linux-kernel - Re: [PATCH] swiotlb: check dynamically allocated TLB address before decrypting

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] [day] [month] [year] [list]

Message-ID: <20231030150904.07a32eaa@meshulam.tesarici.cz>
Date:   Mon, 30 Oct 2023 15:09:04 +0100
From:   Petr Tesařík <petr@...arici.cz>
To:     Christoph Hellwig <hch@....de>
Cc:     Petr Tesarik <petrtesarik@...weicloud.com>,
        Marek Szyprowski <m.szyprowski@...sung.com>,
        Robin Murphy <robin.murphy@....com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Petr Tesarik <petr.tesarik1@...wei-partners.com>,
        "open list:DMA MAPPING HELPERS" <iommu@...ts.linux.dev>,
        open list <linux-kernel@...r.kernel.org>,
        Wangkefeng <wangkefeng.wang@...wei.com>,
        Roberto Sassu <roberto.sassu@...weicloud.com>,
        Michael Kelley <mikelley@...rosoft.com>
Subject: Re: [PATCH] swiotlb: check dynamically allocated TLB address before
 decrypting

Hi Christoph,

On Mon, 30 Oct 2023 14:31:12 +0100
Christoph Hellwig <hch@....de> wrote:

> I'm trying to review it properly this week.  It was defintively too big
> to just rush it into 6.6 in the last few days.

Thank you for the answer. This is OK. Let me give a bit of background.

The bug was reported by Michael Kelley to me, while I temporarily lost
access to my @huaweicloud.com mailbox. Then I was not able to add him
in a Reported-by: header, because this was a private email, which could
not be referred by a Closes: header.

Anyway, Michael explained in that private email that the threat is more
or less theoretical, because environments where set_memory_decrypted()
actually does something are unlikely to have physical address
constraints for the bounce buffer.

But maybe we should add a CC: stable@...r.kernel.org nevertheless.

Petr T