IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
Split out two helpers to make the code more readable and easier to verify
for correctness.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Cc: stable@vger.kernel.org
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Michael L Semon reported that generic/069 runtime increased on v5
superblocks by 100% compared to v4 superblocks. his perf-based
analysis pointed directly at the timestamp updates being done by the
write path in this workload. The append writers are doing 4-byte
writes, so there are lots of timestamp updates occurring.
The thing is, they aren't being triggered by timestamp changes -
they are being triggered by the inode change counter needing to be
updated. That is, every write(2) system call needs to bump the inode
version count, and it does that through the timestamp update
mechanism. Hence for v5 filesystems, test generic/069 is running 3
orders of magnitude more timestmap update transactions on v5
filesystems due to the fact it does a huge number of *4 byte*
write(2) calls.
This isn't a real world scenario we really need to address - anyone
doing such sequential IO should be using fwrite(3), not write(2).
i.e. fwrite(3) buffers the writes in userspace to minimise the
number of write(2) syscalls, and the problem goes away.
However, there is a small change we can make to improve the
situation - removing the expensive lock operation on the change
counter update. All inode version counter changes in XFS occur
under the ip->i_ilock during a transaction, and therefore we
don't actually need the spin lock that provides exclusive access to
it through inc_inode_iversion().
Hence avoid the lock and just open code the increment ourselves when
logging the inode.
Reported-by: Michael L. Semon <mlsemon35@gmail.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
v5 filesystems use 512 byte inodes as a minimum, so read inodes in
clusters that are effectively half the size of a v4 filesystem with
256 byte inodes. For v5 fielsystems, scale the inode cluster size
with the size of the inode so that we keep a constant 32 inodes per
cluster ratio for all inode IO.
This only works if mkfs.xfs sets the inode alignment appropriately
for larger inode clusters, so this functionality is made conditional
on mkfs doing the right thing. xfs_repair needs to know about
the inode alignment changes, too.
Wall time:
create bulkstat find+stat ls -R unlink
v4 237s 161s 173s 201s 299s
v5 235s 163s 205s 31s 356s
patched 234s 160s 182s 29s 317s
System time:
create bulkstat find+stat ls -R unlink
v4 2601s 2490s 1653s 1656s 2960s
v5 2637s 2497s 1681s 20s 3216s
patched 2613s 2451s 1658s 20s 3007s
So, wall time same or down across the board, system time same or
down across the board, and cache hit rates all improve except for
the ls -R case which is a pure cold cache directory read workload
on v5 filesystems...
So, this patch removes most of the performance and CPU usage
differential between v4 and v5 filesystems on traversal related
workloads.
Note: while this patch is currently for v5 filesystems only, there
is no reason it can't be ported back to v4 filesystems. This hasn't
been done here because bringing the code back to v4 requires
forwards and backwards kernel compatibility testing. i.e. to
deterine if older kernels(*) do the right thing with larger inode
alignments but still only using 8k inode cluster sizes. None of this
testing and validation on v4 filesystems has been done, so for the
moment larger inode clusters is limited to v5 superblocks.
(*) a current default config v4 filesystem should mount just fine on
2.6.23 (when lazy-count support was introduced), and so if we change
the alignment emitted by mkfs without a feature bit then we have to
make sure it works properly on all kernels since 2.6.23. And if we
allow it to be changed when the lazy-count bit is not set, then it's
all kernels since v2 logs were introduced that need to be tested for
compatibility...
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Ben Myers <bpm@sgi.com>
xfs_trans_ijoin() activates the inode in a transaction and
also can specify which lock to free when the transaction is
committed or canceled.
xfs_bmap_add_attrfork call locks and adds the lock to the
transaction but also manually removes the lock. Change the
routine to not add the lock to the transaction and manually
remove lock on completion.
While here, clean up the xfs_trans_cancel flags and goto names.
Signed-off-by: Mark Tinguely <tinguely@sgi.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ben Myers <bpm@sgi.com>
This is required to properly calculate the tiling parameters
in userspace.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This is required to properly calculate the tiling parameters
in userspace.
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Should we need to share dma buffers using prime, let's make them prime
aware.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
Add prime exporting and imporing operations to surfaces
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
Also provide a completely dumb dma-buf ops implementation.
Once we have other virtual dma-buf aware devices, we need to provide
something better.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
We're using the ACPI interface to detect whether we're dealing with a Thinkpad
or not. This way we're not loading the thinkpad_acpi module when we're not on
a Thinkpad, but at the same time, we give the opportunity to check for, and
potentially enable, both present and future Thinkpad with mute/micmute LEDs.
At least those running the ALC269 family (269 to 299) of Realtek codecs.
Cc: Alex Hung <alex.hung@canonical.com>
Tested-by: Hui Wang <hui.wang@canonical.com>
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The patch 0ca743a55991: "netfilter: nf_tables: add compatibility
layer for x_tables", leads to the following Smatch
warning: "net/netfilter/nft_compat.c:140 nft_parse_compat()
warn: signedness bug returning '(-34)'"
This nft_parse_compat function returns error codes but the return
type is u8 so the error codes are transformed into small positive
values. The callers don't check the return.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
In commit 41d73ec053d2, sequence number adjustments were moved to a
separate file. Unfortunately, the sequence numbers that are stored
in the nf_ct_seqadj structure are expressed in host byte order. The
necessary ntohl call was removed when the call to adjust_tcp_sequence
was collapsed into nf_ct_seqadj_set. This broke the FTP NAT helper.
Fix it by adding back the byte order conversions.
Reported-by: Dawid Stawiarski <dawid.stawiarski@netart.pl>
Signed-off-by: Phil Oester <kernel@linuxace.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Timestamp are used to store additional syncookie parameters such as sack,
ecn, and wscale. The wscale value we need to encode is the client's
wscale, since we can't recover that later in the session. Next overwrite
the wscale option so the later synproxy_send_client_synack will send
the backend's wscale to the client.
Signed-off-by: Martin Topholm <mph@one.com>
Reviewed-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
When the synproxy_parse_options is called on the client ack the mss
option will not be present. Consequently mss wont be included in the
backend syn packet, which falls back to 536 bytes mss.
Therefore XT_SYNPROXY_OPT_MSS is explicitly flagged when recovering mss
value from cookie.
Signed-off-by: Martin Topholm <mph@one.com>
Reviewed-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
rsnd_dai_connect() is using mod before NULL checking.
This patch fixes it up
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
A lockdep warning is hit when evicting surfaces and reserving the backup
buffer. Since this buffer can only be reserved by the process holding the
surface reservation or by the buffer eviction processes that use tryreserve,
there is no real deadlock here, but there's no other way to silence lockdep
than to use a tryreserve. This means the reservation might fail if the buffer
is about to be evicted or swapped out, but we now have code in place to
handle that reasonably well.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
If no reservation ticket is given to the execbuf reservation utilities,
try reservation with non-blocking semantics.
This is intended for eviction paths that use the execbuf reservation
utilities for convenience rather than for deadlock avoidance.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Jakob Bornecrantz <jakob@vmware.com>
Some BIOS just leak the forcewak bits, which we clean up.
Unfortunately this has been broken in
commit 521198a2e7095c8c7daa8d7d3a76a110c346be6f
Author: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date: Fri Aug 23 16:52:30 2013 +0300
drm/i915: sanitize forcewake registers on reset
To make this work both for resets and for BIOS takeover just add the
forcewake clearing call back to intel_uncore_early_sanitize.
We need to clear the forcewake in early sanitize so that the forcewak
dance in intel_uncore_init (to figure out whether we have mt or legacy
forcewake on ivb) works. That cleanup fits in nicely with the general
topic of early_sanitize to prepare for the very first mmio ops.
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reported-by: Jörg Otte <jrg.otte@gmail.com>
Cc: Jörg Otte <jrg.otte@gmail.com>
References: https://lkml.org/lkml/2013/11/16/40
Cc: stable@vger.kernel.org (for 3.12 only)
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This reverts commit 69f0554ec261fd686ac7fa1c598cc9eb27b83a80.
This patch breaks randconfig on at least the x86-64 architecture, and
most likely on others. There is work underway to support uncompressed
kernels in a generic way, but it looks like it will amount to
rewriting the support from scratch; see the LKML thread in the Link:
for info.
Therefore, revert this change and wait for the fix.
Reported-by: Pavel Roskin <proski@gnu.org>
Cc: Christian Ruppert <christian.ruppert@abilis.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/20131113113418.167b8ffd@IRBT4585
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove mpc85xx_pci_err_remove(...) which is obsolete, this removes the
compiler warning which can be seen when building the driver either
statically or as a module.
Signed-off-by: Johannes Thumshirn <morbidrsa@gmail.com>
Link: https://lkml.kernel.org/r/20131112161901.GA15637@jtlinux
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@men.de>
Signed-off-by: Borislav Petkov <bp@suse.de>
Add drivers/edac/mpc85xx_edac.[ch] to MAINTAINERS file and me as
maintainer.
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@men.de>
Link: https://lkml.kernel.org/r/20131112161901.GA15637@jtlinux
Cc: Doug Thompson <dougthompson@xmission.com>
Cc: Dave Jiang <dave.jiang@gmail.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
All OMAP IP blocks expect LE data, but CPU may operate in BE mode.
Need to use endian neutral functions to read/write h/w registers.
I.e instead of __raw_read[lw] and __raw_write[lw] functions code
need to use read[lw]_relaxed and write[lw]_relaxed functions.
If the first simply reads/writes register, the second will byteswap
it if host operates in BE mode.
Changes are trivial sed like replacement of __raw_xxx functions
with xxx_relaxed variant.
Signed-off-by: Victor Kamensky <victor.kamensky@linaro.org>
Signed-off-by: Taras Kondratiuk <taras.kondratiuk@linaro.org>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
As CSR SiRF is converted to multi platform CLOCK_TICK_RATE is a dummy
value that seems to match the right value is used.
(arch/arm/mach-prima2/include/mach/timex.h which defined CLOCK_TICK_RATE
to 1000000 was removed in commit cf82e0e (ARM: sirf: enable
multiplatform support); marco used the same file.)
To not depend on that dummy value use a local #define instead.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
We changed "buf" from being an array of 6 chars to being a pointer this
sizeof(buf) needs to be updated as well.
Fixes: 2ddb8089a7e5 ('watchdog: pcwd_usb: Use allocated buffer for usb_control_msg')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
In case of error, the function devm_request_and_ioremap() returns NULL
pointer not ERR_PTR(). Fix it by using devm_ioremap_resource() instead
of devm_request_and_ioremap().
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
I just can't find any value in MODULE_ALIAS_MISCDEV(WATCHDOG_MINOR)
and MODULE_ALIAS_MISCDEV(TEMP_MINOR) statements.
Either the device is enumerated and the driver already has a module
alias (e.g. PCI, USB etc.) that will get the right driver loaded
automatically.
Or the device is not enumerated and loading its driver will lead to
more or less intrusive hardware poking. Such hardware poking should be
limited to a bare minimum, so the user should really decide which
drivers should be tried and in what order. Trying them all in
arbitrary order can't do any good.
On top of that, loading that many drivers at once bloats the kernel
log. Also many drivers will stay loaded afterward, bloating the output
of "lsmod" and wasting memory. Some modules (cs5535_mfgpt which gets
loaded as a dependency) can't even be unloaded!
If defining char-major-10-130 is needed then it should happen in
user-space.
Signed-off-by: Jean Delvare <jdelvare@suse.de>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Cc: Stephen Warren <swarren@wwwdotorg.org>
Cc: Mike Frysinger <vapier.adi@gmail.com>
Cc: Wan ZongShun <mcuos.com@gmail.com>
Cc: Ben Dooks <ben-linux@fluff.org>
Cc: Kukjin Kim <kgene.kim@samsung.com>
Cc: Zwane Mwaikambo <zwane@arm.linux.org.uk>
Cc: Jim Cromie <jim.cromie@gmail.com>
timeout_to_regval() returns a valid error code. Might as well use it.
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
usb_control_msg() must use a dma-capable buffer.
This fixes the following error reported by smatch:
drivers/watchdog/pcwd_usb.c:257 usb_pcwd_send_command() error: doing dma on the
stack (buf)
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Driver core clears the driver data to NULL after device_release
or on probe failure, so just remove it from here.
Driver core change:
"device-core: Ensure drvdata = NULL when no driver is bound"
(sha1: 0998d0631001288a5974afc0b2a5f568bcdecb4d)
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
On CSR SiRFprimaII and SiRFatlasVI, the 6th timer can act as a watchdog
timer when the Watchdog mode is enabled.
watchdog occur when TIMER watchdog counter matches the value software
pre-set, when this event occurs, the effect is the same as the system
software reset.
Signed-off-by: Xianglong Du <Xianglong.Du@csr.com>
Signed-off-by: Barry Song <Baohua.Song@csr.com>
Cc: Romain Izard <romain.izard.pro@gmail.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
of_match_ptr() is a macro used to avoid undefined reference error if
CONFIG_OF is used to selectively compile in or out the
data structure. It is defined as follows:
#ifdef CONFIG_OF
#define of_match_ptr(ptr) ptr
#else
#define of_match_ptr(ptr) NULL
#endif
In the case of this series, none of the drivers use CONFIG_OF macro to
compile out the data structure (i.e., the data structure is always
defined).
Hence the use of of_match_ptr() does not make any sense. Thus removing
it to make the code look simpler for readability.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
There seems to be some confusion here which functions return positive
numbers and which return negative error codes.
copy_to_user() returns the number of bytes remaining to be copied but we
want to return -EFAULT.
The rest is just clean up. get_user() actually returns zero on success
and -EFAULT on error so we can preserve the error code. The
timeout_to_regval() function returns -EINVAL on failure, but we can
propogate that back instead of hardcoding -EINVAL ourselves.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
--
I accidently put the devicetree bindings for the MEN A21 watchdog driver in
Documentation/devicetree/bindings/gpio instead of
Documentation/devicetree/bindings/watchdog, this patch addresses this error.
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@men.de>
Acked-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Cc: Rob Herring <rob.herring@calxeda.com>
Cc: Pawel Moll <pawel.moll@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Cc: Rob Landley <rob@landley.net>
This patch adds a watchdog driver for the main hardware watchdog timer
found on MOXA ART SoCs.
The MOXA ART SoC provides one writable timer register, restarting
the hardware once it reaches zero. The register is auto decremented
every APB clock cycle.
Signed-off-by: Jonas Jensen <jonas.jensen@gmail.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Added __user annotation to fix the following sparse warnings.
Also, it makes 'kempld_prescaler' static because it is used
only in this file.
drivers/watchdog/kempld_wdt.c:70:11: warning: symbol 'kempld_prescaler' was not declared. Should it be static?
drivers/watchdog/kempld_wdt.c:364:23: warning: incorrect type in initializer (different address spaces)
drivers/watchdog/kempld_wdt.c:364:23: expected int const [noderef] <asn:1>*register __p
drivers/watchdog/kempld_wdt.c:364:23: got int *<noident>
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Use the wrapper function for retrieving the platform data instead of
accessing dev->platform_data directly.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Enable auto loading by udev when imx2_wdt is compiled as a module.
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Acked-by: Shawn Guo <shawn.guo@linaro.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
This is necessary to make the driver work with platforms using the
common clock framework.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
The dw_wdt only provides PM_SLEEP operations, so convert the driver
to use SIMPLE_DEV_PM_OPS instead of populating the struct manually.
This has the added effect of simplifying the CONFIG_PM ifdefs.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
Using the i2c-eg20t driver and call i2cdetect or probe on the bus,
the driver will print a lot of error messages if there was no ACK
received.
i2cdetect normally print a table with all the available devices. If there
is no device on the address, the table will be empty.
Currently with the i2c-eg20t driver, the table is not visible because
the error messages destroy the table.
Error message: pch_i2c_getack return -71
This patch prevent the driver to print the messages to syslog.
The pch_i2c_wait_for_check_xfer function is the only one who is
calling pch_i2c_getack, so we can delete the function and add the
read to pch_i2c_wait_for_check_xfer.
If no ACK is received, the Message will be printed as a dbg
message.
Fixed print message to be a one liner so we can grep for the
error message.
Tested on Intel Atom E6xx and Eg20t Chipset.
Signed-off-by: Andreas Werner <wernerandy@gmx.de>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
This commit reverts commit 7afbddfae993 ("IB/core: Temporarily disable
create_flow/destroy_flow uverbs"). Since the uverbs extensions
functionality was experimental for v3.12, this patch re-enables the
support for them and flow-steering for v3.13.
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Commit 400dbc96583f ("IB/core: Infrastructure for extensible uverbs
commands") added an infrastructure for extensible uverbs commands
while later commit 436f2ad05a0b ("IB/core: Export ib_create/destroy_flow
through uverbs") exported ib_create_flow()/ib_destroy_flow() functions
using this new infrastructure.
According to the commit 400dbc96583f, the purpose of this
infrastructure is to support passing around provider (eg. hardware)
specific buffers when userspace issue commands to the kernel, so that
it would be possible to extend uverbs (eg. core) buffers independently
from the provider buffers.
But the new kernel command function prototypes were not modified to
take advantage of this extension. This issue was exposed by Roland
Dreier in a previous review[1].
So the following patch is an attempt to a revised extensible command
infrastructure.
This improved extensible command infrastructure distinguish between
core (eg. legacy)'s command/response buffers from provider
(eg. hardware)'s command/response buffers: each extended command
implementing function is given a struct ib_udata to hold core
(eg. uverbs) input and output buffers, and another struct ib_udata to
hold the hw (eg. provider) input and output buffers.
Having those buffers identified separately make it easier to increase
one buffer to support extension without having to add some code to
guess the exact size of each command/response parts: This should make
the extended functions more reliable.
Additionally, instead of relying on command identifier being greater
than IB_USER_VERBS_CMD_THRESHOLD, the proposed infrastructure rely on
unused bits in command field: on the 32 bits provided by command
field, only 6 bits are really needed to encode the identifier of
commands currently supported by the kernel. (Even using only 6 bits
leaves room for about 23 new commands).
So this patch makes use of some high order bits in command field to
store flags, leaving enough room for more command identifiers than one
will ever need (eg. 256).
The new flags are used to specify if the command should be processed
as an extended one or a legacy one. While designing the new command
format, care was taken to make usage of flags itself extensible.
Using high order bits of the commands field ensure that newer
libibverbs on older kernel will properly fail when trying to call
extended commands. On the other hand, older libibverbs on newer kernel
will never be able to issue calls to extended commands.
The extended command header includes the optional response pointer so
that output buffer length and output buffer pointer are located
together in the command, allowing proper parameters checking. This
should make implementing functions easier and safer.
Additionally the extended header ensure 64bits alignment, while making
all sizes multiple of 8 bytes, extending the maximum buffer size:
legacy extended
Maximum command buffer: 256KBytes 1024KBytes (512KBytes + 512KBytes)
Maximum response buffer: 256KBytes 1024KBytes (512KBytes + 512KBytes)
For the purpose of doing proper buffer size accounting, the headers
size are no more taken in account in "in_words".
One of the odds of the current extensible infrastructure, reading
twice the "legacy" command header, is fixed by removing the "legacy"
command header from the extended command header: they are processed as
two different parts of the command: memory is read once and
information are not duplicated: it's making clear that's an extended
command scheme and not a different command scheme.
The proposed scheme will format input (command) and output (response)
buffers this way:
- command:
legacy header +
extended header +
command data (core + hw):
+----------------------------------------+
| flags | 00 00 | command |
| in_words | out_words |
+----------------------------------------+
| response |
| response |
| provider_in_words | provider_out_words |
| padding |
+----------------------------------------+
| |
. <uverbs input> .
. (in_words * 8) .
| |
+----------------------------------------+
| |
. <provider input> .
. (provider_in_words * 8) .
| |
+----------------------------------------+
- response, if present:
+----------------------------------------+
| |
. <uverbs output space> .
. (out_words * 8) .
| |
+----------------------------------------+
| |
. <provider output space> .
. (provider_out_words * 8) .
| |
+----------------------------------------+
The overall design is to ensure that the extensible infrastructure is
itself extensible while begin more reliable with more input and bound
checking.
Note:
The unused field in the extended header would be perfect candidate to
hold the command "comp_mask" (eg. bit field used to handle
compatibility). This was suggested by Roland Dreier in a previous
review[2]. But "comp_mask" field is likely to be present in the uverb
input and/or provider input, likewise for the response, as noted by
Matan Barak[3], so it doesn't make sense to put "comp_mask" in the
header.
[1]:
http://marc.info/?i=CAL1RGDWxmM17W2o_era24A-TTDeKyoL6u3NRu_=t_dhV_ZA9MA@mail.gmail.com
[2]:
http://marc.info/?i=CAL1RGDXJtrc849M6_XNZT5xO1+ybKtLWGq6yg6LhoSsKpsmkYA@mail.gmail.com
[3]:
http://marc.info/?i=525C1149.6000701@mellanox.com
Signed-off-by: Yann Droneaud <ydroneaud@opteya.com>
Link: http://marc.info/?i=cover.1383773832.git.ydroneaud@opteya.com
[ Convert "ret ? ret : 0" to the equivalent "ret". - Roland ]
Signed-off-by: Roland Dreier <roland@purestorage.com>
