netdev - Re: [PATCH v3] igb: Allocate MSI-X vector when testing

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: <1434bd67-1707-7e43-96b6-d2294ff7f04d@daynix.com>
Date:   Fri, 25 Nov 2022 18:50:45 +0900
From:   Akihiko Odaki <akihiko.odaki@...nix.com>
To:     Maciej Fijalkowski <maciej.fijalkowski@...el.com>
Cc:     linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
        intel-wired-lan@...ts.osuosl.org,
        Jesse Brandeburg <jesse.brandeburg@...el.com>,
        Tony Nguyen <anthony.l.nguyen@...el.com>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        Yan Vugenfirer <yan@...nix.com>,
        Yuri Benditovich <yuri.benditovich@...nix.com>
Subject: Re: [PATCH v3] igb: Allocate MSI-X vector when testing

On 2022/11/25 18:48, Akihiko Odaki wrote:
> 
> 
> On 2022/11/24 0:41, Maciej Fijalkowski wrote:
>> On Wed, Nov 23, 2022 at 10:09:26AM +0900, Akihiko Odaki wrote:
>>> Without this change, the interrupt test fail with MSI-X environment:
>>>
>>> $ sudo ethtool -t enp0s2 offline
>>> [   43.921783] igb 0000:00:02.0: offline testing starting
>>> [   44.855824] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Down
>>> [   44.961249] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Up 
>>> 1000 Mbps Full Duplex, Flow Control: RX/TX
>>> [   51.272202] igb 0000:00:02.0: testing shared interrupt
>>> [   56.996975] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Up 
>>> 1000 Mbps Full Duplex, Flow Control: RX/TX
>>> The test result is FAIL
>>> The test extra info:
>>> Register test  (offline)     0
>>> Eeprom test    (offline)     0
>>> Interrupt test (offline)     4
>>> Loopback test  (offline)     0
>>> Link test   (on/offline)     0
>>>
>>> Here, "4" means an expected interrupt was not delivered.
>>>
>>> To fix this, route IRQs correctly to the first MSI-X vector by setting
>>> IVAR_MISC. Also, set bit 0 of EIMS so that the vector will not be
>>> masked. The interrupt test now runs properly with this change:
>>
>> Much better!
>>
>>>
>>> $ sudo ethtool -t enp0s2 offline
>>> [   42.762985] igb 0000:00:02.0: offline testing starting
>>> [   50.141967] igb 0000:00:02.0: testing shared interrupt
>>> [   56.163957] igb 0000:00:02.0 enp0s2: igb: enp0s2 NIC Link is Up 
>>> 1000 Mbps Full Duplex, Flow Control: RX/TX
>>> The test result is PASS
>>> The test extra info:
>>> Register test  (offline)     0
>>> Eeprom test    (offline)     0
>>> Interrupt test (offline)     0
>>> Loopback test  (offline)     0
>>> Link test   (on/offline)     0
>>>
>>> Signed-off-by: Akihiko Odaki <akihiko.odaki@...nix.com>
>>
>> Same comment as on other patch - justify why there is no fixes tag and
>> specify the tree in subject.
> 
> I couldn't identify what commit introduced the problem. Please see:
> https://lore.kernel.org/netdev/f2457229-865a-57a0-94a1-c5c63b2f30a5@daynix.com/

Sorry, the URL was wrong. The correct URL is:
https://lore.kernel.org/netdev/be5617fe-d332-447a-b836-bec9a6c6d42d@daynix.com/

Regards,
Akihiko Odaki

> 
> Regards,
> Akihiko Odaki
> 
>>
>>> ---
>>>   drivers/net/ethernet/intel/igb/igb_ethtool.c | 2 ++
>>>   1 file changed, 2 insertions(+)
>>>
>>> diff --git a/drivers/net/ethernet/intel/igb/igb_ethtool.c 
>>> b/drivers/net/ethernet/intel/igb/igb_ethtool.c
>>> index e5f3e7680dc6..ff911af16a4b 100644
>>> --- a/drivers/net/ethernet/intel/igb/igb_ethtool.c
>>> +++ b/drivers/net/ethernet/intel/igb/igb_ethtool.c
>>> @@ -1413,6 +1413,8 @@ static int igb_intr_test(struct igb_adapter 
>>> *adapter, u64 *data)
>>>               *data = 1;
>>>               return -1;
>>>           }
>>> +        wr32(E1000_IVAR_MISC, E1000_IVAR_VALID << 8);
>>> +        wr32(E1000_EIMS, BIT(0));
>>>       } else if (adapter->flags & IGB_FLAG_HAS_MSI) {
>>>           shared_int = false;
>>>           if (request_irq(irq,
>>> -- 
>>> 2.38.1
>>>