linux-kernel - Re: [PATCH v8 4/7] x86/mm: Simplify clear_page

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: <87ikfxguhd.fsf@oracle.com>
Date: Wed, 29 Oct 2025 16:26:38 -0700
From: Ankur Arora <ankur.a.arora@...cle.com>
To: Borislav Petkov <bp@...en8.de>
Cc: Ankur Arora <ankur.a.arora@...cle.com>, linux-kernel@...r.kernel.org,
        linux-mm@...ck.org, x86@...nel.org, akpm@...ux-foundation.org,
        david@...hat.com, dave.hansen@...ux.intel.com, hpa@...or.com,
        mingo@...hat.com, mjguzik@...il.com, luto@...nel.org,
        peterz@...radead.org, acme@...nel.org, namhyung@...nel.org,
        tglx@...utronix.de, willy@...radead.org, raghavendra.kt@....com,
        boris.ostrovsky@...cle.com, konrad.wilk@...cle.com
Subject: Re: [PATCH v8 4/7] x86/mm: Simplify clear_page_*


Borislav Petkov <bp@...en8.de> writes:

> On Mon, Oct 27, 2025 at 01:21:06PM -0700, Ankur Arora wrote:
>> +/**
>> + * clear_page() - clear a page using a kernel virtual address.
>> + * @addr: address of kernel page
>> + *
>> + * Switch between three implementations of page clearing based on CPU
>> + * capabilities:
>> + *
>> + *  - memzero_page_aligned_unrolled(): the oldest, slowest and universally
>
> So I don't see how a memzero_<bla> name shows that it belongs to the
> clear_page "stack" of functions? clear_page_orig() kinda conveys better what
> this thing is. In any case, having "clear_page" somewhere there in the name
> should stay.

After this change the only remaining functions in x86/lib/clear_page_64.S are
this one and rep_stos_alternative.

And so the _orig suffix seemed to be a little confusing. That's why the
change to the more functional name.

>> + *    supported method. Zeroes via 8-byte MOV instructions unrolled 8x
>> + *    to write a 64-byte cacheline in each loop iteration..
> 							    ^
>
> one fullstop is enough.
>
>> + *
>> + *  - "rep stosq": really old CPUs had crummy REP implementations.
>
> We spell all x86 insns in ALL CAPS. Like you've almost done.
>
> Also, it is
>
> 	REP; STOSQ
>
> with a ;
>
> Otherwise the idea for the cleanup makes sense.

Thanks. Will fix the above.

--
ankur