netdev - Re: [Intel-wired-lan] [PATCH 1/1] ixgbe: sync the first fragment unconditionally

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: <CAKgT0UctiHjKxGXHJEiJX4_aEJt1swBAxMOKPT9huXBOcf+2VQ@mail.gmail.com>
Date:   Tue, 6 Aug 2019 12:52:51 -0700
From:   Alexander Duyck <alexander.duyck@...il.com>
To:     David Miller <davem@...emloft.net>
Cc:     firo.yang@...e.com, Netdev <netdev@...r.kernel.org>,
        intel-wired-lan <intel-wired-lan@...ts.osuosl.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [Intel-wired-lan] [PATCH 1/1] ixgbe: sync the first fragment unconditionally

On Tue, Aug 6, 2019 at 11:36 AM David Miller <davem@...emloft.net> wrote:
>
> From: Firo Yang <firo.yang@...e.com>
> Date: Tue, 6 Aug 2019 09:29:51 +0000
>
> > In Xen environment, if Xen-swiotlb is enabled, ixgbe driver
> > could possibly allocate a page, DMA memory buffer, for the first
> > fragment which is not suitable for Xen-swiotlb to do DMA operations.
> > Xen-swiotlb will internally allocate another page for doing DMA
> > operations. It requires syncing between those two pages. Otherwise,
> > we may get an incomplete skb. To fix this problem, sync the first
> > fragment no matter the first fargment is makred as "page_released"
> > or not.
> >
> > Signed-off-by: Firo Yang <firo.yang@...e.com>
>
> I don't understand, an unmap operation implies a sync operation.

Actually it doesn't because ixgbe is mapping and unmapping with
DMA_ATTR_SKIP_CPU_SYNC.

The patch description isn't very good. The issue is that the sync in
this case is being skipped in ixgbe_get_rx_buffer for a frame where
the buffer spans more then a single page. As such we need to do both
the sync and the unmap call on the last frame when we encounter the
End Of Packet.