Re: [PATCH v3 3/8] x86: detect and initialize Intel CAT feature

From: Andrew Cooper <andrew.cooper3@citrix.com>
To: Jan Beulich <JBeulich@suse.com>,
	Chao Peng <chao.p.peng@linux.intel.com>,
	xen-devel@lists.xen.org
Cc: wei.liu2@citrix.com, Ian.Campbell@citrix.com,
	stefano.stabellini@eu.citrix.com, Ian.Jackson@eu.citrix.com,
	will.auld@intel.com, keir@xen.org, dgdegra@tycho.nsa.gov
Subject: Re: [PATCH v3 3/8] x86: detect and initialize Intel CAT feature
Date: Mon, 13 Apr 2015 11:58:02 +0100	[thread overview]
Message-ID: <552BA13A.6060100@citrix.com> (raw)
In-Reply-To: <552BBBE20200007800071685@mail.emea.novell.com>

On 13/04/15 11:51, Jan Beulich wrote:
>>>> On 27.03.15 at 19:31, <andrew.cooper3@citrix.com> wrote:
>> On 26/03/15 12:38, Chao Peng wrote:
>>> +    cpuid_count(PSR_CPUID_LEVEL_CAT, 0, &eax, &ebx, &ecx, &edx);
>>> +    if ( ebx & PSR_RESOURCE_TYPE_L3 )
>>> +    {
>>> +        cpuid_count(PSR_CPUID_LEVEL_CAT, 1, &eax, &ebx, &ecx, &edx);
>>> +        info->cbm_len = (eax & 0x1f) + 1;
>>> +        info->cos_max = (edx & 0xffff);
>>> +
>>> +        info->enabled = 1;
>>> +        printk(XENLOG_DEBUG "CAT: enabled on socket %u, cos_max:%u, cbm_len:%u\n",
>>> +               socket, info->cos_max, info->cbm_len);
>> I would bump this to INFO, at least for socket 0.  Similar to CMT, a
>> user which has gone to the effort of enabling CAT will want some
>> indication in the boot log without also having to bump the logging level.
> Which INFO wouldn't be sufficient for (at least not with the xen.org
> defaults, not sure about XenServer's).

Huh - you are right.  XenServer doesn't patch this, but I spend most of
my time using debug builds so I tend not to notice.

~Andrew