leandrof/linux
1
0
Fork 0
forked from Minki/linux
Code Issues Pull Requests Packages Projects Releases Wiki Activity
74d4affde8
linux/arch/x86/kernel
History
Jeremy Fitzhardinge 74d4affde8 x86/paravirt: add hooks for spinlock operations 
Ticket spinlocks have absolutely ghastly worst-case performance
characteristics in a virtual environment.  If there is any contention
for physical CPUs (ie, there are more runnable vcpus than cpus), then
ticket locks can cause the system to end up spending 90+% of its time
spinning.

The problem is that (v)cpus waiting on a ticket spinlock will be
granted access to the lock in strict order they got their tickets.  If
the hypervisor scheduler doesn't give the vcpus time in that order,
they will burn timeslices waiting for the scheduler to give the right
vcpu some time.  In the worst case it could take O(n^2) vcpu scheduler
timeslices for everyone waiting on the lock to get it, not counting
new cpus trying to take the lock while the log-jam is sorted out.

These hooks allow a paravirt backend to replace the spinlock
implementation.

At the very least, this could revert the implementation back to the
old lock algorithm, which allows the next scheduled vcpu to take the
lock, and has basically fairly good performance.

It also allows the spinlocks to take advantages of the hypervisor
features to make locks more efficient (spin and block, for example).

The cost to native execution is an extra direct call when using a
spinlock function.  There's no overhead if CONFIG_PARAVIRT is turned
off.

The lock structure is fixed at a single "unsigned int", initialized to
zero, but the spinlock implementation can use it as it wishes.

Thanks to Thomas Friebel's Xen Summit talk "Preventing Guests from
Spinning Around" for pointing out this problem.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Christoph Lameter <clameter@linux-foundation.org>
Cc: Petr Tesarik <ptesarik@suse.cz>
Cc: Virtualization <virtualization@lists.linux-foundation.org>
Cc: Xen devel <xen-devel@lists.xensource.com>
Cc: Thomas Friebel <thomas.friebel@amd.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-07-16 11:15:52 +02:00
..
acpi Merge branch 'linus' into x86/urgent 2008-07-15 13:45:59 +02:00
cpu x86_64: further cleanup of 32-bit compat syscall mechanisms 2008-07-16 11:08:27 +02:00
.gitignore arch/x86/kernel/.gitignore: Added vmlinux.lds to .gitignore file because it shouldn't be tracked. 2008-07-10 10:13:51 -07:00
alternative.c x86: fix SMP alternatives: use mutex instead of spinlock, text_poke is sleepable 2008-05-23 21:56:52 +02:00
amd_iommu_init.c x86, AMD IOMMU: remove unnecessary code from the iommu_enable function 2008-07-04 11:44:44 +02:00
amd_iommu.c x86, AMD IOMMU: flush domain TLB when there is more than one page to flush 2008-07-04 11:44:40 +02:00
aperture_64.c x86: remove end_pfn in 64bit 2008-07-08 13:10:38 +02:00
apic_32.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
apic_64.c x86: nmi_watchdog - use NMI_NONE by default 2008-07-08 12:51:41 +02:00
apm_32.c Merge branch 'bkl-removal' of git://git.lwn.net/linux-2.6 2008-07-14 14:48:31 -07:00
asm-offsets_32.c x86/paravirt: split sysret and sysexit 2008-07-08 13:13:15 +02:00
asm-offsets_64.c xen64: add asm-offsets 2008-07-16 10:58:55 +02:00
asm-offsets.c
audit_64.c
bootflag.c x86: coding style cleanup for kernel/bootflag.c 2008-01-30 13:32:31 +01:00
cpuid.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
crash_dump_32.c
crash_dump_64.c
crash.c x86: allow machine_crash_shutdown to be replaced 2008-04-27 12:00:29 +03:00
doublefault_32.c x86: unify tss_struct 2008-01-30 13:31:31 +01:00
ds.c x86: debug Store - call kfree if only we really need it 2008-04-17 17:41:34 +02:00
e820.c x86, e820: remove end_user_pfn 2008-07-13 08:19:40 +02:00
early_printk.c x86, generic: mark early_printk as asmlinkage 2008-06-18 13:11:01 +02:00
early-quirks.c x86: remove end_pfn in 64bit 2008-07-08 13:10:38 +02:00
efi_32.c x86: fix Intel Mac booting with EFI 2008-07-03 08:19:18 +02:00
efi_64.c x86: add flags parameter to reserve_bootmem_generic() 2008-07-08 11:49:49 +02:00
efi_stub_32.S
efi_stub_64.S x86: EFI runtime service support 2008-01-30 13:31:19 +01:00
efi.c x86: introduce max_low_pfn_mapped for 64-bit 2008-07-11 10:24:04 +02:00
entry_32.S Merge branch 'auto-ftrace-next' into tracing/for-linus 2008-07-14 16:11:52 +02:00
entry_64.S x86_64: adjust exception frame on paranoid exceptions 2008-07-16 11:08:58 +02:00
ftrace.c ftrace: store mcount address in rec->ip 2008-06-23 22:10:56 +02:00
genapic_64.c x86: introduce max_physical_apicid for bigsmp switching 2008-06-10 11:32:09 +02:00
genapic_flat_64.c x86: use cpumask_of_cpu() 2008-04-17 17:41:36 +02:00
genx2apic_uv_x.c x86: map UV chipset space - UV support 2008-07-09 07:43:23 +02:00
geode_32.c x86, geode: add a VSA2 ID for General Software 2008-06-19 14:19:03 +02:00
head32.c x86: extract common part of head32.c and head64.c into head.c 2008-06-05 15:10:02 +02:00
head64.c x86_64: add workaround for no %gs-based percpu 2008-07-16 10:58:13 +02:00
head_32.S x86: boot secondary cpus through initial_code 2008-07-08 12:48:18 +02:00
head_64.S Revert "x86_64: there's no need to preallocate level1_fixmap_pgt" 2008-07-16 11:07:30 +02:00
head.c x86: move reserve_setup_data to setup.c 2008-07-08 13:16:14 +02:00
hpet.c x86: make 64bit hpet_set_mapping to use ioremap too, v2 2008-07-14 09:24:17 +02:00
i386_ksyms_32.c ftrace: store mcount address in rec->ip 2008-06-23 22:10:56 +02:00
i387.c x86 ptrace: fix PTRACE_GETFPXREGS error 2008-07-01 11:03:31 +02:00
i8237.c
i8253.c x86: cleanup div_sc() usage 2008-04-26 17:35:47 +02:00
i8259.c i8259: fix final ugliness 2008-06-02 11:55:52 +02:00
init_task.c [PATCH] take init_files to fs/file.c 2008-05-16 17:22:20 -04:00
io_apic_32.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
io_apic_64.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
io_delay.c x86: add dmi quirk for io_delay 2008-03-26 22:23:40 +01:00
ioport.c x86: refactor ioport unification 2008-01-30 13:33:10 +01:00
ipi.c x86: ipi.c: removed duplicated include 2008-05-25 08:58:32 +02:00
irq_32.c x86: use __page_aligned_data/bss 2008-07-16 10:54:39 +02:00
irq_64.c x86: make /proc/stat account for all interrupts 2008-05-25 07:11:49 +02:00
irqinit_32.c x86: rename the i8259_32/64.c leftovers to irqinit_32/64.c 2008-05-24 16:47:39 +02:00
irqinit_64.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
k8.c
kdebugfs.c x86, boot: export linked list of struct setup_data via debugfs 2008-04-26 21:34:42 +02:00
kgdb.c x86: KGDB build fix 2008-04-19 19:19:54 +02:00
kprobes.c x86: replace most VM86 flags with flags from processor-flags.h 2008-04-17 17:41:33 +02:00
kvm.c x86: KVM guest: hypercall batching 2008-04-27 12:00:28 +03:00
kvmclock.c x86: KVM guest: Use the paravirt clocksource structs and functions 2008-06-24 21:02:33 +03:00
ldt.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
machine_kexec_32.c Merge branch 'auto-ftrace-next' into tracing/for-linus 2008-07-14 16:11:52 +02:00
machine_kexec_64.c Merge branch 'auto-ftrace-next' into tracing/for-linus 2008-07-14 16:11:52 +02:00
Makefile Merge branch 'auto-ftrace-next' into tracing/for-linus 2008-07-14 16:11:52 +02:00
mca_32.c x86: coding style fixes to arch/x86/kernel/mca_32.c 2008-04-17 17:40:49 +02:00
mfgpt_32.c geode: fix modular build 2008-06-12 21:25:51 +02:00
microcode.c Merge branch 'for-2.6.27' of git://git.infradead.org/users/dwmw2/firmware-2.6 2008-07-14 16:54:07 -07:00
mmconf-fam10h_64.c x86: move mmconfig declarations to header 2008-05-30 15:45:17 -07:00
module_32.c
module_64.c
mpparse.c x86: add early quirk support 2008-07-10 18:55:31 +02:00
msr.c x86: cdev lock_kernel() pushdown 2008-05-18 15:43:40 -06:00
nmi.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
numaq_32.c x86: fix numaq_tsc_disable calling 2008-07-13 08:19:45 +02:00
olpc.c x86: olpc: add One Laptop Per Child architecture support 2008-04-29 08:06:07 -07:00
paravirt_patch_32.c x86/paravirt: split sysret and sysexit 2008-07-08 13:13:15 +02:00
paravirt_patch_64.c x86/paravirt: add sysret/sysexit pvops for returning to 32-bit compatibility userspace 2008-07-08 13:15:52 +02:00
paravirt.c x86/paravirt: add hooks for spinlock operations 2008-07-16 11:15:52 +02:00
pci-calgary_64.c x86: remove end_pfn in 64bit 2008-07-08 13:10:38 +02:00
pci-dma.c x86: remove end_pfn in 64bit 2008-07-08 13:10:38 +02:00
pci-gart_64.c x86: max_low_pfn_mapped fix #4 2008-07-14 09:24:16 +02:00
pci-nommu.c x86: unify pci-nommu 2008-04-19 19:19:57 +02:00
pci-swiotlb_64.c x86: remove end_pfn in 64bit 2008-07-08 13:10:38 +02:00
pcspeaker.c
pmtimer_64.c x86: move pmtmr related declarations 2008-01-30 13:30:18 +01:00
probe_roms_32.c x86: seperate probe_roms into another file 2008-07-08 12:50:05 +02:00
process_32.c Merge branch 'auto-ftrace-next' into tracing/for-linus 2008-07-14 16:11:52 +02:00
process_64.c x86: clean up formatting of __switch_to 2008-07-16 10:54:25 +02:00
process.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
ptrace.c x86 ptrace: fix PTRACE_GETFPXREGS error 2008-07-01 11:03:31 +02:00
pvclock.c x86: Add structs and functions for paravirt clocksource 2008-06-24 21:02:31 +03:00
quirks.c Merge branch 'timers/for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2008-07-15 10:39:57 -07:00
reboot_fixups_32.c x86: constify data in reboot.c 2008-05-25 08:58:30 +02:00
reboot.c x86: constify data in reboot.c 2008-05-25 08:58:30 +02:00
relocate_kernel_32.S x86: relocate_kernel - use predefined macroses for page attributes 2008-04-17 17:41:29 +02:00
relocate_kernel_64.S x86: relocate_kernel - use predefined macroses for page attributes 2008-04-17 17:41:29 +02:00
rtc.c provide rtc_cmos platform device 2008-06-12 18:05:42 -07:00
scx200_32.c x86: fix sparse warning in kernel/scx200_32.c 2008-01-31 22:05:45 +01:00
setup_percpu.c x86: change _node_to_cpumask_ptr to return const ptr 2008-07-13 19:11:58 +02:00
setup.c pvops-64: call paravirt_post_allocator_init() on setup_arch() 2008-07-16 10:53:57 +02:00
sigframe.h x86: move struct definitions to unifed sigframe.h 2008-04-17 17:40:46 +02:00
signal_32.c signals: x86 TS_RESTORE_SIGMASK 2008-04-30 08:29:37 -07:00
signal_64.c signals: x86 TS_RESTORE_SIGMASK 2008-04-30 08:29:37 -07:00
smp.c generic-ipi: fixlet 2008-07-06 14:01:50 +02:00
smpboot.c x86_64: unstatic get_local_pda 2008-07-16 10:55:07 +02:00
smpcommon_32.c x86: create smpcommon.c 2008-04-17 17:40:55 +02:00
smpcommon.c x86: convert to generic helpers for IPI function calls 2008-06-26 11:21:54 +02:00
stacktrace.c stacktrace: fix modular build, export print_stack_trace and save_stack_trace 2008-06-30 09:20:55 +02:00
step.c x86: prevent unconditional writes to DebugCtl MSR 2008-04-17 17:40:58 +02:00
summit_32.c x86: make generic arch support NUMAQ 2008-06-10 11:34:42 +02:00
sys_i386_32.c x86: coding style fixes to arch/x86/kernel/sys_i386_32.c 2008-06-10 12:34:54 +02:00
sys_x86_64.c unified (weak) sys_pipe implementation 2008-05-03 13:50:33 -07:00
syscall_64.c x86: coding style fixes to arch/x86/kernel/syscall_64.c 2008-04-17 17:40:48 +02:00
syscall_table_32.S timerfd: wire the new timerfd API to the x86 family 2008-02-05 09:44:07 -08:00
tce_64.c
test_nx.c x86: Explicitly include required header files. 2008-04-17 17:41:15 +02:00
test_rodata.c x86: include proper prototypes for rodata_test 2008-02-14 23:30:20 +01:00
time_32.c x86: merge sched_clock handling 2008-07-09 07:43:25 +02:00
time_64.c x86: merge tsc_init and clocksource code 2008-07-09 07:43:27 +02:00
tlb_32.c on_each_cpu(): kill unused 'retry' parameter 2008-06-26 11:24:38 +02:00
tlb_64.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
tlb_uv.c x86, SGI UV: uv_ptc_proc_write fix 2008-07-08 12:23:31 +02:00
tls.c asmlinkage_protect replaces prevent_tail_call 2008-04-10 17:28:26 -07:00
tls.h x86: x86 user_regset TLS 2008-01-30 13:31:52 +01:00
topology.c x86: fix section mismatch warning in topology.c:arch_register_cpu 2008-02-19 16:18:30 +01:00
trampoline_32.S x86: trampoline_32.S - switch to .cpuinit.data 2008-04-26 17:35:47 +02:00
trampoline_64.S x86: move suspend wakeup code to C 2008-04-17 17:41:37 +02:00
trampoline.c x86: extend e820 ealy_res support 32bit 2008-05-25 10:55:11 +02:00
traps_32.c x86: traps_xx: various small changes 2008-07-09 08:18:31 +02:00
traps_64.c Merge branch 'x86/for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2008-07-14 13:43:24 -07:00
tsc_sync.c x86: add warning to check_tsc_warp() 2008-01-30 13:33:24 +01:00
tsc.c x86: fix TSC build error on 32bit 2008-07-15 22:46:47 +02:00
verify_cpu_64.S
visws_quirks.c x86, VisWS: build fix 2008-07-10 19:45:01 +02:00
vm86_32.c x86: replace most VM86 flags with flags from processor-flags.h 2008-04-17 17:41:33 +02:00
vmi_32.c x86: rename paravirtualized TSC functions 2008-07-09 07:43:28 +02:00
vmiclock_32.c x86: rename paravirtualized TSC functions 2008-07-09 07:43:28 +02:00
vmlinux_32.lds.S Merge branch 'core/rodata' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2008-07-14 15:28:10 -07:00
vmlinux_64.lds.S Merge branch 'core/rodata' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip 2008-07-14 15:28:10 -07:00
vmlinux.lds.S
vsmp_64.c x86: vsmp_64 add missing includes 2008-05-25 08:58:24 +02:00
vsyscall_64.c Merge branch 'generic-ipi' into generic-ipi-for-linus 2008-07-15 21:55:59 +02:00
x8664_ksyms_64.c Merge branch 'auto-ftrace-next' into tracing/for-linus 2008-07-14 16:11:52 +02:00