[<prev] [next>] [day] [month] [year] [list]
Message-ID: <be5617fe-d332-447a-b836-bec9a6c6d42d@daynix.com>
Date: Tue, 22 Nov 2022 22:17:41 +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] igb: Allocate MSI-X vector when testing
On 2022/11/22 21:15, Maciej Fijalkowski wrote:
> On Tue, Nov 22, 2022 at 09:13:12PM +0900, Akihiko Odaki wrote:
>> Allocate MSI-X vector when testing interrupts, otherwise the tests will
>> not work.
>
> Hi,
>
> can you say a bit more about why current code was broken? And also what is
> the current result of that ethtool test?
>
> Also this is a fix, please provide Fixes: tag and route it to net tree.
Hi, I have just sent v2, please check it out.
Regarding Fixes: tag, I couldn't tell when the bug appeared. The
modified function, igb_intr_test() lacked the interrupt allocation code
from the start. My guess is that some code in igb_reset() or after the
function had code to allocate interrupts in the past and later removed.
But I couldn't find such code.
Regards,
Akihiko Odaki
>
>>
>> Signed-off-by: Akihiko Odaki <akihiko.odaki@...nix.com>
>> ---
>> 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
>>
Powered by blists - more mailing lists