| Message ID | 1509371843-22931-8-git-send-email-mmaddireddy@nvidia.com (mailing list archive) |
|---|---|
| State | New, archived |
| Delegated to: | Bjorn Helgaas |
| Headers | show |
On Mon, Oct 30, 2017 at 07:27:18PM +0530, Manikanta Maddireddy wrote: > When there are 32 outstanding writes from AFI to memory, the outstanding > write counter overflows and indicates that there are "0" outstanding write > transactions. This outstanding write counter is used to generate IDLE > signal to dynamically gate the AFI clock. > > When memory controller is under heavy load, its possible that write > completions will come back to AFI after long delay and AFI write counter > overflows. AFI clock gets gated even when there are outstanding > transactions towards memory controller resutling in system hang. > > Disable dynamic clock gating of AFI clock to avoid system hang. > > Signed-off-by: Manikanta Maddireddy <mmaddireddy@nvidia.com> > --- > V3: > * no change in this patch > V2: > * no change in this patch > > drivers/pci/host/pci-tegra.c | 10 ++++++---- > 1 file changed, 6 insertions(+), 4 deletions(-) > > diff --git a/drivers/pci/host/pci-tegra.c b/drivers/pci/host/pci-tegra.c > index 4562b0c113a8..c264037112cb 100644 > --- a/drivers/pci/host/pci-tegra.c > +++ b/drivers/pci/host/pci-tegra.c > @@ -103,8 +103,9 @@ > #define AFI_MSI_EN_VEC6 0xa4 > #define AFI_MSI_EN_VEC7 0xa8 > > -#define AFI_CONFIGURATION 0xac > -#define AFI_CONFIGURATION_EN_FPCI (1 << 0) > +#define AFI_CONFIGURATION 0xac > +#define AFI_CONFIGURATION_EN_FPCI (1 << 0) > +#define AFI_CONFIGURATION_CLKEN_OVERRIDE (1 << 31) > > #define AFI_FPCI_ERROR_MASKS 0xb0 > > @@ -1059,9 +1060,10 @@ static int tegra_pcie_enable_controller(struct tegra_pcie *pcie) > } > } > > - /* finally enable PCIe */ > + /* Disable AFI dynamic clock gating and enable PCIe */ > value = afi_readl(pcie, AFI_CONFIGURATION); > - value |= AFI_CONFIGURATION_EN_FPCI; > + value |= (AFI_CONFIGURATION_EN_FPCI | > + AFI_CONFIGURATION_CLKEN_OVERRIDE); I think this is cleaner (and results in a smaller diff) if you simply put this on a new line: value |= AFI_CONFIGURATION_CLKEN_OVERRIDE; Thierry
diff --git a/drivers/pci/host/pci-tegra.c b/drivers/pci/host/pci-tegra.c index 4562b0c113a8..c264037112cb 100644 --- a/drivers/pci/host/pci-tegra.c +++ b/drivers/pci/host/pci-tegra.c @@ -103,8 +103,9 @@ #define AFI_MSI_EN_VEC6 0xa4 #define AFI_MSI_EN_VEC7 0xa8 -#define AFI_CONFIGURATION 0xac -#define AFI_CONFIGURATION_EN_FPCI (1 << 0) +#define AFI_CONFIGURATION 0xac +#define AFI_CONFIGURATION_EN_FPCI (1 << 0) +#define AFI_CONFIGURATION_CLKEN_OVERRIDE (1 << 31) #define AFI_FPCI_ERROR_MASKS 0xb0 @@ -1059,9 +1060,10 @@ static int tegra_pcie_enable_controller(struct tegra_pcie *pcie) } } - /* finally enable PCIe */ + /* Disable AFI dynamic clock gating and enable PCIe */ value = afi_readl(pcie, AFI_CONFIGURATION); - value |= AFI_CONFIGURATION_EN_FPCI; + value |= (AFI_CONFIGURATION_EN_FPCI | + AFI_CONFIGURATION_CLKEN_OVERRIDE); afi_writel(pcie, value, AFI_CONFIGURATION); value = AFI_INTR_EN_INI_SLVERR | AFI_INTR_EN_INI_DECERR |
When there are 32 outstanding writes from AFI to memory, the outstanding write counter overflows and indicates that there are "0" outstanding write transactions. This outstanding write counter is used to generate IDLE signal to dynamically gate the AFI clock. When memory controller is under heavy load, its possible that write completions will come back to AFI after long delay and AFI write counter overflows. AFI clock gets gated even when there are outstanding transactions towards memory controller resutling in system hang. Disable dynamic clock gating of AFI clock to avoid system hang. Signed-off-by: Manikanta Maddireddy <mmaddireddy@nvidia.com> --- V3: * no change in this patch V2: * no change in this patch drivers/pci/host/pci-tegra.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)