linux_dsm_epyc7002/arch/powerpc
Paul Mackerras 6af27c847a KVM: PPC: Book3S HV: Streamline guest entry and exit
On entry to the guest, secondary threads now wait for the primary to
switch the MMU after loading up most of their state, rather than before.
This means that the secondary threads get into the guest sooner, in the
common case where the secondary threads get to kvmppc_hv_entry before
the primary thread.

On exit, the first thread out increments the exit count and interrupts
the other threads (to get them out of the guest) before saving most
of its state, rather than after.  That means that the other threads
exit sooner and means that the first thread doesn't spend so much
time waiting for the other threads at the point where the MMU gets
switched back to the host.

This pulls out the code that increments the exit count and interrupts
other threads into a separate function, kvmhv_commence_exit().
This also makes sure that r12 and vcpu->arch.trap are set correctly
in some corner cases.

Statistics from /sys/kernel/debug/kvm/vm*/vcpu*/timings show the
improvement.  Aggregating across vcpus for a guest with 32 vcpus,
8 threads/vcore, running on a POWER8, gives this before the change:

 rm_entry:     avg 4537.3ns (222 - 48444, 1068878 samples)
  rm_exit:     avg 4787.6ns (152 - 165490, 1010717 samples)
  rm_intr:     avg 1673.6ns (12 - 341304, 3818691 samples)

and this after the change:

 rm_entry:     avg 3427.7ns (232 - 68150, 1118921 samples)
  rm_exit:     avg 4716.0ns (12 - 150720, 1119477 samples)
  rm_intr:     avg 1614.8ns (12 - 522436, 3850432 samples)

showing a substantial reduction in the time spent per guest entry in
the real-mode guest entry code, and smaller reductions in the real
mode guest exit and interrupt handling times.  (The test was to start
the guest and boot Fedora 20 big-endian to the login prompt.)

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2015-04-21 15:21:33 +02:00
..
boot powerpc: dts: pq3/85xx: Fix GPIO address 2015-01-29 23:04:32 -06:00
configs The clock framework changes for 3.20 contain the usual driver additions, 2015-02-21 12:30:30 -08:00
crypto crypto: add missing crypto module aliases 2015-01-13 22:29:11 +11:00
include KVM: PPC: Book3S HV: Use bitmap of active threads rather than count 2015-04-21 15:21:33 +02:00
kernel KVM: PPC: Book3S HV: Use bitmap of active threads rather than count 2015-04-21 15:21:33 +02:00
kvm KVM: PPC: Book3S HV: Streamline guest entry and exit 2015-04-21 15:21:33 +02:00
lib powerpc: Export __spin_yield 2015-04-21 15:21:28 +02:00
math-emu powerpc: Correct emulated mtfsf instruction 2014-04-07 10:33:11 +10:00
mm powerpc: drop _PAGE_FILE and pte_file()-related helpers 2015-02-16 17:56:05 -08:00
net module: remove mod arg from module_free, rename module_memfree(). 2015-01-20 11:38:33 +10:30
oprofile powerpc updates for 3.19 2014-12-11 17:48:14 -08:00
perf Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/scottwood/linux into next 2015-02-04 12:03:21 +11:00
platforms KVM: PPC: Book3S HV: Add fast real-mode H_RANDOM implementation. 2015-04-21 15:21:29 +02:00
sysdev powerpc: use %*pb[l] to print bitmaps including cpumasks and nodemasks 2015-02-13 21:21:36 -08:00
xmon powerpc updates for 3.20 2015-02-11 18:15:38 -08:00
Kconfig powerpc/mm: fix undefined reference to `.__kernel_map_pages' on FSL PPC64 2015-01-28 14:22:22 +11:00
Kconfig.debug Patch queue for ppc - 2014-08-01 2014-08-05 09:58:11 +02:00
Makefile kbuild: do not add $(call ...) to invoke cc-version or cc-fullversion 2015-01-09 17:25:44 +01:00
relocs_check.pl Fix warning typo "CONFIG_RELCOATABLE" 2013-05-29 15:11:30 +02:00