getting increasingly common, especially in the
Kubernetes world.
* tag 'ceph-for-6.7-rc1' of https://github.com/ceph/ceph-client:
ceph: allow idmapped mounts
ceph: allow idmapped atomic_open inode op
ceph: allow idmapped set_acl inode op
ceph: allow idmapped setattr inode op
ceph: pass idmap to __ceph_setattr
ceph: allow idmapped permission inode op
ceph: allow idmapped getattr inode op
ceph: pass an idmapping to mknod/symlink/mkdir
ceph: add enable_unsafe_idmap module parameter
ceph: handle idmapped mounts in create_request_message()
ceph: stash idmapping in mdsc request
fs: export mnt_idmap_get/mnt_idmap_put
libceph, ceph: move mdsmap.h to fs/ceph
ceph: print cluster fsid and client global_id in all debug logs
ceph: rename _to_client() to _to_fs_client()
ceph: pass the mdsc to several helpers
libceph: add doutc and *_client debug macros support
2023-11-10 10:52:56 -07:00
|
|
|
doutc(cl, "%p %llx.%llx mtime %lld.%09ld -> %lld.%09ld\n",
|
|
|
|
inode, ceph_vinop(inode),
|
|
|
|
mtime.tv_sec, mtime.tv_nsec,
|
|
|
|
attr->ia_mtime.tv_sec, attr->ia_mtime.tv_nsec);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (issued & CEPH_CAP_FILE_EXCL) {
|
|
|
|
ci->i_time_warp_seq++;
|
2023-10-04 11:52:09 -07:00
|
|
|
inode_set_mtime_to_ts(inode, attr->ia_mtime);
|
2009-10-06 11:31:08 -07:00
|
|
|
dirtied |= CEPH_CAP_FILE_EXCL;
|
|
|
|
} else if ((issued & CEPH_CAP_FILE_WR) &&
|
2023-10-04 11:52:09 -07:00
|
|
|
timespec64_compare(&mtime, &attr->ia_mtime) < 0) {
|
|
|
|
inode_set_mtime_to_ts(inode, attr->ia_mtime);
|
2009-10-06 11:31:08 -07:00
|
|
|
dirtied |= CEPH_CAP_FILE_WR;
|
|
|
|
} else if ((issued & CEPH_CAP_FILE_SHARED) == 0 ||
|
2023-10-04 11:52:09 -07:00
|
|
|
!timespec64_equal(&mtime, &attr->ia_mtime)) {
|
2018-07-13 13:18:36 -07:00
|
|
|
ceph_encode_timespec64(&req->r_args.setattr.mtime,
|
|
|
|
&attr->ia_mtime);
|
2009-10-06 11:31:08 -07:00
|
|
|
mask |= CEPH_SETATTR_MTIME;
|
2017-11-23 03:22:06 -07:00
|
|
|
release |= CEPH_CAP_FILE_SHARED |
|
|
|
|
CEPH_CAP_FILE_RD | CEPH_CAP_FILE_WR;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* these do nothing */
|
|
|
|
if (ia_valid & ATTR_CTIME) {
|
|
|
|
bool only = (ia_valid & (ATTR_SIZE|ATTR_MTIME|ATTR_ATIME|
|
|
|
|
ATTR_MODE|ATTR_UID|ATTR_GID)) == 0;
|
Two items:
- support for idmapped mounts in CephFS (Christian Brauner, Alexander
Mikhalitsyn). The series was originally developed by Christian and
later picked up and brought over the finish line by Alexander, who
also contributed an enabler on the MDS side (separate owner_{u,g}id
fields on the wire). The required exports for mnt_idmap_{get,put}()
in VFS have been acked by Christian and received no objection from
Christoph.
- a churny change in CephFS logging to include cluster and client
identifiers in log and debug messages (Xiubo Li). This would help
in scenarios with dozens of CephFS mounts on the same node which are
getting increasingly common, especially in the Kubernetes world.
-----BEGIN PGP SIGNATURE-----
iQFHBAABCAAxFiEEydHwtzie9C7TfviiSn/eOAIR84sFAmVNDyMTHGlkcnlvbW92
QGdtYWlsLmNvbQAKCRBKf944AhHziy6xCACmUikhgo+pZDO7oQS7HChdZFz5Q2Jb
5K9K7sJr7Zb2gFKLAV93cyFrz0JRljmgZuA3DmDTH1omrkrVAJfHJ6Md1UFG3W7o
LaowP3kECsykOiz+YtVU2957sfoFqds/q6KCXdp1Pc8WfNnL1vysCim6EGYtCqUm
c6vv0zkvEDQp+3kjTN01LHzHFZdHg8+STNM4BMiB0sO/NqbADnaQBEtDpYrgzwh7
YPwVIKcXutUmAKlb7vjUF9ptSICYkGV7B+loPS4BDva1I6dT42xqLQOu89tTKU0D
DiQAfE2oRgU+fl+mMooFJNSqD25Q6IkXrPs0HuiSHS4wtLGqYZAwwfHB
=F9o1
-----END PGP SIGNATURE-----
Merge tag 'ceph-for-6.7-rc1' of https://github.com/ceph/ceph-client
Pull ceph updates from Ilya Dryomov:
- support for idmapped mounts in CephFS (Christian Brauner, Alexander
Mikhalitsyn).
The series was originally developed by Christian and later picked up
and brought over the finish line by Alexander, who also contributed
an enabler on the MDS side (separate owner_{u,g}id fields on the
wire).
The required exports for mnt_idmap_{get,put}() in VFS have been acked
by Christian and received no objection from Christoph.
- a churny change in CephFS logging to include cluster and client
identifiers in log and debug messages (Xiubo Li).
This would help in scenarios with dozens of CephFS mounts on the same
node which are getting increasingly common, especially in the
Kubernetes world.
* tag 'ceph-for-6.7-rc1' of https://github.com/ceph/ceph-client:
ceph: allow idmapped mounts
ceph: allow idmapped atomic_open inode op
ceph: allow idmapped set_acl inode op
ceph: allow idmapped setattr inode op
ceph: pass idmap to __ceph_setattr
ceph: allow idmapped permission inode op
ceph: allow idmapped getattr inode op
ceph: pass an idmapping to mknod/symlink/mkdir
ceph: add enable_unsafe_idmap module parameter
ceph: handle idmapped mounts in create_request_message()
ceph: stash idmapping in mdsc request
fs: export mnt_idmap_get/mnt_idmap_put
libceph, ceph: move mdsmap.h to fs/ceph
ceph: print cluster fsid and client global_id in all debug logs
ceph: rename _to_client() to _to_fs_client()
ceph: pass the mdsc to several helpers
libceph: add doutc and *_client debug macros support
2023-11-10 10:52:56 -07:00
|
|
|
doutc(cl, "%p %llx.%llx ctime %lld.%09ld -> %lld.%09ld (%s)\n",
|
|
|
|
inode, ceph_vinop(inode),
|
|
|
|
inode_get_ctime_sec(inode),
|
|
|
|
inode_get_ctime_nsec(inode),
|
2023-06-11 18:04:07 -07:00
|
|
|
attr->ia_ctime.tv_sec, attr->ia_ctime.tv_nsec,
|
|
|
|
only ? "ctime only" : "ignored");
|
2009-10-06 11:31:08 -07:00
|
|
|
if (only) {
|
|
|
|
/*
|
|
|
|
* if kernel wants to dirty ctime but nothing else,
|
|
|
|
* we need to choose a cap to dirty under, or do
|
|
|
|
* a almost-no-op setattr
|
|
|
|
*/
|
|
|
|
if (issued & CEPH_CAP_AUTH_EXCL)
|
|
|
|
dirtied |= CEPH_CAP_AUTH_EXCL;
|
|
|
|
else if (issued & CEPH_CAP_FILE_EXCL)
|
|
|
|
dirtied |= CEPH_CAP_FILE_EXCL;
|
|
|
|
else if (issued & CEPH_CAP_XATTR_EXCL)
|
|
|
|
dirtied |= CEPH_CAP_XATTR_EXCL;
|
|
|
|
else
|
|
|
|
mask |= CEPH_SETATTR_CTIME;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (ia_valid & ATTR_FILE)
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx ATTR_FILE ... hrm!\n", inode,
|
|
|
|
ceph_vinop(inode));
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (dirtied) {
|
2015-06-10 02:26:13 -07:00
|
|
|
inode_dirty_flags = __ceph_mark_dirty_caps(ci, dirtied,
|
|
|
|
&prealloc_cf);
|
2023-07-05 12:00:55 -07:00
|
|
|
inode_set_ctime_to_ts(inode, attr->ia_ctime);
|
2022-08-24 06:24:42 -07:00
|
|
|
inode_inc_iversion_raw(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
release &= issued;
|
2011-11-30 10:47:09 -07:00
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
2022-08-25 06:31:11 -07:00
|
|
|
if (lock_snap_rwsem) {
|
2015-05-01 02:49:16 -07:00
|
|
|
up_read(&mdsc->snap_rwsem);
|
2022-08-25 06:31:11 -07:00
|
|
|
lock_snap_rwsem = false;
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2011-05-04 11:33:47 -07:00
|
|
|
if (inode_dirty_flags)
|
|
|
|
__mark_inode_dirty(inode, inode_dirty_flags);
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
if (mask) {
|
2011-05-27 09:24:26 -07:00
|
|
|
req->r_inode = inode;
|
|
|
|
ihold(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
req->r_inode_drop = release;
|
|
|
|
req->r_args.setattr.mask = cpu_to_le32(mask);
|
|
|
|
req->r_num_caps = 1;
|
2018-07-13 13:18:38 -07:00
|
|
|
req->r_stamp = attr->ia_ctime;
|
2022-08-25 06:31:11 -07:00
|
|
|
if (fill_fscrypt) {
|
|
|
|
err = fill_fscrypt_truncate(inode, req, attr);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The truncate request will return -EAGAIN when the
|
|
|
|
* last block has been updated just before the MDS
|
|
|
|
* successfully gets the xlock for the FILE lock. To
|
|
|
|
* avoid corrupting the file contents we need to retry
|
|
|
|
* it.
|
|
|
|
*/
|
2013-02-05 14:52:29 -07:00
|
|
|
err = ceph_mdsc_do_request(mdsc, NULL, req);
|
2022-08-25 06:31:11 -07:00
|
|
|
if (err == -EAGAIN && truncate_retry--) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx result=%d (%s locally, %d remote), retry it!\n",
|
|
|
|
inode, ceph_vinop(inode), err,
|
|
|
|
ceph_cap_string(dirtied), mask);
|
2022-08-25 06:31:11 -07:00
|
|
|
ceph_mdsc_put_request(req);
|
|
|
|
ceph_free_cap_flush(prealloc_cf);
|
|
|
|
goto retry;
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
2020-07-27 07:16:09 -07:00
|
|
|
out:
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "%p %llx.%llx result=%d (%s locally, %d remote)\n", inode,
|
|
|
|
ceph_vinop(inode), err, ceph_cap_string(dirtied), mask);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
ceph_mdsc_put_request(req);
|
2015-06-10 02:26:13 -07:00
|
|
|
ceph_free_cap_flush(prealloc_cf);
|
2017-04-18 19:01:48 -07:00
|
|
|
|
|
|
|
if (err >= 0 && (mask & CEPH_SETATTR_SIZE))
|
|
|
|
__ceph_do_pending_vmtruncate(inode);
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2016-04-13 15:30:16 -07:00
|
|
|
/*
|
|
|
|
* setattr
|
|
|
|
*/
|
2023-01-13 04:49:11 -07:00
|
|
|
int ceph_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
|
2021-01-21 06:19:43 -07:00
|
|
|
struct iattr *attr)
|
2016-04-13 15:30:16 -07:00
|
|
|
{
|
2016-05-26 07:10:38 -07:00
|
|
|
struct inode *inode = d_inode(dentry);
|
2023-06-11 19:50:38 -07:00
|
|
|
struct ceph_fs_client *fsc = ceph_inode_to_fs_client(inode);
|
2016-05-26 07:10:38 -07:00
|
|
|
int err;
|
|
|
|
|
|
|
|
if (ceph_snap(inode) != CEPH_NOSNAP)
|
|
|
|
return -EROFS;
|
|
|
|
|
2021-08-31 10:39:13 -07:00
|
|
|
if (ceph_inode_is_shutdown(inode))
|
|
|
|
return -ESTALE;
|
|
|
|
|
2021-07-01 11:40:51 -07:00
|
|
|
err = fscrypt_prepare_setattr(dentry, attr);
|
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
|
2023-08-07 06:26:23 -07:00
|
|
|
err = setattr_prepare(idmap, dentry, attr);
|
2016-05-26 07:10:38 -07:00
|
|
|
if (err != 0)
|
|
|
|
return err;
|
|
|
|
|
2018-08-05 04:33:01 -07:00
|
|
|
if ((attr->ia_valid & ATTR_SIZE) &&
|
2021-04-09 12:58:35 -07:00
|
|
|
attr->ia_size > max(i_size_read(inode), fsc->max_file_size))
|
2018-08-05 04:33:01 -07:00
|
|
|
return -EFBIG;
|
|
|
|
|
2018-01-05 03:47:21 -07:00
|
|
|
if ((attr->ia_valid & ATTR_SIZE) &&
|
|
|
|
ceph_quota_is_max_bytes_exceeded(inode, attr->ia_size))
|
|
|
|
return -EDQUOT;
|
|
|
|
|
2023-08-07 06:26:22 -07:00
|
|
|
err = __ceph_setattr(idmap, inode, attr, NULL);
|
2017-04-18 19:01:48 -07:00
|
|
|
|
|
|
|
if (err >= 0 && (attr->ia_valid & ATTR_MODE))
|
2023-08-07 06:26:23 -07:00
|
|
|
err = posix_acl_chmod(idmap, dentry, attr->ia_mode);
|
2017-04-18 19:01:48 -07:00
|
|
|
|
|
|
|
return err;
|
2016-04-13 15:30:16 -07:00
|
|
|
}
|
|
|
|
|
2022-04-20 20:26:40 -07:00
|
|
|
int ceph_try_to_choose_auth_mds(struct inode *inode, int mask)
|
|
|
|
{
|
|
|
|
int issued = ceph_caps_issued(ceph_inode(inode));
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If any 'x' caps is issued we can just choose the auth MDS
|
|
|
|
* instead of the random replica MDSes. Because only when the
|
|
|
|
* Locker is in LOCK_EXEC state will the loner client could
|
|
|
|
* get the 'x' caps. And if we send the getattr requests to
|
|
|
|
* any replica MDS it must auth pin and tries to rdlock from
|
|
|
|
* the auth MDS, and then the auth MDS need to do the Locker
|
|
|
|
* state transition to LOCK_SYNC. And after that the lock state
|
|
|
|
* will change back.
|
|
|
|
*
|
|
|
|
* This cost much when doing the Locker state transition and
|
|
|
|
* usually will need to revoke caps from clients.
|
2022-05-26 21:39:17 -07:00
|
|
|
*
|
|
|
|
* And for the 'Xs' caps for getxattr we will also choose the
|
|
|
|
* auth MDS, because the MDS side code is buggy due to setxattr
|
|
|
|
* won't notify the replica MDSes when the values changed and
|
|
|
|
* the replica MDS will return the old values. Though we will
|
|
|
|
* fix it in MDS code, but this still makes sense for old ceph.
|
2022-04-20 20:26:40 -07:00
|
|
|
*/
|
|
|
|
if (((mask & CEPH_CAP_ANY_SHARED) && (issued & CEPH_CAP_ANY_EXCL))
|
2022-05-26 21:39:17 -07:00
|
|
|
|| (mask & (CEPH_STAT_RSTAT | CEPH_STAT_CAP_XATTR)))
|
2022-04-20 20:26:40 -07:00
|
|
|
return USE_AUTH_MDS;
|
|
|
|
else
|
|
|
|
return USE_ANY_MDS;
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
|
|
|
* Verify that we have a lease on the given mask. If not,
|
|
|
|
* do a getattr against an mds.
|
|
|
|
*/
|
2014-11-14 06:56:29 -07:00
|
|
|
int __ceph_do_getattr(struct inode *inode, struct page *locked_page,
|
|
|
|
int mask, bool force)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2023-06-11 19:50:38 -07:00
|
|
|
struct ceph_fs_client *fsc = ceph_sb_to_fs_client(inode->i_sb);
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = fsc->client;
|
2010-04-06 15:14:15 -07:00
|
|
|
struct ceph_mds_client *mdsc = fsc->mdsc;
|
2009-10-06 11:31:08 -07:00
|
|
|
struct ceph_mds_request *req;
|
2018-04-25 02:30:23 -07:00
|
|
|
int mode;
|
2009-10-06 11:31:08 -07:00
|
|
|
int err;
|
|
|
|
|
|
|
|
if (ceph_snap(inode) == CEPH_SNAPDIR) {
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "inode %p %llx.%llx SNAPDIR\n", inode,
|
|
|
|
ceph_vinop(inode));
|
2009-10-06 11:31:08 -07:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "inode %p %llx.%llx mask %s mode 0%o\n", inode,
|
|
|
|
ceph_vinop(inode), ceph_cap_string(mask), inode->i_mode);
|
2020-03-19 20:45:00 -07:00
|
|
|
if (!force && ceph_caps_issued_mask_metric(ceph_inode(inode), mask, 1))
|
|
|
|
return 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2022-04-20 20:26:40 -07:00
|
|
|
mode = ceph_try_to_choose_auth_mds(inode, mask);
|
2018-04-25 02:30:23 -07:00
|
|
|
req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_GETATTR, mode);
|
2009-10-06 11:31:08 -07:00
|
|
|
if (IS_ERR(req))
|
|
|
|
return PTR_ERR(req);
|
2011-05-27 09:24:26 -07:00
|
|
|
req->r_inode = inode;
|
|
|
|
ihold(inode);
|
2009-10-06 11:31:08 -07:00
|
|
|
req->r_num_caps = 1;
|
|
|
|
req->r_args.getattr.mask = cpu_to_le32(mask);
|
2014-11-14 06:56:29 -07:00
|
|
|
req->r_locked_page = locked_page;
|
2009-10-06 11:31:08 -07:00
|
|
|
err = ceph_mdsc_do_request(mdsc, NULL, req);
|
2014-11-14 06:56:29 -07:00
|
|
|
if (locked_page && err == 0) {
|
|
|
|
u64 inline_version = req->r_reply_info.targeti.inline_version;
|
|
|
|
if (inline_version == 0) {
|
|
|
|
/* the reply is supposed to contain inline data */
|
|
|
|
err = -EINVAL;
|
2022-06-06 19:13:53 -07:00
|
|
|
} else if (inline_version == CEPH_INLINE_NONE ||
|
|
|
|
inline_version == 1) {
|
2014-11-14 06:56:29 -07:00
|
|
|
err = -ENODATA;
|
|
|
|
} else {
|
|
|
|
err = req->r_reply_info.targeti.inline_len;
|
|
|
|
}
|
|
|
|
}
|
2009-10-06 11:31:08 -07:00
|
|
|
ceph_mdsc_put_request(req);
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "result=%d\n", err);
|
2009-10-06 11:31:08 -07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2022-02-13 22:01:01 -07:00
|
|
|
int ceph_do_getvxattr(struct inode *inode, const char *name, void *value,
|
|
|
|
size_t size)
|
|
|
|
{
|
2023-06-11 19:50:38 -07:00
|
|
|
struct ceph_fs_client *fsc = ceph_sb_to_fs_client(inode->i_sb);
|
2023-06-11 18:04:07 -07:00
|
|
|
struct ceph_client *cl = fsc->client;
|
2022-02-13 22:01:01 -07:00
|
|
|
struct ceph_mds_client *mdsc = fsc->mdsc;
|
|
|
|
struct ceph_mds_request *req;
|
|
|
|
int mode = USE_AUTH_MDS;
|
|
|
|
int err;
|
|
|
|
char *xattr_value;
|
|
|
|
size_t xattr_value_len;
|
|
|
|
|
|
|
|
req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_GETVXATTR, mode);
|
|
|
|
if (IS_ERR(req)) {
|
|
|
|
err = -ENOMEM;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2022-07-26 21:29:10 -07:00
|
|
|
req->r_feature_needed = CEPHFS_FEATURE_OP_GETVXATTR;
|
2022-02-13 22:01:01 -07:00
|
|
|
req->r_path2 = kstrdup(name, GFP_NOFS);
|
|
|
|
if (!req->r_path2) {
|
|
|
|
err = -ENOMEM;
|
|
|
|
goto put;
|
|
|
|
}
|
|
|
|
|
|
|
|
ihold(inode);
|
|
|
|
req->r_inode = inode;
|
|
|
|
err = ceph_mdsc_do_request(mdsc, NULL, req);
|
|
|
|
if (err < 0)
|
|
|
|
goto put;
|
|
|
|
|
|
|
|
xattr_value = req->r_reply_info.xattr_info.xattr_value;
|
|
|
|
xattr_value_len = req->r_reply_info.xattr_info.xattr_value_len;
|
|
|
|
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "xattr_value_len:%zu, size:%zu\n", xattr_value_len, size);
|
2022-02-13 22:01:01 -07:00
|
|
|
|
|
|
|
err = (int)xattr_value_len;
|
|
|
|
if (size == 0)
|
|
|
|
goto put;
|
|
|
|
|
|
|
|
if (xattr_value_len > size) {
|
|
|
|
err = -ERANGE;
|
|
|
|
goto put;
|
|
|
|
}
|
|
|
|
|
|
|
|
memcpy(value, xattr_value, xattr_value_len);
|
|
|
|
put:
|
|
|
|
ceph_mdsc_put_request(req);
|
|
|
|
out:
|
2023-06-11 18:04:07 -07:00
|
|
|
doutc(cl, "result=%d\n", err);
|
2022-02-13 22:01:01 -07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Check inode permissions. We verify we have a valid value for
|
|
|
|
* the AUTH cap, then call the generic handler.
|
|
|
|
*/
|
2023-01-13 04:49:22 -07:00
|
|
|
int ceph_permission(struct mnt_idmap *idmap, struct inode *inode,
|
2021-01-21 06:19:43 -07:00
|
|
|
int mask)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
2011-01-06 23:49:58 -07:00
|
|
|
int err;
|
|
|
|
|
2011-06-20 16:28:19 -07:00
|
|
|
if (mask & MAY_NOT_BLOCK)
|
2011-01-06 23:49:58 -07:00
|
|
|
return -ECHILD;
|
|
|
|
|
2014-09-16 06:46:17 -07:00
|
|
|
err = ceph_do_getattr(inode, CEPH_CAP_AUTH_SHARED, false);
|
2009-10-06 11:31:08 -07:00
|
|
|
|
|
|
|
if (!err)
|
2023-08-07 06:26:21 -07:00
|
|
|
err = generic_permission(idmap, inode, mask);
|
2009-10-06 11:31:08 -07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2019-04-11 12:27:03 -07:00
|
|
|
/* Craft a mask of needed caps given a set of requested statx attrs. */
|
2020-11-09 11:14:07 -07:00
|
|
|
static int statx_to_caps(u32 want, umode_t mode)
|
2019-04-11 12:27:03 -07:00
|
|
|
{
|
|
|
|
int mask = 0;
|
|
|
|
|
2022-08-05 10:50:02 -07:00
|
|
|
if (want & (STATX_MODE|STATX_UID|STATX_GID|STATX_CTIME|STATX_BTIME|STATX_CHANGE_COOKIE))
|
2019-04-11 12:27:03 -07:00
|
|
|
mask |= CEPH_CAP_AUTH_SHARED;
|
|
|
|
|
2022-08-05 10:50:02 -07:00
|
|
|
if (want & (STATX_NLINK|STATX_CTIME|STATX_CHANGE_COOKIE)) {
|
2020-11-09 11:14:07 -07:00
|
|
|
/*
|
|
|
|
* The link count for directories depends on inode->i_subdirs,
|
|
|
|
* and that is only updated when Fs caps are held.
|
|
|
|
*/
|
|
|
|
if (S_ISDIR(mode))
|
|
|
|
mask |= CEPH_CAP_FILE_SHARED;
|
|
|
|
else
|
|
|
|
mask |= CEPH_CAP_LINK_SHARED;
|
|
|
|
}
|
2019-04-11 12:27:03 -07:00
|
|
|
|
2022-08-05 10:50:02 -07:00
|
|
|
if (want & (STATX_ATIME|STATX_MTIME|STATX_CTIME|STATX_SIZE|STATX_BLOCKS|STATX_CHANGE_COOKIE))
|
2019-04-11 12:27:03 -07:00
|
|
|
mask |= CEPH_CAP_FILE_SHARED;
|
|
|
|
|
2022-08-05 10:50:02 -07:00
|
|
|
if (want & (STATX_CTIME|STATX_CHANGE_COOKIE))
|
2019-04-11 12:27:03 -07:00
|
|
|
mask |= CEPH_CAP_XATTR_SHARED;
|
|
|
|
|
|
|
|
return mask;
|
|
|
|
}
|
|
|
|
|
2009-10-06 11:31:08 -07:00
|
|
|
/*
|
2019-04-11 12:27:03 -07:00
|
|
|
* Get all the attributes. If we have sufficient caps for the requested attrs,
|
|
|
|
* then we can avoid talking to the MDS at all.
|
2009-10-06 11:31:08 -07:00
|
|
|
*/
|
2023-01-13 04:49:12 -07:00
|
|
|
int ceph_getattr(struct mnt_idmap *idmap, const struct path *path,
|
2021-01-21 06:19:43 -07:00
|
|
|
struct kstat *stat, u32 request_mask, unsigned int flags)
|
2009-10-06 11:31:08 -07:00
|
|
|
{
|
statx: Add a system call to make enhanced file info available
Add a system call to make extended file information available, including
file creation and some attribute flags where available through the
underlying filesystem.
The getattr inode operation is altered to take two additional arguments: a
u32 request_mask and an unsigned int flags that indicate the
synchronisation mode. This change is propagated to the vfs_getattr*()
function.
Functions like vfs_stat() are now inline wrappers around new functions
vfs_statx() and vfs_statx_fd() to reduce stack usage.
========
OVERVIEW
========
The idea was initially proposed as a set of xattrs that could be retrieved
with getxattr(), but the general preference proved to be for a new syscall
with an extended stat structure.
A number of requests were gathered for features to be included. The
following have been included:
(1) Make the fields a consistent size on all arches and make them large.
(2) Spare space, request flags and information flags are provided for
future expansion.
(3) Better support for the y2038 problem [Arnd Bergmann] (tv_sec is an
__s64).
(4) Creation time: The SMB protocol carries the creation time, which could
be exported by Samba, which will in turn help CIFS make use of
FS-Cache as that can be used for coherency data (stx_btime).
This is also specified in NFSv4 as a recommended attribute and could
be exported by NFSD [Steve French].
(5) Lightweight stat: Ask for just those details of interest, and allow a
netfs (such as NFS) to approximate anything not of interest, possibly
without going to the server [Trond Myklebust, Ulrich Drepper, Andreas
Dilger] (AT_STATX_DONT_SYNC).
(6) Heavyweight stat: Force a netfs to go to the server, even if it thinks
its cached attributes are up to date [Trond Myklebust]
(AT_STATX_FORCE_SYNC).
And the following have been left out for future extension:
(7) Data version number: Could be used by userspace NFS servers [Aneesh
Kumar].
Can also be used to modify fill_post_wcc() in NFSD which retrieves
i_version directly, but has just called vfs_getattr(). It could get
it from the kstat struct if it used vfs_xgetattr() instead.
(There's disagreement on the exact semantics of a single field, since
not all filesystems do this the same way).
(8) BSD stat compatibility: Including more fields from the BSD stat such
as creation time (st_btime) and inode generation number (st_gen)
[Jeremy Allison, Bernd Schubert].
(9) Inode generation number: Useful for FUSE and userspace NFS servers
[Bernd Schubert].
(This was asked for but later deemed unnecessary with the
open-by-handle capability available and caused disagreement as to
whether it's a security hole or not).
(10) Extra coherency data may be useful in making backups [Andreas Dilger].
(No particular data were offered, but things like last backup
timestamp, the data version number and the DOS archive bit would come
into this category).
(11) Allow the filesystem to indicate what it can/cannot provide: A
filesystem can now say it doesn't support a standard stat feature if
that isn't available, so if, for instance, inode numbers or UIDs don't
exist or are fabricated locally...
(This requires a separate system call - I have an fsinfo() call idea
for this).
(12) Store a 16-byte volume ID in the superblock that can be returned in
struct xstat [Steve French].
(Deferred to fsinfo).
(13) Include granularity fields in the time data to indicate the
granularity of each of the times (NFSv4 time_delta) [Steve French].
(Deferred to fsinfo).
(14) FS_IOC_GETFLAGS value. These could be translated to BSD's st_flags.
Note that the Linux IOC flags are a mess and filesystems such as Ext4
define flags that aren't in linux/fs.h, so translation in the kernel
may be a necessity (or, possibly, we provide the filesystem type too).
(Some attributes are made available in stx_attributes, but the general
feeling was that the IOC flags were to ext[234]-specific and shouldn't
be exposed through statx this way).
(15) Mask of features available on file (eg: ACLs, seclabel) [Brad Boyer,
Michael Kerrisk].
(Deferred, probably to fsinfo. Finding out if there's an ACL or
seclabal might require extra filesystem operations).
(16) Femtosecond-resolution timestamps [Dave Chinner].
(A __reserved field has been left in the statx_timestamp struct for
this - if there proves to be a need).
(17) A set multiple attributes syscall to go with this.
===============
NEW SYSTEM CALL
===============
The new system call is:
int ret = statx(int dfd,
const char *filename,
unsigned int flags,
unsigned int mask,
struct statx *buffer);
The dfd, filename and flags parameters indicate the file to query, in a
similar way to fstatat(). There is no equivalent of lstat() as that can be
emulated with statx() by passing AT_SYMLINK_NOFOLLOW in flags. There is
also no equivalent of fstat() as that can be emulated by passing a NULL
filename to statx() with the fd of interest in dfd.
Whether or not statx() synchronises the attributes with the backing store
can be controlled by OR'ing a value into the flags argument (this typically
only affects network filesystems):
(1) AT_STATX_SYNC_AS_STAT tells statx() to behave as stat() does in this
respect.
(2) AT_STATX_FORCE_SYNC will require a network filesystem to synchronise
its attributes with the server - which might require data writeback to
occur to get the timestamps correct.
(3) AT_STATX_DONT_SYNC will suppress synchronisation with the server in a
network filesystem. The resulting values should be considered
approximate.
mask is a bitmask indicating the fields in struct statx that are of
interest to the caller. The user should set this to STATX_BASIC_STATS to
get the basic set returned by stat(). It should be noted that asking for
more information may entail extra I/O operations.
buffer points to the destination for the data. This must be 256 bytes in
size.
======================
MAIN ATTRIBUTES RECORD
======================
The following structures are defined in which to return the main attribute
set:
struct statx_timestamp {
__s64 tv_sec;
__s32 tv_nsec;
__s32 __reserved;
};
struct statx {
__u32 stx_mask;
__u32 stx_blksize;
__u64 stx_attributes;
__u32 stx_nlink;
__u32 stx_uid;
__u32 stx_gid;
__u16 stx_mode;
__u16 __spare0[1];
__u64 stx_ino;
__u64 stx_size;
__u64 stx_blocks;
__u64 __spare1[1];
struct statx_timestamp stx_atime;
struct statx_timestamp stx_btime;
struct statx_timestamp stx_ctime;
struct statx_timestamp stx_mtime;
__u32 stx_rdev_major;
__u32 stx_rdev_minor;
__u32 stx_dev_major;
__u32 stx_dev_minor;
__u64 __spare2[14];
};
The defined bits in request_mask and stx_mask are:
STATX_TYPE Want/got stx_mode & S_IFMT
STATX_MODE Want/got stx_mode & ~S_IFMT
STATX_NLINK Want/got stx_nlink
STATX_UID Want/got stx_uid
STATX_GID Want/got stx_gid
STATX_ATIME Want/got stx_atime{,_ns}
STATX_MTIME Want/got stx_mtime{,_ns}
STATX_CTIME Want/got stx_ctime{,_ns}
STATX_INO Want/got stx_ino
STATX_SIZE Want/got stx_size
STATX_BLOCKS Want/got stx_blocks
STATX_BASIC_STATS [The stuff in the normal stat struct]
STATX_BTIME Want/got stx_btime{,_ns}
STATX_ALL [All currently available stuff]
stx_btime is the file creation time, stx_mask is a bitmask indicating the
data provided and __spares*[] are where as-yet undefined fields can be
placed.
Time fields are structures with separate seconds and nanoseconds fields
plus a reserved field in case we want to add even finer resolution. Note
that times will be negative if before 1970; in such a case, the nanosecond
fields will also be negative if not zero.
The bits defined in the stx_attributes field convey information about a
file, how it is accessed, where it is and what it does. The following
attributes map to FS_*_FL flags and are the same numerical value:
STATX_ATTR_COMPRESSED File is compressed by the fs
STATX_ATTR_IMMUTABLE File is marked immutable
STATX_ATTR_APPEND File is append-only
STATX_ATTR_NODUMP File is not to be dumped
STATX_ATTR_ENCRYPTED File requires key to decrypt in fs
Within the kernel, the supported flags are listed by:
KSTAT_ATTR_FS_IOC_FLAGS
[Are any other IOC flags of sufficient general interest to be exposed
through this interface?]
New flags include:
STATX_ATTR_AUTOMOUNT Object is an automount trigger
These are for the use of GUI tools that might want to mark files specially,
depending on what they are.
Fields in struct statx come in a number of classes:
(0) stx_dev_*, stx_blksize.
These are local system information and are always available.
(1) stx_mode, stx_nlinks, stx_uid, stx_gid, stx_[amc]time, stx_ino,
stx_size, stx_blocks.
These will be returned whether the caller asks for them or not. The
corresponding bits in stx_mask will be set to indicate whether they
actually have valid values.
If the caller didn't ask for them, then they may be approximated. For
example, NFS won't waste any time updating them from the server,
unless as a byproduct of updating something requested.
If the values don't actually exist for the underlying object (such as
UID or GID on a DOS file), then the bit won't be set in the stx_mask,
even if the caller asked for the value. In such a case, the returned
value will be a fabrication.
Note that there are instances where the type might not be valid, for
instance Windows reparse points.
(2) stx_rdev_*.
This will be set only if stx_mode indicates we're looking at a
blockdev or a chardev, otherwise will be 0.
(3) stx_btime.
Similar to (1), except this will be set to 0 if it doesn't exist.
=======
TESTING
=======
The following test program can be used to test the statx system call:
samples/statx/test-statx.c
Just compile and run, passing it paths to the files you want to examine.
The file is built automatically if CONFIG_SAMPLES is enabled.
Here's some example output. Firstly, an NFS directory that crosses to
another FSID. Note that the AUTOMOUNT attribute is set because transiting
this directory will cause d_automount to be invoked by the VFS.
[root@andromeda ~]# /tmp/test-statx -A /warthog/data
statx(/warthog/data) = 0
results=7ff
Size: 4096 Blocks: 8 IO Block: 1048576 directory
Device: 00:26 Inode: 1703937 Links: 125
Access: (3777/drwxrwxrwx) Uid: 0 Gid: 4041
Access: 2016-11-24 09:02:12.219699527+0000
Modify: 2016-11-17 10:44:36.225653653+0000
Change: 2016-11-17 10:44:36.225653653+0000
Attributes: 0000000000001000 (-------- -------- -------- -------- -------- -------- ---m---- --------)
Secondly, the result of automounting on that directory.
[root@andromeda ~]# /tmp/test-statx /warthog/data
statx(/warthog/data) = 0
results=7ff
Size: 4096 Blocks: 8 IO Block: 1048576 directory
Device: 00:27 Inode: 2 Links: 125
Access: (3777/drwxrwxrwx) Uid: 0 Gid: 4041
Access: 2016-11-24 09:02:12.219699527+0000
Modify: 2016-11-17 10:44:36.225653653+0000
Change: 2016-11-17 10:44:36.225653653+0000
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2017-01-31 09:46:22 -07:00
|
|
|
struct inode *inode = d_inode(path->dentry);
|
2022-08-30 21:13:28 -07:00
|
|
|
struct super_block *sb = inode->i_sb;
|
2009-10-21 11:21:49 -07:00
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
2019-06-05 09:08:32 -07:00
|
|
|
u32 valid_mask = STATX_BASIC_STATS;
|
2019-04-11 12:27:03 -07:00
|
|
|
int err = 0;
|
2009-10-06 11:31:08 -07:00
|
|
|
|
2021-08-31 10:39:13 -07:00
|
|
|
if (ceph_inode_is_shutdown(inode))
|
|
|
|
return -ESTALE;
|
|
|
|
|
2019-04-11 12:27:03 -07:00
|
|
|
/* Skip the getattr altogether if we're asked not to sync */
|
2022-04-10 18:59:09 -07:00
|
|
|
if ((flags & AT_STATX_SYNC_TYPE) != AT_STATX_DONT_SYNC) {
|
2020-11-09 11:14:07 -07:00
|
|
|
err = ceph_do_getattr(inode,
|
|
|
|
statx_to_caps(request_mask, inode->i_mode),
|
|
|
|
flags & AT_STATX_FORCE_SYNC);
|
2019-04-11 12:27:03 -07:00
|
|
|
if (err)
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2023-08-07 06:26:20 -07:00
|
|
|
generic_fillattr(idmap, request_mask, inode, stat);
|
ceph: fix inode number handling on arches with 32-bit ino_t
Tuan and Ulrich mentioned that they were hitting a problem on s390x,
which has a 32-bit ino_t value, even though it's a 64-bit arch (for
historical reasons).
I think the current handling of inode numbers in the ceph driver is
wrong. It tries to use 32-bit inode numbers on 32-bit arches, but that's
actually not a problem. 32-bit arches can deal with 64-bit inode numbers
just fine when userland code is compiled with LFS support (the common
case these days).
What we really want to do is just use 64-bit numbers everywhere, unless
someone has mounted with the ino32 mount option. In that case, we want
to ensure that we hash the inode number down to something that will fit
in 32 bits before presenting the value to userland.
Add new helper functions that do this, and only do the conversion before
presenting these values to userland in getattr and readdir.
The inode table hashvalue is changed to just cast the inode number to
unsigned long, as low-order bits are the most likely to vary anyway.
While it's not strictly required, we do want to put something in
inode->i_ino. Instead of basing it on BITS_PER_LONG, however, base it on
the size of the ino_t type.
NOTE: This is a user-visible change on 32-bit arches:
1/ inode numbers will be seen to have changed between kernel versions.
32-bit arches will see large inode numbers now instead of the hashed
ones they saw before.
2/ any really old software not built with LFS support may start failing
stat() calls with -EOVERFLOW on inode numbers >2^32. Nothing much we
can do about these, but hopefully the intersection of people running
such code on ceph will be very small.
The workaround for both problems is to mount with "-o ino32".
[ idryomov: changelog tweak ]
URL: https://tracker.ceph.com/issues/46828
Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
Reported-and-Tested-by: Tuan Hoang1 <Tuan.Hoang1@ibm.com>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2020-08-18 05:03:48 -07:00
|
|
|
stat->ino = ceph_present_inode(inode);
|
2019-06-05 09:08:32 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* btime on newly-allocated inodes is 0, so if this is still set to
|
|
|
|
* that, then assume that it's not valid.
|
|
|
|
*/
|
|
|
|
if (ci->i_btime.tv_sec || ci->i_btime.tv_nsec) {
|
|
|
|
stat->btime = ci->i_btime;
|
|
|
|
valid_mask |= STATX_BTIME;
|
|
|
|
}
|
|
|
|
|
2022-08-05 10:50:02 -07:00
|
|
|
if (request_mask & STATX_CHANGE_COOKIE) {
|
|
|
|
stat->change_cookie = inode_peek_iversion_raw(inode);
|
|
|
|
valid_mask |= STATX_CHANGE_COOKIE;
|
|
|
|
}
|
|
|
|
|
2019-04-11 12:27:03 -07:00
|
|
|
if (ceph_snap(inode) == CEPH_NOSNAP)
|
2022-08-30 21:13:28 -07:00
|
|
|
stat->dev = sb->s_dev;
|
2019-04-11 12:27:03 -07:00
|
|
|
else
|
|
|
|
stat->dev = ci->i_snapid_map ? ci->i_snapid_map->dev : 0;
|
|
|
|
|
|
|
|
if (S_ISDIR(inode->i_mode)) {
|
2023-06-11 19:50:38 -07:00
|
|
|
if (ceph_test_mount_opt(ceph_sb_to_fs_client(sb), RBYTES)) {
|
2019-04-11 12:27:03 -07:00
|
|
|
stat->size = ci->i_rbytes;
|
2022-08-30 21:13:28 -07:00
|
|
|
} else if (ceph_snap(inode) == CEPH_SNAPDIR) {
|
|
|
|
struct ceph_inode_info *pci;
|
|
|
|
struct ceph_snap_realm *realm;
|
|
|
|
struct inode *parent;
|
|
|
|
|
|
|
|
parent = ceph_lookup_inode(sb, ceph_ino(inode));
|
2022-10-14 02:36:53 -07:00
|
|
|
if (IS_ERR(parent))
|
2022-08-30 21:13:28 -07:00
|
|
|
return PTR_ERR(parent);
|
|
|
|
|
|
|
|
pci = ceph_inode(parent);
|
|
|
|
spin_lock(&pci->i_ceph_lock);
|
|
|
|
realm = pci->i_snap_realm;
|
|
|
|
if (realm)
|
|
|
|
stat->size = realm->num_snaps;
|
|
|
|
else
|
|
|
|
stat->size = 0;
|
|
|
|
spin_unlock(&pci->i_ceph_lock);
|
|
|
|
iput(parent);
|
|
|
|
} else {
|
2019-04-11 12:27:03 -07:00
|
|
|
stat->size = ci->i_files + ci->i_subdirs;
|
2022-08-30 21:13:28 -07:00
|
|
|
}
|
2019-04-11 12:27:03 -07:00
|
|
|
stat->blocks = 0;
|
|
|
|
stat->blksize = 65536;
|
|
|
|
/*
|
|
|
|
* Some applications rely on the number of st_nlink
|
|
|
|
* value on directories to be either 0 (if unlinked)
|
|
|
|
* or 2 + number of subdirectories.
|
|
|
|
*/
|
|
|
|
if (stat->nlink == 1)
|
|
|
|
/* '.' + '..' + subdirs */
|
|
|
|
stat->nlink = 1 + 1 + ci->i_subdirs;
|
2009-10-06 11:31:08 -07:00
|
|
|
}
|
2019-04-11 12:27:03 -07:00
|
|
|
|
2022-08-05 10:50:02 -07:00
|
|
|
stat->attributes |= STATX_ATTR_CHANGE_MONOTONIC;
|
2021-07-01 11:40:51 -07:00
|
|
|
if (IS_ENCRYPTED(inode))
|
|
|
|
stat->attributes |= STATX_ATTR_ENCRYPTED;
|
|
|
|
stat->attributes_mask |= (STATX_ATTR_CHANGE_MONOTONIC |
|
|
|
|
STATX_ATTR_ENCRYPTED);
|
|
|
|
|
2019-06-05 09:08:32 -07:00
|
|
|
stat->result_mask = request_mask & valid_mask;
|
2009-10-06 11:31:08 -07:00
|
|
|
return err;
|
|
|
|
}
|
2021-08-31 10:39:13 -07:00
|
|
|
|
|
|
|
void ceph_inode_shutdown(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct ceph_inode_info *ci = ceph_inode(inode);
|
|
|
|
struct rb_node *p;
|
|
|
|
int iputs = 0;
|
|
|
|
bool invalidate = false;
|
|
|
|
|
|
|
|
spin_lock(&ci->i_ceph_lock);
|
|
|
|
ci->i_ceph_flags |= CEPH_I_SHUTDOWN;
|
|
|
|
p = rb_first(&ci->i_caps);
|
|
|
|
while (p) {
|
|
|
|
struct ceph_cap *cap = rb_entry(p, struct ceph_cap, ci_node);
|
|
|
|
|
|
|
|
p = rb_next(p);
|
|
|
|
iputs += ceph_purge_inode_cap(inode, cap, &invalidate);
|
|
|
|
}
|
|
|
|
spin_unlock(&ci->i_ceph_lock);
|
|
|
|
|
|
|
|
if (invalidate)
|
|
|
|
ceph_queue_invalidate(inode);
|
|
|
|
while (iputs--)
|
|
|
|
iput(inode);
|
|
|
|
}
|