Jeff Layton [Tue, 7 Dec 2021 13:44:51 +0000 (08:44 -0500)]
ceph: add fscache writeback support
When updating the backing store from the pagecache (a'la writepage or
writepages), write to the cache first. This allows us to keep caching
files even when they are being written, as long as we have appropriate
caps.
Jeff Layton [Tue, 7 Dec 2021 13:44:50 +0000 (08:44 -0500)]
ceph: conversion to new fscache API
Now that the fscache API has been reworked and simplified, change ceph
over to use it.
With the old API, we would only instantiate a cookie when the file was
open for reads. Change it to instantiate the cookie when the inode is
instantiated and call use/unuse when the file is opened/closed.
Also, ensure we resize the cached data on truncates, and invalidate the
cache in response to the appropriate events. This will allow us to
plumb in write support later.
David Howells [Fri, 27 Aug 2021 14:19:34 +0000 (15:19 +0100)]
nfs: Implement cache I/O by accessing the cache directly
Move NFS to using fscache DIO API instead of the old upstream I/O API as
that has been removed. This is a stopgap solution as the intention is that
at sometime in the future, the cache will move to using larger blocks and
won't be able to store individual pages in order to deal with the potential
for data corruption due to the backing filesystem being able insert/remove
bridging blocks of zeros into its extent list[1].
NFS then reads and writes cache pages synchronously and one page at a time.
The preferred change would be to use the netfs lib, but the new I/O API can
be used directly. It's just that as the cache now needs to track data for
itself, caching blocks may exceed page size...
This code is somewhat borrowed from my "fallback I/O" patchset[2].
Changes
=======
ver #3:
- Restore lost =n fallback for nfs_fscache_release_page()[2].
Dave Wysochanski [Sat, 14 Nov 2020 18:43:54 +0000 (13:43 -0500)]
nfs: Convert to new fscache volume/cookie API
Change the nfs filesystem to support fscache's indexing rewrite and
reenable caching in nfs.
The following changes have been made:
(1) The fscache_netfs struct is no more, and there's no need to register
the filesystem as a whole.
(2) The session cookie is now an fscache_volume cookie, allocated with
fscache_acquire_volume(). That takes three parameters: a string
representing the "volume" in the index, a string naming the cache to
use (or NULL) and a u64 that conveys coherency metadata for the
volume.
For nfs, I've made it render the volume name string as:
(3) The fscache_cookie_def is no more and needed information is passed
directly to fscache_acquire_cookie(). The cache no longer calls back
into the filesystem, but rather metadata changes are indicated at
other times.
fscache_acquire_cookie() is passed the same keying and coherency
information as before.
(4) fscache_enable/disable_cookie() have been removed.
Call fscache_use_cookie() and fscache_unuse_cookie() when a file is
opened or closed to prevent a cache file from being culled and to keep
resources to hand that are needed to do I/O.
If a file is opened for writing, we invalidate it with
FSCACHE_INVAL_DIO_WRITE in lieu of doing writeback to the cache,
thereby making it cease caching until all currently open files are
closed. This should give the same behaviour as the uptream code.
Making the cache store local modifications isn't straightforward for
NFS, so that's left for future patches.
(5) fscache_invalidate() now needs to be given uptodate auxiliary data and
a file size. It also takes a flag to indicate if this was due to a
DIO write.
(6) Call nfs_fscache_invalidate() with FSCACHE_INVAL_DIO_WRITE on a file
to which a DIO write is made.
(7) Call fscache_note_page_release() from nfs_release_page().
(8) Use a killable wait in nfs_vm_page_mkwrite() when waiting for
PG_fscache to be cleared.
(9) The functions to read and write data to/from the cache are stubbed out
pending a conversion to use netfslib.
Changes
=======
ver #3:
- Added missing =n fallback for nfs_fscache_release_file()[1][2].
ver #2:
- Use gfpflags_allow_blocking() rather than using flag directly.
- fscache_acquire_volume() now returns errors.
- Remove NFS_INO_FSCACHE as it's no longer used.
- Need to unuse a cookie on file-release, not inode-clear.
David Howells [Wed, 18 Nov 2020 09:06:42 +0000 (09:06 +0000)]
9p: Copy local writes to the cache when writing to the server
When writing to the server from v9fs_vfs_writepage(), copy the data to the
cache object too.
To make this possible, the cookie must have its active users count
incremented when the page is dirtied and kept incremented until we manage
to clean up all the pages. This allows the writeback to take place after
the last file struct is released.
This is done by taking a use on the cookie in v9fs_set_page_dirty() if we
haven't already done so (controlled by the I_PINNING_FSCACHE_WB flag) and
dropping the pin in v9fs_write_inode() if __writeback_single_inode() clears
all the outstanding dirty pages (conveyed by the unpinned_fscache_wb flag
in the writeback_control struct).
Inode eviction must also clear the flag after truncating away all the
outstanding pages.
In the future this will be handled more gracefully by netfslib.
Changes
=======
ver #3:
- Canonicalise the coherency data to make it endianness-independent.
ver #2:
- Fix an unused-var warning due to CONFIG_9P_FSCACHE=n[1].
David Howells [Wed, 18 Nov 2020 09:06:42 +0000 (09:06 +0000)]
9p: Use fscache indexing rewrite and reenable caching
Change the 9p filesystem to take account of the changes to fscache's
indexing rewrite and reenable caching in 9p.
The following changes have been made:
(1) The fscache_netfs struct is no more, and there's no need to register
the filesystem as a whole.
(2) The session cookie is now an fscache_volume cookie, allocated with
fscache_acquire_volume(). That takes three parameters: a string
representing the "volume" in the index, a string naming the cache to
use (or NULL) and a u64 that conveys coherency metadata for the
volume.
For 9p, I've made it render the volume name string as:
"9p,<devname>,<cachetag>"
where the cachetag is replaced by the aname if it wasn't supplied.
This probably needs rethinking a bit as the aname can have slashes in
it. It might be better to hash the cachetag and use the hash or I
could substitute commas for the slashes or something.
(3) The fscache_cookie_def is no more and needed information is passed
directly to fscache_acquire_cookie(). The cache no longer calls back
into the filesystem, but rather metadata changes are indicated at
other times.
fscache_acquire_cookie() is passed the same keying and coherency
information as before.
(4) The functions to set/reset/flush cookies are removed and
fscache_use_cookie() and fscache_unuse_cookie() are used instead.
fscache_use_cookie() is passed a flag to indicate if the cookie is
opened for writing. fscache_unuse_cookie() is passed updates for the
metadata if we changed it (ie. if the file was opened for writing).
These are called when the file is opened or closed.
(5) wait_on_page_bit[_killable]() is replaced with the specific wait
functions for the bits waited upon.
(6) I've got rid of some of the 9p-specific cache helper functions and
called things like fscache_relinquish_cookie() directly as they'll
optimise away if v9fs_inode_cookie() returns an unconditional NULL
(which will be the case if CONFIG_9P_FSCACHE=n).
(7) v9fs_vfs_setattr() is made to call fscache_resize() to change the size
of the cache object.
Notes:
(A) We should call fscache_invalidate() if we detect that the server's
copy of a file got changed by a third party, but I don't know where to
do that. We don't need to do that when allocating the cookie as we
get a check-and-invalidate when we initially bind to the cache object.
(B) The copy-to-cache-on-writeback side of things will be handled in
separate patch.
Changes
=======
ver #3:
- Canonicalise the cookie key and coherency data to make them
endianness-independent.
ver #2:
- Use gfpflags_allow_blocking() rather than using flag directly.
- fscache_acquire_volume() now returns errors.
David Howells [Thu, 6 Feb 2020 14:22:30 +0000 (14:22 +0000)]
afs: Copy local writes to the cache when writing to the server
When writing to the server from afs_writepage() or afs_writepages(), copy
the data to the cache object too.
To make this possible, the cookie must have its active users count
incremented when the page is dirtied and kept incremented until we manage
to clean up all the pages. This allows the writeback to take place after
the last file struct is released.
David Howells [Thu, 6 Feb 2020 14:22:21 +0000 (14:22 +0000)]
afs: Convert afs to use the new fscache API
Change the afs filesystem to support the new afs driver.
The following changes have been made:
(1) The fscache_netfs struct is no more, and there's no need to register
the filesystem as a whole. There's also no longer a cell cookie.
(2) The volume cookie is now an fscache_volume cookie, allocated with
fscache_acquire_volume(). This function takes three parameters: a
string representing the "volume" in the index, a string naming the
cache to use (or NULL) and a u64 that conveys coherency metadata for
the volume.
For afs, I've made it render the volume name string as:
"afs,<cell>,<volume_id>"
and the coherency data is currently 0.
(3) The fscache_cookie_def is no more and needed information is passed
directly to fscache_acquire_cookie(). The cache no longer calls back
into the filesystem, but rather metadata changes are indicated at
other times.
fscache_acquire_cookie() is passed the same keying and coherency
information as before, except that these are now stored in big endian
form instead of cpu endian. This makes the cache more copyable.
(4) fscache_use_cookie() and fscache_unuse_cookie() are called when a file
is opened or closed to prevent a cache file from being culled and to
keep resources to hand that are needed to do I/O.
fscache_use_cookie() is given an indication if the cache is likely to
be modified locally (e.g. the file is open for writing).
fscache_unuse_cookie() is given a coherency update if we had the file
open for writing and will update that.
(5) fscache_invalidate() is now given uptodate auxiliary data and a file
size. It can also take a flag to indicate if this was due to a DIO
write. This is wrapped into afs_fscache_invalidate() now for
convenience.
(6) fscache_resize() now gets called from the finalisation of
afs_setattr(), and afs_setattr() does use/unuse of the cookie around
the call to support this.
(7) fscache_note_page_release() is called from afs_release_page().
(8) Use a killable wait in nfs_vm_page_mkwrite() when waiting for
PG_fscache to be cleared.
Render the parts of the cookie key for an afs inode cookie as big endian.
Changes
=======
ver #2:
- Use gfpflags_allow_blocking() rather than using flag directly.
- fscache_acquire_volume() now returns errors.
David Howells [Fri, 22 Oct 2021 08:17:58 +0000 (09:17 +0100)]
fscache, cachefiles: Display stat of culling events
Add a stat counter of culling events whereby the cache backend culls a file
to make space (when asked by cachefilesd in this case) and display in
/proc/fs/fscache/stats.
David Howells [Tue, 14 Dec 2021 09:51:43 +0000 (09:51 +0000)]
fscache, cachefiles: Store the volume coherency data
Store the volume coherency data in an xattr and check it when we rebind the
volume. If it doesn't match the cache volume is moved to the graveyard and
rebuilt anew.
Changes
=======
ver #4:
- Remove a couple of debugging prints.
David Howells [Wed, 20 Oct 2021 23:19:46 +0000 (00:19 +0100)]
cachefiles: Implement the I/O routines
Implement the I/O routines for cachefiles. There are two sets of routines
here: preparation and actual I/O.
Preparation for read involves looking to see whether there is data present,
and how much. Netfslib tells us what it wants us to do and we have the
option of adjusting shrinking and telling it whether to read from the
cache, download from the server or simply clear a region.
Preparation for write involves checking for space and defending against
possibly running short of space, if necessary punching out a hole in the
file so that we don't leave old data in the cache if we update the
coherency information.
Then there's a read routine and a write routine. They wait for the cookie
state to move to something appropriate and then start a potentially
asynchronous direct I/O operation upon it.
Changes
=======
ver #2:
- Fix a misassigned variable[1].
David Howells [Thu, 21 Oct 2021 10:05:53 +0000 (11:05 +0100)]
cachefiles: Implement begin and end I/O operation
Implement the methods for beginning and ending an I/O operation.
When called to begin an I/O operation, we are guaranteed that the cookie
has reached a certain stage (we're called by fscache after it has done a
suitable wait).
If a file is available, we paste a ref over into the cache resources for
the I/O routines to use. This means that the object can be invalidated
whilst the I/O is ongoing without the need to synchronise as the file
pointer in the object is replaced, but the file pointer in the cache
resources is unaffected.
Ending the operation just requires ditching any refs we have and dropping
the access guarantee that fscache got for us on the cookie.
David Howells [Thu, 21 Oct 2021 07:50:10 +0000 (08:50 +0100)]
cachefiles: Implement backing file wrangling
Implement the wrangling of backing files, including the following pieces:
(1) Lookup and creation of a file on disk, using a tmpfile if the file
isn't yet present. The file is then opened, sized for DIO and the
file handle is attached to the cachefiles_object struct. The inode is
marked to indicate that it's in use by a kernel service.
(2) Invalidation of an object, creating a tmpfile and switching the file
pointer in the cachefiles object.
(3) Committing a file to disk, including setting the coherency xattr on it
and, if necessary, creating a hard link to it.
Note that this would be a good place to use Omar Sandoval's vfs_link()
with AT_LINK_REPLACE[1] as I may have to unlink an old file before I
can link a tmpfile into place.
(4) Withdrawal of open objects when a cache is being withdrawn or a cookie
is relinquished. This involves committing or discarding the file.
Changes
=======
ver #2:
- Fix logging of wrong error[1].
David Howells [Thu, 21 Oct 2021 07:50:10 +0000 (08:50 +0100)]
cachefiles: Implement culling daemon commands
Implement the ability for the userspace daemon to try and cull a file or
directory in the cache. Two daemon commands are implemented:
(1) The "inuse" command. This queries if a file is in use or whether it
can be deleted. It checks the S_KERNEL_FILE flag on the inode
referred to by the specified filename.
(2) The "cull" command. This asks for a file or directory to be removed,
where removal means either unlinking it or moving it to the graveyard
directory for userspace to dismantle.
Changes
=======
ver #2:
- Fix logging of wrong error[1].
- Need to unmark an inode we've moved to the graveyard before unlocking.
David Howells [Thu, 18 Nov 2021 08:58:08 +0000 (08:58 +0000)]
cachefiles: Mark a backing file in use with an inode flag
Use an inode flag, S_KERNEL_FILE, to mark that a backing file is in use by
the kernel to prevent cachefiles or other kernel services from interfering
with that file.
Using S_SWAPFILE instead isn't really viable as that has other effects in
the I/O paths.
David Howells [Wed, 17 Nov 2021 16:11:07 +0000 (16:11 +0000)]
cachefiles: Implement metadata/coherency data storage in xattrs
Use an xattr on each backing file in the cache to store some metadata, such
as the content type and the coherency data.
Five content types are defined:
(0) No content stored.
(1) The file contains a single monolithic blob and must be all or nothing.
This would be used for something like an AFS directory or a symlink.
(2) The file is populated with content completely up to a point with
nothing beyond that.
(3) The file has a map attached and is sparsely populated. This would be
stored in one or more additional xattrs.
(4) The file is dirty, being in the process of local modification and the
contents are not necessarily represented correctly by the metadata.
The file should be deleted if this is seen on binding.
David Howells [Wed, 17 Nov 2021 15:48:06 +0000 (15:48 +0000)]
cachefiles: Implement key to filename encoding
Implement a function to encode a binary cookie key as something that can be
used as a filename. Four options are considered:
(1) All printable chars with no '/' characters. Prepend a 'D' to indicate
the encoding but otherwise use as-is.
(2) Appears to be an array of __be32. Encode as 'S' plus a list of
hex-encoded 32-bit ints separated by commas. If a number is 0, it is
rendered as "" instead of "0".
(3) Appears to be an array of __le32. Encoded as (2) but with a 'T'
encoding prefix.
(4) Encoded as base64 with an 'E' prefix plus a second char indicating how
much padding is involved. A non-standard base64 encoding is used
because '/' cannot be used in the encoded form.
If (1) is not possible, whichever of (2), (3) or (4) produces the shortest
string is selected (hex-encoding a number may be less dense than base64
encoding it).
Note that the prefix characters have to be selected from the set [DEIJST@]
lest cachefilesd remove the files because it recognise the name.
Changes
=======
ver #2:
- Fix a short allocation that didn't allow for a string terminator[1]
David Howells [Thu, 21 Oct 2021 08:55:21 +0000 (09:55 +0100)]
cachefiles: Implement volume support
Implement support for creating the directory layout for a volume on disk
and setting up and withdrawing volume caching.
Each volume has a directory named for the volume key under the root of the
cache (prefixed with an 'I' to indicate to cachefilesd that it's an index)
and then creates a bunch of hash bucket subdirectories under that (named as
'@' plus a hex number) in which cookie files will be created.
David Howells [Thu, 21 Oct 2021 07:34:55 +0000 (08:34 +0100)]
cachefiles: Implement a function to get/create a directory in the cache
Implement a function to get/create structural directories in the cache.
This is used for setting up a cache and creating volume substructures. The
directory in memory are marked with the S_KERNEL_FILE inode flag whilst
they're in use to tell rmdir to reject attempts to remove them.
Changes
=======
ver #3:
- Return an indication as to whether the directory was freshly created.
David Howells [Thu, 18 Nov 2021 08:58:08 +0000 (08:58 +0000)]
vfs, cachefiles: Mark a backing file in use with an inode flag
Use an inode flag, S_KERNEL_FILE, to mark that a backing file is in use by
the kernel to prevent cachefiles or other kernel services from interfering
with that file.
Alter rmdir to reject attempts to remove a directory marked with this flag.
This is used by cachefiles to prevent cachefilesd from removing them.
Using S_SWAPFILE instead isn't really viable as that has other effects in
the I/O paths.
Changes
=======
ver #3:
- Check for the object pointer being NULL in the tracepoints rather than
the caller.
David Howells [Thu, 21 Oct 2021 07:59:46 +0000 (08:59 +0100)]
cachefiles: Provide a function to check how much space there is
Provide a function to check how much space there is. This also flips the
state on the cache and will signal the daemon to inform it of the change
and to ask it to do some culling if necessary.
We will also need to subtract the amount of data currently being written to
the cache (cache->b_writing) from the amount of available space to avoid
hitting ENOSPC accidentally.
David Howells [Fri, 26 Nov 2021 15:12:07 +0000 (15:12 +0000)]
cachefiles: Register a miscdev and parse commands over it
Register a misc device with which to talk to the daemon. The misc device
holds a cache set up through it around and closing the device kills the
cache.
cachefilesd communicates with the kernel by passing it single-line text
commands. Parse these and use them to parameterise the cache state. This
does not implement the command to actually bring a cache online. That's
left for later.
David Howells [Fri, 26 Nov 2021 14:59:10 +0000 (14:59 +0000)]
cachefiles: Add security derivation
Implement code to derive a new set of creds for the cachefiles to use when
making VFS or I/O calls and to change the auditing info since the
application interacting with the network filesystem is not accessing the
cache directly. Cachefiles uses override_creds() to change the effective
creds temporarily.
set_security_override_from_ctx() is called to derive the LSM 'label' that
the cachefiles driver will act with. set_create_files_as() is called to
determine the LSM 'label' that will be applied to files and directories
created in the cache. These functions alter the new creds.
Also implement a couple of functions to wrap the calls to begin/end cred
overriding.
David Howells [Thu, 21 Oct 2021 07:15:26 +0000 (08:15 +0100)]
cachefiles: Add some error injection support
Add support for injecting ENOSPC or EIO errors. This needs to be enabled
by CONFIG_CACHEFILES_ERROR_INJECTION=y. Once enabled, ENOSPC on things
like write and mkdir can be triggered by:
David Howells [Fri, 26 Nov 2021 14:45:38 +0000 (14:45 +0000)]
cachefiles: Define structs
Define the cachefiles_cache struct that's going to carry the cache-level
parameters and state of a cache.
Define the beginning of the cachefiles_object struct that's going to carry
the state for a data storage object. For the moment this is just a
debugging ID for logging purposes.
David Howells [Wed, 20 Oct 2021 13:06:34 +0000 (14:06 +0100)]
fscache: Provide a function to resize a cookie
Provide a function to change the size of the storage attached to a cookie,
to match the size of the file being cached when it's changed by truncate or
fallocate:
David Howells [Wed, 20 Oct 2021 13:06:34 +0000 (14:06 +0100)]
fscache: Provide a function to note the release of a page
Provide a function to be called from a network filesystem's releasepage
method to indicate that a page has been released that might have been a
reflection of data upon the server - and now that data must be reloaded
from the server or the cache.
This is used to end an optimisation for empty files, in particular files
that have just been created locally, whereby we know there cannot yet be
any data that we would need to read from the server or the cache.
David Howells [Wed, 20 Oct 2021 22:50:01 +0000 (23:50 +0100)]
vfs, fscache: Implement pinning of cache usage for writeback
Cachefiles has a problem in that it needs to keep the backing file for a
cookie open whilst there are local modifications pending that need to be
written to it. However, we don't want to keep the file open indefinitely,
as that causes EMFILE/ENFILE/ENOMEM problems.
Reopening the cache file, however, is a problem if this is being done due
to writeback triggered by exit(). Some filesystems will oops if we try to
open a file in that context because they want to access current->fs or
other resources that have already been dismantled.
To get around this, I added the following:
(1) An inode flag, I_PINNING_FSCACHE_WB, to be set on a network filesystem
inode to indicate that we have a usage count on the cookie caching
that inode.
(2) A flag in struct writeback_control, unpinned_fscache_wb, that is set
when __writeback_single_inode() clears the last dirty page from
i_pages - at which point it clears I_PINNING_FSCACHE_WB and sets this
flag.
This has to be done here so that clearing I_PINNING_FSCACHE_WB can be
done atomically with the check of PAGECACHE_TAG_DIRTY that clears
I_DIRTY_PAGES.
(3) A function, fscache_set_page_dirty(), which if it is not set, sets
I_PINNING_FSCACHE_WB and calls fscache_use_cookie() to pin the cache
resources.
(4) A function, fscache_unpin_writeback(), to be called by ->write_inode()
to unuse the cookie.
(5) A function, fscache_clear_inode_writeback(), to be called when the
inode is evicted, before clear_inode() is called. This cleans up any
lingering I_PINNING_FSCACHE_WB.
The network filesystem can then use these tools to make sure that
fscache_write_to_cache() can write locally modified data to the cache as
well as to the server.
For the future, I'm working on write helpers for netfs lib that should
allow this facility to be removed by keeping track of the dirty regions
separately - but that's incomplete at the moment and is also going to be
affected by folios, one way or another, since it deals with pages
Provide a higher-level function than fscache_write() to perform a write
from an inode's pagecache to the cache, whilst fending off concurrent
writes by means of the PG_fscache mark on a page:
If caching is false, this function does nothing except call (*term_func)()
if given. It assumes that, in such a case, PG_fscache will not have been
set on the pages.
Otherwise, if caching is true, this function requires the source pages to
have had PG_fscache set on them before calling. start and len define the
region of the file to be modified and i_size indicates the new file size.
The source pages are extracted from the mapping.
term_func and term_func_priv work as for fscache_write(). The PG_fscache
marks will be cleared at the end of the operation, before term_func is
called or the function otherwise returns.
There is an additonal helper function to clear the PG_fscache bits from a
range of pages:
If caching is true, the pages to be managed are expected to be located on
mapping in the range defined by start and len. If caching is false, it
does nothing.
David Howells [Wed, 20 Oct 2021 13:06:34 +0000 (14:06 +0100)]
fscache: Implement raw I/O interface
Provide a pair of functions to perform raw I/O on the cache. The first
function allows an arbitrary asynchronous direct-IO read to be made against
a cache object, though the read should be aligned and sized appropriately
for the backing device:
The cache resources must have been previously initialised by
fscache_begin_read_operation(). A read operation is sent to the backing
filesystem, starting at start_pos within the file. The size of the read is
specified by the iterator, as is the location of the output buffer.
If there is a hole in the data it can be ignored and left to the backing
filesystem to deal with (NETFS_READ_HOLE_IGNORE), a hole at the beginning
can be skipped over and the buffer padded with zeros
(NETFS_READ_HOLE_CLEAR) or -ENODATA can be given (NETFS_READ_HOLE_FAIL).
If term_func is not NULL, the operation may be performed asynchronously.
Upon completion, successful or otherwise, (*term_func)() will be called and
passed term_func_priv, along with an error or the amount of data
transferred. If the op is run asynchronously, fscache_read() will return
-EIOCBQUEUED.
The second function allows an arbitrary asynchronous direct-IO write to be
made against a cache object, though the write should be aligned and sized
appropriately for the backing device:
This will update the auxiliary data and/or the size of the object attached
to a cookie if either pointer is not-NULL and flag that the disk needs to
be updated.
Note that fscache_unuse_cookie() also allows this to be done.
David Howells [Thu, 11 Nov 2021 23:14:29 +0000 (23:14 +0000)]
fscache: Count data storage objects in a cache
Count the data storage objects that are currently allocated in a cache.
This is used to pin certain cache structures until cache withdrawal is
complete.
Three helpers are provided to manage and make use of the count:
This should be called by the backend to note that an object has been
destroyed. This sends a wakeup event that allows cache withdrawal to
proceed if it was waiting for that object.
David Howells [Wed, 20 Oct 2021 13:06:34 +0000 (14:06 +0100)]
fscache: Provide a means to begin an operation
Provide a function to begin a read operation:
int fscache_begin_read_operation(
struct netfs_cache_resources *cres,
struct fscache_cookie *cookie)
This is primarily intended to be called by network filesystems on behalf of
netfslib, but may also be called to use the I/O access functions directly.
It attaches the resources required by the cache to cres struct from the
supplied cookie.
This holds access to the cache behind the cookie for the duration of the
operation and forces cache withdrawal and cookie invalidation to perform
synchronisation on the operation. cres->inval_counter is set from the
cookie at this point so that it can be compared at the end of the
operation.
Note that this does not guarantee that the cache state is fully set up and
able to perform I/O immediately; looking up and creation may be left in
progress in the background. The operations intended to be called by the
network filesystem, such as reading and writing, are expected to wait for
the cookie to move to the correct state.
This will, however, potentially sleep, waiting for a certain minimum state
to be set or for operations such as invalidate to advance far enough that
I/O can resume.
Also provide a function for the cache to call to wait for the cache object
to get to a state where it can be used for certain things:
This looks at the cache resources provided by the begin function and waits
for them to get to an appropriate stage. There's a choice of wanting just
some parameters (FSCACHE_WANT_PARAM) or the ability to do I/O
(FSCACHE_WANT_READ or FSCACHE_WANT_WRITE).
This causes any cached data for the specified cookie to be discarded. If
the cookie is marked as being in use, a new cache object will be created if
possible and future I/O will use that instead. In-flight I/O should be
abandoned (writes) or reconsidered (reads). Each time it is called
cookie->inval_counter is incremented and this can be used to detect
invalidation at the end of an I/O operation.
The coherency data attached to the cookie can be updated and the cookie
size should be reset. One flag is available, FSCACHE_INVAL_DIO_WRITE,
which should be used to indicate invalidation due to a DIO write on a
file. This will temporarily disable caching for this cookie.
Changes
=======
ver #2:
- Should only change to inval state if can get access to cache.
David Howells [Wed, 20 Oct 2021 14:53:34 +0000 (15:53 +0100)]
fscache: Implement cookie user counting and resource pinning
Provide a pair of functions to count the number of users of a cookie (open
files, writeback, invalidation, resizing, reads, writes), to obtain and pin
resources for the cookie and to prevent culling for the whilst there are
users.
The first function marks a cookie as being in use:
The caller should indicate the cookie to use and whether or not the caller
is in a context that may modify the cookie (e.g. a file open O_RDWR).
If the cookie is not already resourced, fscache will ask the cache backend
in the background to do whatever it needs to look up, create or otherwise
obtain the resources necessary to access data. This is pinned to the
cookie and may not be culled, though it may be withdrawn if the cache as a
whole is withdrawn.
The second function removes the in-use mark from a cookie and, optionally,
updates the coherency data:
If non-NULL, the aux_data buffer and/or the object_size will be saved into
the cookie and will be set on the backing store when the object is
committed.
If this removes the last usage on a cookie, the cookie is placed onto an
LRU list from which it will be removed and closed after a couple of seconds
if it doesn't get reused. This prevents resource overload in the cache -
in particular it prevents it from holding too many files open.
Changes
=======
ver #2:
- Fix fscache_unuse_cookie() to use atomic_dec_and_lock() to avoid a
potential race if the cookie gets reused before it completes the
unusement.
- Added missing transition to LRU_DISCARDING state.
David Howells [Wed, 20 Oct 2021 14:53:34 +0000 (15:53 +0100)]
fscache: Implement simple cookie state machine
Implement a very simple cookie state machine to handle lookup,
invalidation, withdrawal, relinquishment and, to be added later, commit on
LRU discard.
Three cache methods are provided: ->lookup_cookie() to look up and, if
necessary, create a data storage object; ->withdraw_cookie() to free the
resources associated with that object and potentially delete it; and
->prepare_to_write(), to do prepare for changes to the cached data to be
modified locally.
Changes
=======
ver #3:
- Fix a race between LRU discard and relinquishment whereby the former
would override the latter and thus the latter would never happen[1].
ver #2:
- Don't hold n_accesses elevated whilst cache is bound to a cookie, but
rather add a flag that prevents the state machine from being queued when
n_accesses reaches 0.
David Howells [Wed, 20 Oct 2021 14:26:17 +0000 (15:26 +0100)]
fscache: Provide and use cache methods to lookup/create/free a volume
Add cache methods to lookup, create and remove a volume.
Looking up or creating the volume requires the cache pinning for access;
freeing the volume requires the volume pinning for access. The
->acquire_volume() method is used to ask the cache backend to lookup and,
if necessary, create a volume; the ->free_volume() method is used to free
the resources for a volume.
David Howells [Wed, 20 Oct 2021 14:53:34 +0000 (15:53 +0100)]
fscache: Implement cookie-level access helpers
Add a number of helper functions to manage access to a cookie, pinning the
cache object in place for the duration to prevent cache withdrawal from
removing it:
This function initialises the access count when a cache binds to a
cookie. An extra ref is taken on the access count to prevent wakeups
while the cache is active. We're only interested in the wakeup when a
cookie is being withdrawn and we're waiting for it to quiesce - at
which point the counter will be decremented before the wait.
The FSCACHE_COOKIE_NACC_ELEVATED flag is set on the cookie to keep
track of the extra ref in order to handle a race between
relinquishment and withdrawal both trying to drop the extra ref.
This function attempts to begin access upon a cookie, pinning it in
place if it's cached. If successful, it returns true and leaves a the
access count incremented.
This function drops the access count obtained by (2), permitting
object withdrawal to take place when it reaches zero.
A tracepoint is provided to track changes to the access counter on a
cookie.
Changes
=======
ver #2:
- Don't hold n_accesses elevated whilst cache is bound to a cookie, but
rather add a flag that prevents the state machine from being queued when
n_accesses reaches 0.
David Howells [Wed, 20 Oct 2021 14:26:17 +0000 (15:26 +0100)]
fscache: Implement volume-level access helpers
Add a pair of helper functions to manage access to a volume, pinning the
volume in place for the duration to prevent cache withdrawal from removing
it:
The way the access gate on the volume works/will work is:
(1) If the cache tests as not live (state is not FSCACHE_CACHE_IS_ACTIVE),
then we return false to indicate access was not permitted.
(2) If the cache tests as live, then we increment the volume's n_accesses
count and then recheck the cache liveness, ending the access if it
ceased to be live.
(3) When we end the access, we decrement the volume's n_accesses and wake
up the any waiters if it reaches 0.
(4) Whilst the cache is caching, the volume's n_accesses is kept
artificially incremented to prevent wakeups from happening.
(5) When the cache is taken offline, the state is changed to prevent new
accesses, the volume's n_accesses is decremented and we wait for it to
become 0.
David Howells [Wed, 20 Oct 2021 14:53:34 +0000 (15:53 +0100)]
fscache: Implement cookie registration
Add functions to the fscache API to allow data file cookies to be acquired
and relinquished by the network filesystem. It is intended that the
filesystem will create such cookies per-inode under a volume.
The filesystem must first have created a volume cookie, which is passed in
here. If it passes in NULL then the function will just return a NULL
cookie.
A binary key should be passed in index_key and is of size index_key_len.
This is saved in the cookie and is used to locate the associated data in
the cache.
A coherency data buffer of size aux_data_len will be allocated and
initialised from the buffer pointed to by aux_data. This is used to
validate cache objects when they're opened and is stored on disk with them
when they're committed. The data is stored in the cookie and will be
updateable by various functions in later patches.
The object_size must also be given. This is also used to perform a
coherency check and to size the backing storage appropriately.
This function disallows a cookie from being acquired twice in parallel,
though it will cause the second user to wait if the first is busy
relinquishing its cookie.
When a network filesystem has finished with a cookie, it should call:
If retire is true, any backing data will be discarded immediately.
Changes
=======
ver #3:
- fscache_hash()'s size parameter is now in bytes. Use __le32 as the unit
to round up to.
- When comparing cookies, simply see if the attributes are the same rather
than subtracting them to produce a strcmp-style return[1].
- Add a check to see if the cookie is still hashed at the point of
freeing.
ver #2:
- Don't hold n_accesses elevated whilst cache is bound to a cookie, but
rather add a flag that prevents the state machine from being queued when
n_accesses reaches 0.
- Remove the unused cookie pointer field from the fscache_acquire
tracepoint.
David Howells [Wed, 20 Oct 2021 14:26:17 +0000 (15:26 +0100)]
fscache: Implement volume registration
Add functions to the fscache API to allow volumes to be acquired and
relinquished by the network filesystem. A volume is an index of data
storage cache objects. A volume is represented by a volume cookie in the
API. A filesystem would typically create a volume for a superblock and
then create per-inode cookies within it.
The volume_key is a printable string used to match the volume in the cache.
It should not contain any '/' characters. For AFS, for example, this would
be "afs,<cellname>,<volume_id>", e.g. "afs,example.com,523001".
The cache_name can be NULL, but if not it should be a string indicating the
name of the cache to use if there's more than one available.
The coherency data, if given, is an arbitrarily-sized blob that's attached
to the volume and is compared when the volume is looked up. If it doesn't
match, the old volume is judged to be out of date and it and everything
within it is discarded.
Acquiring a volume twice concurrently is disallowed, though the function
will wait if an old volume cookie is being relinquishing.
When a network filesystem has finished with a volume, it should return the
volume cookie by calling:
If invalidate is true, the entire volume will be discarded; if false, the
volume will be synced and the coherency data will be updated.
Changes
=======
ver #4:
- Removed an extraneous param from kdoc on fscache_relinquish_volume()[3].
ver #3:
- fscache_hash()'s size parameter is now in bytes. Use __le32 as the unit
to round up to.
- When comparing cookies, simply see if the attributes are the same rather
than subtracting them to produce a strcmp-style return[2].
- Make the coherency data an arbitrary blob rather than a u64, but don't
store it for the moment.
ver #2:
- Fix error check[1].
- Make a fscache_acquire_volume() return errors, including EBUSY if a
conflicting volume cookie already exists. No error is printed now -
that's left to the netfs.
This gets the cache cookie for a cache of the specified name and moves
it to the preparation state. If a nameless cache cookie exists, that
will be given this name and used.
David Howells [Wed, 20 Oct 2021 14:45:28 +0000 (15:45 +0100)]
fscache: Implement a hash function
Implement a function to generate hashes. It needs to be stable over time
and endianness-independent as the hashes will appear on disk in future
patches. It can assume that its input is a multiple of four bytes in size
and alignment.
This is borrowed from the VFS and simplified. le32_to_cpu() is added to
make it endianness-independent.
Changes
=======
ver #3:
- Read the data being hashed in an endianness-independent way[1].
- Change the size parameter to be in bytes rather than words.
David Howells [Wed, 20 Oct 2021 13:30:37 +0000 (14:30 +0100)]
netfs: Pass a flag to ->prepare_write() to say if there's no alloc'd space
Pass a flag to ->prepare_write() to indicate if there's definitely no
space allocated in the cache yet (for instance if we've already checked as
we were asked to do a read).
David Howells [Mon, 25 Oct 2021 20:53:44 +0000 (21:53 +0100)]
fscache: Remove the contents of the fscache driver, pending rewrite
Remove the code that comprises the fscache driver as it's going to be
substantially rewritten, with the majority of the code being erased in the
rewrite.
A small piece of linux/fscache.h is left as that is #included by a bunch of
network filesystems.
Jeffle Xu [Tue, 7 Dec 2021 03:14:49 +0000 (11:14 +0800)]
netfs: fix parameter of cleanup()
The order of these two parameters is just reversed. gcc didn't warn on
that, probably because 'void *' can be converted from or to other
pointer types without warning.
Cc: stable@vger.kernel.org Fixes: f04d087afe94 ("netfs: Provide readahead and readpage netfs helpers") Fixes: 4784311c1fa0 ("netfs: Add write_begin helper") Signed-off-by: Jeffle Xu <jefflexu@linux.alibaba.com> Signed-off-by: David Howells <dhowells@redhat.com> Reviewed-by: Jeff Layton <jlayton@redhat.com> Link: https://lore.kernel.org/r/20211207031449.100510-1-jefflexu@linux.alibaba.com/
David Howells [Tue, 7 Dec 2021 09:53:24 +0000 (09:53 +0000)]
netfs: Fix lockdep warning from taking sb_writers whilst holding mmap_lock
Taking sb_writers whilst holding mmap_lock isn't allowed and will result in
a lockdep warning like that below. The problem comes from cachefiles
needing to take the sb_writers lock in order to do a write to the cache,
but being asked to do this by netfslib called from readpage, readahead or
write_begin[1].
Fix this by always offloading the write to the cache off to a worker
thread. The main thread doesn't need to wait for it, so deadlock can be
avoided.
This can be tested by running the quick xfstests on something like afs or
ceph with lockdep enabled.
WARNING: possible circular locking dependency detected
5.15.0-rc1-build2+ #292 Not tainted
------------------------------------------------------
holetest/65517 is trying to acquire lock: ffff88810c81d730 (mapping.invalidate_lock#3){.+.+}-{3:3}, at: filemap_fault+0x276/0x7a5
but task is already holding lock: ffff8881595b53e8 (&mm->mmap_lock#2){++++}-{3:3}, at: do_user_addr_fault+0x28d/0x59c
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
Linus Torvalds [Sun, 5 Dec 2021 20:58:18 +0000 (12:58 -0800)]
Merge tag 'for-5.16/parisc-6' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux
Pull parisc fixes from Helge Deller:
"Some bug and warning fixes:
- Fix "make install" to use debians "installkernel" script which is
now in /usr/sbin
- Fix the bindeb-pkg make target by giving the correct KBUILD_IMAGE
file name
- Fix compiler warnings by annotating parisc agp init functions with
__init
- Fix timekeeping on SMP machines with dual-core CPUs
- Enable some more config options in the 64-bit defconfig"
* tag 'for-5.16/parisc-6' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
parisc: Mark cr16 CPU clocksource unstable on all SMP machines
parisc: Fix "make install" on newer debian releases
parisc/agp: Annotate parisc agp init functions with __init
parisc: Enable sata sil, audit and usb support on 64-bit defconfig
parisc: Fix KBUILD_IMAGE for self-extracting kernel
Linus Torvalds [Sun, 5 Dec 2021 17:34:57 +0000 (09:34 -0800)]
Merge tag 'usb-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB fixes from Greg KH:
"Here are some small USB fixes for a few reported issues. Included in
here are:
- xhci fix for a _much_ reported regression. I don't think there's a
community distro that has not reported this problem yet :(
- new USB quirk addition
- cdns3 minor fixes
- typec regression fix.
All of these have been in linux-next with no reported problems, and
the xhci fix has been reported by many to resolve their reported
problem"
* tag 'usb-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
usb: cdnsp: Fix a NULL pointer dereference in cdnsp_endpoint_init()
usb: cdns3: gadget: fix new urb never complete if ep cancel previous requests
usb: typec: tcpm: Wait in SNK_DEBOUNCED until disconnect
USB: NO_LPM quirk Lenovo Powered USB-C Travel Hub
xhci: Fix commad ring abort, write all 64 bits to CRCR register.
Linus Torvalds [Sun, 5 Dec 2021 17:13:20 +0000 (09:13 -0800)]
Merge tag 'tty-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty
Pull tty/serial fixes from Greg KH:
"Here are some small TTY and Serial driver fixes for 5.16-rc4 to
resolve a number of reported problems.
They include:
- liteuart serial driver fixes
- 8250_pci serial driver fixes for pericom devices
- 8250 RTS line control fix while in RS-485 mode
- tegra serial driver fix
- msm_serial driver fix
- pl011 serial driver new id
- fsl_lpuart revert of broken change
- 8250_bcm7271 serial driver fix
- MAINTAINERS file update for rpmsg tty driver that came in 5.16-rc1
- vgacon fix for reported problem
All of these, except for the 8250_bcm7271 fix have been in linux-next
with no reported problem. The 8250_bcm7271 fix was added to the tree
on Friday so no chance to be linux-next yet. But it should be fine as
the affected developers submitted it"
* tag 'tty-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
serial: 8250_bcm7271: UART errors after resuming from S2
serial: 8250_pci: rewrite pericom_do_set_divisor()
serial: 8250_pci: Fix ACCES entries in pci_serial_quirks array
serial: 8250: Fix RTS modem control while in rs485 mode
Revert "tty: serial: fsl_lpuart: drop earlycon entry for i.MX8QXP"
serial: tegra: Change lower tolerance baud rate limit for tegra20 and tegra30
serial: liteuart: relax compile-test dependencies
serial: liteuart: fix minor-number leak on probe errors
serial: liteuart: fix use-after-free and memleak on unbind
serial: liteuart: Fix NULL pointer dereference in ->remove()
vgacon: Propagate console boot parameters before calling `vc_resize'
tty: serial: msm_serial: Deactivate RX DMA for polling support
serial: pl011: Add ACPI SBSA UART match id
serial: core: fix transmit-buffer reset and memleak
MAINTAINERS: Add rpmsg tty driver maintainer
Linus Torvalds [Sun, 5 Dec 2021 16:58:52 +0000 (08:58 -0800)]
Merge tag 'timers_urgent_for_v5.16_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull timer fix from Borislav Petkov:
- Prevent a tick storm when a dedicated timekeeper CPU in nohz_full
mode runs for prolonged periods with interrupts disabled and ends up
programming the next tick in the past, leading to that storm
* tag 'timers_urgent_for_v5.16_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timers/nohz: Last resort update jiffies on nohz_full IRQ entry
Linus Torvalds [Sun, 5 Dec 2021 16:53:31 +0000 (08:53 -0800)]
Merge tag 'sched_urgent_for_v5.16_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull scheduler fixes from Borislav Petkov:
- Properly init uclamp_flags of a runqueue, on first enqueuing
- Fix preempt= callback return values
- Correct utime/stime resource usage reporting on nohz_full to return
the proper times instead of shorter ones
* tag 'sched_urgent_for_v5.16_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/uclamp: Fix rq->uclamp_max not set on first enqueue
preempt/dynamic: Fix setup_preempt_mode() return value
sched/cputime: Fix getrusage(RUSAGE_THREAD) with nohz_full
Linus Torvalds [Sun, 5 Dec 2021 16:43:35 +0000 (08:43 -0800)]
Merge tag 'x86_urgent_for_v5.16_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Borislav Petkov:
- Fix a couple of SWAPGS fencing issues in the x86 entry code
- Use the proper operand types in __{get,put}_user() to prevent
truncation in SEV-ES string io
- Make sure the kernel mappings are present in trampoline_pgd in order
to prevent any potential accesses to unmapped memory after switching
to it
- Fix a trivial list corruption in objtool's pv_ops validation
- Disable the clocksource watchdog for TSC on platforms which claim
that the TSC is constant, doesn't stop in sleep states, CPU has TSC
adjust and the number of sockets of the platform are max 2, to
prevent erroneous markings of the TSC as unstable.
- Make sure TSC adjust is always checked not only when going idle
- Prevent a stack leak by initializing struct _fpx_sw_bytes properly in
the FPU code
- Fix INTEL_FAM6_RAPTORLAKE define naming to adhere to the convention
* tag 'x86_urgent_for_v5.16_rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/xen: Add xenpv_restore_regs_and_return_to_usermode()
x86/entry: Use the correct fence macro after swapgs in kernel CR3
x86/entry: Add a fence for kernel entry SWAPGS in paranoid_entry()
x86/sev: Fix SEV-ES INS/OUTS instructions for word, dword, and qword
x86/64/mm: Map all kernel memory into trampoline_pgd
objtool: Fix pv_ops noinstr validation
x86/tsc: Disable clocksource watchdog for TSC on qualified platorms
x86/tsc: Add a timer to make sure TSC_adjust is always checked
x86/fpu/signal: Initialize sw_bytes in save_xstate_epilog()
x86/cpu: Drop spurious underscore from RAPTOR_LAKE #define
Linus Torvalds [Sun, 5 Dec 2021 16:25:33 +0000 (08:25 -0800)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull more kvm fixes from Paolo Bonzini:
- Static analysis fix
- New SEV-ES protocol for communicating invalid VMGEXIT requests
- Ensure APICv is considered inactive if there is no APIC
- Fix reserved bits for AMD PerfEvtSeln register
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: SVM: Do not terminate SEV-ES guests on GHCB validation failure
KVM: SEV: Fall back to vmalloc for SEV-ES scratch area if necessary
KVM: SEV: Return appropriate error codes if SEV-ES scratch setup fails
KVM: x86/mmu: Retry page fault if root is invalidated by memslot update
KVM: VMX: Set failure code in prepare_vmcs02()
KVM: ensure APICv is considered inactive if there is no APIC
KVM: x86/pmu: Fix reserved bits for AMD PerfEvtSeln register
Tom Lendacky [Thu, 2 Dec 2021 18:52:05 +0000 (12:52 -0600)]
KVM: SVM: Do not terminate SEV-ES guests on GHCB validation failure
Currently, an SEV-ES guest is terminated if the validation of the VMGEXIT
exit code or exit parameters fails.
The VMGEXIT instruction can be issued from userspace, even though
userspace (likely) can't update the GHCB. To prevent userspace from being
able to kill the guest, return an error through the GHCB when validation
fails rather than terminating the guest. For cases where the GHCB can't be
updated (e.g. the GHCB can't be mapped, etc.), just return back to the
guest.
The new error codes are documented in the lasest update to the GHCB
specification.
Fixes: 2a4a6e417d97 ("KVM: SVM: Add initial support for a VMGEXIT VMEXIT") Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
Message-Id: <b57280b5562893e2616257ac9c2d4525a9aeeb42.1638471124.git.thomas.lendacky@amd.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: SEV: Fall back to vmalloc for SEV-ES scratch area if necessary
Use kvzalloc() to allocate KVM's buffer for SEV-ES's GHCB scratch area so
that KVM falls back to __vmalloc() if physically contiguous memory isn't
available. The buffer is purely a KVM software construct, i.e. there's
no need for it to be physically contiguous.
Cc: Tom Lendacky <thomas.lendacky@amd.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211109222350.2266045-3-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Return appropriate error codes if setting up the GHCB scratch area for an
SEV-ES guest fails. In particular, returning -EINVAL instead of -ENOMEM
when allocating the kernel buffer could be confusing as userspace would
likely suspect a guest issue.
Fixes: 38454143baff ("KVM: SVM: Support MMIO for an SEV-ES guest") Cc: Tom Lendacky <thomas.lendacky@amd.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20211109222350.2266045-2-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Linus Torvalds [Sat, 4 Dec 2021 21:43:52 +0000 (13:43 -0800)]
Merge tag '5.16-rc3-smb3-fixes' of git://git.samba.org/sfrench/cifs-2.6
Pull cifs fixes from Steve French:
"Three SMB3 multichannel/fscache fixes and a DFS fix.
In testing multichannel reconnect scenarios recently various problems
with the cifs.ko implementation of fscache were found (e.g. incorrect
initialization of fscache cookies in some cases)"
* tag '5.16-rc3-smb3-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: avoid use of dstaddr as key for fscache client cookie
cifs: add server conn_id to fscache client cookie
cifs: wait for tcon resource_id before getting fscache super
cifs: fix missed refcounting of ipc tcon
Helge Deller [Sat, 4 Dec 2021 20:21:46 +0000 (21:21 +0100)]
parisc: Mark cr16 CPU clocksource unstable on all SMP machines
In commit cc1957920d1f ("parisc: Enhance detection of synchronous cr16
clocksources") I assumed that CPUs on the same physical core are syncronous.
While booting up the kernel on two different C8000 machines, one with a
dual-core PA8800 and one with a dual-core PA8900 CPU, this turned out to be
wrong. The symptom was that I saw a jump in the internal clocks printed to the
syslog and strange overall behaviour. On machines which have 4 cores (2
dual-cores) the problem isn't visible, because the current logic already marked
the cr16 clocksource unstable in this case.
This patch now marks the cr16 interval timers unstable if we have more than one
CPU in the system, and it fixes this issue.
Helge Deller [Sat, 4 Dec 2021 20:14:40 +0000 (21:14 +0100)]
parisc: Fix "make install" on newer debian releases
On newer debian releases the debian-provided "installkernel" script is
installed in /usr/sbin. Fix the kernel install.sh script to look for the
script in this directory as well.
Linus Torvalds [Sat, 4 Dec 2021 16:34:59 +0000 (08:34 -0800)]
Merge tag 'io_uring-5.16-2021-12-03' of git://git.kernel.dk/linux-block
Pull io_uring fix from Jens Axboe:
"Just a single fix preventing repeated retries of task_work based io-wq
thread creation, fixing a regression from when io-wq was made more (a
bit too much) resilient against signals"
* tag 'io_uring-5.16-2021-12-03' of git://git.kernel.dk/linux-block:
io-wq: don't retry task_work creation failure on fatal conditions
Linus Torvalds [Sat, 4 Dec 2021 16:28:42 +0000 (08:28 -0800)]
Merge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI fixes from James Bottomley:
"Two patches, both in drivers.
One is a fix to FC recovery (lpfc) and the other is an enhancement to
support the Intel Alder Motherboard with the UFS driver which comes
under the -rc exception process for hardware enabling"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: ufs: ufs-pci: Add support for Intel ADL
scsi: lpfc: Fix non-recovery of remote ports following an unsolicited LOGO
Linus Torvalds [Sat, 4 Dec 2021 16:13:20 +0000 (08:13 -0800)]
Merge tag 'gfs2-v5.16-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2
Pull gfs2 fixes from Andreas Gruenbacher:
- Since commit 473c32889d32 ("gfs2: Cancel remote delete work
asynchronously"), inode create and lookup-by-number can overlap more
easily and we can end up with temporary duplicate inodes. Fix the
code to prevent that.
- Fix a BUG demoting weak glock holders from a remote node.
* tag 'gfs2-v5.16-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2:
gfs2: gfs2_create_inode rework
gfs2: gfs2_inode_lookup rework
gfs2: gfs2_inode_lookup cleanup
gfs2: Fix remote demote of weak glock holders
Qais Yousef [Thu, 2 Dec 2021 11:20:33 +0000 (11:20 +0000)]
sched/uclamp: Fix rq->uclamp_max not set on first enqueue
Commit f555fba8c72e ("sched/uclamp: Fix initialization of struct
uclamp_rq") introduced a bug where uclamp_max of the rq is not reset to
match the woken up task's uclamp_max when the rq is idle.
The code was relying on rq->uclamp_max initialized to zero, so on first
enqueue
if (uc_se->value > READ_ONCE(uc_rq->value))
WRITE_ONCE(uc_rq->value, uc_se->value);
}
was actually resetting it. But since commit f555fba8c72e changed the
default to 1024, this no longer works. And since rq->uclamp_flags is
also initialized to 0, neither above code path nor uclamp_idle_reset()
update the rq->uclamp_max on first wake up from idle.
This is only visible from first wake up(s) until the first dequeue to
idle after enabling the static key. And it only matters if the
uclamp_max of this task is < 1024 since only then its uclamp_max will be
effectively ignored.
Fix it by properly initializing rq->uclamp_flags = UCLAMP_FLAG_IDLE to
ensure uclamp_idle_reset() is called which then will update the rq
uclamp_max value as expected.
Andrew Halaney [Fri, 3 Dec 2021 23:32:03 +0000 (17:32 -0600)]
preempt/dynamic: Fix setup_preempt_mode() return value
__setup() callbacks expect 1 for success and 0 for failure. Correct the
usage here to reflect that.
Fixes: 8eb0bd8701cd ("preempt/dynamic: Support dynamic preempt with preempt= boot option") Reported-by: Mark Rutland <mark.rutland@arm.com> Signed-off-by: Andrew Halaney <ahalaney@redhat.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/20211203233203.133581-1-ahalaney@redhat.com
Linus Torvalds [Fri, 3 Dec 2021 20:22:56 +0000 (12:22 -0800)]
Merge tag 'pm-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management fixes from Rafael Wysocki:
"These fix a CPU hot-add issue in the cpufreq core, fix a comment in
the cpufreq core code and update its documentation, and disable the
DTPM (Dynamic Thermal Power Management) code for the time being to
prevent it from causing issues to appear.
Specifics:
- Disable DTPM for this cycle to prevent it from causing issues to
appear on otherwise functional systems (Daniel Lezcano)
- Fix cpufreq sysfs interface failure related to physical CPU hot-add
(Xiongfeng Wang)
- Fix comment in cpufreq core and update its documentation (Tang
Yizhou)"
* tag 'pm-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
powercap: DTPM: Drop unused local variable from init_dtpm()
cpufreq: docs: Update core.rst
cpufreq: Fix a comment in cpufreq_policy_free
powercap/drivers/dtpm: Disable DTPM at boot time
cpufreq: Fix get_cpu_device() failure in add_cpu_dev_symlink()
Linus Torvalds [Fri, 3 Dec 2021 19:46:20 +0000 (11:46 -0800)]
Merge tag 's390-5.16-4' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Pull s390 fixes from Heiko Carstens:
- Fix potential overlap of pseudo-MMIO addresses with MIO addresses
- Fix stack unwinder test case inline assembly compile error that
happens with LLVM's integrated assembler
- Update defconfigs
* tag 's390-5.16-4' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
s390: update defconfigs
s390/pci: move pseudo-MMIO to prevent MIO overlap
s390/test_unwind: use raw opcode instead of invalid instruction
Linus Torvalds [Fri, 3 Dec 2021 18:50:14 +0000 (10:50 -0800)]
Merge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Pull arm64 fixes from Will Deacon:
"Three arm64 fixes for -rc4.
One of them is just a trivial documentation fix, whereas the other two
address a warning in the kexec code and a crash in ftrace on systems
implementing BTI.
The latter patch has a couple of ugly ifdefs which Mark plans to clean
up separately, but as-is the patch is straightforward for backporting
to stable kernels.
Summary:
- Add missing BTI landing instructions to the ftrace*_caller
trampolines
- Fix kexec() WARN when DEBUG_VIRTUAL is enabled
- Fix PAC documentation by removing stale references to compiler
flags"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: ftrace: add missing BTIs
arm64: kexec: use __pa_symbol(empty_zero_page)
arm64: update PAC description for kernel
Linus Torvalds [Fri, 3 Dec 2021 18:44:16 +0000 (10:44 -0800)]
Merge branch 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux
Pull i2c fixes from Wolfram Sang:
"I2C has another set of driver bugfixes, mostly for the stm32f7 driver"
* 'i2c/for-current' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
i2c: rk3x: Handle a spurious start completion interrupt flag
i2c: stm32f7: use proper DMAENGINE API for termination
i2c: stm32f7: stop dma transfer in case of NACK
i2c: stm32f7: recover the bus on access timeout
i2c: stm32f7: flush TX FIFO upon transfer errors
i2c: cbus-gpio: set atomic transfer callback
Linus Torvalds [Fri, 3 Dec 2021 18:38:45 +0000 (10:38 -0800)]
Merge tag 'libata-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata
Pull libata fixes from Damien Le Moal:
"Two sparse warning fixes and a couple of patches to fix an issue with
sata_fsl driver module removal:
- A couple of patches to avoid sparse warnings in libata-sata and in
the pata_falcon driver (from Yang and Finn).
- A couple of sata_fsl driver patches fixing IRQ free and proc
unregister on module removal (from Baokun)"
* tag 'libata-5.16-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata:
ata: replace snprintf in show functions with sysfs_emit
sata_fsl: fix warning in remove_proc_entry when rmmod sata_fsl
sata_fsl: fix UAF in sata_fsl_port_stop when rmmod sata_fsl
pata_falcon: Avoid type warnings from sparse
Shyam Prasad N [Thu, 2 Dec 2021 07:46:54 +0000 (07:46 +0000)]
cifs: avoid use of dstaddr as key for fscache client cookie
server->dstaddr can change when the DNS mapping for the
server hostname changes. But conn_id is a u64 counter
that is incremented each time a new TCP connection
is setup. So use only that as a key.
Signed-off-by: Shyam Prasad N <sprasad@microsoft.com> Reviewed-by: Paulo Alcantara (SUSE) <pc@cjr.nz> Signed-off-by: Steve French <stfrench@microsoft.com>
Shyam Prasad N [Thu, 2 Dec 2021 07:30:00 +0000 (07:30 +0000)]
cifs: add server conn_id to fscache client cookie
The fscache client cookie uses the server address
(and port) as the cookie key. This is a problem when
nosharesock is used. Two different connections will
use duplicate cookies. Avoid this by adding
server->conn_id to the key, so that it's guaranteed
that cookie will not be duplicated.
Also, for secondary channels of a session, copy the
fscache pointer from the primary channel. The primary
channel is guaranteed not to go away as long as secondary
channels are in use. Also addresses minor problem found
by kernel test robot.
Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Shyam Prasad N <sprasad@microsoft.com> Reviewed-by: Paulo Alcantara (SUSE) <pc@cjr.nz> Signed-off-by: Steve French <stfrench@microsoft.com>
Shyam Prasad N [Thu, 2 Dec 2021 07:14:42 +0000 (07:14 +0000)]
cifs: wait for tcon resource_id before getting fscache super
The logic for initializing tcon->resource_id is done inside
cifs_root_iget. fscache super cookie relies on this for aux
data. So we need to push the fscache initialization to this
later point during mount.
Signed-off-by: Shyam Prasad N <sprasad@microsoft.com> Reviewed-by: Paulo Alcantara (SUSE) <pc@cjr.nz> Signed-off-by: Steve French <stfrench@microsoft.com>
Paulo Alcantara [Thu, 2 Dec 2021 18:29:35 +0000 (15:29 -0300)]
cifs: fix missed refcounting of ipc tcon
Fix missed refcounting of IPC tcon used for getting domain-based DFS
root referrals. We want to keep it alive as long as mount is active
and can be refreshed. For standalone DFS root referrals it wouldn't
be a problem as the client ends up having an IPC tcon for both mount
and cache.
Fixes: 631f7087f623 ("cifs: support nested dfs links over reconnect") Signed-off-by: Paulo Alcantara (SUSE) <pc@cjr.nz> Reviewed-by: Enzo Matsumiya <ematsumiya@suse.de> Signed-off-by: Steve French <stfrench@microsoft.com>
In the native case, PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is the
trampoline stack. But XEN pv doesn't use trampoline stack, so
PER_CPU_VAR(cpu_tss_rw + TSS_sp0) is also the kernel stack.
In that case, source and destination stacks are identical, which means
that reusing swapgs_restore_regs_and_return_to_usermode() in XEN pv
would cause %rsp to move up to the top of the kernel stack and leave the
IRET frame below %rsp.
This is dangerous as it can be corrupted if #NMI / #MC hit as either of
these events occurring in the middle of the stack pushing would clobber
data on the (original) stack.
And, with XEN pv, swapgs_restore_regs_and_return_to_usermode() pushing
the IRET frame on to the original address is useless and error-prone
when there is any future attempt to modify the code.
[ bp: Massage commit message. ]
Fixes: a70e6a0fb8d2 ("x86/entry/64: Use a per-CPU trampoline stack for IDT entries") Signed-off-by: Lai Jiangshan <laijs@linux.alibaba.com> Signed-off-by: Borislav Petkov <bp@suse.de> Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Link: https://lkml.kernel.org/r/20211126101209.8613-4-jiangshanlai@gmail.com
removed a CR3 write in the faulting path of load_gs_index().
But the path's FENCE_SWAPGS_USER_ENTRY has no fence operation if PTI is
enabled, see spectre_v1_select_mitigation().
Rather, it depended on the serializing CR3 write of SWITCH_TO_KERNEL_CR3
and since it got removed, add a FENCE_SWAPGS_KERNEL_ENTRY call to make
sure speculation is blocked.
Linus Torvalds [Wed, 1 Dec 2021 18:06:14 +0000 (10:06 -0800)]
fget: check that the fd still exists after getting a ref to it
Jann Horn points out that there is another possible race wrt Unix domain
socket garbage collection, somewhat reminiscent of the one fixed in
commit 7c5fb3d5e93e ("af_unix: fix garbage collect vs MSG_PEEK").
See the extended comment about the garbage collection requirements added
to unix_peek_fds() by that commit for details.
The race comes from how we can locklessly look up a file descriptor just
as it is in the process of being closed, and with the right artificial
timing (Jann added a few strategic 'mdelay(500)' calls to do that), the
Unix domain socket garbage collector could see the reference count
decrement of the close() happen before fget() took its reference to the
file and the file was attached onto a new file descriptor.
This is all (intentionally) correct on the 'struct file *' side, with
RCU lookups and lockless reference counting very much part of the
design. Getting that reference count out of order isn't a problem per
se.
But the garbage collector can get confused by seeing this situation of
having seen a file not having any remaining external references and then
seeing it being attached to an fd.
In commit 7c5fb3d5e93e ("af_unix: fix garbage collect vs MSG_PEEK") the
fix was to serialize the file descriptor install with the garbage
collector by taking and releasing the unix_gc_lock.
That's not really an option here, but since this all happens when we are
in the process of looking up a file descriptor, we can instead simply
just re-check that the file hasn't been closed in the meantime, and just
re-do the lookup if we raced with a concurrent close() of the same file
descriptor.
projects / kernel.git / log