[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.DEB.2.20.1610131736441.7777@nanos>
Date: Thu, 13 Oct 2016 17:39:34 +0200 (CEST)
From: Thomas Gleixner <tglx@...utronix.de>
To: Grzegorz Andrejczuk <grzegorz.andrejczuk@...el.com>
cc: mingo@...hat.com, hpa@...or.com, x86@...nel.org, bp@...e.de,
dave.hansen@...ux.intel.com, linux-kernel@...r.kernel.org,
lukasz.daniluk@...el.com, james.h.cownie@...el.com,
jacob.jun.pan@...el.com
Subject: Re: [PATCH v3 4/4] Add R3MWAIT to CPU features
On Thu, 13 Oct 2016, Grzegorz Andrejczuk wrote:
> --- a/arch/x86/kernel/cpu/common.c
> +++ b/arch/x86/kernel/cpu/common.c
> @@ -1108,6 +1108,9 @@ static void identify_cpu(struct cpuinfo_x86 *c)
> #endif
> /* The boot/hotplug time assigment got cleared, restore it */
> c->logical_proc_id = topology_phys_to_logical_pkg(c->phys_proc_id);
> +
> + if (cpu_has(c, X86_FEATURE_PHIR3MWAIT))
> + elf_hwcap2 |= HWCAP2_PHIR3MWAIT;
Why can't you just set that bit in probe_xeon_phi_r3mwait()?
Just because its so much fun to sprinkle code at random places, right?
Sigh.
tglx
Powered by blists - more mailing lists