linux-kernel - Re: [PATCH V2 01/10] x86/CPU: Expose if cache is inclusive of lower level caches

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: <18004821-577d-b0dd-62b8-13b6f9264e72@intel.com>
Date:   Tue, 6 Aug 2019 13:22:22 -0700
From:   Reinette Chatre <reinette.chatre@...el.com>
To:     Borislav Petkov <bp@...en8.de>
Cc:     tglx@...utronix.de, fenghua.yu@...el.com, tony.luck@...el.com,
        kuo-lang.tseng@...el.com, mingo@...hat.com, hpa@...or.com,
        x86@...nel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH V2 01/10] x86/CPU: Expose if cache is inclusive of lower
 level caches

Hi Borislav,

On 8/6/2019 12:16 PM, Borislav Petkov wrote:
> On Tue, Aug 06, 2019 at 11:53:40AM -0700, Reinette Chatre wrote:
>> In get_prefetch_disable_bits() the platforms that support cache
>> pseudo-locking are hardcoded as part of configuring the hardware
>> prefetch disable bits to use.
> 
> Ok, so there is already a way to check pseudo-locking support.

Correct. This check is per platform though.

> Now, why
> do we have to look at cache inclusivity too?

... because some platforms differ in which SKUs support cache
pseudo-locking. On these platforms only the SKUs with inclusive cache
support cache pseudo-locking, thus the additional check.

> 
> Your 0/10 mail says:
> 
> "Only systems with L3 inclusive cache is supported at this time because
> if the L3 cache is not inclusive then pseudo-locked memory within the L3
> cache would be evicted when migrated to L2."
> 
> but then a couple of mails earlier you said:
> 
> "... this seems to be different between L2 and L3. On the Atom systems
> where L2 pseudo-locking works well the L2 cache is not inclusive. We are
> also working on supporting cache pseudo-locking on L3 cache that is not
> inclusive."
> 
> which leads me to still think that we don't really need L3 cache
> inclusivity and theoretically, you could do without it.
> 
> Or are you saying that cache pseudo-locking on non-inclusive L3 is not
> supported yet so no need to enable it yet?

Correct. Hardware and software changes will be needed to support this.

Reinette