linux-kernel - Re: [PATCH v4 07/18] x86/intel_rdt: Add Haswell feature discovery

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <alpine.DEB.2.20.1610171257451.4912@nanos>
Date:   Mon, 17 Oct 2016 13:03:01 +0200 (CEST)
From:   Thomas Gleixner <tglx@...utronix.de>
To:     Fenghua Yu <fenghua.yu@...el.com>
cc:     "H. Peter Anvin" <h.peter.anvin@...el.com>,
        Ingo Molnar <mingo@...e.hu>, Tony Luck <tony.luck@...el.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Stephane Eranian <eranian@...gle.com>,
        Borislav Petkov <bp@...e.de>,
        Dave Hansen <dave.hansen@...el.com>,
        Nilay Vaish <nilayvaish@...il.com>, Shaohua Li <shli@...com>,
        David Carrillo-Cisneros <davidcc@...gle.com>,
        Ravi V Shankar <ravi.v.shankar@...el.com>,
        Sai Prakhya <sai.praneeth.prakhya@...el.com>,
        Vikas Shivappa <vikas.shivappa@...ux.intel.com>,
        linux-kernel <linux-kernel@...r.kernel.org>, x86 <x86@...nel.org>
Subject: Re: [PATCH v4 07/18] x86/intel_rdt: Add Haswell feature discovery

On Fri, 14 Oct 2016, Fenghua Yu wrote:
> +static inline bool cache_alloc_hsw_probe(void)
> +{
> +	u32 l, h;
> +	u32 max_cbm = BIT_MASK(20) - 1;

Two options here:

+	u32 l, h, max_cbm = BIT_MASK(20) - 1;

or

+	u32 max_cbm = BIT_MASK(20) - 1;
+	u32 l, h;

I personally prefer #1, but I can accept #2 as well. Both are quick to
parse while the one you chose is stopping the reading flow.

> +
> +	if (wrmsr_safe(IA32_L3_CBM_BASE, max_cbm, 0))
> +		return false;
> +	rdmsr(IA32_L3_CBM_BASE, l, h);
> +	if (l != max_cbm)
> +		return false;
> +
> +	return true;

  	return l == max_cbm;

Hmm?

> +}
> +
>  static inline bool get_rdt_resources(void)
>  {
>  	bool ret = false;
>  
> +	if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL &&
> +	    boot_cpu_data.x86 == 6 &&
> +	    boot_cpu_data.x86_model == INTEL_FAM6_HASWELL_X)
> +		return cache_alloc_hsw_probe();

Can you please stick that model check into the probe function and do:

    	if (cache_alloc_hsw_probe())
		return true;
> +
>  	if (!boot_cpu_has(X86_FEATURE_RDT_A))
>  		return false;
>  	if (boot_cpu_has(X86_FEATURE_CAT_L3))

Thanks,

	tglx