kernel_optimize_test/include
Christian Borntraeger 3491caf275 KVM: halt_polling: provide a way to qualify wakeups during poll
Some wakeups should not be considered a sucessful poll. For example on
s390 I/O interrupts are usually floating, which means that _ALL_ CPUs
would be considered runnable - letting all vCPUs poll all the time for
transactional like workload, even if one vCPU would be enough.
This can result in huge CPU usage for large guests.
This patch lets architectures provide a way to qualify wakeups if they
should be considered a good/bad wakeups in regard to polls.

For s390 the implementation will fence of halt polling for anything but
known good, single vCPU events. The s390 implementation for floating
interrupts does a wakeup for one vCPU, but the interrupt will be delivered
by whatever CPU checks first for a pending interrupt. We prefer the
woken up CPU by marking the poll of this CPU as "good" poll.
This code will also mark several other wakeup reasons like IPI or
expired timers as "good". This will of course also mark some events as
not sucessful. As  KVM on z runs always as a 2nd level hypervisor,
we prefer to not poll, unless we are really sure, though.

This patch successfully limits the CPU usage for cases like uperf 1byte
transactional ping pong workload or wakeup heavy workload like OLTP
while still providing a proper speedup.

This also introduced a new vcpu stat "halt_poll_no_tuning" that marks
wakeups that are considered not good for polling.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Acked-by: Radim Krčmář <rkrcmar@redhat.com> (for an earlier version)
Cc: David Matlack <dmatlack@google.com>
Cc: Wanpeng Li <kernellwp@gmail.com>
[Rename config symbol. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-05-13 17:29:23 +02:00
..
acpi
asm-generic arch, ftrace: for KASAN put hard/soft IRQ entries into separate sections 2016-03-25 16:37:42 -07:00
clocksource clocksource: arm_arch_timer: Remove arch_timer_get_timecounter 2016-05-03 12:54:21 +02:00
crypto
drm drm/ttm: use phys_addr_t for ttm_bus_placement 2016-04-04 17:00:01 -04:00
dt-bindings The clk changes for this release cycle are mostly dominated by 2016-03-23 06:06:45 -07:00
keys
kvm KVM: arm/arm64: vgic: Rely on the GIC driver to parse the firmware tables 2016-05-03 12:54:21 +02:00
linux KVM: halt_polling: provide a way to qualify wakeups during poll 2016-05-13 17:29:23 +02:00
math-emu
media Merge branch 'for-linus' of git://ftp.arm.linux.org.uk/~rmk/linux-arm 2016-03-19 16:31:54 -07:00
memory
misc
net For the current RC series, we have the following fixes: 2016-04-08 16:41:28 -04:00
pcmcia
ras
rdma Round two of 4.6 merge window patches 2016-03-22 15:48:44 -07:00
rxrpc
scsi Merge branch 'fixes-base' into fixes 2016-04-05 06:56:47 -04:00
soc IOMMU Updates for Linux v4.6 2016-03-22 11:57:43 -07:00
sound
target target: add a new add_wwn_groups fabrics method 2016-03-30 20:06:44 -07:00
trace KVM: halt_polling: provide a way to qualify wakeups during poll 2016-05-13 17:29:23 +02:00
uapi kvm: introduce KVM_MAX_VCPU_ID 2016-05-11 22:37:54 +02:00
video gpu: ipu-v3: ipu-dmfc: Rename ipu_dmfc_init_channel to ipu_dmfc_config_wait4eot 2016-03-31 11:24:33 +02:00
xen
Kbuild