| 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
| ||
|
Message-ID: <7f95eede-888e-9e81-eedf-7664d91e2e16@linux.dev>
Date: Wed, 15 Mar 2023 15:56:05 +0000
From: Vadim Fedorenko <vadim.fedorenko@...ux.dev>
To: Yan Jiang <Yan.Jiang@...visolutions.com>,
"davem@...emloft.net" <davem@...emloft.net>,
"edumazet@...gle.com" <edumazet@...gle.com>,
"kuba@...nel.org" <kuba@...nel.org>,
"pabeni@...hat.com" <pabeni@...hat.com>,
"richardcochran@...il.com" <richardcochran@...il.com>
Cc: "netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: ioctl SIOCSHWTSTAMP failed on a vlan interface within non-default
network namespace
On 15/03/2023 15:17, Yan Jiang wrote:
> HI Vadim,
>
> Thank you for your reply. This explains why ioctl failed on vlan interface.
>
> I am ignorant here, but in my example below both base interface and vlan
> interface are in same namespace, and I feel that the defense you
> pointed to me is not protecting anything.
> We can even take the default namespace as example: why we allow ioctl
> SIOCSHWTSTAMP to work on a VLAN interface in default namespace?
>
> I think it is common to assign a sriov VF to a container. In this case,
> both base interface and VLAN interface created on top of the VF, are in
> same network namespace.
I've just proposed the patch to address the situation, hope it will find
some support :)
>
> Possibly there are some situations that I am not aware. Hope to have
> your thoughts. Thank you!
>
>
> ------------------------------------------------------------------------
> *From:* Vadim Fedorenko <vadim.fedorenko@...ux.dev>
> *Sent:* Wednesday, March 15, 2023 10:38 PM
> *To:* Yan Jiang <Yan.Jiang@...visolutions.com>; davem@...emloft.net
> <davem@...emloft.net>; edumazet@...gle.com <edumazet@...gle.com>;
> kuba@...nel.org <kuba@...nel.org>; pabeni@...hat.com
> <pabeni@...hat.com>; richardcochran@...il.com <richardcochran@...il.com>
> *Cc:* netdev@...r.kernel.org <netdev@...r.kernel.org>
> *Subject:* Re: ioctl SIOCSHWTSTAMP failed on a vlan interface within
> non-default network namespace
> EXTERNAL EMAIL: Do not click links or open attachments unless you know
> and trust the sender.
>
>
> On 15/03/2023 05:35, Yan Jiang wrote:
>> Hi all,
>>
>> I'd like to report an issue, which I suspect it's related to kernel network namespace handling.
>> It's my first report so please bear with me if I missed something. Feel free to ask for more information please.
>>
>> What's the issue:
>> Ptp4l works perfectly on a VLAN interface in default network namespace. But it doesn't work if the VLAN interface is in non-default network namespace: ioctl(fd, SIOCSHWTSTAMP, &ifreq) failed due to error "Operation not supported".
>> "ethtool -T" shows that the VLAN interface has all required capabilities. And ptp4l works fine on the base interface in that network namespace.
>
>
> That was explicitly forbidden in
> https://urldefense.com/v3/__https://lore.kernel.org/netdev/20190509065507.23991-1-liuhangbin@gmail.com/__;!!Niha4SQ!_NTn9qN2nCq-QapNeIln9DvJwmpQWK7Y4EJOdXjbOq6w7Bhh-ZQp-vJu8TPyI9QMigOIDPeN2IuZSJamNFZI4DAfQvMIWyjqEMj9$ <https://urldefense.com/v3/__https://lore.kernel.org/netdev/20190509065507.23991-1-liuhangbin@gmail.com/__;!!Niha4SQ!_NTn9qN2nCq-QapNeIln9DvJwmpQWK7Y4EJOdXjbOq6w7Bhh-ZQp-vJu8TPyI9QMigOIDPeN2IuZSJamNFZI4DAfQvMIWyjqEMj9$>
>
>
>> How to reproduce this issue:
>> # create a new network namespace for test purpose
>> root@...vi-PowerEdge-R740:~# ip netns add mytest
>>
>> # move eno4 into the namespace
>> root@...vi-PowerEdge-R740:~# ip link set eno4 netns mytest
>>
>> # add vlan interface and turn interfaces up
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ifconfig eno4 up
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ip link add link eno4 name eno4.4000 type vlan id 4000
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ifconfig eno4.4000 up
>>
>> # ptp4l runs ok on eno4
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ptp4l -2 -m -i eno4
>> ptp4l[1670814.457]: selected /dev/ptp1 as PTP clock
>> ptp4l[1670814.491]: port 1: INITIALIZING to LISTENING on INIT_COMPLETE
>> ptp4l[1670814.492]: port 0: INITIALIZING to LISTENING on INIT_COMPLETE
>>
>> #ptp4l cannot run on eno4.4000
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ptp4l -2 -m -i eno4.4000
>> ptp4l[1670819.969]: selected /dev/ptp1 as PTP clock
>> ptp4l[1670820.003]: driver rejected most general HWTSTAMP filter
>> ptp4l[1670820.003]: ioctl SIOCSHWTSTAMP failed: Operation not supported
>> ptp4l[1670820.039]: port 1: INITIALIZING to FAULTY on FAULT_DETECTED (FT_UNSPECIFIED)
>> ptp4l[1670820.039]: port 0: INITIALIZING to LISTENING on INIT_COMPLETE
>>
>> #ethtool -T shows that eno4.4000 has same capability as eno4:
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ethtool -T eno4.4000
>> Time stamping parameters for eno4.4000:
>> Capabilities:
>> hardware-transmit (SOF_TIMESTAMPING_TX_HARDWARE)
>> software-transmit (SOF_TIMESTAMPING_TX_SOFTWARE)
>> hardware-receive (SOF_TIMESTAMPING_RX_HARDWARE)
>> software-receive (SOF_TIMESTAMPING_RX_SOFTWARE)
>> software-system-clock (SOF_TIMESTAMPING_SOFTWARE)
>> hardware-raw-clock (SOF_TIMESTAMPING_RAW_HARDWARE)
>> PTP Hardware Clock: 1
>> Hardware Transmit Timestamp Modes:
>> off (HWTSTAMP_TX_OFF)
>> on (HWTSTAMP_TX_ON)
>> Hardware Receive Filter Modes:
>> none (HWTSTAMP_FILTER_NONE)
>> ptpv1-l4-event (HWTSTAMP_FILTER_PTP_V1_L4_EVENT)
>> ptpv2-l4-event (HWTSTAMP_FILTER_PTP_V2_L4_EVENT)
>> ptpv2-l2-event (HWTSTAMP_FILTER_PTP_V2_L2_EVENT)
>>
>> #OS/kernel version
>> root@...vi-PowerEdge-R740:~# uname -a
>> Linux viavi-PowerEdge-R740 5.15.0-60-generic #66~20.04.1-Ubuntu SMP Wed Jan 25 09:41:30 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux
>>
>> #driver info:
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ethtool -i eno4
>> driver: tg3
>> version: 5.15.0-60-generic
>> firmware-version: FFV21.40.21 bc 5720-v1.39
>> expansion-rom-version:
>> bus-info: 0000:01:00.1
>> supports-statistics: yes
>> supports-test: yes
>> supports-eeprom-access: yes
>> supports-register-dump: yes
>> supports-priv-flags: no
>> root@...vi-PowerEdge-R740:~# ip netns exec mytest ethtool -i eno4.4000
>> driver: 802.1Q VLAN Support
>> version: 1.8
>> firmware-version: N/A
>> expansion-rom-version:
>> bus-info:
>> supports-statistics: no
>> supports-test: no
>> supports-eeprom-access: no
>> supports-register-dump: no
>> supports-priv-flags: no
>>
>> #(PS: can reproduce same issue with a Mellanox NIC, so I guess it's not caused by a specific NIC model)
>>
>> Could you kindly take a look at this and see if this is a kernel issue? This blocks linuxptp running in cloud environment, if VLAN is required.
>> Hope to get feedback from you soon, thanks!
>>
>> /Yan
>
Powered by blists - more mailing lists