diff options
| author | Mauro Carvalho Chehab <mchehab@redhat.com> | 2013-04-30 09:00:33 -0300 |
|---|---|---|
| committer | Mauro Carvalho Chehab <mchehab@redhat.com> | 2013-04-30 09:00:33 -0300 |
| commit | aad797c89903d570c17f6affc770eb98afd74e62 (patch) | |
| tree | bddefd4242b0efba1068b5260f831c2697ba4037 /kernel/sched/clock.c | |
| parent | c95789ecd5a979fd718ae09763df3fa50dd97a91 (diff) | |
| parent | c1be5a5b1b355d40e6cf79cc979eb66dafa24ad1 (diff) | |
Merge tag 'v3.9' into v4l_for_linus
Linux 3.9
* tag 'v3.9': (1099 commits)
Linux 3.9
vm: add no-mmu vm_iomap_memory() stub
efivars: only check for duplicates on the registered list
TTY: fix atime/mtime regression
aio: fix possible invalid memory access when DEBUG is enabled
parisc: use spin_lock_irqsave/spin_unlock_irqrestore for PTE updates
parisc: disable -mlong-calls compiler option for kernel modules
parisc: uaccess: fix compiler warnings caused by __put_user casting
parisc: Change kunmap macro to static inline function
parisc: Provide __ucmpdi2 to resolve undefined references in 32 bit builds.
sparc64: Fix missing put_cpu_var() in tlb_batch_add_one() when not batching.
Revert "gpio: pxa: set initcall level to module init"
efi: Check EFI revision in setup_efi_vars
x86, efi: Fix a build warning
Revert "MIPS: page.h: Provide more readable definition for PAGE_MASK."
kernel/hz.bc: ignore.
Linux 3.9-rc8
events: Protect access via task_subsys_state_check()
net: fix incorrect credentials passing
x86, microcode: Verify the family before dispatching microcode patching
...
Diffstat (limited to 'kernel/sched/clock.c')
| -rw-r--r-- | kernel/sched/clock.c | 26 |
1 files changed, 26 insertions, 0 deletions
diff --git a/kernel/sched/clock.c b/kernel/sched/clock.c index c685e31492d..c3ae1446461 100644 --- a/kernel/sched/clock.c +++ b/kernel/sched/clock.c @@ -176,10 +176,36 @@ static u64 sched_clock_remote(struct sched_clock_data *scd) u64 this_clock, remote_clock; u64 *ptr, old_val, val; +#if BITS_PER_LONG != 64 +again: + /* + * Careful here: The local and the remote clock values need to + * be read out atomic as we need to compare the values and + * then update either the local or the remote side. So the + * cmpxchg64 below only protects one readout. + * + * We must reread via sched_clock_local() in the retry case on + * 32bit as an NMI could use sched_clock_local() via the + * tracer and hit between the readout of + * the low32bit and the high 32bit portion. + */ + this_clock = sched_clock_local(my_scd); + /* + * We must enforce atomic readout on 32bit, otherwise the + * update on the remote cpu can hit inbetween the readout of + * the low32bit and the high 32bit portion. + */ + remote_clock = cmpxchg64(&scd->clock, 0, 0); +#else + /* + * On 64bit the read of [my]scd->clock is atomic versus the + * update, so we can avoid the above 32bit dance. + */ sched_clock_local(my_scd); again: this_clock = my_scd->clock; remote_clock = scd->clock; +#endif /* * Use the opportunity that we have both locks |
