This hashtable implementation is using hlist buckets to provide a simple
hashtable to prevent it from getting reimplemented all over the kernel.
Signed-off-by: Sasha Levin <levinsasha928@gmail.com>
[ Merging this now, so that subsystems can start applying Sasha's
patches that use this - Linus ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Have the ring buffer commit function use the irq_work infrastructure to
wake up any waiters waiting on the ring buffer for new data. The irq_work
was created for such a purpose, where doing the actual wake up at the
time of adding data is too dangerous, as an event or function trace may
be in the midst of the work queue locks and cause deadlocks. The irq_work
will either delay the action to the next timer interrupt, or trigger an IPI
to itself forcing an interrupt to do the work (in a safe location).
With irq_work, all ring buffer commits can safely do wakeups, removing
the need for the ring buffer commit "nowake" variants, which were used
by events and function tracing. All commits can now safely use the
normal commit, and the "nowake" variants can be removed.
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
This patch adds a way to configure the CPTS input clock scaling factors
via the device tree.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Because time stamping on both external ports of the switch simultaneously
is positively useless from the application's point of view, this patch
provides a DT configuration method to choose the active port.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes the timecompare code from the kernel. The top five
reasons to do this are:
1. There are no more users of this code.
2. The original idea was a bit weak.
3. The original author has disappeared.
4. The code was not general purpose but tuned to a particular hardware,
5. There are better ways to accomplish clock synchronization.
Signed-off-by: Richard Cochran <richardcochran@gmail.com>
Acked-by: John Stultz <john.stultz@linaro.org>
Tested-by: Bob Liu <lliubbo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The SO_ATTACH_FILTER option is set only. I propose to add the get
ability by using SO_ATTACH_FILTER in getsockopt. To be less
irritating to eyes the SO_GET_FILTER alias to it is declared. This
ability is required by checkpoint-restore project to be able to
save full state of a socket.
There are two issues with getting filter back.
First, kernel modifies the sock_filter->code on filter load, thus in
order to return the filter element back to user we have to decode it
into user-visible constants. Fortunately the modification in question
is interconvertible.
Second, the BPF_S_ALU_DIV_K code modifies the command argument k to
speed up the run-time division by doing kernel_k = reciprocal(user_k).
Bad news is that different user_k may result in same kernel_k, so we
can't get the original user_k back. Good news is that we don't have
to do it. What we need to is calculate a user2_k so, that
reciprocal(user2_k) == reciprocal(user_k) == kernel_k
i.e. if it's re-loaded back the compiled again value will be exactly
the same as it was. That said, the user2_k can be calculated like this
user2_k = reciprocal(kernel_k)
with an exception, that if kernel_k == 0, then user2_k == 1.
The optlen argument is treated like this -- when zero, kernel returns
the amount of sock_fprog elements in filter, otherwise it should be
large enough for the sock_fprog array.
changes since v1:
* Declared SO_GET_FILTER in all arch headers
* Added decode of vlan-tag codes
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Define and export function ceph_pg_pool_name_by_id() to supply
the name of a pg pool whose id is given. This will be used by
the next patch.
Signed-off-by: Alex Elder <elder@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
Add support for getting the the information identifying the parent
image for rbd images that have them. The child image holds a
reference to its parent image specification structure. Create a new
entry "parent" in /sys/bus/rbd/image/N/ to report the identifying
information for the parent image, if any.
Signed-off-by: Alex Elder <elder@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
Attempts to retrigger nested threaded IRQs currently fail because they
have no primary handler. In order to support retrigger of nested
IRQs, the parent IRQ needs to be retriggered.
To fix, when an IRQ needs to be resent, if the interrupt has a parent
IRQ and runs in the context of the parent IRQ, then resend the parent.
Also, handle_nested_irq() needs to clear the replay flag like the
other handlers, otherwise check_irq_resend() will set it and it will
never be cleared. Without clearing, it results in the first resend
working fine, but check_irq_resend() returning early on subsequent
resends because the replay flag is still set.
Problem discovered on ARM/OMAP platforms where a nested IRQ that's
also a wakeup IRQ happens late in suspend and needed to be retriggered
during the resume process.
[khilman@ti.com: changelog edits, clear IRQS_REPLAY in handle_nested_irq()]
Reported-by: Kevin Hilman <khilman@ti.com>
Tested-by: Kevin Hilman <khilman@ti.com>
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lkml.kernel.org/r/1350425269-11489-1-git-send-email-khilman@deeprootsystems.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
After commit b3356bf0db (KVM: emulator: optimize "rep ins" handling),
the pieces of io data can be collected and write them to the guest memory
or MMIO together
Unfortunately, kvm splits the mmio access into 8 bytes and store them to
vcpu->mmio_fragments. If the guest uses "rep ins" to move large data, it
will cause vcpu->mmio_fragments overflow
The bug can be exposed by isapc (-M isapc):
[23154.818733] general protection fault: 0000 [#1] SMP DEBUG_PAGEALLOC
[ ......]
[23154.858083] Call Trace:
[23154.859874] [<ffffffffa04f0e17>] kvm_get_cr8+0x1d/0x28 [kvm]
[23154.861677] [<ffffffffa04fa6d4>] kvm_arch_vcpu_ioctl_run+0xcda/0xe45 [kvm]
[23154.863604] [<ffffffffa04f5a1a>] ? kvm_arch_vcpu_load+0x17b/0x180 [kvm]
Actually, we can use one mmio_fragment to store a large mmio access then
split it when we pass the mmio-exit-info to userspace. After that, we only
need two entries to store mmio info for the cross-mmio pages access
Signed-off-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
The existing 'overrun' counter is incremented when the ring
buffer wraps around, with overflow on (the default). We wanted
a way to count requests lost from the buffer filling up with
overflow off, too. I decided to add a new counter instead
of retro-fitting the existing one because it seems like a
different statistic to count conceptually, and also because
of how the code was structured.
Link: http://lkml.kernel.org/r/1310765038-26399-1-git-send-email-slavapestov@google.com
Signed-off-by: Slava Pestov <slavapestov@google.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Matching on device and interface class with with unspecified
subclass and protocol is sometimes useful. This is slightly
different from USB_DEVICE_AND_INTERFACE_INFO which requires
the full interface class/subclass/protocol triplet.
Signed-off-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch (as1623) removes the ehci_port_power() routine and all the
places that call it. There's no reason for ehci-hcd to change the
port power settings; the hub driver takes care of all that stuff.
There is one exception: When the controller is resumed from
hibernation or following a loss of power, the ports that are supposed
to be handed over to a companion controller must be powered on first.
Otherwise the handover won't work. This process is not visible to the
hub driver, so it has to be handled in ehci-hcd.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Jeff Kirsher says:
====================
This series contains updates to ixgbe, ixgbevf, igbvf, igb and
networking core (bridge). Most notably is the addition of support
for local link multicast addresses in SR-IOV mode to the networking
core.
Also note, the ixgbe patch "ixgbe: Add support for pipeline reset" and
"ixgbe: Fix return value from macvlan filter function" is revised based
on community feedback.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
BPF filters lack ability to access skb->vlan_tci
This patch adds two new ancillary accessors :
SKF_AD_VLAN_TAG (44) mapped to vlan_tx_tag_get(skb)
SKF_AD_VLAN_TAG_PRESENT (48) mapped to vlan_tx_tag_present(skb)
This allows libpcap/tcpdump to use a kernel filter instead of
having to fallback to accept all packets, then filter them in
user space.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Suggested-by: Ani Sinha <ani@aristanetworks.com>
Suggested-by: Daniel Borkmann <danborkmann@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Network device drivers can communicate a Toeplitz hash in skb->rxhash,
but devices differ in their hashing capabilities. All compute a 5-tuple
hash for TCP over IPv4, but for other connection-oriented protocols,
they may compute only a 3-tuple. This breaks RPS load balancing, e.g.,
for TCP over IPv6 flows. Additionally, for GRE and other tunnels,
the kernel computes a 5-tuple hash over the inner packet if possible,
but devices do not.
This patch recomputes the rxhash in software in all cases where it
cannot be certain that a 5-tuple was computed. Device drivers can avoid
recomputation by setting the skb->l4_rxhash flag.
Recomputing adds cycles to each packet when RPS is enabled or the
packet arrives over a tunnel. A comparison of 200x TCP_STREAM between
two servers running unmodified netnext with rxhash computation
in hardware vs software (using ethtool -K eth0 rxhash [on|off]) shows
how much time is spent in __skb_get_rxhash in this worst case:
0.03% swapper [kernel.kallsyms] [k] __skb_get_rxhash
0.03% swapper [kernel.kallsyms] [k] __skb_get_rxhash
0.05% swapper [kernel.kallsyms] [k] __skb_get_rxhash
With 200x TCP_RR it increases to
0.10% netperf [kernel.kallsyms] [k] __skb_get_rxhash
0.10% netperf [kernel.kallsyms] [k] __skb_get_rxhash
0.10% netperf [kernel.kallsyms] [k] __skb_get_rxhash
I considered having the patch explicitly skips recomputation when it knows
that it will not improve the hash (TCP over IPv4), but that conditional
complicates code without saving many cycles in practice, because it has
to take place after flow dissector.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This adds support for the net device ops to manage the embedded
hardware bridge on ixgbe devices. With this patch the bridge
mode can be toggled between VEB and VEPA to support stacking
macvlan devices or using the embedded switch without any SW
component in 802.1Qbg/br environments.
Additionally, this adds source address pruning to the ixgbevf
driver to prune any frames sent back from a reflective relay on
the switch. This is required because the existing hardware does
not support this. Without it frames get pushed into the stack
with its own src mac which is invalid per 802.1Qbg VEPA
definition.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The PF_BRIDGE:RTM_{GET|SET}LINK nlmsg family and type are
currently embedded in the ./net/bridge module. This prohibits
them from being used by other bridging devices. One example
of this being hardware that has embedded bridging components.
In order to use these nlmsg types more generically this patch
adds two net_device_ops hooks. One to set link bridge attributes
and another to dump the current bride attributes.
ndo_bridge_setlink()
ndo_bridge_getlink()
CC: Lennert Buytenhek <buytenh@wantstofly.org>
CC: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
HS and SS descriptors are staticaly created. They are updated during the
bind process with the endpoint address, string id or interface numbers.
After that, the descriptor chain is linked to struct usb_function which
is used by composite in order to serve the GET_DESCRIPTOR requests,
number of available configs and so on.
There is no need to assign the HS descriptor only if the UDC supports
HS speed because composite won't report those to the host if HS support
has not been reached. The same reasoning is valid for SS.
This patch makes sure each function updates HS/SS descriptors
unconditionally and uses the newly introduced helper function to create a
copy the descriptors for the speed which is supported by the UDC.
While at that, also rename f->descriptors to f->fs_descriptors in order
to make it more explicit what that means.
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
The 5+4 magic for HS tries to aim 32ms which is also what is intended
with 1 << 5 for FS. This little macro should make this easier to
understand.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
This function is identically defined in all three allocators
and it's trivial to move it to slab.h
Since now it's static, inline, header-defined function
this patch also drops the EXPORT_SYMBOL tag.
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Matt Mackall <mpm@selenic.com>
Acked-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Ezequiel Garcia <elezegarcia@gmail.com>
Signed-off-by: Pekka Enberg <penberg@kernel.org>
Pull md fixes from NeilBrown:
"Some fixes for md in 3.7
- one recently introduced crash for dm-raid10 with discard
- one bug in new functionality that has been around for a few
releases.
- minor bug in md's 'faulty' personality
and UAPI disintegration for md."
* tag 'md-3.7-fixes' of git://neil.brown.name/md:
MD RAID10: Fix oops when creating RAID10 arrays via dm-raid.c
md/raid1: Fix assembling of arrays containing Replacements.
md faulty: use disk_stack_limits()
UAPI: (Scripted) Disintegrate include/linux/raid
It is main portion of pci_rescan_bus().
Separate it out and prepare to use it for PCI root bus hot add later.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Remove usage of plat/cpu.h and get information from platform data
instead. This enables omapdrm to be built with ARCH_MULTIPLATFORM.
Signed-off-by: Rob Clark <rob@ti.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently swiotlb is the only consumer for swiotlb_bounce. Since that is the
case it doesn't make much sense to be exporting it so make it a static
function only.
In addition we can save a few more lines of code by making it so that it
accepts the DMA address as a physical address instead of a virtual one. This
is the last piece in essentially pushing all of the DMA address values to use
physical addresses in swiotlb.
In order to clarify things since we now have 2 physical addresses in use
inside of swiotlb_bounce I am renaming phys to orig_addr, and dma_addr to
tlb_addr. This way is should be clear that orig_addr is contained within
io_orig_addr and tlb_addr is an address within the io_tlb_addr buffer.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
This change makes it so that the sync functionality also uses physical
addresses. This helps to further reduce the use of virt_to_phys and
phys_to_virt functions.
In order to clarify things since we now have 2 physical addresses in use
inside of swiotlb_tbl_sync_single I am renaming phys to orig_addr, and
dma_addr to tlb_addr. This way is should be clear that orig_addr is
contained within io_orig_addr and tlb_addr is an address within the
io_tlb_addr buffer.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
This change makes it so that the unmap functionality also uses physical
addresses. This helps to further reduce the use of virt_to_phys and
phys_to_virt functions.
In order to clarify things since we now have 2 physical addresses in use
inside of swiotlb_tbl_unmap_single I am renaming phys to orig_addr, and
dma_addr to tlb_addr. This way is should be clear that orig_addr is
contained within io_orig_addr and tlb_addr is an address within the
io_tlb_addr buffer.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
This change makes it so that swiotlb_tbl_map_single will return a physical
address instead of a virtual address when called. The advantage to this once
again is that we are avoiding a number of virt_to_phys and phys_to_virt
translations by working with everything as a physical address.
One change I had to make in order to support using physical addresses is that
I could no longer trust 0 to be a invalid physical address on all platforms.
So instead I made it so that ~0 is returned on error. This should never be a
valid return value as it implies that only one byte would be available for
use.
In order to clarify things since we now have 2 physical addresses in use
inside of swiotlb_tbl_map_single I am renaming phys to orig_addr, and
dma_addr to tlb_addr. This way is should be clear that orig_addr is
contained within io_orig_addr and tlb_addr is an address within the
io_tlb_addr buffer.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Other than ix86, x86-64 on EFI so far didn't set the
{g,s}et_wallclock accessors to the EFI routines, thus
incorrectly using raw RTC accesses instead.
Simply removing the #ifdef around the respective code isn't
enough, however: While so far early get-time calls were done in
physical mode, this doesn't work properly for x86-64, as virtual
addresses would still need to be set up for all runtime regions
(which wasn't the case on the system I have access to), so
instead the patch moves the call to efi_enter_virtual_mode()
ahead (which in turn allows to drop all code related to calling
efi-get-time in physical mode).
Additionally the earlier calling of efi_set_executable()
requires the CPA code to cope, i.e. during early boot it must be
avoided to call cpa_flush_array(), as the first thing this
function does is a BUG_ON(irqs_disabled()).
Also make the two EFI functions in question here static -
they're not being referenced elsewhere.
History:
This commit was originally merged as bacef661ac ("x86-64/efi:
Use EFI to deal with platform wall clock") but it resulted in some
ASUS machines no longer booting due to a firmware bug, and so was
reverted in f026cfa82f. A pre-emptive fix for the buggy ASUS
firmware was merged in 03a1c254975e ("x86, efi: 1:1 pagetable
mapping for virtual EFI calls") so now this patch can be
reapplied.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Tested-by: Matt Fleming <matt.fleming@intel.com>
Acked-by: Matthew Garrett <mjg@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Matt Fleming <matt.fleming@intel.com> [added commit history]
UEFI variable filesystem need a new mount point, so this patch add
efivars kobject to efi_kobj for create a /sys/firmware/efi/efivars
folder.
Cc: Matthew Garrett <mjg@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Lee, Chun-Yi <jlee@suse.com>
Signed-off-by: Jeremy Kerr <jeremy.kerr@canonical.com>
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
The existing EFI variables code only supports variables of up to 1024
bytes. This limitation existed in version 0.99 of the EFI specification,
but was removed before any full releases. Since variables can now be
larger than a single page, sysfs isn't the best interface for this. So,
instead, let's add a filesystem. Variables can be read, written and
created, with the first 4 bytes of each variable representing its UEFI
attributes. The create() method doesn't actually commit to flash since
zero-length variables can't exist per-spec.
Updates from Jeremy Kerr <jeremy.kerr@canonical.com>.
Signed-off-by: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Jeremy Kerr <jeremy.kerr@canonical.com>
Signed-off-by: Matt Fleming <matt.fleming@intel.com>
We now can schedule only a specific range of sectors for online verify,
or interrupt a running verify without interrupting the connection.
Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
In SR-IOV mode the PF driver acts as the uplink port and is
used to send control packets e.g. lldpad, stp, etc.
eth0.1 eth0.2 eth0
VF VF PF
| | | <-- stand-in for uplink
| | |
--------------------------
| Embedded Switch |
--------------------------
|
MAC <-- uplink
But the embedded switch is setup to forward multicast addresses
to all interfaces both VFs and PF and onto the physical link.
This results in reserved MAC addresses used by control protocols
to be forwarded over the switch onto the VF.
In the LLDP case the PF sends an LLDPDU and it is currently
being forwarded to all the VFs who then see the PF as a peer.
This is incorrect.
This patch adds the multicast addresses to the RAR table in the
hardware to prevent this behavior.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Tested-by: Phil Schmitt <phillip.j.schmitt@intel.com>
Tested-by: Sibai Li <sibai.li@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
The OMAP watchdog timer driver directly calls a function exported by
code in arch/arm/mach-omap2. This is not good; it tightly couples
this driver to the mach-omap2 integration code. Instead, add a
temporary platform_data function pointer to abstract this function
call. A subsequent patch will convert the watchdog driver to use this
function pointer.
This patch also moves the device creation code out of
arch/arm/mach-omap2/devices.c and into arch/arm/mach-omap2/wd_timer.c.
This is another step towards the removal of
arch/arm/mach-omap2/devices.c.
Cc: Wim Van Sebroeck <wim@iguana.be>
Acked-by: Wim Van Sebroeck <wim@iguana.be>
[paul@pwsan.com: skip wd_timer device creation when DT blob is present]
Signed-off-by: Paul Walmsley <paul@pwsan.com>
This patch filters noslot pfn out from error pfns based on Marcelo comment:
noslot pfn is not a error pfn
After this patch,
- is_noslot_pfn indicates that the gfn is not in slot
- is_error_pfn indicates that the gfn is in slot but the error is occurred
when translate the gfn to pfn
- is_error_noslot_pfn indicates that the pfn either it is error pfns or it
is noslot pfn
And is_invalid_pfn can be removed, it makes the code more clean
Signed-off-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Merge reason: development work has dependency on kvm patches merged
upstream.
Conflicts:
arch/powerpc/include/asm/Kbuild
arch/powerpc/include/asm/kvm_para.h
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
vtime_account() doesn't have the same role in
CONFIG_VIRT_CPU_ACCOUNTING and CONFIG_IRQ_TIME_ACCOUNTING.
In the first case it handles time accounting in any context. In
the second case it only handles irq time accounting.
So when vtime_account() is called from outside vtime_account_irq_*()
this call is pointless to CONFIG_IRQ_TIME_ACCOUNTING.
To fix the confusion, change vtime_account() to irqtime_account_irq()
in CONFIG_IRQ_TIME_ACCOUNTING. This way we ensure future account_vtime()
calls won't waste useless cycles in the irqtime APIs.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
With CONFIG_VIRT_CPU_ACCOUNTING, when vtime_account()
is called in irq entry/exit, we perform a check on the
context: if we are interrupting the idle task we
account the pending cputime to idle, otherwise account
to system time or its sub-areas: tsk->stime, hardirq time,
softirq time, ...
However this check for idle only concerns the hardirq entry
and softirq entry:
* Hardirq may directly interrupt the idle task, in which case
we need to flush the pending CPU time to idle.
* The idle task may be directly interrupted by a softirq if
it calls local_bh_enable(). There is probably no such call
in any idle task but we need to cover every case. Ksoftirqd
is not concerned because the idle time is flushed on context
switch and softirq in the end of hardirq have the idle time
already flushed from the hardirq entry.
In the other cases we always account to system/irq time:
* On hardirq exit we account the time to hardirq time.
* On softirq exit we account the time to softirq time.
To optimize this and avoid the indirect call to vtime_account()
and the checks it performs, specialize the vtime irq APIs and
only perform the check on irq entry. Irq exit can directly call
vtime_account_system().
CONFIG_IRQ_TIME_ACCOUNTING behaviour doesn't change and directly
maps to its own vtime_account() implementation. One may want
to take benefits from the new APIs to optimize irq time accounting
as well in the future.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Switching to or from guest context is done on ioctl context.
So by the time we call kvm_guest_enter() or kvm_guest_exit()
we know we are not running the idle task.
As a result, we can directly account the cputime using
vtime_account_system().
There are two good reasons to do this:
* We avoid some useless checks on guest switch. It optimizes
a bit this fast path.
* In the case of CONFIG_IRQ_TIME_ACCOUNTING, calling vtime_account()
checks for irq time to account. This is pointless since we know
we are not in an irq on guest switch. This is wasting cpu cycles
for no good reason. vtime_account_system() OTOH is a no-op in
this config option.
* We can remove the irq disable/enable around kvm guest switch in s390.
A further optimization may consist in introducing a vtime_account_guest()
that directly calls account_guest_time().
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Joerg Roedel <joerg.roedel@amd.com>
Cc: Alexander Graf <agraf@suse.de>
Cc: Xiantao Zhang <xiantao.zhang@intel.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Cornelia Huck <cornelia.huck@de.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Some clock drivers can be simplified if devres takes care of
unregistering any registered clocks along error paths. Introduce
devm_clk_register() so that clock drivers get unregistration for
free along with simplified error paths.
Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Signed-off-by: Mike Turquette <mturquette@linaro.org>
