diff options
| author | Andi Kleen <ak@suse.de> | 2006-12-07 02:14:12 +0100 |
|---|---|---|
| committer | Andi Kleen <andi@basil.nowhere.org> | 2006-12-07 02:14:12 +0100 |
| commit | 616779656989cb8c59177e35cb13e87028b1edc8 (patch) | |
| tree | afdd8188e3a845b879e5048eda6ea6909684d015 /arch | |
| parent | a1a70c25bed75ed36ed48bbe18b9029428d2452d (diff) | |
[PATCH] x86-64: Synchronize RDTSC on single core AMD
There is no guarantee that two RDTSCs in a row are monotonic,
so don't assume it on single core AMD systems.
This will make gettimeofday slower again
Signed-off-by: Andi Kleen <ak@suse.de>
Diffstat (limited to 'arch')
| -rw-r--r-- | arch/x86_64/kernel/setup.c | 7 |
1 files changed, 2 insertions, 5 deletions
diff --git a/arch/x86_64/kernel/setup.c b/arch/x86_64/kernel/setup.c index a570c81c8316..05eaca41802b 100644 --- a/arch/x86_64/kernel/setup.c +++ b/arch/x86_64/kernel/setup.c @@ -732,11 +732,8 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) /* Fix cpuid4 emulation for more */ num_cache_leaves = 3; - /* When there is only one core no need to synchronize RDTSC */ - if (num_possible_cpus() == 1) - set_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability); - else - clear_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability); + /* RDTSC can be speculated around */ + clear_bit(X86_FEATURE_SYNC_RDTSC, &c->x86_capability); } static void __cpuinit detect_ht(struct cpuinfo_x86 *c) |