On 8/25/2022 1:59 AM, Bjorn Helgaas wrote:
On Wed, Aug 03, 2022 at 04:58:51PM +0530, Krishna chaitanya chundru wrote:
If the endpoint device state is D0 and irq's are not freed, then
kernel try to mask interrupts in system suspend path by writing in to
the vector table (for MSIX interrupts) and config space (for MSI's).
If clocks are being turned off while the PCI core is still accessing
the device, I think that means qcom suspend is not implemented
correctly.
we are registering the suspend and resume ops as NO_IRQ pm ops and in
those suspend ops we are disbaling clks there.
NO_IRQ ops is the last the pm ops that are getting called. But we are
getting pcie access nearly at end of the suspend and near cpu disable.
The pcie access is nothing but the interrupts masks to endpoint to
disable the interrupts.
These transactions are initiated in the pm suspend after pcie clocks got
disabled as part of platform driver pm suspend call. Due to it, these
transactions are resulting in un-clocked access and eventually to crashes.
So added a logic in qcom driver to restrict these unclocked access.
And updated the logic to check the link state before masking
or unmasking the interrupts.
And some devices are taking time to settle the link in L1ss, so added a
retry logic in the suspend ops.
Krishna chaitanya chundru (3):
PCI: qcom: Add system PM support
PCI: qcom: Restrict pci transactions after pci suspend
PCI: qcom: Add retry logic for link to be stable in L1ss
drivers/pci/controller/dwc/pcie-designware-host.c | 14 ++-
drivers/pci/controller/dwc/pcie-qcom.c | 117 +++++++++++++++++++++-
2 files changed, 127 insertions(+), 4 deletions(-)
--
2.7.4
