linux-kernel - Re: [PATCH v3] x86: mm: Do not use set_{pud,pmd}

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: <ae2c9081-f61c-a33a-67f0-2e05ef702e61@intel.com>
Date:   Wed, 17 Apr 2019 10:24:24 -0700
From:   Dave Hansen <dave.hansen@...el.com>
To:     "Singh, Brijesh" <brijesh.singh@....com>,
        "x86@...nel.org" <x86@...nel.org>
Cc:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Dan Williams <dan.j.williams@...el.com>,
        "Kirill A . Shutemov" <kirill.shutemov@...ux.intel.com>,
        Andy Lutomirski <luto@...nel.org>,
        Borislav Petkov <bp@...en8.de>,
        "H . Peter Anvin" <hpa@...or.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        "Lendacky, Thomas" <Thomas.Lendacky@....com>
Subject: Re: [PATCH v3] x86: mm: Do not use set_{pud,pmd}_safe when splitting
 the large page

On 4/17/19 8:41 AM, Singh, Brijesh wrote:
> Changes since v2:
> - rename variable safe->init
> - rename __set_* -> set_*_init()
> - remame __*_populate -> *_populate_init()

"init" in this context really means "setting non-present PTEs is OK, but
changing present ones is not".  It still doesn't do a great job of
conveying that because both the init=0 and init=1 variants are used
during init-time.

> +/*
> + * Create page table mapping for the physical memory for specific physical
> + * addresses. The virtual and physical addresses have to be aligned on PMD level
> + * down. It returns the last physical address mapped.
> + */
> +unsigned long __meminit
> +kernel_physical_mapping_init(unsigned long paddr_start,
> +			     unsigned long paddr_end,
> +			     unsigned long page_size_mask)
> +{
> +	return __kernel_physical_mapping_init(paddr_start, paddr_end,
> +					      page_size_mask, true);
> +}

The comment here is missing one key point:
kernel_physical_mapping_init() can only be used to populate non-present
entries.  It kinda infers that from "Create...", but I really think we
need to be explicit.

Anyway, it's better than it was, and it does fix a bug, so:

Reviewed-by: Dave Hansen <dave.hansen@...el.com>

But please flesh out that comment if you do another version for some reason.