If the admin has specified the "noresvport" option for an NFS mount
point, the kernel's NFS client uses an unprivileged source port for
the main NFS transport. The kernel's lockd client should use an
unprivileged port in this case as well.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The standard default security setting for NFS is AUTH_SYS. An NFS
client connects to NFS servers via a privileged source port and a
fixed standard destination port (2049). The client sends raw uid and
gid numbers to identify users making NFS requests, and the server
assumes an appropriate authority on the client has vetted these
values because the source port is privileged.
On Linux, by default in-kernel RPC services use a privileged port in
the range between 650 and 1023 to avoid using source ports of well-
known IP services. Using such a small range limits the number of NFS
mount points and the number of unique NFS servers to which a client
can connect concurrently.
An NFS client can use unprivileged source ports to expand the range of
source port numbers, allowing more concurrent server connections and
more NFS mount points. Servers must explicitly allow NFS connections
from unprivileged ports for this to work.
In the past, bumping the value of the sunrpc.max_resvport sysctl on
the client would permit the NFS client to use unprivileged ports.
Bumping this setting also changes the maximum port number used by
other in-kernel RPC services, some of which still required a port
number less than 1023.
This is exacerbated by the way source port numbers are chosen by the
Linux RPC client, which starts at the top of the range and works
downwards. It means that bumping the maximum means all RPC services
requesting a source port will likely get an unprivileged port instead
of a privileged one.
Changing this setting effects all NFS mount points on a client. A
sysadmin could not selectively choose which mount points would use
non-privileged ports and which could not.
Lastly, this mechanism of expanding the limit on the number of NFS
mount points was entirely undocumented.
To address the need for the NFS client to use a large range of source
ports without interfering with the activity of other in-kernel RPC
services, we introduce a new NFS mount option. This option explicitly
tells only the NFS client to use a non-privileged source port when
communicating with the NFS server for one specific mount point.
This new mount option is called "resvport," like the similar NFS mount
option on FreeBSD and Mac OS X. A sister patch for nfs-utils will be
submitted that documents this new option in nfs(5).
The default setting for this new mount option requires the NFS client
to use a privileged port, as before. Explicitly specifying the
"noresvport" mount option allows the NFS client to use an unprivileged
source port for this mount point when connecting to the NFS server
port.
This mount option is supported only for text-based NFS mounts.
[ Sidebar: it is widely known that security mechanisms based on the
use of privileged source ports are ineffective. However, the NFS
client can combine the use of unprivileged ports with the use of
secure authentication mechanisms, such as Kerberos. This allows a
large number of connections and mount points while ensuring a useful
level of security.
Eventually we may change the default setting for this option
depending on the security flavor used for the mount. For example,
if the mount is using only AUTH_SYS, then the default setting will
be "resvport;" if the mount is using a strong security flavor such
as krb5, the default setting will be "noresvport." ]
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
[Trond.Myklebust@netapp.com: Fixed a bug whereby nfs4_init_client()
was being called with incorrect arguments.]
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Clean up: The nfs_mount() function is not to be used outside of the
NFS client. Move its public declaration to fs/nfs/internal.h.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Somehow, this escaped the previous purge. There should be no need to keep
any extra locks in the XDR callbacks.
The NFS client XDR code only writes into private objects, whereas all reads
of shared objects are confined to fields that do not change, such as
filehandles...
Ditto for lockd, the NFSv2/v3 client mount code, and rpcbind.
The nfsd XDR code may require the BKL, but since it does a synchronous RPC
call from a thread that already holds the lock, that issue is moot.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
When the napi api was changed to separate its 1:1 binding to the net_device
struct, the netif_rx_[prep|schedule|complete] api failed to remove the now
vestigual net_device structure parameter. This patch cleans up that api by
properly removing it..
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When using MSI-X mode, create a completion event queue for each CPU.
Report the number of completion EQs in a new struct mlx4_caps member,
num_comp_vectors, and extend the mlx4_cq_alloc() interface with a
vector parameter so that consumers can specify which completion EQ
should be used to report events for the CQ being created.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch adds sh_mobile_lcdcfb deferred io support for SYS panels.
The LCDC hardware block managed by the sh_mobile_lcdcfb driver supports
RGB or SYS panel configurations. SYS panels come with an external display
controller that is resposible for refreshing the actual LCD panel. RGB
panels are controlled directly by the LCDC and they need to be refreshed
by the LCDC hardware.
In the case of SYS panels we can save some power by configuring the LCDC
hardware block in one-shot mode. In this one-shot mode panel refresh is
managed by software. This works well together with deferred io since it
allows us to stop clocks for most of the time and only enable clocks when
we actually want to trigger an update. When there is no fbdev activity
the clocks are kept stopped which allows us to deep sleep.
The refresh rate in deferred io mode is set using platform data. The same
platform data can also be used to disable deferred io mode.
As with other deferred io frame buffers user space code should use fsync()
on the frame buffer device to trigger an update.
Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Adds the Backward Congestion Notification Address (BCNA) attribute to the
Backward Congestion Notification (BCN) interface for Data Center Bridging
(DCB), which was missing. Receive the BCNA attribute in the ixgbe driver.
The BCNA attribute is for a switch to inform the endstation about the physical
port identification in order to support BCN on aggregated links.
Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com>
Signed-off-by: Eric W Multanen <eric.w.multanen@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Data Center Bridging (DCB) had no way to know if setstate had failed in the
driver. This patch enables dcb netlink code to handle the status for the DCB
setstate interface. Likewise it allows the driver to return a failed status
if MSI-X isn't enabled.
Signed-off-by: Don Skidmore <donald.c.skidmore@intel.com>
Signed-off-by: Eric W Multanen <eric.w.multanen@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Impact: broaden gcc printf format checks for ftrace_printk()
format arguments checking for ftrace_printk() is __printf(1, 2),
not __printf(1, 0).
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This function is used to count how many GPIOs are specified for
a device node.
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
This drive has been tested on ARM9 based SoC - MV86XX.
Signed-off-by: Kwangwoo Lee <kwangwoo.lee@gmail.com>
Acked-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Impact: move the BTS buffer accounting to the mlock bucket
Add alloc_locked_buffer() and free_locked_buffer() functions to mm/mlock.c
to kalloc a buffer and account the locked memory to current.
Account the memory for the BTS buffer to the tracer.
Signed-off-by: Markus Metzger <markus.t.metzger@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: introduce new ptrace facility
Add arch_ptrace_untrace() function that is called when the tracer
detaches (either voluntarily or when the tracing task dies);
ptrace_disable() is only called on a voluntary detach.
Add ptrace_fork() and arch_ptrace_fork(). They are called when a
traced task is forked.
Clear DS and BTS related fields on fork.
Release DS resources and reclaim memory in ptrace_untrace(). This
releases resources already when the tracing task dies. We used to do
that when the traced task dies.
Signed-off-by: Markus Metzger <markus.t.metzger@intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: Cleanup and branch hints only.
Move the track and untrack pfn stub routines from memory.c to asm-generic.
Also add unlikely to pfnmap related calls in fork and exit path.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Cleanup - removes a new function in favor of a recently modified older one.
Replace follow_pfnmap_pte in pat code with follow_phys. follow_phys lso
returns protection eliminating the need of pte_pgprot call. Using follow_phys
also eliminates the need for pte_pa.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Changes and globalizes an existing static interface.
Follow_phys does similar things as follow_pfnmap_pte. Make a minor change
to follow_phys so that it can be used in place of follow_pfnmap_pte.
Physical address return value with 0 as error return does not work in
follow_phys as the actual physical address 0 mapping may exist in pte.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Documentation only
Incremental patches to address the review comments from Nick Piggin
for v3 version of x86 PAT pfnmap changes patchset here
http://lkml.indiana.edu/hypermail/linux/kernel/0812.2/01330.html
This patch:
Clarify is_linear_pfn_mapping() and its usage.
It is used by x86 PAT code for performance reasons. Identifying pfnmap
as linear over entire vma helps speedup reserve and free of memtype
for the region.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Pass mount flags to security_sb_kern_mount(), so security modules
can determine if a mount operation is being performed by the kernel.
Signed-off-by: James Morris <jmorris@namei.org>
Acked-by: Stephen Smalley <sds@tycho.nsa.gov>
This patch implements dynamic power save for mac80211. Basically it
means enabling power save mode after an idle period. Implementing it
dynamically gives a good compromise of low power consumption and low
latency. Some hardware have support for this in firmware, but some
require the host to do it.
The dynamic power save is implemented by adding an timeout to
ieee80211_subif_start_xmit(). The timeout can be enabled from userspace
with Wireless Extensions. For example, the command below enables the
dynamic power save and sets the time timeout to 500 ms:
iwconfig wlan0 power timeout 500m
Power save now only works with devices which handle power save in firmware.
It's also disabled by default and the heuristics when and how to enable is
considered as a policy decision and will be left for the userspace to handle.
In case the firmware has support for this, drivers can disable this feature
with IEEE80211_HW_NO_STACK_DYNAMIC_PS.
Big thanks to Johannes Berg for the help with the design and code.
Signed-off-by: Kalle Valo <kalle.valo@nokia.com>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This patch adds option for HT-enabled drivers to report HT rates
(HT20/HT40, short GI, MCS index) to mac80211. These rates are
currently not in the rate table, so the rate_idx is used to indicate
MCS index.
Signed-off-by: Jouni Malinen <jouni.malinen@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
HT management is done differently for AP and STA modes, unify
to just the ->config() callback since HT is fundamentally a
PHY property and cannot be per-BSS.
Rename enum nl80211_sec_chan_offset as nl80211_channel_type to denote
the channel type ( NO_HT, HT20, HT40+, HT40- ).
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Sujith <Sujith.Manoharan@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
This patch adds signal strength and transmission bitrate
to the station_info of nl80211.
Signed-off-by: Henning Rogge <rogge@fgan.de>
Acked-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
The ACPI interpreter usually runs with irqs enabled.
However, during suspend/resume it runs with
irqs disabled to evaluate _GTS/_BFS, as well as
by irqrouter_resume() which evaluates _CRS, _PRS, _SRS.
http://bugzilla.kernel.org/show_bug.cgi?id=12252
Signed-off-by: Wu Fengguang <wfg@linux.intel.com>
Signed-off-by: Len Brown <len.brown@intel.com>
acpi_early_init() was changed to over-write the cmdline param,
making it really inconvenient to set debug flags at boot-time.
Also,
This sets the default level to "info", which is what all the ACPI
drivers use. So to enable messages from drivers, you only have to
supply the "layer" (a.k.a. "component"). For non-"info" ACPI core
and ACPI interpreter messages, you have to supply both level and
layer masks, as before.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Len Brown <len.brown@intel.com>
Impact: change task balancing to save power more agressively
Add SD_BALANCE_NEWIDLE flag at MC level and CPU level
if sched_mc is set. This helps power savings and
will not affect performance when sched_mc=0
Ingo and Mike Galbraith have optimised the SD flags by
removing SD_BALANCE_NEWIDLE at MC and CPU level. This
helps performance but hurts power savings since this
slows down task consolidation by reducing the number
of times load_balance is run.
sched: fine-tune SD_MC_INIT
commit 1480098470
Author: Mike Galbraith <efault@gmx.de>
Date: Fri Nov 7 15:26:50 2008 +0100
sched: re-tune balancing -- revert
commit 9fcd18c9e6
Author: Ingo Molnar <mingo@elte.hu>
Date: Wed Nov 5 16:52:08 2008 +0100
This patch selectively enables SD_BALANCE_NEWIDLE flag
only when sched_mc is set to 1 or 2. This helps power savings
by task consolidation and also does not hurt performance at
sched_mc=0 where all power saving optimisations are turned off.
Signed-off-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: extend range of /sys/devices/system/cpu/sched_mc_power_savings
Currently the sched_mc/smt_power_savings variable is a boolean,
which either enables or disables topology based power savings.
This patch extends the behaviour of the variable from boolean to
multivalued, such that based on the value, we decide how
aggressively do we want to perform powersavings balance at
appropriate sched domain based on topology.
Variable levels of power saving tunable would benefit end user to
match the required level of power savings vs performance
trade-off depending on the system configuration and workloads.
This version makes the sched_mc_power_savings global variable to
take more values (0,1,2). Later versions can have a single
tunable called sched_power_savings instead of
sched_{mc,smt}_power_savings.
Signed-off-by: Gautham R Shenoy <ego@in.ibm.com>
Signed-off-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
BALANCE_FOR_MC_POWER and similar macros defined in sched.h are
not constants and have various condition checks and significant
amount of code that is not suitable to be contain in a macro.
Also there could be side effects on the expressions passed to
some of them like test_sd_parent().
This patch converts all complex macros related to power savings
balance to inline functions.
Signed-off-by: Vaidyanathan Srinivasan <svaidy@linux.vnet.ibm.com>
Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: add new sysfs files.
Add sysfs files "kernel_max" and "offline" to display the max CPU index
allowed (NR_CPUS-1), and the map of cpus that are offline.
Cpus can be offlined via HOTPLUG, disabled by the BIOS ACPI tables, or
if they exceed the number of cpus allowed by the NR_CPUS config option,
or the "maxcpus=NUM" kernel start parameter.
The "possible_cpus=NUM" parameter can also extend the number of possible
cpus allowed, in which case the cpus not present at startup will be
in the offline state. (These cpus can be HOTPLUGGED ON after system
startup [pending a follow-on patch to provide the capability via the
/sys/devices/sys/cpu/cpuN/online mechanism to bring them online.])
By design, the "offlined cpus > possible cpus" display will always
use the following formats:
* all possible cpus online: "x$" or "x-y$"
* some possible cpus offline: ".*,x$" or ".*,x-y$"
where:
x == number of possible cpus (nr_cpu_ids); and
y == number of cpus >= NR_CPUS or maxcpus (if y > x).
One use of this feature is for distros to select (or configure) the
appropriate kernel to install for the resident system.
Notes:
* cpus offlined <= possible cpus will be printed for all architectures.
* cpus offlined > possible cpus will only be printed for arches that
set 'total_cpus' [X86 only in this patch].
Based on tip/cpus4096 + .../rusty/linux-2.6-for-ingo.git/master +
x86-only-patches sent 12/15.
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Impact: New API
This will be needed in x86 code to allocate the domain and old_domain
cpumasks on the same node as where the containing irq_cfg struct is
allocated.
(Also fixes double-dump_stack on rare CONFIG_DEBUG_PER_CPU_MAPS case)
Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> (re-impl alloc_cpumask_var)
these warnings:
kernel/trace/trace_sched_switch.c: In function ‘tracing_sched_register’:
kernel/trace/trace_sched_switch.c:96: warning: passing argument 1 of ‘register_trace_sched_wakeup_new’ from incompatible pointer type
kernel/trace/trace_sched_switch.c:112: warning: passing argument 1 of ‘unregister_trace_sched_wakeup_new’ from incompatible pointer type
kernel/trace/trace_sched_switch.c: In function ‘tracing_sched_unregister’:
kernel/trace/trace_sched_switch.c:121: warning: passing argument 1 of ‘unregister_trace_sched_wakeup_new’ from incompatible pointer type
Trigger because sched_wakeup_new tracepoints need the same trace
signature as sched_wakeup - which was changed recently.
Fix it.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: New mm functionality.
Add pgprot_writecombine. pgprot_writecombine will be aliased to
pgprot_noncached when not supported by the architecture.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Introduces new hooks, which are currently null.
Introduce generic hooks in remap_pfn_range and vm_insert_pfn and
corresponding copy and free routines with reserve and free tracking.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: New currently unused interface.
Add a generic interface to follow pfn in a pfnmap vma range. This is used by
one of the subsequent x86 PAT related patch to keep track of memory types
for vma regions across vma copy and free.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Impact: Code transformation, new functions added should have no effect.
Drivers use mmap followed by pgprot_* and remap_pfn_range or vm_insert_pfn,
in order to export reserved memory to userspace. Currently, such mappings are
not tracked and hence not kept consistent with other mappings (/dev/mem,
pci resource, ioremap) for the sme memory, that may exist in the system.
The following patchset adds x86 PAT attribute tracking and untracking for
pfnmap related APIs.
First three patches in the patchset are changing the generic mm code to fit
in this tracking. Last four patches are x86 specific to make things work
with x86 PAT code. The patchset aso introduces pgprot_writecombine interface,
which gives writecombine mapping when enabled, falling back to
pgprot_noncached otherwise.
This patch:
While working on x86 PAT, we faced some hurdles with trackking
remap_pfn_range() regions, as we do not have any information to say
whether that PFNMAP mapping is linear for the entire vma range or
it is smaller granularity regions within the vma.
A simple solution to this is to use vm_pgoff as an indicator for
linear mapping over the vma region. Currently, remap_pfn_range
only sets vm_pgoff for COW mappings. Below patch changes the
logic and sets the vm_pgoff irrespective of COW. This will still not
be enough for the case where pfn is zero (vma region mapped to
physical address zero). But, for all the other cases, we can look at
pfnmap VMAs and say whether the mappng is for the entire vma region
or not.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
This patch fixes a long-standing performance bug in classic RCU that
results in massive internal-to-RCU lock contention on systems with
more than a few hundred CPUs. Although this patch creates a separate
flavor of RCU for ease of review and patch maintenance, it is intended
to replace classic RCU.
This patch still handles stress better than does mainline, so I am still
calling it ready for inclusion. This patch is against the -tip tree.
Nevertheless, experience on an actual 1000+ CPU machine would still be
most welcome.
Most of the changes noted below were found while creating an rcutiny
(which should permit ejecting the current rcuclassic) and while doing
detailed line-by-line documentation.
Updates from v9 (http://lkml.org/lkml/2008/12/2/334):
o Fixes from remainder of line-by-line code walkthrough,
including comment spelling, initialization, undesirable
narrowing due to type conversion, removing redundant memory
barriers, removing redundant local-variable initialization,
and removing redundant local variables.
I do not believe that any of these fixes address the CPU-hotplug
issues that Andi Kleen was seeing, but please do give it a whirl
in case the machine is smarter than I am.
A writeup from the walkthrough may be found at the following
URL, in case you are suffering from terminal insomnia or
masochism:
http://www.kernel.org/pub/linux/kernel/people/paulmck/tmp/rcutree-walkthrough.2008.12.16a.pdf
o Made rcutree tracing use seq_file, as suggested some time
ago by Lai Jiangshan.
o Added a .csv variant of the rcudata debugfs trace file, to allow
people having thousands of CPUs to drop the data into
a spreadsheet. Tested with oocalc and gnumeric. Updated
documentation to suit.
Updates from v8 (http://lkml.org/lkml/2008/11/15/139):
o Fix a theoretical race between grace-period initialization and
force_quiescent_state() that could occur if more than three
jiffies were required to carry out the grace-period
initialization. Which it might, if you had enough CPUs.
o Apply Ingo's printk-standardization patch.
o Substitute local variables for repeated accesses to global
variables.
o Fix comment misspellings and redundant (but harmless) increments
of ->n_rcu_pending (this latter after having explicitly added it).
o Apply checkpatch fixes.
Updates from v7 (http://lkml.org/lkml/2008/10/10/291):
o Fixed a number of problems noted by Gautham Shenoy, including
the cpu-stall-detection bug that he was having difficulty
convincing me was real. ;-)
o Changed cpu-stall detection to wait for ten seconds rather than
three in order to reduce false positive, as suggested by Ingo
Molnar.
o Produced a design document (http://lwn.net/Articles/305782/).
The act of writing this document uncovered a number of both
theoretical and "here and now" bugs as noted below.
o Fix dynticks_nesting accounting confusion, simplify WARN_ON()
condition, fix kerneldoc comments, and add memory barriers
in dynticks interface functions.
o Add more data to tracing.
o Remove unused "rcu_barrier" field from rcu_data structure.
o Count calls to rcu_pending() from scheduling-clock interrupt
to use as a surrogate timebase should jiffies stop counting.
o Fix a theoretical race between force_quiescent_state() and
grace-period initialization. Yes, initialization does have to
go on for some jiffies for this race to occur, but given enough
CPUs...
Updates from v6 (http://lkml.org/lkml/2008/9/23/448):
o Fix a number of checkpatch.pl complaints.
o Apply review comments from Ingo Molnar and Lai Jiangshan
on the stall-detection code.
o Fix several bugs in !CONFIG_SMP builds.
o Fix a misspelled config-parameter name so that RCU now announces
at boot time if stall detection is configured.
o Run tests on numerous combinations of configurations parameters,
which after the fixes above, now build and run correctly.
Updates from v5 (http://lkml.org/lkml/2008/9/15/92, bad subject line):
o Fix a compiler error in the !CONFIG_FANOUT_EXACT case (blew a
changeset some time ago, and finally got around to retesting
this option).
o Fix some tracing bugs in rcupreempt that caused incorrect
totals to be printed.
o I now test with a more brutal random-selection online/offline
script (attached). Probably more brutal than it needs to be
on the people reading it as well, but so it goes.
o A number of optimizations and usability improvements:
o Make rcu_pending() ignore the grace-period timeout when
there is no grace period in progress.
o Make force_quiescent_state() avoid going for a global
lock in the case where there is no grace period in
progress.
o Rearrange struct fields to improve struct layout.
o Make call_rcu() initiate a grace period if RCU was
idle, rather than waiting for the next scheduling
clock interrupt.
o Invoke rcu_irq_enter() and rcu_irq_exit() only when
idle, as suggested by Andi Kleen. I still don't
completely trust this change, and might back it out.
o Make CONFIG_RCU_TRACE be the single config variable
manipulated for all forms of RCU, instead of the prior
confusion.
o Document tracing files and formats for both rcupreempt
and rcutree.
Updates from v4 for those missing v5 given its bad subject line:
o Separated dynticks interface so that NMIs and irqs call separate
functions, greatly simplifying it. In particular, this code
no longer requires a proof of correctness. ;-)
o Separated dynticks state out into its own per-CPU structure,
avoiding the duplicated accounting.
o The case where a dynticks-idle CPU runs an irq handler that
invokes call_rcu() is now correctly handled, forcing that CPU
out of dynticks-idle mode.
o Review comments have been applied (thank you all!!!).
For but one example, fixed the dynticks-ordering issue that
Manfred pointed out, saving me much debugging. ;-)
o Adjusted rcuclassic and rcupreempt to handle dynticks changes.
Attached is an updated patch to Classic RCU that applies a hierarchy,
greatly reducing the contention on the top-level lock for large machines.
This passes 10-hour concurrent rcutorture and online-offline testing on
128-CPU ppc64 without dynticks enabled, and exposes some timekeeping
bugs in presence of dynticks (exciting working on a system where
"sleep 1" hangs until interrupted...), which were fixed in the
2.6.27 kernel. It is getting more reliable than mainline by some
measures, so the next version will be against -tip for inclusion.
See also Manfred Spraul's recent patches (or his earlier work from
2004 at http://marc.info/?l=linux-kernel&m=108546384711797&w=2).
We will converge onto a common patch in the fullness of time, but are
currently exploring different regions of the design space. That said,
I have already gratefully stolen quite a few of Manfred's ideas.
This patch provides CONFIG_RCU_FANOUT, which controls the bushiness
of the RCU hierarchy. Defaults to 32 on 32-bit machines and 64 on
64-bit machines. If CONFIG_NR_CPUS is less than CONFIG_RCU_FANOUT,
there is no hierarchy. By default, the RCU initialization code will
adjust CONFIG_RCU_FANOUT to balance the hierarchy, so strongly NUMA
architectures may choose to set CONFIG_RCU_FANOUT_EXACT to disable
this balancing, allowing the hierarchy to be exactly aligned to the
underlying hardware. Up to two levels of hierarchy are permitted
(in addition to the root node), allowing up to 16,384 CPUs on 32-bit
systems and up to 262,144 CPUs on 64-bit systems. I just know that I
am going to regret saying this, but this seems more than sufficient
for the foreseeable future. (Some architectures might wish to set
CONFIG_RCU_FANOUT=4, which would limit such architectures to 64 CPUs.
If this becomes a real problem, additional levels can be added, but I
doubt that it will make a significant difference on real hardware.)
In the common case, a given CPU will manipulate its private rcu_data
structure and the rcu_node structure that it shares with its immediate
neighbors. This can reduce both lock and memory contention by multiple
orders of magnitude, which should eliminate the need for the strange
manipulations that are reported to be required when running Linux on
very large systems.
Some shortcomings:
o More bugs will probably surface as a result of an ongoing
line-by-line code inspection.
Patches will be provided as required.
o There are probably hangs, rcutorture failures, &c. Seems
quite stable on a 128-CPU machine, but that is kind of small
compared to 4096 CPUs. However, seems to do better than
mainline.
Patches will be provided as required.
o The memory footprint of this version is several KB larger
than rcuclassic.
A separate UP-only rcutiny patch will be provided, which will
reduce the memory footprint significantly, even compared
to the old rcuclassic. One such patch passes light testing,
and has a memory footprint smaller even than rcuclassic.
Initial reaction from various embedded guys was "it is not
worth it", so am putting it aside.
Credits:
o Manfred Spraul for ideas, review comments, and bugs spotted,
as well as some good friendly competition. ;-)
o Josh Triplett, Ingo Molnar, Peter Zijlstra, Mathieu Desnoyers,
Lai Jiangshan, Andi Kleen, Andy Whitcroft, and Andrew Morton
for reviews and comments.
o Thomas Gleixner for much-needed help with some timer issues
(see patches below).
o Jon M. Tollefson, Tim Pepper, Andrew Theurer, Jose R. Santos,
Andy Whitcroft, Darrick Wong, Nishanth Aravamudan, Anton
Blanchard, Dave Kleikamp, and Nathan Lynch for keeping machines
alive despite my heavy abuse^Wtesting.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
