| Message ID | 20230608190941.29063-1-philip.cox@canonical.com |
|---|---|
| Headers | show |
| Series | TSN - Cache HW timestamp | expand |
On 6/8/23 1:09 PM, Philip Cox wrote: > > BugLink: https://bugs.launchpad.net/bugs/2023325 > > SRU Justification: > > [Impact] > In the i225/TSN driver the HW timestamp needs to be saved. > > > [Fix] > Enable the Rx HW timestamp for the TSN driver. > > [Test Plan] > Intel ahs tested this patch, and passed it along. > > I am not able to test the TSN networking, but I've tested other networking with this code, and did not see any other networking related regressions. > > [Where problems could occur] > The patch is fairly narrow, as it is only saving a timestamp. And this code only caches a timestamp inside an existing struct, the risk of a regression should be fairly low. > > When the SAUCE patch eventually makes it into the upstream kernel, we should cherry-pick the upstreamed version. > > This may cause a problem if other commits make it into the kernel via the SRU process which conflict with the SAUCE patch. But, as the TSN patches likely won't be picked in to the SRU stable kernes, this risk is fairly low as well. > > > ----- > > mohd faizal abdul rahim (1): > UBUNTU: SAUCE: (no-up) igc: Add RX HW timestamp to support XDP zero > copy > > drivers/net/ethernet/intel/igc/igc_main.c | 4 ++++ > 1 file changed, 4 insertions(+) > Acked-by: Tim Gardner <tim.gardner@canonical.com>
On 23/06/08 03:09PM, Philip Cox wrote: > > BugLink: https://bugs.launchpad.net/bugs/2023325 > > SRU Justification: > > [Impact] > In the i225/TSN driver the HW timestamp needs to be saved. > > > [Fix] > Enable the Rx HW timestamp for the TSN driver. > > [Test Plan] > Intel ahs tested this patch, and passed it along. > > I am not able to test the TSN networking, but I've tested other networking with this code, and did not see any other networking related regressions. > > [Where problems could occur] > The patch is fairly narrow, as it is only saving a timestamp. And this code only caches a timestamp inside an existing struct, the risk of a regression should be fairly low. > > When the SAUCE patch eventually makes it into the upstream kernel, we should cherry-pick the upstreamed version. > > This may cause a problem if other commits make it into the kernel via the SRU process which conflict with the SAUCE patch. But, as the TSN patches likely won't be picked in to the SRU stable kernes, this risk is fairly low as well. > > > ----- > > mohd faizal abdul rahim (1): > UBUNTU: SAUCE: (no-up) igc: Add RX HW timestamp to support XDP zero > copy > > drivers/net/ethernet/intel/igc/igc_main.c | 4 ++++ > 1 file changed, 4 insertions(+) > > -- > 2.34.1 Acked-by: Andrei Gherzan <andrei.gherzan@canonical.com>
On Thu, Jun 8, 2023 at 3:12 PM Philip Cox <philip.cox@canonical.com> wrote: > > BugLink: https://bugs.launchpad.net/bugs/2023325 > > SRU Justification: > > [Impact] > In the i225/TSN driver the HW timestamp needs to be saved. > > > [Fix] > Enable the Rx HW timestamp for the TSN driver. > > [Test Plan] > Intel ahs tested this patch, and passed it along. > > I am not able to test the TSN networking, but I've tested other networking > with this code, and did not see any other networking related regressions. > > [Where problems could occur] > The patch is fairly narrow, as it is only saving a timestamp. And this > code only caches a timestamp inside an existing struct, the risk of a > regression should be fairly low. > > When the SAUCE patch eventually makes it into the upstream kernel, we > should cherry-pick the upstreamed version. > > This may cause a problem if other commits make it into the kernel via the > SRU process which conflict with the SAUCE patch. But, as the TSN patches > likely won't be picked in to the SRU stable kernes, this risk is fairly low > as well. > > > ----- > > mohd faizal abdul rahim (1): > UBUNTU: SAUCE: (no-up) igc: Add RX HW timestamp to support XDP zero > copy > > drivers/net/ethernet/intel/igc/igc_main.c | 4 ++++ > 1 file changed, 4 insertions(+) > > -- > 2.34.1 > >
BugLink: https://bugs.launchpad.net/bugs/2023325 SRU Justification: [Impact] In the i225/TSN driver the HW timestamp needs to be saved. [Fix] Enable the Rx HW timestamp for the TSN driver. [Test Plan] Intel ahs tested this patch, and passed it along. I am not able to test the TSN networking, but I've tested other networking with this code, and did not see any other networking related regressions. [Where problems could occur] The patch is fairly narrow, as it is only saving a timestamp. And this code only caches a timestamp inside an existing struct, the risk of a regression should be fairly low. When the SAUCE patch eventually makes it into the upstream kernel, we should cherry-pick the upstreamed version. This may cause a problem if other commits make it into the kernel via the SRU process which conflict with the SAUCE patch. But, as the TSN patches likely won't be picked in to the SRU stable kernes, this risk is fairly low as well. ----- mohd faizal abdul rahim (1): UBUNTU: SAUCE: (no-up) igc: Add RX HW timestamp to support XDP zero copy drivers/net/ethernet/intel/igc/igc_main.c | 4 ++++ 1 file changed, 4 insertions(+)