q3k/linux
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
linux/arch/powerpc/kernel
History
Paul Mackerras 19ccb76a19 KVM: PPC: Implement H_CEDE hcall for book3s_hv in real-mode code 
With a KVM guest operating in SMT4 mode (i.e. 4 hardware threads per
core), whenever a CPU goes idle, we have to pull all the other
hardware threads in the core out of the guest, because the H_CEDE
hcall is handled in the kernel.  This is inefficient.

This adds code to book3s_hv_rmhandlers.S to handle the H_CEDE hcall
in real mode.  When a guest vcpu does an H_CEDE hcall, we now only
exit to the kernel if all the other vcpus in the same core are also
idle.  Otherwise we mark this vcpu as napping, save state that could
be lost in nap mode (mainly GPRs and FPRs), and execute the nap
instruction.  When the thread wakes up, because of a decrementer or
external interrupt, we come back in at kvm_start_guest (from the
system reset interrupt vector), find the `napping' flag set in the
paca, and go to the resume path.

This has some other ramifications.  First, when starting a core, we
now start all the threads, both those that are immediately runnable and
those that are idle.  This is so that we don't have to pull all the
threads out of the guest when an idle thread gets a decrementer interrupt
and wants to start running.  In fact the idle threads will all start
with the H_CEDE hcall returning; being idle they will just do another
H_CEDE immediately and go to nap mode.

This required some changes to kvmppc_run_core() and kvmppc_run_vcpu().
These functions have been restructured to make them simpler and clearer.
We introduce a level of indirection in the wait queue that gets woken
when external and decrementer interrupts get generated for a vcpu, so
that we can have the 4 vcpus in a vcore using the same wait queue.
We need this because the 4 vcpus are being handled by one thread.

Secondly, when we need to exit from the guest to the kernel, we now
have to generate an IPI for any napping threads, because an HDEC
interrupt doesn't wake up a napping thread.

Thirdly, we now need to be able to handle virtual external interrupts
and decrementer interrupts becoming pending while a thread is napping,
and deliver those interrupts to the guest when the thread wakes.
This is done in kvmppc_cede_reentry, just before fast_guest_return.

Finally, since we are not using the generic kvm_vcpu_block for book3s_hv,
and hence not calling kvm_arch_vcpu_runnable, we can remove the #ifdef
from kvm_arch_vcpu_runnable.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-09-25 19:52:30 +03:00
..
vdso32 Fix common misspellings 2011-03-31 11:26:23 -03:00
vdso64 Fix common misspellings 2011-03-31 11:26:23 -03:00
.gitignore
align.c powerpc: Remove fpscr use from [kvm_]cvt_{fd,df} 2010-09-02 14:07:32 +10:00
asm-offsets.c KVM: PPC: Implement H_CEDE hcall for book3s_hv in real-mode code 2011-09-25 19:52:30 +03:00
audit.c
btext.c Fix common misspellings 2011-03-31 11:26:23 -03:00
cacheinfo.c include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
cacheinfo.h
clock.c
compat_audit.c
cpu_setup_6xx.S powerpc: Fix some 6xx/7xxx CPU setup functions 2011-02-07 12:57:11 +11:00
cpu_setup_44x.S powerpc/44x: Add support for the AMCC APM821xx SoC 2010-10-13 08:47:09 -04:00
cpu_setup_a2.S powerpc: Add A2 cpu support 2011-04-27 13:02:02 +10:00
cpu_setup_fsl_booke.S powerpc/e5500: set non-base IVORs 2011-05-19 00:36:43 -05:00
cpu_setup_pa6t.S
cpu_setup_power7.S powerpc, KVM: Split HVMODE_206 cpu feature bit into separate HV and architecture bits 2011-07-12 13:16:58 +03:00
cpu_setup_ppc970.S powerpc, KVM: Split HVMODE_206 cpu feature bit into separate HV and architecture bits 2011-07-12 13:16:58 +03:00
cputable.c powerpc: Return the_cpu_ spec from identify_cpu 2011-08-05 14:47:54 +10:00
crash.c powerpc/kdump: Fix timeout in crash_kexec_wait_realmode 2011-07-19 15:13:05 +10:00
crash_dump.c crash_dump: export is_kdump_kernel to modules, consolidate elfcorehdr_addr, setup_elfcorehdr and saved_max_pfn 2011-03-23 19:47:19 -07:00
dbell.c powerpc: Consolidate ipi message mux and demux 2011-05-19 15:31:03 +10:00
dma-iommu.c powerpc/iommu: Use coherent_dma_mask for alloc_coherent 2010-12-09 15:17:50 +11:00
dma-swiotlb.c of: Merge of_platform_bus_type with platform_bus_type 2010-07-24 09:57:51 -06:00
dma.c powerpc: Dont require a dma_ops struct to set dma mask 2011-07-08 00:21:36 -05:00
e500-pmu.c perf, arch: Add generic NODE cache events 2011-07-01 11:06:38 +02:00
entry_32.S powerpc/ppc32/tracing: Add stack frame to calls of trace_hardirqs_on/off 2011-01-21 14:08:33 +11:00
entry_64.S powerpc: Free up some CPU feature bits by moving out MMU-related features 2011-04-27 14:18:52 +10:00
exceptions-64e.S powerpc/book3e-64: Reraise doorbell when masked by soft-irq-disable 2011-06-29 16:40:59 +10:00
exceptions-64s.S KVM: PPC: Assemble book3s{,_hv}_rmhandlers.S separately 2011-09-25 19:52:28 +03:00
firmware.c
fpu.S powerpc: Remove second definition of STACK_FRAME_OVERHEAD 2010-11-29 15:48:23 +11:00
fsl_booke_entry_mapping.S powerpc/fsl-booke: Fix address issue when using relocatable kernels 2010-07-11 11:04:08 -05:00
ftrace.c powerpc/ftrace: Implement raw syscall tracepoints on PowerPC 2011-05-26 13:38:57 +10:00
head_8xx.S powerpc: Remove second definition of STACK_FRAME_OVERHEAD 2010-11-29 15:48:23 +11:00
head_32.S powerpc: Remove last piece of GEMINI 2011-05-19 17:32:29 +10:00
head_40x.S Fix common misspellings 2011-03-31 11:26:23 -03:00
head_44x.S powerpc/47x: allow kernel to be loaded in higher physical memory 2011-07-12 10:34:24 -04:00
head_64.S powerpc: Fix early boot accounting of CPUs 2011-06-17 16:19:51 +10:00
head_booke.h powerpc/85xx: Save scratch registers to thread info instead of using SPRGs. 2011-06-22 21:44:55 -05:00
head_fsl_booke.S Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
hw_breakpoint.c powerpc, hw_breakpoint: Tell generic code we have no instruction breakpoints 2010-06-30 13:54:58 +10:00
ibmebus.c PM / Hibernate: Introduce CONFIG_HIBERNATE_CALLBACKS 2011-04-11 22:54:42 +02:00
idle.c powerpc: Re-enable preemption before cpu_die() 2010-08-24 15:26:29 +10:00
idle_6xx.S
idle_book3e.S powerpc/book3e: Add generic 64-bit idle powersave support 2010-07-14 14:13:18 +10:00
idle_e500.S powerpc/e500mc: Add support for the wait instruction in e500_idle 2011-06-27 08:36:15 -05:00
idle_power4.S powerpc/pmac/smp: Properly NAP offlined CPU on G5 2011-04-01 15:37:25 +11:00
idle_power7.S KVM: PPC: Allow book3s_hv guests to use SMT processor modes 2011-07-12 13:16:57 +03:00
init_task.c
io-workarounds.c powerpc/pci: Properly initialize IO workaround "private" 2011-04-27 14:18:33 +10:00
io.c
iomap.c powerpc: Really fix build without CONFIG_PCI 2011-08-11 01:15:44 +10:00
iommu.c powerpc: iommu: Add device name to iommu error printks 2010-12-09 15:35:32 +11:00
irq.c powerpc: Copy back TIF flags on return from softirq stack 2011-07-22 13:38:58 +10:00
isa-bridge.c
jump_label.c powerpc: Add jump label support 2011-07-01 13:48:55 +10:00
kgdb.c Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial 2011-05-23 09:12:26 -07:00
kprobes.c powerpc/kprobes: Remove resume_execution() in kprobes 2010-06-02 17:50:37 +10:00
kvm.c PPC: KVM: Book E doesn't have __end_interrupts. 2010-11-05 14:42:27 -02:00
kvm_emul.S KVM: PPC: Make PV mtmsrd L=1 work with r30 and r31 2010-10-24 10:52:14 +02:00
l2cr_6xx.S Fix common misspellings 2011-03-31 11:26:23 -03:00
legacy_serial.c powerpc: Check device status before adding serial device 2011-04-12 06:29:21 -05:00
lparcfg.c powerpc/pseries: Add page coalescing support 2011-05-04 16:02:21 +10:00
machine_kexec.c powerpc: Move kdump default base address to half RMO size on 64bit 2011-08-05 14:47:56 +10:00
machine_kexec_32.c powerpc/kexec: make masking/disabling interrupts generic 2010-10-14 00:52:46 -05:00
machine_kexec_64.c powerpc/kexec: Fix orphaned offline CPUs across kexec 2010-07-31 15:05:22 +10:00
Makefile powerpc: Add jump label support 2011-07-01 13:48:55 +10:00
misc.S powerpc: Remove unneeded cpu_setup/restore from POWER7 cputable entry 2010-11-29 15:48:22 +11:00
misc_32.S powerpc: Fix 32-bit SMP build 2011-05-20 16:23:19 -07:00
misc_64.S powerpc/maple: Enable scom access functions on Maple 2011-06-29 17:48:20 +10:00
module.c modules: make arch's use default loader hooks 2011-07-24 22:06:04 +09:30
module_32.c modules: make arch's use default loader hooks 2011-07-24 22:06:04 +09:30
module_64.c modules: make arch's use default loader hooks 2011-07-24 22:06:04 +09:30
mpc7450-pmu.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
msi.c
nvram_64.c powerpc/nvram: Generalize code for OS partitions in NVRAM 2011-03-04 18:19:04 +11:00
of_platform.c atomic: use <linux/atomic.h> 2011-07-26 16:49:47 -07:00
paca.c powerpc, KVM: Split HVMODE_206 cpu feature bit into separate HV and architecture bits 2011-07-12 13:16:58 +03:00
pci-common.c Merge branch 'next/cross-platform' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/linux-arm-soc 2011-07-26 17:12:10 -07:00
pci_32.c Merge branch 'next/cross-platform' of git://git.kernel.org/pub/scm/linux/kernel/git/arm/linux-arm-soc 2011-07-26 17:12:10 -07:00
pci_64.c powerpc: rename ppc_pci_*_flags to pci_*_flags 2011-07-12 09:28:04 -05:00
pci_dn.c pci/of: Match PCI devices to OF nodes dynamically 2011-06-08 09:08:17 +10:00
pci_of_scan.c pci/of: Match PCI devices to OF nodes dynamically 2011-06-08 09:08:17 +10:00
perf_callchain.c powerpc/perf: Disable pagefaults during callchain stack read 2011-08-05 14:47:56 +10:00
perf_event.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
perf_event_fsl_emb.c perf: Remove the nmi parameter from the swevent and overflow interface 2011-07-01 11:06:35 +02:00
pmc.c powerpc: Convert pmc_owner_lock to raw_spinlock 2010-02-19 14:52:33 +11:00
power4-pmu.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
power5+-pmu.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
power5-pmu.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
power6-pmu.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
power7-pmu.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
ppc32.h
ppc970-pmu.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
ppc_ksyms.c atomic: use <linux/atomic.h> 2011-07-26 16:49:47 -07:00
ppc_save_regs.S Fix common misspellings 2011-03-31 11:26:23 -03:00
proc_powerpc.c include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
process.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
prom.c Merge remote branch 'origin/master' into next 2011-06-30 15:23:59 +10:00
prom_init.c powerpc: Clean up some panic messages in prom_init 2011-08-05 14:47:55 +10:00
prom_init_check.sh powerpc: Fix compile errors in prom_init_check for gcc 4.5 2010-07-08 18:11:39 +10:00
prom_parse.c of/pci: move of_irq_map_pci() into generic code 2011-02-04 11:46:50 -07:00
ptrace.c perf: Add context field to perf_event 2011-07-01 11:06:38 +02:00
ptrace32.c powerpc: Update compat_arch_ptrace 2010-12-09 15:35:32 +11:00
reloc_64.S
rtas-proc.c
rtas-rtc.c powerpc/rtas-rtc: remove sideeffects of printk_ratelimit 2011-06-29 15:30:43 +10:00
rtas.c atomic: use <linux/atomic.h> 2011-07-26 16:49:47 -07:00
rtas_flash.c notifiers: sys: move reboot notifiers into reboot.h 2011-07-25 20:57:14 -07:00
rtas_pci.c powerpc: rename ppc_pci_*_flags to pci_*_flags 2011-07-12 09:28:04 -05:00
rtasd.c atomic: use <linux/atomic.h> 2011-07-26 16:49:47 -07:00
setup-common.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
setup.h
setup_32.c Merge remote-tracking branch 'jwb/next' into next 2011-07-22 13:16:41 +10:00
setup_64.c Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc 2011-07-25 22:59:39 -07:00
signal.c powerpc: fix double syscall restarts 2010-09-22 09:33:50 -07:00
signal.h
signal_32.c arch/powerpc: use printk_ratelimited instead of printk_ratelimit 2011-06-29 15:31:01 +10:00
signal_64.c arch/powerpc: use printk_ratelimited instead of printk_ratelimit 2011-06-29 15:31:01 +10:00
smp-tbsync.c atomic: use <linux/atomic.h> 2011-07-26 16:49:47 -07:00
smp.c atomic: use <linux/atomic.h> 2011-07-26 16:49:47 -07:00
softemu8xx.c include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h 2010-03-30 22:02:32 +09:00
stacktrace.c
suspend.c update email address 2010-07-19 10:56:54 +02:00
swsusp.c PM / Hibernate: Remove arch_prepare_suspend() 2011-05-24 23:35:55 +02:00
swsusp_32.S Fix common misspellings 2011-03-31 11:26:23 -03:00
swsusp_64.c
swsusp_asm64.S
swsusp_booke.S powerpc/fsl-booke: Add hibernation support for FSL BookE processors 2010-05-21 07:41:53 -05:00
sys_ppc32.c BKL: remove extraneous #include <smp_lock.h> 2010-11-17 08:59:32 -08:00
syscalls.c Add generic sys_olduname() 2010-03-12 15:52:32 -08:00
sysfs.c powerpc: Per process DSCR + some fixes (try#4) 2011-04-27 14:18:19 +10:00
systbl.S
systbl_chk.c
systbl_chk.sh
tau_6xx.c
time.c irq_work, ppc: Fix up arch hooks 2011-07-01 11:02:22 +02:00
traps.c powerpc/e500: Save SPEFCSR in flush_spe_to_thread() 2011-07-12 13:16:30 +03:00
udbg.c powerpc/pseries: Re-implement HVSI as part of hvc_vio 2011-06-29 17:48:35 +10:00
udbg_16550.c powerpc: Add early debug for WSP platforms 2011-05-06 13:32:41 +10:00
vdso.c mm: arch: rename in_gate_area_no_task to in_gate_area_no_mm 2011-03-23 16:36:55 -04:00
vecemu.c
vector.S powerpc: Remove static branch hint in giveup_altivec 2011-05-19 14:30:42 +10:00
vio.c powerpc/iommu: Use coherent_dma_mask for alloc_coherent 2010-12-09 15:17:50 +11:00
vmlinux.lds.S percpu: Always align percpu output section to PAGE_SIZE 2011-03-24 18:50:09 +01:00