| Message ID | af076f1f-a034-82e5-8f76-f3ec32a14eaa@gmail.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | r8169: disable ASPM during NAPI poll | expand |
On Sun, Feb 26, 2023 at 5:43 AM Heiner Kallweit <hkallweit1@gmail.com> wrote: > > This is a rework of ideas from Kai-Heng on how to avoid the known > ASPM issues whilst still allowing for a maximum of ASPM-related power > savings. As a prerequisite some locking is added first. > > This change affects a bigger number of supported chip versions, > therefore this series comes as RFC first for further testing. Thanks for the series. Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> > > Heiner Kallweit (6): > r8169: use spinlock to protect mac ocp register access > r8169: use spinlock to protect access to registers Config2 and Config5 > r8169: enable cfg9346 config register access in atomic context > r8169: prepare rtl_hw_aspm_clkreq_enable for usage in atomic context > r8169: disable ASPM during NAPI poll > r8169: remove ASPM restrictions now that ASPM is disabled during NAPI > poll > > drivers/net/ethernet/realtek/r8169_main.c | 145 +++++++++++++++------- > 1 file changed, 100 insertions(+), 45 deletions(-) > > -- > 2.39.2 >
On Wed, Mar 01, 2023 at 02:50:43PM +0800, Kai-Heng Feng wrote: > On Sun, Feb 26, 2023 at 5:43 AM Heiner Kallweit <hkallweit1@gmail.com> wrote: > > > > This is a rework of ideas from Kai-Heng on how to avoid the known > > ASPM issues whilst still allowing for a maximum of ASPM-related power > > savings. As a prerequisite some locking is added first. > > > > This change affects a bigger number of supported chip versions, > > therefore this series comes as RFC first for further testing. > > Thanks for the series. > > Tested-by: Kai-Heng Feng <kai.heng.feng@canonical.com> FWIIW, I also looked over this series and it looks good to me. Reviewed-by: Simon Horman <simon.horman@corigine.com>
On 2023-02-25 22:43, Heiner Kallweit wrote: > This is a rework of ideas from Kai-Heng on how to avoid the known > ASPM issues whilst still allowing for a maximum of ASPM-related power > savings. As a prerequisite some locking is added first. > > This change affects a bigger number of supported chip versions, > therefore this series comes as RFC first for further testing. > > Heiner Kallweit (6): > r8169: use spinlock to protect mac ocp register access > r8169: use spinlock to protect access to registers Config2 and Config5 > r8169: enable cfg9346 config register access in atomic context > r8169: prepare rtl_hw_aspm_clkreq_enable for usage in atomic context > r8169: disable ASPM during NAPI poll > r8169: remove ASPM restrictions now that ASPM is disabled during NAPI > poll > > drivers/net/ethernet/realtek/r8169_main.c | 145 +++++++++++++++------- > 1 file changed, 100 insertions(+), 45 deletions(-) > I've been running this 24/7 for a few days now without any problems, so for the whole series: Tested-by: Holger Hoffstätte <holger@applied-asynchrony.com> Thanks! Holger
This is a rework of ideas from Kai-Heng on how to avoid the known ASPM issues whilst still allowing for a maximum of ASPM-related power savings. As a prerequisite some locking is added first. This change affects a bigger number of supported chip versions, therefore this series comes as RFC first for further testing. Heiner Kallweit (6): r8169: use spinlock to protect mac ocp register access r8169: use spinlock to protect access to registers Config2 and Config5 r8169: enable cfg9346 config register access in atomic context r8169: prepare rtl_hw_aspm_clkreq_enable for usage in atomic context r8169: disable ASPM during NAPI poll r8169: remove ASPM restrictions now that ASPM is disabled during NAPI poll drivers/net/ethernet/realtek/r8169_main.c | 145 +++++++++++++++------- 1 file changed, 100 insertions(+), 45 deletions(-)