Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1604661
| From | Mark Rutland <mark.rutland@arm.com> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: [PATCH 15/17] arm64: arch_timer: Workaround for Cortex-A73 erratum 858921 |
| Date | 2017-03-20 16:00 +0100 |
| Message-ID | <tn6WS-6aK-21@gated-at.bofh.it> (permalink) |
| References | <thYZX-7uY-5@gated-at.bofh.it> <thYZY-7uY-39@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
On Mon, Mar 06, 2017 at 11:26:20AM +0000, Marc Zyngier wrote:
> Cortex-A73 (all versions) counter read can return a wrong value
> when the counter crosses a 32bit boundary.
>
> The workaround involves performing the read twice, and to return
> one or the other depending on whether a transition has taken place.
> +#ifdef CONFIG_ARM64_ERRATUM_858921
> +static u64 notrace arm64_858921_read_cntvct_el0(void)
> +{
> + u64 _old, _new;
> +
> + _old = read_sysreg(cntvct_el0);
> + _new = read_sysreg(cntvct_el0);
> + return (((_old ^ _new) >> 32) & 1) ? _old : _new;
> +}
> +#endif
Given this is a static inline, I think we can drop the underscores here.
Thanks,
Mark.
Back to linux.kernel | Previous | Next | Find similar | Unroll thread
Re: [PATCH 15/17] arm64: arch_timer: Workaround for Cortex-A73 erratum 858921 Mark Rutland <mark.rutland@arm.com> - 2017-03-20 16:00 +0100
csiph-web