Hi Leif,
在 2017/4/10 21:53, Leif Lindholm 写道:
On Mon, Apr 10, 2017 at 08:33:14PM +0800, Chenhui Sun wrote:
The M3(the coprocessor)PCIe driver will read Option Rom header durning enumeration, this operation will cause a completion error when there is no device inserted to the RC port, and the Option rom is uesless now. So we need to disable the RC Option Rom.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Heyi Guo heyi.guo@linaro.org Signed-off-by: Yi Li phoenix.liyi@huawei.com Signed-off-by: Chenhui Sun chenhui.sun@linaro.org Reviewed-by: Leif Lindholm leif.lindholm@linaro.org
.../Hi1610/Drivers/PcieInit1610/PcieInitLib.c | 46 ++++++++++++++++++++++ 1 file changed, 46 insertions(+)
diff --git a/Chips/Hisilicon/Hi1610/Drivers/PcieInit1610/PcieInitLib.c b/Chips/Hisilicon/Hi1610/Drivers/PcieInit1610/PcieInitLib.c index a9b3d74..1df7a90 100644 --- a/Chips/Hisilicon/Hi1610/Drivers/PcieInit1610/PcieInitLib.c +++ b/Chips/Hisilicon/Hi1610/Drivers/PcieInit1610/PcieInitLib.c @@ -901,6 +901,50 @@ void PcieConfigContextHi1610(UINT32 soctype, UINT32 HostBridgeNum, UINT32 Port) return; } +UINT32 +SysRegRead (
- IN UINT32 SocType,
- IN UINT32 HostBridgeNum,
- IN UINT32 Port,
- IN UINTN Reg
- )
+{
- UINT32 Value;
- if (SocType == 0x1610) {
- RegRead (PCIE_APB_SLAVE_BASE_1610[HostBridgeNum][Port] + Reg, Value);
- } else {
- //PCIE_APB_SLVAE_BASE is for 660,and each PCIe Ccontroller has the same APB_SLVAE_BASE
- //in the same hostbridge.
- RegRead (PCIE_APB_SLVAE_BASE[HostBridgeNum] + Reg, Value);
This error was not introduced by this patch, so we don't need to fix it for 2017.04, but it would be nice if after the release you could submit a patch fixing the above spelling error (PCIE_APB_SLVAE_BASE -> PCIE_APB_SLAVE_BASE). Extra visible here because PCIE_APB_SLAVE_BASE_1610 gets it right.
Thanks for pointing the spelling error.
Actually, even better would be if those could be modified to conform to coding style: gPcieApbSlaveBase and gPcieApbSlaveBase1610.
ok, will do that.
(I already signed off on this patch, I just didn't push it since I think it conflicted with other patches I had comments on.)
ok,
Thanks an Regards, Chenhui
Regards,
Leif
- }
- return Value;
+}
+VOID +DisableRcOptionRom (
- IN UINT32 Soctype,
- IN UINT32 HostBridgeNum,
- IN UINT32 Port,
- IN PCIE_PORT_TYPE PcieType
+) +{
- UINT32 Value = 0;
- if (PcieType == PCIE_ROOT_COMPLEX) {
- Value = SysRegRead (Soctype, HostBridgeNum, Port, PCIE_SYS_REG_OFFSET + PCIE_SYS_CTRL21_REG);
- Value |= BIT2; //cs2 enable
- SysRegWrite (Soctype, HostBridgeNum, Port, PCIE_SYS_REG_OFFSET + PCIE_SYS_CTRL21_REG, Value);
- Value = SysRegRead (Soctype, HostBridgeNum, Port, PCIE_SYS_REG_OFFSET + PCIE_EP_PCI_CFG_HDR12_REG);
- Value &= ~BIT0; //disable option rom
- SysRegWrite (Soctype, HostBridgeNum, Port, PCIE_SYS_REG_OFFSET + PCIE_EP_PCI_CFG_HDR12_REG, Value);
- Value = SysRegRead (Soctype, HostBridgeNum, Port, PCIE_SYS_REG_OFFSET + PCIE_SYS_CTRL21_REG);
- Value &= ~BIT2; //cs2 disable
- SysRegWrite (Soctype, HostBridgeNum, Port, PCIE_SYS_REG_OFFSET + PCIE_SYS_CTRL21_REG, Value);
- }
- return;
+}
- EFI_STATUS EFIAPI PciePortInit (
@@ -961,6 +1005,8 @@ PciePortInit ( /* Pcie Equalization*/ (VOID)PcieEqualization(soctype ,HostBridgeNum, PortIndex);
/* Disable RC Option Rom */DisableRcOptionRom (soctype, HostBridgeNum, PortIndex, PcieCfg->PortInfo.PortType); /* assert LTSSM enable */ (VOID)PcieEnableItssm(soctype, HostBridgeNum, PortIndex); if (FeaturePcdGet(PcdIsPciPerfTuningEnable)) {-- 1.9.1