Roberto Sassu [Fri, 7 Jun 2013 10:16:29 +0000 (12:16 +0200)]
ima: new templates management mechanism
The original 'ima' template is fixed length, containing the filedata hash
and pathname. The filedata hash is limited to 20 bytes (md5/sha1). The
pathname is a null terminated string, limited to 255 characters. To
overcome these limitations and to add additional file metadata, it is
necessary to extend the current version of IMA by defining additional
templates.
The main reason to introduce this feature is that, each time a new
template is defined, the functions that generate and display the
measurement list would include the code for handling a new format and,
thus, would significantly grow over time.
This patch set solves this problem by separating the template management
from the remaining IMA code. The core of this solution is the definition
of two new data structures: a template descriptor, to determine which
information should be included in the measurement list, and a template
field, to generate and display data of a given type.
To define a new template field, developers define the field identifier
and implement two functions, init() and show(), respectively to generate
and display measurement entries. Initially, this patch set defines the
following template fields (support for additional data types will be
added later):
- 'd': the digest of the event (i.e. the digest of a measured file),
calculated with the SHA1 or MD5 hash algorithm;
- 'n': the name of the event (i.e. the file name), with size up to
255 bytes;
- 'd-ng': the digest of the event, calculated with an arbitrary hash
algorithm (field format: [<hash algo>:]digest, where the digest
prefix is shown only if the hash algorithm is not SHA1 or MD5);
- 'n-ng': the name of the event, without size limitations.
Defining a new template descriptor requires specifying the template format,
a string of field identifiers separated by the '|' character. This patch
set defines the following template descriptors:
- "ima": its format is 'd|n';
- "ima-ng" (default): its format is 'd-ng|n-ng'
Further details about the new template architecture can be found in
Documentation/security/IMA-templates.txt.
Changelog:
- don't defer calling ima_init_template() - Mimi
- don't define ima_lookup_template_desc() until used - Mimi
- squashed with documentation patch - Mimi
Roberto Sassu [Fri, 7 Jun 2013 10:16:28 +0000 (12:16 +0200)]
ima: define new function ima_alloc_init_template() to API
Instead of allocating and initializing the template entry from multiple
places (eg. boot aggregate, violation, and regular measurements), this
patch defines a new function called ima_alloc_init_template(). The new
function allocates and initializes the measurement entry with the inode
digest and the filename.
In respect to the current behavior, it truncates the file name passed
in the 'filename' argument if the latter's size is greater than 255 bytes
and the passed file descriptor is NULL.
Changelog:
- initialize 'hash' variable for non TPM case - Mimi
- conform to expectation for 'iint' to be defined as a pointer. - Mimi
- add missing 'file' dependency for recalculating file hash. - Mimi
Roberto Sassu [Fri, 7 Jun 2013 10:16:27 +0000 (12:16 +0200)]
ima: pass the filename argument up to ima_add_template_entry()
Pass the filename argument to ima_add_template_entry() in order to
eliminate a dependency on template specific data (third argument of
integrity_audit_msg).
This change is required because, with the new template management
mechanism, the generation of a new measurement entry will be performed
by new specific functions (introduced in next patches) and the current IMA
code will not be aware anymore of how data is stored in the entry payload.
Roberto Sassu [Fri, 7 Jun 2013 10:16:26 +0000 (12:16 +0200)]
ima: pass the file descriptor to ima_add_violation()
Pass the file descriptor instead of the inode to ima_add_violation(),
to make the latter consistent with ima_store_measurement() in
preparation for the new template architecture.
Mimi Zohar [Tue, 12 Mar 2013 00:29:47 +0000 (20:29 -0400)]
ima: differentiate between template hash and file data hash sizes
The TPM v1.2 limits the template hash size to 20 bytes. This
patch differentiates between the template hash size, as defined
in the ima_template_entry, and the file data hash size, as
defined in the ima_template_data. Subsequent patches add support
for different file data hash algorithms.
Change log:
- hash digest definition in ima_store_template() should be TPM_DIGEST_SIZE
For each inode in the IMA policy, an iint is allocated. To support
larger hash digests, the iint digest size changed from 20 bytes to
the maximum supported hash digest size. Instead of allocating the
maximum size, which most likely is not needed, this patch dynamically
allocates the needed hash storage.
All files on the filesystem, currently, are hashed using the same hash
algorithm. In preparation for files from different packages being
signed using different hash algorithms, this patch adds support for
reading the signature hash algorithm from the 'security.ima' extended
attribute and calculates the appropriate file data hash based on it.
Changelog:
- fix scripts Lindent and checkpatch msgs - Mimi
- fix md5 support for older version, which occupied 20 bytes in the
xattr, not the expected 16 bytes. Fix the comparison to compare
only the first 16 bytes.
ima: provide support for arbitrary hash algorithms
In preparation of supporting more hash algorithms with larger hash sizes
needed for signature verification, this patch replaces the 20 byte sized
digest, with a more flexible structure. The new structure includes the
hash algorithm, digest size, and digest.
Changelog:
- recalculate filedata hash for the measurement list, if the signature
hash digest size is greater than 20 bytes.
- use generic HASH_ALGO_
- make ima_calc_file_hash static
- scripts lindent and checkpatch fixes
Dmitry Kasatkin [Mon, 6 May 2013 12:40:01 +0000 (15:40 +0300)]
crypto: provide single place for hash algo information
This patch provides a single place for information about hash algorithms,
such as hash sizes and kernel driver names, which will be used by IMA
and the public key code.
Changelog:
- Fix sparse and checkpatch warnings
- Move hash algo enums to uapi for userspace signing functions.
Jason Gunthorpe [Sun, 6 Oct 2013 18:43:36 +0000 (12:43 -0600)]
tpm: Add support for Atmel I2C TPMs
This is based on the work of Teddy Reed <teddy@prosauce.org> published
on GitHub:
https://github.com/theopolis/tpm-i2c-atmel.git 34894b988b67e0ae55088d6388e77b0dbf10c07d
That driver was never merged, I have taken it as a starting port,
forward ported, tested and revised the driver:
- Make it broadly textually similar to the Infineon and Nuvoton I2C
driver
- Place everything in a format suitable for mainline inclusion
- Use high level I2C functions i2c_master_send and
i2c_master_recv for data xfer
- Use the timeout system from the core code, by faking out a status
register
- Only I2C transfer the number of bytes in the reply, not a fixed
message size.
- checkpatch cleanups
- Testing on ARM Kirkwood, with this device tree, using a
AT97SC3204T-X1A180
tpm@29 {
compatible = "atmel,at97sc3204t";
reg = <0x29>;
};
Signed-off-by: Teddy Reed <teddy@prosauce.org>
[jgg: revised and tested] Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
[phuewe: minor whitespace changes]
Jason Gunthorpe [Sun, 6 Oct 2013 18:43:13 +0000 (12:43 -0600)]
tpm: Add support for the Nuvoton NPCT501 I2C TPM
This chip is/was also branded as a Winbond WPCT301.
Originally written by Dan Morav <dmorav@nuvoton.com> and posted to LKML:
https://lkml.org/lkml/2011/9/7/206
The original posting was not merged, I have taken it as a
starting point, forward ported, tested and revised the driver:
- Rework interrupt handling to work properly with level triggered
interrupts. The old version just locked up.
- Synchronize various items with Peter Huewe's Infineon driver:
* Add durations/timeouts sysfs calls
* Remove I2C device auto-detection
* Don't fiddle with chip->release
* Call tpm_dev_vendor_release in the probe error path
* Use MODULE_DEVICE_TABLE for the I2C ids
* Provide OF compatible strings for DT support
* Use SIMPLE_DEV_PM_OPS
* Use module_i2c_driver
- checkpatch cleanups
- Testing on ARM Kirkwood with GPIO interrupts, with this device tree:
tpm@57 {
compatible = "nuvoton,npct501";
reg = <0x57>;
interrupt-parent = <&gpio1>;
interrupts = <6 IRQ_TYPE_LEVEL_LOW>;
};
Signed-off-by: Dan Morav <dmorav@nuvoton.com>
[jgg: revised and tested] Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
[phuewe: minor whitespace changes, fixed module name in kconfig]
Jason Gunthorpe [Thu, 3 Oct 2013 04:51:09 +0000 (22:51 -0600)]
tpm: Merge the tpm-bios module with tpm.o
Now that we can have multiple .c files in the tpm module there is
no reason for tpm-bios.
tpm-bios exported several functions: tpm_bios_log_setup,
tpm_bios_log_teardown, tpm_add_ppi, and tpm_remove_ppi.
They are only used by tpm, and if tpm-bios is built then
tpm will unconditionally require them. Further, tpm-bios does
nothing on its own, it has no module_init function.
Thus we remove the exports and merge the modules to simplify things.
The Makefile conditions are changed slightly to match the code,
tpm_ppi is always required if CONFIG_ACPI is set.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Jason Gunthorpe [Sun, 6 Oct 2013 18:00:51 +0000 (12:00 -0600)]
tpm: Rename tpm.c to tpm-interface.c
This is preparation for making the tpm module multi-file. kbuild does
not like having a .c file with the same name as a module. We wish to
keep the tpm module name so that userspace doesn't see this change.
tpm-interface.c is chosen because the next several commits in the series
migrate items into tpm-sysfs.c, tpm-dev.c and tpm-class.c. All that will
be left is tpm command processing and interfacing code.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Peter Huewe [Mon, 21 Oct 2013 22:29:14 +0000 (00:29 +0200)]
tpm: cleanup checkpatch warnings
before we rename the file it might be a good idea to cleanup the long
persisting checkpatch warnings.
Since everything is really trivial, splitting the patch up would only
result in noise.
For the interested reader - here the checkpatch warnings:
(regrouped for easer readability)
WARNING: please, no space before tabs
+ * @chip_num: ^Itpm idx # or ANY$
+ * @res_buf: ^ITPM_PCR value$
+ * ^I^Isize of res_buf is 20 bytes (or NULL if you don't care)$
+ * @chip_num: ^Itpm idx # or AN&$
+ * @hash: ^Ihash value used to extend pcr value$
ERROR: code indent should use tabs where possible
+^I TPM_ORD_CONTINUE_SELFTEST);$
WARNING: line over 80 characters
+static bool wait_for_tpm_stat_cond(struct tpm_chip *chip, u8 mask, bool check_cancel,
ERROR: trailing whitespace
+ * Called from tpm_<specific>.c probe function only for devices $
Jason Gunthorpe [Wed, 18 Sep 2013 00:19:44 +0000 (18:19 -0600)]
tpm: Remove tpm_show_caps_1_2
The version of the TPM should not depend on the bus it is connected
through. 1.1, 1.2 and soon 2.0 TPMS will be all be able to use the
same bus interfaces.
Make tpm_show_caps try the 1.2 capability first. If that fails then
fall back to the 1.1 capability. This effectively auto-detects what
interface the TPM supports at run-time.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Reviewed-by: Joel Schopp <jschopp@linux.vnet.ibm.com> Reviewed-by: Peter Huewe <peterhuewe@gmx.de> Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Jason Gunthorpe [Sun, 22 Sep 2013 22:03:28 +0000 (16:03 -0600)]
tpm: st33: Remove chip->data_buffer access from this driver
For some reason this driver thinks that chip->data_buffer needs
to be set before it can call tpm_pm_*. This is not true. data_buffer
is used only by /dev/tpmX, which is why it is managed exclusively
by the fops functions.
Cc: Mathias Leblanc <mathias.leblanc@st.com> Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Reviewed-by: Joel Schopp <jschopp@linux.vnet.ibm.com>
Jason Gunthorpe [Sat, 5 Oct 2013 17:40:19 +0000 (11:40 -0600)]
tpm: Remove redundant dev_set_drvdata
TPM drivers should not call dev_set_drvdata (or aliases), only the core
code is allowed to call dev_set_drvdata, and it does it during
tpm_register_hardware.
These extra sets are harmless, but are an anti-pattern that many drivers
have copied.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Reviewed-by: Joel Schopp <jschopp@linux.vnet.ibm.com> Reviewed-by: Peter Huewe <peterhuewe@gmx.de> Acked-by: Ashley Lai <adlai@linux.vnet.ibm.com> Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Jason Gunthorpe [Sat, 5 Oct 2013 17:38:02 +0000 (11:38 -0600)]
tpm: Use container_of to locate the tpm_chip in tpm_open
misc_open sets the file->private_date to the misc_dev when calling
open. We can use container_of to go from the misc_dev back to the
tpm_chip.
Future clean ups will move tpm_open into a new file and this change
means we do not have to export the tpm_chip list.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Reviewed-by: Joel Schopp <jschopp@linux.vnet.ibm.com> Reviewed-by: Peter Huewe <peterhuewe@gmx.de> Acked-by: Ashley Lai <adlai@linux.vnet.ibm.com> Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Jason Gunthorpe [Sun, 22 Sep 2013 20:19:18 +0000 (14:19 -0600)]
tpm: Store devname in the tpm_chip
Just put the memory directly in the chip structure, rather than
in a 2nd dedicated kmalloc.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Reviewed-by: Joel Schopp <jschopp@linux.vnet.ibm.com> Reviewed-by: Peter Huewe <peterhuewe@gmx.de> Signed-off-by: Peter Huewe <peterhuewe@gmx.de> Acked-by: Ashley Lai <adlai@linux.vnet.ibm.com>
Jason Gunthorpe [Sat, 14 Sep 2013 23:36:29 +0000 (17:36 -0600)]
tpm atmel: Call request_region with the correct base
Commit e0dd03caf20d040a0a86 ("tpm: return chip from
tpm_register_hardware") changed the code path here so that
ateml_get_base_addr no longer directly altered the tpm_vendor_specific
structure, and instead placed the base address on the stack.
The commit missed updating the request_region call, which would have
resulted in request_region being called with 0 as the base address.
I don't know if request_region(0, ..) will fail, if so the
driver has been broken since 2006 and we should remove it
from the tree as it has no users.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Reviewed-by: Joel Schopp <jschopp@linux.vnet.ibm.com> Reviewed-by: Peter Huewe <peterhuewe@gmx.de> Signed-off-by: Peter Huewe <peterhuewe@gmx.de>
Jason Gunthorpe [Sat, 14 Sep 2013 22:57:58 +0000 (16:57 -0600)]
tpm: ibmvtpm: Use %zd formatting for size_t format arguments
This suppresses compile warnings on 32 bit builds.
Signed-off-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Reviewed-by: Joel Schopp <jschopp@linux.vnet.ibm.com> Reviewed-by: Peter Huewe <peterhuewe@gmx.de> Signed-off-by: Peter Huewe <peterhuewe@gmx.de> Acked-by: Ashley Lai <adlai@linux.vnet.ibm.com>
Peter Huewe [Tue, 22 Oct 2013 17:28:30 +0000 (19:28 +0200)]
tpm: MAINTAINERS: Add myself as tpm maintainer
Since I'm actively maintaining the tpm subsystem for a few months now,
it's time to step up and be an official maintainer for the tpm subsystem,
atleast until I hear something different from my company.
The maintaining is done solely in my private time, out of private interest.
Speaking only on behalf of myself, trying to be as vendor neutral as possible.
Paul Moore [Thu, 26 Sep 2013 21:00:46 +0000 (17:00 -0400)]
selinux: correct locking in selinux_netlbl_socket_connect)
The SELinux/NetLabel glue code has a locking bug that affects systems
with NetLabel enabled, see the kernel error message below. This patch
corrects this problem by converting the bottom half socket lock to a
more conventional, and correct for this call-path, lock_sock() call.
Mimi Zohar [Tue, 20 Aug 2013 18:36:27 +0000 (14:36 -0400)]
KEYS: verify a certificate is signed by a 'trusted' key
Only public keys, with certificates signed by an existing
'trusted' key on the system trusted keyring, should be added
to a trusted keyring. This patch adds support for verifying
a certificate's signature.
This is derived from David Howells pkcs7_request_asymmetric_key() patch.
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com> Signed-off-by: David Howells <dhowells@redhat.com>
Mimi Zohar [Tue, 20 Aug 2013 18:36:26 +0000 (14:36 -0400)]
KEYS: Make the system 'trusted' keyring viewable by userspace
Give the root user the ability to read the system keyring and put read
permission on the trusted keys added during boot. The latter is actually more
theoretical than real for the moment as asymmetric keys do not currently
provide a read operation.
Signed-off-by: Mimi Zohar <zohar@us.ibm.com> Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Wed, 4 Sep 2013 18:28:03 +0000 (19:28 +0100)]
KEYS: Set the asymmetric-key type default search method
The keyring expansion patches introduces a new search method by which
key_search() attempts to walk directly to the key that has exactly the same
description as the requested one.
However, this causes inexact matching of asymmetric keys to fail. The
solution to this is to select iterative rather than direct search as the
default search type for asymmetric keys.
As an example, the kernel might have a key like this:
David Howells [Fri, 30 Aug 2013 15:07:37 +0000 (16:07 +0100)]
KEYS: Add a 'trusted' flag and a 'trusted only' flag
Add KEY_FLAG_TRUSTED to indicate that a key either comes from a trusted source
or had a cryptographic signature chain that led back to a trusted key the
kernel already possessed.
Add KEY_FLAGS_TRUSTED_ONLY to indicate that a keyring will only accept links to
keys marked with KEY_FLAGS_TRUSTED.
Signed-off-by: David Howells <dhowells@redhat.com> Reviewed-by: Kees Cook <keescook@chromium.org>
David Howells [Tue, 18 Jun 2013 16:40:44 +0000 (17:40 +0100)]
X.509: Remove certificate date checks
Remove the certificate date checks that are performed when a certificate is
parsed. There are two checks: a valid from and a valid to. The first check is
causing a lot of problems with system clocks that don't keep good time and the
second places an implicit expiry date upon the kernel when used for module
signing, so do we really need them?
Signed-off-by: David Howells <dhowells@redhat.com>
cc: David Woodhouse <dwmw2@infradead.org>
cc: Rusty Russell <rusty@rustcorp.com.au>
cc: Josh Boyer <jwboyer@redhat.com>
cc: Alexander Holler <holler@ahsoftware.de>
cc: stable@vger.kernel.org
David Howells [Fri, 30 Aug 2013 15:18:02 +0000 (16:18 +0100)]
X.509: Embed public_key_signature struct and create filler function
Embed a public_key_signature struct in struct x509_certificate, eliminating
now unnecessary fields, and split x509_check_signature() to create a filler
function for it that attaches a digest of the signed data and an MPI that
represents the signature data. x509_free_certificate() is then modified to
deal with these.
Whilst we're at it, export both x509_check_signature() and the new
x509_get_sig_params().
David Howells [Fri, 30 Aug 2013 15:15:37 +0000 (16:15 +0100)]
KEYS: Store public key algo ID in public_key_signature struct
Store public key algorithm ID in public_key_signature struct for reference
purposes. This allows a public_key_signature struct to be embedded in
struct x509_certificate and other places more easily.
David Howells [Fri, 30 Aug 2013 15:15:30 +0000 (16:15 +0100)]
KEYS: Split public_key_verify_signature() and make available
Modify public_key_verify_signature() so that it now takes a public_key struct
rather than a key struct and supply a wrapper that takes a key struct. The
wrapper is then used by the asymmetric key subtype and the modified function is
used by X.509 self-signature checking and can be used by other things also.
David Howells [Fri, 30 Aug 2013 15:15:24 +0000 (16:15 +0100)]
KEYS: Store public key algo ID in public_key struct
Store public key algo ID in public_key struct for reference purposes. This
allows it to be removed from the x509_certificate struct and used to find a
default in public_key_verify_signature().
David Howells [Fri, 30 Aug 2013 15:15:18 +0000 (16:15 +0100)]
KEYS: Move the algorithm pointer array from x509 to public_key.c
Move the public-key algorithm pointer array from x509_public_key.c to
public_key.c as it isn't X.509 specific.
Note that to make this configure correctly, the public key part must be
dependent on the RSA module rather than the other way round. This needs a
further patch to make use of the crypto module loading stuff rather than using
a fixed table.
David Howells [Fri, 30 Aug 2013 15:15:10 +0000 (16:15 +0100)]
KEYS: Rename public key parameter name arrays
Rename the arrays of public key parameters (public key algorithm names, hash
algorithm names and ID type names) so that the array name ends in "_name".
David Howells [Tue, 24 Sep 2013 09:35:19 +0000 (10:35 +0100)]
KEYS: Add per-user_namespace registers for persistent per-UID kerberos caches
Add support for per-user_namespace registers of persistent per-UID kerberos
caches held within the kernel.
This allows the kerberos cache to be retained beyond the life of all a user's
processes so that the user's cron jobs can work.
The kerberos cache is envisioned as a keyring/key tree looking something like:
struct user_namespace
\___ .krb_cache keyring - The register
\___ _krb.0 keyring - Root's Kerberos cache
\___ _krb.5000 keyring - User 5000's Kerberos cache
\___ _krb.5001 keyring - User 5001's Kerberos cache
\___ tkt785 big_key - A ccache blob
\___ tkt12345 big_key - Another ccache blob
Or possibly:
struct user_namespace
\___ .krb_cache keyring - The register
\___ _krb.0 keyring - Root's Kerberos cache
\___ _krb.5000 keyring - User 5000's Kerberos cache
\___ _krb.5001 keyring - User 5001's Kerberos cache
\___ tkt785 keyring - A ccache
\___ krbtgt/REDHAT.COM@REDHAT.COM big_key
\___ http/REDHAT.COM@REDHAT.COM user
\___ afs/REDHAT.COM@REDHAT.COM user
\___ nfs/REDHAT.COM@REDHAT.COM user
\___ krbtgt/KERNEL.ORG@KERNEL.ORG big_key
\___ http/KERNEL.ORG@KERNEL.ORG big_key
What goes into a particular Kerberos cache is entirely up to userspace. Kernel
support is limited to giving you the Kerberos cache keyring that you want.
The uid is -1 or the user's own UID for the user's own cache or the uid of some
other user's cache (requires CAP_SETUID). This permits rpc.gssd or whatever to
mess with the cache.
The cache returned is a keyring named "_krb.<uid>" that the possessor can read,
search, clear, invalidate, unlink from and add links to. Active LSMs get a
chance to rule on whether the caller is permitted to make a link.
Each uid's cache keyring is created when it first accessed and is given a
timeout that is extended each time this function is called so that the keyring
goes away after a while. The timeout is configurable by sysctl but defaults to
three days.
Each user_namespace struct gets a lazily-created keyring that serves as the
register. The cache keyrings are added to it. This means that standard key
search and garbage collection facilities are available.
The user_namespace struct's register goes away when it does and anything left
in it is then automatically gc'd.
Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Simo Sorce <simo@redhat.com>
cc: Serge E. Hallyn <serge.hallyn@ubuntu.com>
cc: Eric W. Biederman <ebiederm@xmission.com>
David Howells [Tue, 24 Sep 2013 09:35:18 +0000 (10:35 +0100)]
KEYS: Expand the capacity of a keyring
Expand the capacity of a keyring to be able to hold a lot more keys by using
the previously added associative array implementation. Currently the maximum
capacity is:
which, on a 64-bit system, is a little more 500. However, since this is being
used for the NFS uid mapper, we need more than that. The new implementation
gives us effectively unlimited capacity.
With some alterations, the keyutils testsuite runs successfully to completion
after this patch is applied. The alterations are because (a) keyrings that
are simply added to no longer appear ordered and (b) some of the errors have
changed a bit.
Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Tue, 24 Sep 2013 09:35:17 +0000 (10:35 +0100)]
Add a generic associative array implementation.
Add a generic associative array implementation that can be used as the
container for keyrings, thereby massively increasing the capacity available
whilst also speeding up searching in keyrings that contain a lot of keys.
This may also be useful in FS-Cache for tracking cookies.
Documentation is added into Documentation/associative_array.txt
Some of the properties of the implementation are:
(1) Objects are opaque pointers. The implementation does not care where they
point (if anywhere) or what they point to (if anything).
[!] NOTE: Pointers to objects _must_ be zero in the two least significant
bits.
(2) Objects do not need to contain linkage blocks for use by the array. This
permits an object to be located in multiple arrays simultaneously.
Rather, the array is made up of metadata blocks that point to objects.
(3) Objects are labelled as being one of two types (the type is a bool value).
This information is stored in the array, but has no consequence to the
array itself or its algorithms.
(4) Objects require index keys to locate them within the array.
(5) Index keys must be unique. Inserting an object with the same key as one
already in the array will replace the old object.
(6) Index keys can be of any length and can be of different lengths.
(7) Index keys should encode the length early on, before any variation due to
length is seen.
(8) Index keys can include a hash to scatter objects throughout the array.
(9) The array can iterated over. The objects will not necessarily come out in
key order.
(10) The array can be iterated whilst it is being modified, provided the RCU
readlock is being held by the iterator. Note, however, under these
circumstances, some objects may be seen more than once. If this is a
problem, the iterator should lock against modification. Objects will not
be missed, however, unless deleted.
(11) Objects in the array can be looked up by means of their index key.
(12) Objects can be looked up whilst the array is being modified, provided the
RCU readlock is being held by the thread doing the look up.
The implementation uses a tree of 16-pointer nodes internally that are indexed
on each level by nibbles from the index key. To improve memory efficiency,
shortcuts can be emplaced to skip over what would otherwise be a series of
single-occupancy nodes. Further, nodes pack leaf object pointers into spare
space in the node rather than making an extra branch until as such time an
object needs to be added to a full node.
Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Tue, 24 Sep 2013 09:35:16 +0000 (10:35 +0100)]
KEYS: Search for auth-key by name rather than target key ID
Search for auth-key by name rather than by target key ID as, in a future
patch, we'll by searching directly by index key in preference to iteration
over all keys.
Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Tue, 24 Sep 2013 09:35:15 +0000 (10:35 +0100)]
KEYS: Introduce a search context structure
Search functions pass around a bunch of arguments, each of which gets copied
with each call. Introduce a search context structure to hold these.
Whilst we're at it, create a search flag that indicates whether the search
should be directly to the description or whether it should iterate through all
keys looking for a non-description match.
This will be useful when keyrings use a generic data struct with generic
routines to manage their content as the search terms can just be passed
through to the iterator callback function.
Also, for future use, the data to be supplied to the match function is
separated from the description pointer in the search context. This makes it
clear which is being supplied.
Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Tue, 24 Sep 2013 09:35:15 +0000 (10:35 +0100)]
KEYS: Consolidate the concept of an 'index key' for key access
Consolidate the concept of an 'index key' for accessing keys. The index key
is the search term needed to find a key directly - basically the key type and
the key description. We can add to that the description length.
This will be useful when turning a keyring into an associative array rather
than just a pointer block.
Signed-off-by: David Howells <dhowells@redhat.com>
David Howells [Tue, 24 Sep 2013 09:35:13 +0000 (10:35 +0100)]
KEYS: Skip key state checks when checking for possession
Skip key state checks (invalidation, revocation and expiration) when checking
for possession. Without this, keys that have been marked invalid, revoked
keys and expired keys are not given a possession attribute - which means the
possessor is not granted any possession permits and cannot do anything with
them unless they also have one a user, group or other permit.
This causes failures in the keyutils test suite's revocation and expiration
tests now that commit 96b5c8fea6c0861621051290d705ec2e971963f1 reduced the
initial permissions granted to a key.
The failures are due to accesses to revoked and expired keys being given
EACCES instead of EKEYREVOKED or EKEYEXPIRED.
Signed-off-by: David Howells <dhowells@redhat.com>
Eric Paris [Tue, 10 Sep 2013 13:51:50 +0000 (09:51 -0400)]
security: remove erroneous comment about capabilities.o link ordering
Back when we had half ass LSM stacking we had to link capabilities.o
after bigger LSMs so that on initialization the bigger LSM would
register first and the capabilities module would be the one stacked as
the 'seconday'. Somewhere around 6f0f0fd496333777d53 (back in 2008) we
finally removed the last of the kinda module stacking code but this
comment in the makefile still lives today.
Reported-by: Valdis Kletnieks <Valdis.Kletnieks@vt.edu> Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: James Morris <james.l.morris@oracle.com>
Merge tag 'staging-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging
Pull staging fixes from Greg KH:
"Here are a number of small staging tree and iio driver fixes. Nothing
major, just lots of little things"
* tag 'staging-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging: (34 commits)
iio:buffer_cb: Add missing iio_buffer_init()
iio: Prevent race between IIO chardev opening and IIO device free
iio: fix: Keep a reference to the IIO device for open file descriptors
iio: Stop sampling when the device is removed
iio: Fix crash when scan_bytes is computed with active_scan_mask == NULL
iio: Fix mcp4725 dev-to-indio_dev conversion in suspend/resume
iio: Fix bma180 dev-to-indio_dev conversion in suspend/resume
iio: Fix tmp006 dev-to-indio_dev conversion in suspend/resume
iio: iio_device_add_event_sysfs() bugfix
staging: iio: ade7854-spi: Fix return value
staging:iio:hmc5843: Fix measurement conversion
iio: isl29018: Fix uninitialized value
staging:iio:dummy fix kfifo_buf kconfig dependency issue if kfifo modular and buffer enabled for built in dummy driver.
iio: at91: fix adc_clk overflow
staging: line6: add bounds check in snd_toneport_source_put()
Staging: comedi: Fix dependencies for drivers misclassified as PCI
staging: r8188eu: Adjust RX gain
staging: r8188eu: Fix smatch warning in core/rtw_ieee80211.
staging: r8188eu: Fix smatch error in core/rtw_mlme_ext.c
staging: r8188eu: Fix Smatch off-by-one warning in hal/rtl8188e_hal_init.c
...
Merge tag 'usb-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB fixes from Greg KH:
"Here are a number of small USB fixes for 3.12-rc2.
One is a revert of a EHCI change that isn't quite ready for 3.12.
Others are minor things, gadget fixes, Kconfig fixes, and some quirks
and documentation updates.
All have been in linux-next for a bit"
* tag 'usb-3.12-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
USB: pl2303: distinguish between original and cloned HX chips
USB: Faraday fotg210: fix email addresses
USB: fix typo in usb serial simple driver Kconfig
Revert "USB: EHCI: support running URB giveback in tasklet context"
usb: s3c-hsotg: do not disconnect gadget when receiving ErlySusp intr
usb: s3c-hsotg: fix unregistration function
usb: gadget: f_mass_storage: reset endpoint driver data when disabled
usb: host: fsl-mph-dr-of: Staticize local symbols
usb: gadget: f_eem: Staticize eem_alloc
usb: gadget: f_ecm: Staticize ecm_alloc
usb: phy: omap-usb3: Fix return value
usb: dwc3: gadget: avoid memory leak when failing to allocate all eps
usb: dwc3: remove extcon dependency
usb: gadget: add '__ref' for rndis_config_register() and cdc_config_register()
usb: dwc3: pci: add support for BayTrail
usb: gadget: cdc2: fix conversion to new interface of f_ecm
usb: gadget: fix a bug and a WARN_ON in dummy-hcd
usb: gadget: mv_u3d_core: fix violation of locking discipline in mv_u3d_ep_disable()
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Pull drm fixes from Dave Airlie:
- some small fixes for msm and exynos
- a regression revert affecting nouveau users with old userspace
- intel pageflip deadlock and gpu hang fixes, hsw modesetting hangs
* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux: (22 commits)
Revert "drm: mark context support as a legacy subsystem"
drm/i915: Don't enable the cursor on a disable pipe
drm/i915: do not update cursor in crtc mode set
drm/exynos: fix return value check in lowlevel_buffer_allocate()
drm/exynos: Fix address space warnings in exynos_drm_fbdev.c
drm/exynos: Fix address space warning in exynos_drm_buf.c
drm/exynos: Remove redundant OF dependency
drm/msm: drop unnecessary set_need_resched()
drm/i915: kill set_need_resched
drm/msm: fix potential NULL pointer dereference
drm/i915/dvo: set crtc timings again for panel fixed modes
drm/i915/sdvo: Robustify the dtd<->drm_mode conversions
drm/msm: workaround for missing irq
drm/msm: return -EBUSY if bo still active
drm/msm: fix return value check in ERR_PTR()
drm/msm: fix cmdstream size check
drm/msm: hangcheck harder
drm/msm: handle read vs write fences
drm/i915/sdvo: Fully translate sync flags in the dtd->mode conversion
drm/i915: Use proper print format for debug prints
...
Merge branch 'for-3.12/core' of git://git.kernel.dk/linux-block
Pull block IO fixes from Jens Axboe:
"After merge window, no new stuff this time only a collection of neatly
confined and simple fixes"
* 'for-3.12/core' of git://git.kernel.dk/linux-block:
cfq: explicitly use 64bit divide operation for 64bit arguments
block: Add nr_bios to block_rq_remap tracepoint
If the queue is dying then we only call the rq->end_io callout. This leaves bios setup on the request, because the caller assumes when the blk_execute_rq_nowait/blk_execute_rq call has completed that the rq->bios have been cleaned up.
bio-integrity: Fix use of bs->bio_integrity_pool after free
blkcg: relocate root_blkg setting and clearing
block: Convert kmalloc_node(...GFP_ZERO...) to kzalloc_node(...)
block: trace all devices plug operation
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs
Pull btrfs fixes from Chris Mason:
"These are mostly bug fixes and a two small performance fixes. The
most important of the bunch are Josef's fix for a snapshotting
regression and Mark's update to fix compile problems on arm"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (25 commits)
Btrfs: create the uuid tree on remount rw
btrfs: change extent-same to copy entire argument struct
Btrfs: dir_inode_operations should use btrfs_update_time also
btrfs: Add btrfs: prefix to kernel log output
btrfs: refuse to remount read-write after abort
Btrfs: btrfs_ioctl_default_subvol: Revert back to toplevel subvolume when arg is 0
Btrfs: don't leak transaction in btrfs_sync_file()
Btrfs: add the missing mutex unlock in write_all_supers()
Btrfs: iput inode on allocation failure
Btrfs: remove space_info->reservation_progress
Btrfs: kill delay_iput arg to the wait_ordered functions
Btrfs: fix worst case calculator for space usage
Revert "Btrfs: rework the overcommit logic to be based on the total size"
Btrfs: improve replacing nocow extents
Btrfs: drop dir i_size when adding new names on replay
Btrfs: replay dir_index items before other items
Btrfs: check roots last log commit when checking if an inode has been logged
Btrfs: actually log directory we are fsync()'ing
Btrfs: actually limit the size of delalloc range
Btrfs: allocate the free space by the existed max extent size when ENOSPC
...
Merge tag 'iio-fixes-for-3.12a' of git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio into staging-linus
Jonathan writes:
First round of IIO fixes for 3.12
A series of wrong 'struct dev' assumptions in suspend/resume callbacks
following on from this issue being identified in a new driver review.
One to watch out for in future.
A number of driver specific fixes
1) at91 - fix a overflow in clock rate computation
2) dummy - Kconfig dependency issue
3) isl29018 - uninitialized value
4) hmc5843 - measurement conversion bug introduced by recent cleanup.
5) ade7854-spi - wrong return value.
Some IIO core fixes
1) Wrong value picked up for event code creation for a modified channel
2) A null dereference on failure to initialize a buffer after no buffer has
been in use, when using the available_scan_masks approach.
3) Sampling not stopped when a device is removed. Effects forced removal
such as hot unplugging.
4) Prevent device going away if a chrdev is still open in userspace.
5) Prevent race on chardev opening and device being freed.
6) Add a missing iio_buffer_init in the call back buffer.
These last few are the first part of a set from Lars-Peter Clausen who
has been taking a closer look at our removal paths and buffer handling
than anyone has for quite some time.
Adding the number of bios in a remapped request to 'block_rq_remap'
tracepoint.
Request remapper clones bios in a request to track the completion
status of each bio. So the number of bios can be useful information
for investigation.
Related discussions:
http://www.redhat.com/archives/dm-devel/2013-August/msg00084.html
http://www.redhat.com/archives/dm-devel/2013-September/msg00024.html
Josef Bacik [Sat, 21 Sep 2013 02:33:20 +0000 (22:33 -0400)]
Btrfs: create the uuid tree on remount rw
Users have been complaining of the uuid tree stuff warning that there is no uuid
root when trying to do snapshot operations. This is because if you mount -o ro
we will not create the uuid tree. But then if you mount -o rw,remount we will
still not create it and then any subsequent snapshot/subvol operations you try
to do will fail gloriously. Fix this by creating the uuid_root on remount rw if
it was not already there. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Mark Fasheh [Tue, 17 Sep 2013 22:43:54 +0000 (15:43 -0700)]
btrfs: change extent-same to copy entire argument struct
btrfs_ioctl_file_extent_same() uses __put_user_unaligned() to copy some data
back to it's argument struct. Unfortunately, not all architectures provide
__put_user_unaligned(), so compiles break on them if btrfs is selected.
Instead, just copy the whole struct in / out at the start and end of
operations, respectively.
Signed-off-by: Mark Fasheh <mfasheh@suse.de> Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Guangyu Sun [Mon, 16 Sep 2013 17:42:03 +0000 (10:42 -0700)]
Btrfs: dir_inode_operations should use btrfs_update_time also
Commit 2bc5565286121d2a77ccd728eb3484dff2035b58 (Btrfs: don't update atime on
RO subvolumes) ensures that the access time of an inode is not updated when
the inode lives in a read-only subvolume.
However, if a directory on a read-only subvolume is accessed, the atime is
updated. This results in a write operation to a read-only subvolume. I
believe that access times should never be updated on read-only subvolumes.
Reported-by: Koen De Wit <koen.de.wit@oracle.com> Signed-off-by: Guangyu Sun <guangyu.sun@oracle.com> Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Frank Holton [Fri, 13 Sep 2013 15:46:50 +0000 (11:46 -0400)]
btrfs: Add btrfs: prefix to kernel log output
The kernel log entries for device label %s and device fsid %pU
are missing the btrfs: prefix. Add those here.
Signed-off-by: Frank Holton <fholton@gmail.com> Reviewed-by: David Sterba <dsterba@suse.cz> Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
David Sterba [Fri, 13 Sep 2013 15:41:20 +0000 (17:41 +0200)]
btrfs: refuse to remount read-write after abort
It's still possible to flip the filesystem into RW mode after it's
remounted RO due to an abort. There are lots of places that check for
the superblock error bit and will not write data, but we should not let
the filesystem appear read-write.
Signed-off-by: David Sterba <dsterba@suse.cz> Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Btrfs: btrfs_ioctl_default_subvol: Revert back to toplevel subvolume when arg is 0
This patch makes it possible to set BTRFS_FS_TREE_OBJECTID as the default
subvolume by passing a subvolume id of 0.
Signed-off-by: chandan <chandan@linux.vnet.ibm.com> Reviewed-by: David Sterba <dsterba@suse.cz> Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Btrfs: don't leak transaction in btrfs_sync_file()
In btrfs_sync_file(), if the call to btrfs_log_dentry_safe() returns
a negative error (for e.g. -ENOMEM via btrfs_log_inode()), we would
return without ending/freeing the transaction.
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Stefan Behrens [Wed, 11 Sep 2013 07:59:22 +0000 (09:59 +0200)]
Btrfs: add the missing mutex unlock in write_all_supers()
The BUG() was replaced by btrfs_error() and return -EIO with the
patch "get rid of one BUG() in write_all_supers()", but the missing
mutex_unlock() was overlooked.
The 0-DAY kernel build service from Intel reported the missing
unlock which was found by the coccinelle tool:
fs/btrfs/disk-io.c:3422:2-8: preceding lock on line 3374
Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de> Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Tue, 17 Sep 2013 14:55:51 +0000 (10:55 -0400)]
Btrfs: kill delay_iput arg to the wait_ordered functions
This is a left over of how we used to wait for ordered extents, which was to
grab the inode and then run filemap flush on it. However if we have an ordered
extent then we already are holding a ref on the inode, and we just use
btrfs_start_ordered_extent anyway, so there is no reason to have an extra ref on
the inode to start work on the ordered extent. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Tue, 17 Sep 2013 14:50:06 +0000 (10:50 -0400)]
Btrfs: fix worst case calculator for space usage
Forever ago I made the worst case calculator say that we could potentially split
into 3 blocks for every level on the way down, which isn't right. If we split
we're only going to get two new blocks, the one we originally cow'ed and the new
one we're going to split. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Tue, 17 Sep 2013 14:48:00 +0000 (10:48 -0400)]
Revert "Btrfs: rework the overcommit logic to be based on the total size"
This reverts commit 70afa3998c9baed4186df38988246de1abdab56d. It is causing
performance issues and wasn't actually correct. There were problems with the
way we flushed delalloc and that was the real cause of the early enospc.
Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Thu, 12 Sep 2013 20:58:28 +0000 (16:58 -0400)]
Btrfs: improve replacing nocow extents
Various people have hit a deadlock when running btrfs/011. This is because when
replacing nocow extents we will take the i_mutex to make sure nobody messes with
the file while we are replacing the extent. The problem is we are already
holding a transaction open, which is a locking inversion, so instead we need to
save these inodes we find and then process them outside of the transaction.
Further we can't just lock the inode and assume we are good to go. We need to
lock the extent range and then read back the extent cache for the inode to make
sure the extent really still points at the physical block we want. If it
doesn't we don't have to copy it. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Wed, 11 Sep 2013 18:17:00 +0000 (14:17 -0400)]
Btrfs: drop dir i_size when adding new names on replay
So if we have dir_index items in the log that means we also have the inode item
as well, which means that the inode's i_size is correct. However when we
process dir_index'es we call btrfs_add_link() which will increase the
directory's i_size for the new entry. To fix this we need to just set the dir
items i_size to 0, and then as we find dir_index items we adjust the i_size.
btrfs_add_link() will do it for new entries, and if the entry already exists we
can just add the name_len to the i_size ourselves. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Wed, 11 Sep 2013 15:57:23 +0000 (11:57 -0400)]
Btrfs: replay dir_index items before other items
A user reported a bug where his log would not replay because he was getting
-EEXIST back. This was because he had a file moved into a directory that was
logged. What happens is the file had a lower inode number, and so it is
processed first when replaying the log, and so we add the inode ref in for the
directory it was moved to. But then we process the directories DIR_INDEX item
and try to add the inode ref for that inode and it fails because we already
added it when we replayed the inode. To solve this problem we need to just
process any DIR_INDEX items we have in the log first so this all is taken care
of, and then we can replay the rest of the items. With this patch my reproducer
can remount the file system properly instead of erroring out. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Wed, 11 Sep 2013 13:55:42 +0000 (09:55 -0400)]
Btrfs: check roots last log commit when checking if an inode has been logged
Liu introduced a local copy of the last log commit for an inode to make sure we
actually log an inode even if a log commit has already taken place. In order to
make sure we didn't relog the same inode multiple times he set this local copy
to the current trans when we log the inode, because usually we log the inode and
then sync the log. The exception to this is during rename, we will relog an
inode if the name changed and it is already in the log. The problem with this
is then we go to sync the inode, and our check to see if the inode has already
been logged is tripped and we don't sync the log. To fix this we need to _also_
check against the roots last log commit, because it could be less than what is
in our local copy of the log commit. This fixes a bug where we rename a file
into a directory and then fsync the directory and then on remount the directory
is no longer there. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Wed, 11 Sep 2013 13:36:30 +0000 (09:36 -0400)]
Btrfs: actually log directory we are fsync()'ing
If you just create a directory and then fsync that directory and then pull the
power plug you will come back up and the directory will not be there. That is
because we won't actually create directories if we've logged files inside of
them since they will be created on replay, but in this check we will set our
logged_trans of our current directory if it happens to be a directory, making us
think it doesn't need to be logged. Fix the logic to only do this to parent
directories. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Josef Bacik [Fri, 30 Aug 2013 18:38:49 +0000 (14:38 -0400)]
Btrfs: actually limit the size of delalloc range
So forever we have had this thing to limit the amount of delalloc pages we'll
setup to be written out to 128mb. This is because we have to lock all the pages
in this range, so anything above this gets a bit unweildly, and also without a
limit we'll happily allocate gigantic chunks of disk space. Turns out our check
for this wasn't quite right, we wouldn't actually limit the chunk we wanted to
write out, we'd just stop looking for more space after we went over the limit.
So if you do a giant 20gb dd on my box with lots of ram I could get 2gig
extents. This is fine normally, except when you go to relocate these extents
and we can't find enough space to relocate these moster extents, since we have
to be able to allocate exactly the same sized extent to move it around. So fix
this by actually enforcing the limit. With this patch I'm no longer seeing
giant 1.5gb extents. Thanks,
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
Btrfs: allocate the free space by the existed max extent size when ENOSPC
By the current code, if the requested size is very large, and all the extents
in the free space cache are small, we will waste lots of the cpu time to cut
the requested size in half and search the cache again and again until it gets
down to the size the allocator can return. In fact, we can know the max extent
size in the cache after the first search, so we needn't cut the size in half
repeatedly, and just use the max extent size directly. This way can save
lots of cpu time and make the performance grow up when there are only fragments
in the free space cache.
According to my test, if there are only 4KB free space extents in the fs,
and the total size of those extents are 256MB, we can reduce the execute
time of the following test from 5.4s to 1.4s.
dd if=/dev/zero of=<testfile> bs=1MB count=1 oflag=sync
Changelog v2 -> v3:
- fix the problem that we skip the block group with the space which is
less than we need.
Changelog v1 -> v2:
- address the problem that we return a wrong start position when searching
the free space in a bitmap.
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com> Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com>
projects / linux-beck.git / log