| Message ID | 20211020022620.21012-1-saranya.gopal@intel.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | Fix IRQ flood issue in TI PD controller | expand |
Hi, On 10/20/21 04:26, Saranya Gopal wrote: > Hi, > > There was an issue reported that the TI PD controller driver is causing > high CPU load due to a flood of interrupts. So, a patch was added in > the i2c-multi-instantiate driver to stop the TI PD driver from loading > in devices with INT3515 ACPI nodes. > We identified that required event interrupts are not being set in the interrupt > mask register from the driver to the register of the controller. > We enabled only the necessary events like data status update, power status update > and plug events in the interrupt mask register of the TI PD controller. > After enabling these events in the interrupt mask register, there is no interrupt flood. > This patch series contains the fix for the interrupt flood issue > in the TI PD driver and another patch to re-enable the INT3515 platform device. > I prefer this patch series to be taken through usb tree since the fix is in > the TI USB PD driver and the second patch is just a revert patch. > > Hi Hans, > Could I get your Ack to take this series through the usb tree? Since Heikki has reviewed the revert, I'm fine with this and I'm also fine with taking this upstream through the usb tree: Acked-by: Hans de Goede <hdegoede@redhat.com> Regards, Hans > > Saranya Gopal (2): > usb: typec: tipd: Enable event interrupts by default > Revert "platform/x86: i2c-multi-instantiate: Don't create platform > device for INT3515 ACPI nodes" > > drivers/platform/x86/i2c-multi-instantiate.c | 31 +++++--------------- > drivers/usb/typec/tipd/core.c | 8 +++++ > 2 files changed, 16 insertions(+), 23 deletions(-) >
On Wed, Oct 20, 2021 at 01:54:39PM +0200, Hans de Goede wrote: > Hi, > > On 10/20/21 04:26, Saranya Gopal wrote: > > Hi, > > > > There was an issue reported that the TI PD controller driver is causing > > high CPU load due to a flood of interrupts. So, a patch was added in > > the i2c-multi-instantiate driver to stop the TI PD driver from loading > > in devices with INT3515 ACPI nodes. > > We identified that required event interrupts are not being set in the interrupt > > mask register from the driver to the register of the controller. > > We enabled only the necessary events like data status update, power status update > > and plug events in the interrupt mask register of the TI PD controller. > > After enabling these events in the interrupt mask register, there is no interrupt flood. > > This patch series contains the fix for the interrupt flood issue > > in the TI PD driver and another patch to re-enable the INT3515 platform device. > > I prefer this patch series to be taken through usb tree since the fix is in > > the TI USB PD driver and the second patch is just a revert patch. > > > > Hi Hans, > > Could I get your Ack to take this series through the usb tree? > > Since Heikki has reviewed the revert, I'm fine with this and I'm > also fine with taking this upstream through the usb tree: > > Acked-by: Hans de Goede <hdegoede@redhat.com> Thanks, I'll queue it up. greg k-h
Hi, There was an issue reported that the TI PD controller driver is causing high CPU load due to a flood of interrupts. So, a patch was added in the i2c-multi-instantiate driver to stop the TI PD driver from loading in devices with INT3515 ACPI nodes. We identified that required event interrupts are not being set in the interrupt mask register from the driver to the register of the controller. We enabled only the necessary events like data status update, power status update and plug events in the interrupt mask register of the TI PD controller. After enabling these events in the interrupt mask register, there is no interrupt flood. This patch series contains the fix for the interrupt flood issue in the TI PD driver and another patch to re-enable the INT3515 platform device. I prefer this patch series to be taken through usb tree since the fix is in the TI USB PD driver and the second patch is just a revert patch. Hi Hans, Could I get your Ack to take this series through the usb tree? Saranya Gopal (2): usb: typec: tipd: Enable event interrupts by default Revert "platform/x86: i2c-multi-instantiate: Don't create platform device for INT3515 ACPI nodes" drivers/platform/x86/i2c-multi-instantiate.c | 31 +++++--------------- drivers/usb/typec/tipd/core.c | 8 +++++ 2 files changed, 16 insertions(+), 23 deletions(-)