linux-kernel - Re: [PATCH] mm: Handle ksize() vs __alloc

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: <Yhzcc1YfpgEXzKdh@casper.infradead.org>
Date:   Mon, 28 Feb 2022 14:30:11 +0000
From:   Matthew Wilcox <willy@...radead.org>
To:     Marco Elver <elver@...gle.com>
Cc:     Kees Cook <keescook@...omium.org>, llvm@...ts.linux.dev,
        Pekka Enberg <penberg@...nel.org>,
        David Rientjes <rientjes@...gle.com>,
        Joonsoo Kim <iamjoonsoo.kim@....com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Vlastimil Babka <vbabka@...e.cz>, linux-mm@...ck.org,
        stable@...r.kernel.org,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        "Rafael J. Wysocki" <rafael@...nel.org>,
        Christoph Lameter <cl@...ux.com>,
        Nathan Chancellor <nathan@...nel.org>,
        Nick Desaulniers <ndesaulniers@...gle.com>,
        Daniel Micay <danielmicay@...il.com>,
        linux-kernel@...r.kernel.org, linux-hardening@...r.kernel.org
Subject: Re: [PATCH] mm: Handle ksize() vs __alloc_size by forgetting size

On Mon, Feb 28, 2022 at 12:24:51PM +0100, Marco Elver wrote:
> 2. Somehow statically computing the size-class's size (kmalloc_index()
> might help here), removing __alloc_size from allocation functions and
> instead use some wrapper.

I don't think that's computable.  I have been thinking about a slab flag
that would say "speed is more important than size; if the smallest slab
for this size of allocation has no free objects, search larger slabs
to get memory instead of allocating a new slab".  If we did have such
a feature, it would be impossible to know how large ksize() would report.