linux-kernel - Re: [PATCH v3] page cache: Store only head pages in i

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: <863F9255-E992-402F-827D-DA5F4661B9AB@oracle.com>
Date:   Wed, 6 Mar 2019 02:36:35 -0700
From:   William Kucharski <william.kucharski@...cle.com>
To:     Matthew Wilcox <willy@...radead.org>
Cc:     "Kirill A. Shutemov" <kirill@...temov.name>,
        Linux-MM <linux-mm@...ck.org>, linux-fsdevel@...r.kernel.org,
        open list <linux-kernel@...r.kernel.org>,
        Hugh Dickins <hughd@...gle.com>, Jan Kara <jack@...e.cz>,
        Song Liu <liu.song.a23@...il.com>
Subject: Re: [PATCH v3] page cache: Store only head pages in i_pages


Other than the bug Song found in memfd_tag_pins(), I'd like to suggest two quick
but pedantic changes to mm/filemap.c:

Though not modified in this patch, in line 284, the parenthesis should be moved
to after the period:

 * modified.) The function expects only THP head pages to be present in the

> +		 * Move to the next page in the vector if this is a small page
> +		 * or the index is of the last page in this compound page).

A few lines later, there is an extraneous parenthesis, and the comment could be a bit
clearer.

Might I suggest:

                 * Move to the next page in the vector if this is a PAGESIZE
                 * page or if the index is of the last PAGESIZE page within
                 * this compound page.

You can say "base" instead of "PAGESIZE," but "small" seems open to interpretation.

I haven't run across any problems and have been hammering the code for over five days
without issue; all my testing was with transparent_hugepage/enabled set to
"always."

Tested-by: William Kucharski <william.kucharski@...cle.com>
Reviewed-by: William Kucharski <william.kucharski@...cle.com>