linux-kernel - Re: [PATCH net] drivers/net/wan/lapbether: Added needed

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: <CAJht_EOsQ-QLFBeJCytTRSRuor6jnCEE+zMBV+ngtwr25OSCSQ@mail.gmail.com>
Date:   Sat, 15 Aug 2020 19:28:14 -0700
From:   Xie He <xie.he.0141@...il.com>
To:     Willem de Bruijn <willemdebruijn.kernel@...il.com>
Cc:     "David S. Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Network Development <netdev@...r.kernel.org>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        Linux X25 <linux-x25@...r.kernel.org>,
        Martin Schiller <ms@....tdt.de>
Subject: Re: [PATCH net] drivers/net/wan/lapbether: Added needed_tailroom

I took some time to look at the history of needed_tailroom. I found it
was added in this commit:
f5184d267c1a (net: Allow netdevices to specify needed head/tailroom)

The author tried to make use of needed_tailroom at various places in
the kernel by replacing the macro LL_RESERVED_SPACE with his new macro
LL_ALLOCATED_SPACE.

However, the macro LL_ALLOCATED_SPACE was later found to have
problems. So it was removed 3 years later and was replaced by explicit
handling of needed_tailroom. See:
https://lkml.org/lkml/2011/11/18/198

So maybe only those places considered by these two authors have taken
needed_tailroom into account.

Other places might not have taken needed_tailroom into account because
of the rarity of the usage of needed_tailroom.

The second author also said in the commit message of his Patch 5/6
(which changes af_packet.c), that:
    While auditing LL_ALLOCATED_SPACE I noticed that packet_sendmsg_spkt
    did not include needed_tailroom when allocating an skb.  This isn't
    a fatal error as we should always tolerate inadequate tail room but
    it isn't optimal.

This shows not taking needed_tailroom into account is not a bug but
it'd be better to take it into account.