Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-02-01 Thread Jakub Kicinski
On Mon, 1 Feb 2021 20:23:51 + Nick Lowe wrote: > I personally tested with mainline and 5.10, but not 5.4 My preference would be to merge into net-next, and then do the backport after 5.11 is released.

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-02-01 Thread Nick Lowe
I personally tested with mainline and 5.10, but not 5.4 Best, Nick

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-02-01 Thread Matt Corallo
On 2/1/21 2:45 PM, Jakub Kicinski wrote: Matt, would you mind clarifying if this is indeed a regression for i211? Admittedly, I didn't go all the way back to test that it is, indeed, a regression. The Fixes commit that it was tagged with on Tony's tree was something more recent than initia

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-02-01 Thread Nick Lowe
Hi all, It is correct that this is not a regression, it instead has never worked. My suggestion would be to merge it during the 5.11 cycle rather than 5.12 where possible because the patch is non-invasive and should be lower risk. There are significant and tangible real-world benefits to RSS fun

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-02-01 Thread Jakub Kicinski
On Mon, 1 Feb 2021 18:32:50 + Nguyen, Anthony L wrote: > On Mon, 2021-02-01 at 08:47 -0800, Jakub Kicinski wrote: > > On Sun, 31 Jan 2021 18:17:11 -0500 Matt Corallo wrote: > > > Given this fixes a major (albeit ancient) performance regression, > > > is > > > it not a candidate for backport?

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-02-01 Thread Nguyen, Anthony L
On Mon, 2021-02-01 at 08:47 -0800, Jakub Kicinski wrote: > On Sun, 31 Jan 2021 18:17:11 -0500 Matt Corallo wrote: > > Given this fixes a major (albeit ancient) performance regression, > > is > > it not a candidate for backport? It landed on Tony's dev-queue > > branch > > with a Fixes tag but no st

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-02-01 Thread Jakub Kicinski
On Sun, 31 Jan 2021 18:17:11 -0500 Matt Corallo wrote: > Given this fixes a major (albeit ancient) performance regression, is > it not a candidate for backport? It landed on Tony's dev-queue branch > with a Fixes tag but no stable CC. Tony's tree needs to get fed into net, then net into Linus's tr

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2021-01-31 Thread Matt Corallo
Given this fixes a major (albeit ancient) performance regression, is it not a candidate for backport? It landed on Tony's dev-queue branch with a Fixes tag but no stable CC. Thanks, Matt On 12/21/20 5:25 PM, Nick Lowe wrote: The Intel I211 Ethernet Controller supports 2 Receive Side Scaling (R

[PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2020-12-21 Thread Nick Lowe
The Intel I211 Ethernet Controller supports 2 Receive Side Scaling (RSS) queues. It should not be excluded from having this feature enabled. Via commit c883de9fd787b6f49bf825f3de3601aeb78a7114 E1000_MRQC_ENABLE_RSS_4Q was renamed to E1000_MRQC_ENABLE_RSS_MQ to indicate that this is a generic bit f

Re: [PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2020-12-21 Thread Jakub Kicinski
On Mon, 21 Dec 2020 19:15:49 + Nick Lowe wrote: > The Intel I211 Ethernet Controller supports 2 Receive Side Scaling (RSS) > queues. > It should not be excluded from having this feature enabled. > > Via commit c883de9fd787b6f49bf825f3de3601aeb78a7114 > E1000_MRQC_ENABLE_RSS_4Q was renamed to

[PATCH net] igb: Enable RSS for Intel I211 Ethernet Controller

2020-12-21 Thread Nick Lowe
The Intel I211 Ethernet Controller supports 2 Receive Side Scaling (RSS) queues. It should not be excluded from having this feature enabled. Via commit c883de9fd787b6f49bf825f3de3601aeb78a7114 E1000_MRQC_ENABLE_RSS_4Q was renamed to E1000_MRQC_ENABLE_RSS_MQ to indicate that this is a generic bit f