linux-pci.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v2 0/5] Add PCIe support for SM8250 SoC
@ 2020-09-30 15:09 Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 1/5] dt-bindings: phy: qcom,qmp: Add SM8250 PCIe PHY bindings Manivannan Sadhasivam
                   ` (4 more replies)
  0 siblings, 5 replies; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-09-30 15:09 UTC (permalink / raw)
  To: agross, bjorn.andersson, kishon, vkoul, robh
  Cc: svarbanov, bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree, Manivannan Sadhasivam

Hello,

This series adds PCIe support for Qualcomm SM8250 SoC with relevant PHYs.
There are 3 PCIe instances on this SoC each with different PHYs. The PCIe
controller and PHYs are mostly comaptible with the ones found on SDM845
SoC, hence the old drivers are modified to add the support.

This series has been tested on RB5 board with QCA6390 chipset connected
onboard.

NOTE: This series functionally depends on the following patch:
https://lore.kernel.org/linux-arm-kernel/1599814203-14441-3-git-send-email-hayashi.kunihiko@socionext.com/

I've dropped a similar patch in v2.

Thanks,
Mani

Changes in v2:

* Fixed the PHY and PCIe bindings
* Introduced secondary table in PHY driver to abstract out the common configs.
* Used a more generic way of configuring BDF to SID mapping
* Dropped ATU change in favor of a patch spotted by Rob

Manivannan Sadhasivam (5):
  dt-bindings: phy: qcom,qmp: Add SM8250 PCIe PHY bindings
  phy: qcom-qmp: Add SM8250 PCIe QMP PHYs
  dt-bindings: pci: qcom: Document PCIe bindings for SM8250 SoC
  PCI: qcom: Add SM8250 SoC support
  PCI: qcom: Add support for configuring BDF to SID mapping for SM8250

 .../devicetree/bindings/pci/qcom,pcie.txt     |   6 +-
 .../devicetree/bindings/phy/qcom,qmp-phy.yaml |   6 +
 drivers/pci/controller/dwc/Kconfig            |   1 +
 drivers/pci/controller/dwc/pcie-qcom.c        | 149 ++++++++++
 drivers/phy/qualcomm/phy-qcom-qmp.c           | 278 +++++++++++++++++-
 drivers/phy/qualcomm/phy-qcom-qmp.h           |  18 ++
 6 files changed, 454 insertions(+), 4 deletions(-)

-- 
2.17.1


^ permalink raw reply	[flat|nested] 14+ messages in thread

* [PATCH v2 1/5] dt-bindings: phy: qcom,qmp: Add SM8250 PCIe PHY bindings
  2020-09-30 15:09 [PATCH v2 0/5] Add PCIe support for SM8250 SoC Manivannan Sadhasivam
@ 2020-09-30 15:09 ` Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 2/5] phy: qcom-qmp: Add SM8250 PCIe QMP PHYs Manivannan Sadhasivam
                   ` (3 subsequent siblings)
  4 siblings, 0 replies; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-09-30 15:09 UTC (permalink / raw)
  To: agross, bjorn.andersson, kishon, vkoul, robh
  Cc: svarbanov, bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree, Manivannan Sadhasivam

Add the below three PCIe PHYs found in SM8250 to the QMP binding:

QMP GEN3x1 PHY - 1 lane
QMP GEN3x2 PHY - 2 lanes
QMP Modem PHY - 2 lanes

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml b/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml
index 185cdea9cf81..ec05db374645 100644
--- a/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml
+++ b/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml
@@ -31,6 +31,9 @@ properties:
       - qcom,sdm845-qmp-usb3-uni-phy
       - qcom,sm8150-qmp-ufs-phy
       - qcom,sm8250-qmp-ufs-phy
+      - qcom,sm8250-qmp-gen3x1-pcie-phy
+      - qcom,sm8250-qmp-gen3x2-pcie-phy
+      - qcom,sm8250-qmp-modem-pcie-phy
 
   reg:
     items:
@@ -259,6 +262,9 @@ allOf:
             enum:
               - qcom,sdm845-qhp-pcie-phy
               - qcom,sdm845-qmp-pcie-phy
+              - qcom,sm8250-qmp-gen3x1-pcie-phy
+              - qcom,sm8250-qmp-gen3x2-pcie-phy
+              - qcom,sm8250-qmp-modem-pcie-phy
     then:
       properties:
         clocks:
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 14+ messages in thread

* [PATCH v2 2/5] phy: qcom-qmp: Add SM8250 PCIe QMP PHYs
  2020-09-30 15:09 [PATCH v2 0/5] Add PCIe support for SM8250 SoC Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 1/5] dt-bindings: phy: qcom,qmp: Add SM8250 PCIe PHY bindings Manivannan Sadhasivam
@ 2020-09-30 15:09 ` Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 3/5] dt-bindings: pci: qcom: Document PCIe bindings for SM8250 SoC Manivannan Sadhasivam
                   ` (2 subsequent siblings)
  4 siblings, 0 replies; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-09-30 15:09 UTC (permalink / raw)
  To: agross, bjorn.andersson, kishon, vkoul, robh
  Cc: svarbanov, bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree, Manivannan Sadhasivam

SM8250 has multiple different PHY versions:
QMP GEN3x1 PHY - 1 lane
QMP GEN3x2 PHY - 2 lanes
QMP Modem PHY - 2 lanes

Add support for these with relevant init sequence. In order to abstract
the init sequence, this commit introduces secondary tables which can
be used to factor out the unique sequence for each PHY while the former
tables can have the common sequence.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/phy/qualcomm/phy-qcom-qmp.c | 278 +++++++++++++++++++++++++++-
 drivers/phy/qualcomm/phy-qcom-qmp.h |  18 ++
 2 files changed, 294 insertions(+), 2 deletions(-)

diff --git a/drivers/phy/qualcomm/phy-qcom-qmp.c b/drivers/phy/qualcomm/phy-qcom-qmp.c
index 562053ce9455..dd72e29b42f8 100644
--- a/drivers/phy/qualcomm/phy-qcom-qmp.c
+++ b/drivers/phy/qualcomm/phy-qcom-qmp.c
@@ -217,6 +217,13 @@ static const unsigned int sdm845_ufsphy_regs_layout[QPHY_LAYOUT_SIZE] = {
 	[QPHY_PCS_READY_STATUS]		= 0x160,
 };
 
+static const unsigned int sm8250_pcie_regs_layout[QPHY_LAYOUT_SIZE] = {
+	[QPHY_SW_RESET]			= 0x00,
+	[QPHY_START_CTRL]		= 0x44,
+	[QPHY_PCS_STATUS]		= 0x14,
+	[QPHY_PCS_POWER_DOWN_CONTROL]	= 0x40,
+};
+
 static const unsigned int sm8150_ufsphy_regs_layout[QPHY_LAYOUT_SIZE] = {
 	[QPHY_START_CTRL]		= QPHY_V4_PCS_UFS_PHY_START,
 	[QPHY_PCS_READY_STATUS]		= QPHY_V4_PCS_UFS_READY_STATUS,
@@ -1743,6 +1750,149 @@ static const struct qmp_phy_init_tbl sm8250_usb3_uniphy_pcs_tbl[] = {
 	QMP_PHY_INIT_CFG(QPHY_V4_PCS_REFGEN_REQ_CONFIG1, 0x21),
 };
 
+static const struct qmp_phy_init_tbl sm8250_qmp_pcie_serdes_tbl[] = {
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SYSCLK_EN_SEL, 0x08),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_CLK_SELECT, 0x34),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_CORECLK_DIV_MODE1, 0x08),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_PLL_IVCO, 0x0f),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_LOCK_CMP_EN, 0x42),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_VCO_TUNE1_MODE0, 0x24),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_VCO_TUNE2_MODE1, 0x03),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_VCO_TUNE1_MODE1, 0xb4),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_VCO_TUNE_MAP, 0x02),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_BIN_VCOCAL_HSCLK_SEL, 0x11),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DEC_START_MODE0, 0x82),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DIV_FRAC_START3_MODE0, 0x03),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DIV_FRAC_START2_MODE0, 0x55),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DIV_FRAC_START1_MODE0, 0x55),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_LOCK_CMP2_MODE0, 0x1a),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_LOCK_CMP1_MODE0, 0x0a),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DEC_START_MODE1, 0x68),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DIV_FRAC_START3_MODE1, 0x02),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DIV_FRAC_START2_MODE1, 0xaa),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_DIV_FRAC_START1_MODE1, 0xab),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_LOCK_CMP2_MODE1, 0x34),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_LOCK_CMP1_MODE1, 0x14),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_HSCLK_SEL, 0x01),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_CP_CTRL_MODE0, 0x06),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_PLL_RCTRL_MODE0, 0x16),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_PLL_CCTRL_MODE0, 0x36),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_CP_CTRL_MODE1, 0x06),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_PLL_RCTRL_MODE1, 0x16),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_PLL_CCTRL_MODE1, 0x36),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_BIN_VCOCAL_CMP_CODE2_MODE0, 0x1e),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_BIN_VCOCAL_CMP_CODE1_MODE0, 0xca),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_BIN_VCOCAL_CMP_CODE2_MODE1, 0x18),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_BIN_VCOCAL_CMP_CODE1_MODE1, 0xa2),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SSC_EN_CENTER, 0x01),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SSC_PER1, 0x31),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SSC_PER2, 0x01),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SSC_STEP_SIZE1_MODE0, 0xde),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SSC_STEP_SIZE2_MODE0, 0x07),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SSC_STEP_SIZE1_MODE1, 0x4c),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SSC_STEP_SIZE2_MODE1, 0x06),
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_CLK_ENABLE1, 0x90),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x1_pcie_serdes_tbl[] = {
+	QMP_PHY_INIT_CFG(QSERDES_V4_COM_SYSCLK_BUF_ENABLE, 0x07),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_pcie_tx_tbl[] = {
+	QMP_PHY_INIT_CFG(QSERDES_V4_TX_RCV_DETECT_LVL_2, 0x12),
+	QMP_PHY_INIT_CFG(QSERDES_V4_TX_LANE_MODE_1, 0x35),
+	QMP_PHY_INIT_CFG(QSERDES_V4_TX_RES_CODE_LANE_OFFSET_TX, 0x11),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_pcie_rx_tbl[] = {
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_UCDR_FO_GAIN, 0x0c),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_UCDR_SO_GAIN, 0x03),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_GM_CAL, 0x1b),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_IDAC_TSETTLE_HIGH, 0x00),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_IDAC_TSETTLE_LOW, 0xc0),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_AUX_DATA_TCOARSE_TFINE, 0x30),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_VGA_CAL_CNTRL1, 0x04),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_VGA_CAL_CNTRL2, 0x07),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_UCDR_SO_SATURATION_AND_ENABLE, 0x7f),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_UCDR_PI_CONTROLS, 0x70),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL2, 0x0e),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL3, 0x4a),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL4, 0x0f),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_SIGDET_CNTRL, 0x03),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_SIGDET_ENABLES, 0x1c),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_SIGDET_DEGLITCH_CNTRL, 0x1e),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_EQ_OFFSET_ADAPTOR_CNTRL1, 0x17),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_10_LOW, 0xd4),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_10_HIGH, 0x54),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_10_HIGH2, 0xdb),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_10_HIGH3, 0x3b),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_10_HIGH4, 0x31),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_01_LOW, 0x24),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_00_HIGH2, 0xff),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_00_HIGH3, 0x7f),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_DCC_CTRL1, 0x0c),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_01_HIGH, 0xe4),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_01_HIGH2, 0xec),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_01_HIGH3, 0x3b),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_01_HIGH4, 0x36),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x1_pcie_rx_tbl[] = {
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RCLK_AUXDATA_SEL, 0x00),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL1, 0x00),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_DFE_EN_TIMER, 0x04),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_00_LOW, 0x3f),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_00_HIGH4, 0x14),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_DFE_CTLE_POST_CAL_OFFSET, 0x30),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_pcie_pcs_tbl[] = {
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_P2U3_WAKEUP_DLY_TIME_AUXCLK_L, 0x01),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_RX_SIGDET_LVL, 0x77),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_RATE_SLEW_CNTRL1, 0x0b),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x1_pcie_pcs_tbl[] = {
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_REFGEN_REQ_CONFIG1, 0x0d),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_EQ_CONFIG5, 0x12),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_pcie_pcs_misc_tbl[] = {
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_OSC_DTCT_ACTIONS, 0x00),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_L1P1_WAKEUP_DLY_TIME_AUXCLK_L, 0x01),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_L1P2_WAKEUP_DLY_TIME_AUXCLK_L, 0x01),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_PRESET_P6_P7_PRE, 0x33),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_PRESET_P10_PRE, 0x00),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_PRESET_P10_POST, 0x58),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_ENDPOINT_REFCLK_DRIVE, 0xc1),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x1_pcie_pcs_misc_tbl[] = {
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_INT_AUX_CLK_CONFIG1, 0x00),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_EQ_CONFIG2, 0x0f),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x2_pcie_tx_tbl[] = {
+	QMP_PHY_INIT_CFG(QSERDES_V4_TX_PI_QEC_CTRL, 0x20),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x2_pcie_rx_tbl[] = {
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL1, 0x04),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_00_LOW, 0xbf),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_RX_MODE_00_HIGH4, 0x15),
+	QMP_PHY_INIT_CFG(QSERDES_V4_RX_DFE_CTLE_POST_CAL_OFFSET, 0x38),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x2_pcie_pcs_tbl[] = {
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_REFGEN_REQ_CONFIG1, 0x05),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_EQ_CONFIG2, 0x0f),
+};
+
+static const struct qmp_phy_init_tbl sm8250_qmp_gen3x2_pcie_pcs_misc_tbl[] = {
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_POWER_STATE_CONFIG2, 0x0d),
+	QMP_PHY_INIT_CFG(QPHY_V4_PCS_PCIE_POWER_STATE_CONFIG4, 0x07),
+};
+
 /* struct qmp_phy_cfg - per-PHY initialization config */
 struct qmp_phy_cfg {
 	/* phy-type - PCIE/UFS/USB */
@@ -1753,14 +1903,24 @@ struct qmp_phy_cfg {
 	/* Init sequence for PHY blocks - serdes, tx, rx, pcs */
 	const struct qmp_phy_init_tbl *serdes_tbl;
 	int serdes_tbl_num;
+	const struct qmp_phy_init_tbl *serdes_tbl_sec;
+	int serdes_tbl_num_sec;
 	const struct qmp_phy_init_tbl *tx_tbl;
 	int tx_tbl_num;
+	const struct qmp_phy_init_tbl *tx_tbl_sec;
+	int tx_tbl_num_sec;
 	const struct qmp_phy_init_tbl *rx_tbl;
 	int rx_tbl_num;
+	const struct qmp_phy_init_tbl *rx_tbl_sec;
+	int rx_tbl_num_sec;
 	const struct qmp_phy_init_tbl *pcs_tbl;
 	int pcs_tbl_num;
+	const struct qmp_phy_init_tbl *pcs_tbl_sec;
+	int pcs_tbl_num_sec;
 	const struct qmp_phy_init_tbl *pcs_misc_tbl;
 	int pcs_misc_tbl_num;
+	const struct qmp_phy_init_tbl *pcs_misc_tbl_sec;
+	int pcs_misc_tbl_num_sec;
 
 	/* clock ids to be requested */
 	const char * const *clk_list;
@@ -2139,6 +2299,83 @@ static const struct qmp_phy_cfg sdm845_qhp_pciephy_cfg = {
 	.pwrdn_delay_max	= 1005,		/* us */
 };
 
+static const struct qmp_phy_cfg sm8250_qmp_gen3x1_pciephy_cfg = {
+	.type = PHY_TYPE_PCIE,
+	.nlanes = 1,
+
+	.serdes_tbl		= sm8250_qmp_pcie_serdes_tbl,
+	.serdes_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_serdes_tbl),
+	.serdes_tbl_sec		= sm8250_qmp_gen3x1_pcie_serdes_tbl,
+	.serdes_tbl_num_sec	= ARRAY_SIZE(sm8250_qmp_gen3x1_pcie_serdes_tbl),
+	.tx_tbl			= sm8250_qmp_pcie_tx_tbl,
+	.tx_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_tx_tbl),
+	.rx_tbl			= sm8250_qmp_pcie_rx_tbl,
+	.rx_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_rx_tbl),
+	.rx_tbl_sec		= sm8250_qmp_gen3x1_pcie_rx_tbl,
+	.rx_tbl_num_sec		= ARRAY_SIZE(sm8250_qmp_gen3x1_pcie_rx_tbl),
+	.pcs_tbl		= sm8250_qmp_pcie_pcs_tbl,
+	.pcs_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_pcs_tbl),
+	.pcs_tbl_sec		= sm8250_qmp_gen3x1_pcie_pcs_tbl,
+	.pcs_tbl_num_sec		= ARRAY_SIZE(sm8250_qmp_gen3x1_pcie_pcs_tbl),
+	.pcs_misc_tbl		= sm8250_qmp_pcie_pcs_misc_tbl,
+	.pcs_misc_tbl_num	= ARRAY_SIZE(sm8250_qmp_pcie_pcs_misc_tbl),
+	.pcs_misc_tbl_sec		= sm8250_qmp_gen3x1_pcie_pcs_misc_tbl,
+	.pcs_misc_tbl_num_sec	= ARRAY_SIZE(sm8250_qmp_gen3x1_pcie_pcs_misc_tbl),
+	.clk_list		= sdm845_pciephy_clk_l,
+	.num_clks		= ARRAY_SIZE(sdm845_pciephy_clk_l),
+	.reset_list		= sdm845_pciephy_reset_l,
+	.num_resets		= ARRAY_SIZE(sdm845_pciephy_reset_l),
+	.vreg_list		= qmp_phy_vreg_l,
+	.num_vregs		= ARRAY_SIZE(qmp_phy_vreg_l),
+	.regs			= sm8250_pcie_regs_layout,
+
+	.start_ctrl		= PCS_START | SERDES_START,
+	.pwrdn_ctrl		= SW_PWRDN | REFCLK_DRV_DSBL,
+
+	.has_pwrdn_delay	= true,
+	.pwrdn_delay_min	= 995,		/* us */
+	.pwrdn_delay_max	= 1005,		/* us */
+};
+
+static const struct qmp_phy_cfg sm8250_qmp_gen3x2_pciephy_cfg = {
+	.type = PHY_TYPE_PCIE,
+	.nlanes = 2,
+
+	.serdes_tbl		= sm8250_qmp_pcie_serdes_tbl,
+	.serdes_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_serdes_tbl),
+	.tx_tbl			= sm8250_qmp_pcie_tx_tbl,
+	.tx_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_tx_tbl),
+	.tx_tbl_sec		= sm8250_qmp_gen3x2_pcie_tx_tbl,
+	.tx_tbl_num_sec		= ARRAY_SIZE(sm8250_qmp_gen3x2_pcie_tx_tbl),
+	.rx_tbl			= sm8250_qmp_pcie_rx_tbl,
+	.rx_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_rx_tbl),
+	.rx_tbl_sec		= sm8250_qmp_gen3x2_pcie_rx_tbl,
+	.rx_tbl_num_sec		= ARRAY_SIZE(sm8250_qmp_gen3x2_pcie_rx_tbl),
+	.pcs_tbl		= sm8250_qmp_pcie_pcs_tbl,
+	.pcs_tbl_num		= ARRAY_SIZE(sm8250_qmp_pcie_pcs_tbl),
+	.pcs_tbl_sec		= sm8250_qmp_gen3x2_pcie_pcs_tbl,
+	.pcs_tbl_num_sec		= ARRAY_SIZE(sm8250_qmp_gen3x2_pcie_pcs_tbl),
+	.pcs_misc_tbl		= sm8250_qmp_pcie_pcs_misc_tbl,
+	.pcs_misc_tbl_num	= ARRAY_SIZE(sm8250_qmp_pcie_pcs_misc_tbl),
+	.pcs_misc_tbl_sec		= sm8250_qmp_gen3x2_pcie_pcs_misc_tbl,
+	.pcs_misc_tbl_num_sec	= ARRAY_SIZE(sm8250_qmp_gen3x2_pcie_pcs_misc_tbl),
+	.clk_list		= sdm845_pciephy_clk_l,
+	.num_clks		= ARRAY_SIZE(sdm845_pciephy_clk_l),
+	.reset_list		= sdm845_pciephy_reset_l,
+	.num_resets		= ARRAY_SIZE(sdm845_pciephy_reset_l),
+	.vreg_list		= qmp_phy_vreg_l,
+	.num_vregs		= ARRAY_SIZE(qmp_phy_vreg_l),
+	.regs			= sm8250_pcie_regs_layout,
+
+	.start_ctrl		= PCS_START | SERDES_START,
+	.pwrdn_ctrl		= SW_PWRDN | REFCLK_DRV_DSBL,
+
+	.is_dual_lane_phy	= true,
+	.has_pwrdn_delay	= true,
+	.pwrdn_delay_min	= 995,		/* us */
+	.pwrdn_delay_max	= 1005,		/* us */
+};
+
 static const struct qmp_phy_cfg qmp_v3_usb3phy_cfg = {
 	.type			= PHY_TYPE_USB3,
 	.nlanes			= 1,
@@ -2556,6 +2793,9 @@ static int qcom_qmp_phy_com_init(struct qmp_phy *qphy)
 	/* Serdes configuration */
 	qcom_qmp_phy_configure(serdes, cfg->regs, cfg->serdes_tbl,
 			       cfg->serdes_tbl_num);
+	if (cfg->serdes_tbl_sec)
+		qcom_qmp_phy_configure(serdes, cfg->regs, cfg->serdes_tbl_sec,
+				       cfg->serdes_tbl_num_sec);
 
 	if (cfg->has_phy_com_ctrl) {
 		void __iomem *status;
@@ -2693,24 +2933,49 @@ static int qcom_qmp_phy_enable(struct phy *phy)
 	/* Tx, Rx, and PCS configurations */
 	qcom_qmp_phy_configure_lane(tx, cfg->regs,
 				    cfg->tx_tbl, cfg->tx_tbl_num, 1);
+	if (cfg->tx_tbl_sec)
+		qcom_qmp_phy_configure_lane(qphy->tx2, cfg->regs, cfg->tx_tbl_sec,
+					    cfg->tx_tbl_num_sec, 2);
+
 	/* Configuration for other LANE for USB-DP combo PHY */
-	if (cfg->is_dual_lane_phy)
+	if (cfg->is_dual_lane_phy) {
 		qcom_qmp_phy_configure_lane(qphy->tx2, cfg->regs,
 					    cfg->tx_tbl, cfg->tx_tbl_num, 2);
+		if (cfg->tx_tbl_sec)
+			qcom_qmp_phy_configure_lane(qphy->tx2, cfg->regs,
+						    cfg->tx_tbl_sec,
+						    cfg->tx_tbl_num_sec, 2);
+	}
 
 	qcom_qmp_phy_configure_lane(rx, cfg->regs,
 				    cfg->rx_tbl, cfg->rx_tbl_num, 1);
-	if (cfg->is_dual_lane_phy)
+	if (cfg->rx_tbl_sec)
+		qcom_qmp_phy_configure_lane(rx, cfg->regs,
+					    cfg->rx_tbl_sec, cfg->rx_tbl_num_sec, 1);
+
+	if (cfg->is_dual_lane_phy) {
 		qcom_qmp_phy_configure_lane(qphy->rx2, cfg->regs,
 					    cfg->rx_tbl, cfg->rx_tbl_num, 2);
+		if (cfg->rx_tbl_sec)
+			qcom_qmp_phy_configure_lane(rx, cfg->regs,
+						    cfg->rx_tbl_sec,
+						    cfg->rx_tbl_num_sec, 1);
+	}
 
 	qcom_qmp_phy_configure(pcs, cfg->regs, cfg->pcs_tbl, cfg->pcs_tbl_num);
+	if (cfg->pcs_tbl_sec)
+		qcom_qmp_phy_configure(pcs, cfg->regs, cfg->pcs_tbl_sec,
+				       cfg->pcs_tbl_num_sec);
+
 	ret = reset_control_deassert(qmp->ufs_reset);
 	if (ret)
 		goto err_lane_rst;
 
 	qcom_qmp_phy_configure(pcs_misc, cfg->regs, cfg->pcs_misc_tbl,
 			       cfg->pcs_misc_tbl_num);
+	if (cfg->pcs_misc_tbl_sec)
+		qcom_qmp_phy_configure(pcs_misc, cfg->regs, cfg->pcs_misc_tbl_sec,
+				       cfg->pcs_misc_tbl_num_sec);
 
 	/*
 	 * Pull out PHY from POWER DOWN state.
@@ -3232,6 +3497,15 @@ static const struct of_device_id qcom_qmp_phy_of_match_table[] = {
 	}, {
 		.compatible = "qcom,sm8250-qmp-usb3-uni-phy",
 		.data = &sm8250_usb3_uniphy_cfg,
+	}, {
+		.compatible = "qcom,sm8250-qmp-gen3x1-pcie-phy",
+		.data = &sm8250_qmp_gen3x1_pciephy_cfg,
+	}, {
+		.compatible = "qcom,sm8250-qmp-gen3x2-pcie-phy",
+		.data = &sm8250_qmp_gen3x2_pciephy_cfg,
+	}, {
+		.compatible = "qcom,sm8250-qmp-modem-pcie-phy",
+		.data = &sm8250_qmp_gen3x2_pciephy_cfg,
 	},
 	{ },
 };
diff --git a/drivers/phy/qualcomm/phy-qcom-qmp.h b/drivers/phy/qualcomm/phy-qcom-qmp.h
index 4277f592684b..092f1884cba6 100644
--- a/drivers/phy/qualcomm/phy-qcom-qmp.h
+++ b/drivers/phy/qualcomm/phy-qcom-qmp.h
@@ -321,6 +321,7 @@
 #define QSERDES_V4_COM_SSC_STEP_SIZE2_MODE0		0x028
 #define QSERDES_V4_COM_SSC_STEP_SIZE1_MODE1		0x030
 #define QSERDES_V4_COM_SSC_STEP_SIZE2_MODE1		0x034
+#define QSERDES_V4_COM_CLK_ENABLE1			0x048
 #define QSERDES_V4_COM_SYSCLK_BUF_ENABLE		0x050
 #define QSERDES_V4_COM_PLL_IVCO				0x058
 #define QSERDES_V4_COM_CMN_IPTRIM			0x060
@@ -350,6 +351,7 @@
 #define QSERDES_V4_COM_VCO_TUNE1_MODE1			0x118
 #define QSERDES_V4_COM_VCO_TUNE2_MODE1			0x11c
 #define QSERDES_V4_COM_VCO_TUNE_INITVAL2		0x124
+#define QSERDES_V4_COM_CLK_SELECT			0x154
 #define QSERDES_V4_COM_HSCLK_SEL			0x158
 #define QSERDES_V4_COM_HSCLK_HS_SWITCH_SEL		0x15c
 #define QSERDES_V4_COM_CORECLK_DIV_MODE1		0x16c
@@ -389,12 +391,14 @@
 #define QSERDES_V4_RX_UCDR_SB2_GAIN1			0x054
 #define QSERDES_V4_RX_UCDR_SB2_GAIN2			0x058
 #define QSERDES_V4_RX_AUX_DATA_TCOARSE_TFINE			0x060
+#define QSERDES_V4_RX_RCLK_AUXDATA_SEL			0x064
 #define QSERDES_V4_RX_AC_JTAG_ENABLE			0x068
 #define QSERDES_V4_RX_AC_JTAG_MODE			0x078
 #define QSERDES_V4_RX_RX_TERM_BW			0x080
 #define QSERDES_V4_RX_VGA_CAL_CNTRL1			0x0d4
 #define QSERDES_V4_RX_VGA_CAL_CNTRL2			0x0d8
 #define QSERDES_V4_RX_GM_CAL				0x0dc
+#define QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL1		0x0e8
 #define QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL2		0x0ec
 #define QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL3		0x0f0
 #define QSERDES_V4_RX_RX_EQU_ADAPTOR_CNTRL4		0x0f4
@@ -403,6 +407,7 @@
 #define QSERDES_V4_RX_RX_IDAC_MEASURE_TIME		0x100
 #define QSERDES_V4_RX_RX_EQ_OFFSET_ADAPTOR_CNTRL1	0x110
 #define QSERDES_V4_RX_RX_OFFSET_ADAPTOR_CNTRL2		0x114
+#define QSERDES_V4_RX_SIGDET_ENABLES			0x118
 #define QSERDES_V4_RX_SIGDET_CNTRL			0x11c
 #define QSERDES_V4_RX_SIGDET_LVL			0x120
 #define QSERDES_V4_RX_SIGDET_DEGLITCH_CNTRL		0x124
@@ -724,4 +729,17 @@
 #define QPHY_V4_PCS_MISC_TYPEC_STATUS			0x10
 #define QPHY_V4_PCS_MISC_PLACEHOLDER_STATUS		0x14
 
+/* Only for QMP V4 PHY - PCS_PCIE registers (same as PCS_MISC?) */
+#define QPHY_V4_PCS_PCIE_POWER_STATE_CONFIG2		0x0c
+#define QPHY_V4_PCS_PCIE_POWER_STATE_CONFIG4		0x14
+#define QPHY_V4_PCS_PCIE_ENDPOINT_REFCLK_DRIVE		0x1c
+#define QPHY_V4_PCS_PCIE_L1P1_WAKEUP_DLY_TIME_AUXCLK_L	0x40
+#define QPHY_V4_PCS_PCIE_L1P2_WAKEUP_DLY_TIME_AUXCLK_L	0x48
+#define QPHY_V4_PCS_PCIE_INT_AUX_CLK_CONFIG1		0x50
+#define QPHY_V4_PCS_PCIE_OSC_DTCT_ACTIONS		0x90
+#define QPHY_V4_PCS_PCIE_EQ_CONFIG2			0xa4
+#define QPHY_V4_PCS_PCIE_PRESET_P6_P7_PRE		0xb4
+#define QPHY_V4_PCS_PCIE_PRESET_P10_PRE			0xbc
+#define QPHY_V4_PCS_PCIE_PRESET_P10_POST		0xe0
+
 #endif
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 14+ messages in thread

* [PATCH v2 3/5] dt-bindings: pci: qcom: Document PCIe bindings for SM8250 SoC
  2020-09-30 15:09 [PATCH v2 0/5] Add PCIe support for SM8250 SoC Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 1/5] dt-bindings: phy: qcom,qmp: Add SM8250 PCIe PHY bindings Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 2/5] phy: qcom-qmp: Add SM8250 PCIe QMP PHYs Manivannan Sadhasivam
@ 2020-09-30 15:09 ` Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 4/5] PCI: qcom: Add SM8250 SoC support Manivannan Sadhasivam
  2020-09-30 15:09 ` [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250 Manivannan Sadhasivam
  4 siblings, 0 replies; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-09-30 15:09 UTC (permalink / raw)
  To: agross, bjorn.andersson, kishon, vkoul, robh
  Cc: svarbanov, bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree, Manivannan Sadhasivam

Document the PCIe DT bindings for SM8250 SoC. The PCIe IP is similar to
the one used on SDM845, hence just add the compatible along with the
optional "atu" register region.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 Documentation/devicetree/bindings/pci/qcom,pcie.txt | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie.txt b/Documentation/devicetree/bindings/pci/qcom,pcie.txt
index 02bc81bb8b2d..3b55310390a0 100644
--- a/Documentation/devicetree/bindings/pci/qcom,pcie.txt
+++ b/Documentation/devicetree/bindings/pci/qcom,pcie.txt
@@ -13,6 +13,7 @@
 			- "qcom,pcie-ipq8074" for ipq8074
 			- "qcom,pcie-qcs404" for qcs404
 			- "qcom,pcie-sdm845" for sdm845
+			- "qcom,pcie-sm8250" for sm8250
 
 - reg:
 	Usage: required
@@ -27,6 +28,7 @@
 			- "dbi"	   DesignWare PCIe registers
 			- "elbi"   External local bus interface registers
 			- "config" PCIe configuration space
+			- "atu"    ATU address space (optional)
 
 - device_type:
 	Usage: required
@@ -131,7 +133,7 @@
 			- "slave_bus"	AXI Slave clock
 
 -clock-names:
-	Usage: required for sdm845
+	Usage: required for sdm845 and sm8250
 	Value type: <stringlist>
 	Definition: Should contain the following entries
 			- "aux"		Auxiliary clock
@@ -206,7 +208,7 @@
 			- "ahb"			AHB reset
 
 - reset-names:
-	Usage: required for sdm845
+	Usage: required for sdm845 and sm8250
 	Value type: <stringlist>
 	Definition: Should contain the following entries
 			- "pci"			PCIe core reset
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 14+ messages in thread

* [PATCH v2 4/5] PCI: qcom: Add SM8250 SoC support
  2020-09-30 15:09 [PATCH v2 0/5] Add PCIe support for SM8250 SoC Manivannan Sadhasivam
                   ` (2 preceding siblings ...)
  2020-09-30 15:09 ` [PATCH v2 3/5] dt-bindings: pci: qcom: Document PCIe bindings for SM8250 SoC Manivannan Sadhasivam
@ 2020-09-30 15:09 ` Manivannan Sadhasivam
  2020-09-30 21:56   ` Stanimir Varbanov
  2020-09-30 15:09 ` [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250 Manivannan Sadhasivam
  4 siblings, 1 reply; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-09-30 15:09 UTC (permalink / raw)
  To: agross, bjorn.andersson, kishon, vkoul, robh
  Cc: svarbanov, bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree, Manivannan Sadhasivam

The PCIe IP on SM8250 SoC is similar to the one used on SDM845. Hence
the support is added reusing the members of ops_2_7_0. The key
difference between ops_2_7_0 and ops_sm8250 is the config_sid callback,
which will be added in successive commit.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/pci/controller/dwc/pcie-qcom.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
index 3aac77a295ba..44db91861b47 100644
--- a/drivers/pci/controller/dwc/pcie-qcom.c
+++ b/drivers/pci/controller/dwc/pcie-qcom.c
@@ -1359,6 +1359,16 @@ static const struct qcom_pcie_ops ops_2_7_0 = {
 	.post_deinit = qcom_pcie_post_deinit_2_7_0,
 };
 
+/* Qcom IP rev.: 1.9.0 */
+static const struct qcom_pcie_ops ops_sm8250 = {
+	.get_resources = qcom_pcie_get_resources_2_7_0,
+	.init = qcom_pcie_init_2_7_0,
+	.deinit = qcom_pcie_deinit_2_7_0,
+	.ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
+	.post_init = qcom_pcie_post_init_2_7_0,
+	.post_deinit = qcom_pcie_post_deinit_2_7_0,
+};
+
 static const struct dw_pcie_ops dw_pcie_ops = {
 	.link_up = qcom_pcie_link_up,
 };
@@ -1476,6 +1486,7 @@ static const struct of_device_id qcom_pcie_match[] = {
 	{ .compatible = "qcom,pcie-ipq4019", .data = &ops_2_4_0 },
 	{ .compatible = "qcom,pcie-qcs404", .data = &ops_2_4_0 },
 	{ .compatible = "qcom,pcie-sdm845", .data = &ops_2_7_0 },
+	{ .compatible = "qcom,pcie-sm8250", .data = &ops_sm8250 },
 	{ }
 };
 
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 14+ messages in thread

* [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250
  2020-09-30 15:09 [PATCH v2 0/5] Add PCIe support for SM8250 SoC Manivannan Sadhasivam
                   ` (3 preceding siblings ...)
  2020-09-30 15:09 ` [PATCH v2 4/5] PCI: qcom: Add SM8250 SoC support Manivannan Sadhasivam
@ 2020-09-30 15:09 ` Manivannan Sadhasivam
  2020-09-30 21:46   ` Stanimir Varbanov
  4 siblings, 1 reply; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-09-30 15:09 UTC (permalink / raw)
  To: agross, bjorn.andersson, kishon, vkoul, robh
  Cc: svarbanov, bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree, Manivannan Sadhasivam

For SM8250, we need to write the BDF to SID mapping in PCIe controller
register space for proper working. This is accomplished by extracting
the BDF and SID values from "iommu-map" property in DT and writing those
in the register address calculated from the hash value of BDF. In case
of collisions, the index of the next entry will also be written.

For the sake of it, let's introduce a "config_sid" callback and do it
conditionally for SM8250.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/pci/controller/dwc/Kconfig     |   1 +
 drivers/pci/controller/dwc/pcie-qcom.c | 138 +++++++++++++++++++++++++
 2 files changed, 139 insertions(+)

diff --git a/drivers/pci/controller/dwc/Kconfig b/drivers/pci/controller/dwc/Kconfig
index 044a3761c44f..3e9ccdc45ee1 100644
--- a/drivers/pci/controller/dwc/Kconfig
+++ b/drivers/pci/controller/dwc/Kconfig
@@ -169,6 +169,7 @@ config PCIE_QCOM
 	depends on OF && (ARCH_QCOM || COMPILE_TEST)
 	depends on PCI_MSI_IRQ_DOMAIN
 	select PCIE_DW_HOST
+	select CRC8
 	help
 	  Say Y here to enable PCIe controller support on Qualcomm SoCs. The
 	  PCIe controller uses the DesignWare core plus Qualcomm-specific
diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
index 44db91861b47..a7f05b78315b 100644
--- a/drivers/pci/controller/dwc/pcie-qcom.c
+++ b/drivers/pci/controller/dwc/pcie-qcom.c
@@ -9,6 +9,7 @@
  */
 
 #include <linux/clk.h>
+#include <linux/crc8.h>
 #include <linux/delay.h>
 #include <linux/gpio/consumer.h>
 #include <linux/interrupt.h>
@@ -57,6 +58,7 @@
 #define PCIE20_PARF_SID_OFFSET			0x234
 #define PCIE20_PARF_BDF_TRANSLATE_CFG		0x24C
 #define PCIE20_PARF_DEVICE_TYPE			0x1000
+#define PCIE20_PARF_BDF_TO_SID_TABLE_N		0x2000
 
 #define PCIE20_ELBI_SYS_CTRL			0x04
 #define PCIE20_ELBI_SYS_CTRL_LT_ENABLE		BIT(0)
@@ -101,6 +103,9 @@
 
 #define QCOM_PCIE_2_1_0_MAX_SUPPLY	3
 #define QCOM_PCIE_2_1_0_MAX_CLOCKS	5
+
+#define QCOM_PCIE_CRC8_POLYNOMIAL (BIT(2) | BIT(1) | BIT(0))
+
 struct qcom_pcie_resources_2_1_0 {
 	struct clk_bulk_data clks[QCOM_PCIE_2_1_0_MAX_CLOCKS];
 	struct reset_control *pci_reset;
@@ -183,6 +188,16 @@ struct qcom_pcie_ops {
 	void (*deinit)(struct qcom_pcie *pcie);
 	void (*post_deinit)(struct qcom_pcie *pcie);
 	void (*ltssm_enable)(struct qcom_pcie *pcie);
+	int (*config_sid)(struct qcom_pcie *pcie);
+};
+
+/* sid info structure */
+struct qcom_pcie_sid_info_t {
+	u16 bdf;
+	u8 pcie_sid;
+	u8 hash;
+	u32 smmu_sid;
+	u32 value;
 };
 
 struct qcom_pcie {
@@ -193,6 +208,8 @@ struct qcom_pcie {
 	struct phy *phy;
 	struct gpio_desc *reset;
 	const struct qcom_pcie_ops *ops;
+	struct qcom_pcie_sid_info_t *sid_info;
+	u32 sid_info_len;
 	int gen;
 };
 
@@ -1257,6 +1274,120 @@ static int qcom_pcie_link_up(struct dw_pcie *pci)
 	return !!(val & PCI_EXP_LNKSTA_DLLLA);
 }
 
+static int qcom_pcie_get_iommu_map(struct qcom_pcie *pcie)
+{
+	/* iommu map structure */
+	struct {
+		u32 bdf;
+		u32 phandle;
+		u32 smmu_sid;
+		u32 smmu_sid_len;
+	} *map;
+	struct device *dev = pcie->pci->dev;
+	int i, size = 0;
+	u32 smmu_sid_base;
+
+	of_get_property(dev->of_node, "iommu-map", &size);
+	if (!size)
+		return 0;
+
+	map = kzalloc(size, GFP_KERNEL);
+	if (!map)
+		return -ENOMEM;
+
+	of_property_read_u32_array(dev->of_node,
+		"iommu-map", (u32 *)map, size / sizeof(u32));
+
+	pcie->sid_info_len = size / (sizeof(*map));
+	pcie->sid_info = devm_kcalloc(dev, pcie->sid_info_len,
+				sizeof(*pcie->sid_info), GFP_KERNEL);
+	if (!pcie->sid_info) {
+		kfree(map);
+		return -ENOMEM;
+	}
+
+	/* Extract the SMMU SID base from the first entry of iommu-map */
+	smmu_sid_base = map[0].smmu_sid;
+	for (i = 0; i < pcie->sid_info_len; i++) {
+		pcie->sid_info[i].bdf = map[i].bdf;
+		pcie->sid_info[i].smmu_sid = map[i].smmu_sid;
+		pcie->sid_info[i].pcie_sid =
+				pcie->sid_info[i].smmu_sid - smmu_sid_base;
+	}
+
+	kfree(map);
+
+	return 0;
+}
+
+static int qcom_pcie_config_sid_sm8250(struct qcom_pcie *pcie)
+{
+	void __iomem *bdf_to_sid_base = pcie->parf +
+		PCIE20_PARF_BDF_TO_SID_TABLE_N;
+	u8 qcom_pcie_crc8_table[CRC8_TABLE_SIZE];
+	int ret, i;
+
+	ret = qcom_pcie_get_iommu_map(pcie);
+	if (ret)
+		return ret;
+
+	if (!pcie->sid_info)
+		return 0;
+
+	crc8_populate_msb(qcom_pcie_crc8_table, QCOM_PCIE_CRC8_POLYNOMIAL);
+
+	/* Registers need to be zero out first */
+	memset_io(bdf_to_sid_base, 0, CRC8_TABLE_SIZE * sizeof(u32));
+
+	/* Initial setup for boot */
+	for (i = 0; i < pcie->sid_info_len; i++) {
+		struct qcom_pcie_sid_info_t *sid_info = &pcie->sid_info[i];
+		u16 bdf_be = cpu_to_be16(sid_info->bdf);
+		u32 val;
+		u8 hash;
+
+		hash = crc8(qcom_pcie_crc8_table, (u8 *)&bdf_be, sizeof(bdf_be),
+			0);
+
+		val = readl(bdf_to_sid_base + hash * sizeof(u32));
+
+		/* If there is a collision, look for next available entry */
+		while (val) {
+			u8 current_hash = hash++;
+			u8 next_mask = 0xff;
+
+			/* If NEXT field is NULL then update it with next hash */
+			if (!(val & next_mask)) {
+				int j;
+
+				val |= (u32)hash;
+				writel(val, bdf_to_sid_base +
+					current_hash * sizeof(u32));
+
+				/* Look for sid_info of current hash and update it */
+				for (j = 0; j < pcie->sid_info_len; j++) {
+					if (pcie->sid_info[j].hash !=
+						current_hash)
+						continue;
+
+					pcie->sid_info[j].value = val;
+					break;
+				}
+			}
+
+			val = readl(bdf_to_sid_base + hash * sizeof(u32));
+		}
+
+		val = sid_info->bdf << 16 | sid_info->pcie_sid << 8 | 0;
+		writel(val, bdf_to_sid_base + hash * sizeof(u32));
+
+		sid_info->hash = hash;
+		sid_info->value = val;
+	}
+
+	return 0;
+}
+
 static int qcom_pcie_host_init(struct pcie_port *pp)
 {
 	struct dw_pcie *pci = to_dw_pcie_from_pp(pp);
@@ -1290,6 +1421,12 @@ static int qcom_pcie_host_init(struct pcie_port *pp)
 	if (ret)
 		goto err;
 
+	if (pcie->ops->config_sid) {
+		ret = pcie->ops->config_sid(pcie);
+		if (ret)
+			goto err;
+	}
+
 	return 0;
 err:
 	qcom_ep_reset_assert(pcie);
@@ -1367,6 +1504,7 @@ static const struct qcom_pcie_ops ops_sm8250 = {
 	.ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
 	.post_init = qcom_pcie_post_init_2_7_0,
 	.post_deinit = qcom_pcie_post_deinit_2_7_0,
+	.config_sid = qcom_pcie_config_sid_sm8250,
 };
 
 static const struct dw_pcie_ops dw_pcie_ops = {
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250
  2020-09-30 15:09 ` [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250 Manivannan Sadhasivam
@ 2020-09-30 21:46   ` Stanimir Varbanov
  2020-10-01  5:57     ` Manivannan Sadhasivam
  0 siblings, 1 reply; 14+ messages in thread
From: Stanimir Varbanov @ 2020-09-30 21:46 UTC (permalink / raw)
  To: Manivannan Sadhasivam, agross, bjorn.andersson, kishon, vkoul, robh
  Cc: bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree

Hi Mani,

On 9/30/20 6:09 PM, Manivannan Sadhasivam wrote:
> For SM8250, we need to write the BDF to SID mapping in PCIe controller
> register space for proper working. This is accomplished by extracting
> the BDF and SID values from "iommu-map" property in DT and writing those
> in the register address calculated from the hash value of BDF. In case
> of collisions, the index of the next entry will also be written.

This describes what the patch is doing. But why? Is that done in the
other DWC low-level drivers or this is qcom specialty?

> 
> For the sake of it, let's introduce a "config_sid" callback and do it
> conditionally for SM8250.
> 
> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> ---
>  drivers/pci/controller/dwc/Kconfig     |   1 +
>  drivers/pci/controller/dwc/pcie-qcom.c | 138 +++++++++++++++++++++++++
>  2 files changed, 139 insertions(+)
> 
> diff --git a/drivers/pci/controller/dwc/Kconfig b/drivers/pci/controller/dwc/Kconfig
> index 044a3761c44f..3e9ccdc45ee1 100644
> --- a/drivers/pci/controller/dwc/Kconfig
> +++ b/drivers/pci/controller/dwc/Kconfig
> @@ -169,6 +169,7 @@ config PCIE_QCOM
>  	depends on OF && (ARCH_QCOM || COMPILE_TEST)
>  	depends on PCI_MSI_IRQ_DOMAIN
>  	select PCIE_DW_HOST
> +	select CRC8
>  	help
>  	  Say Y here to enable PCIe controller support on Qualcomm SoCs. The
>  	  PCIe controller uses the DesignWare core plus Qualcomm-specific
> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
> index 44db91861b47..a7f05b78315b 100644
> --- a/drivers/pci/controller/dwc/pcie-qcom.c
> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
> @@ -9,6 +9,7 @@
>   */
>  
>  #include <linux/clk.h>
> +#include <linux/crc8.h>
>  #include <linux/delay.h>
>  #include <linux/gpio/consumer.h>
>  #include <linux/interrupt.h>
> @@ -57,6 +58,7 @@
>  #define PCIE20_PARF_SID_OFFSET			0x234
>  #define PCIE20_PARF_BDF_TRANSLATE_CFG		0x24C
>  #define PCIE20_PARF_DEVICE_TYPE			0x1000
> +#define PCIE20_PARF_BDF_TO_SID_TABLE_N		0x2000
>  
>  #define PCIE20_ELBI_SYS_CTRL			0x04
>  #define PCIE20_ELBI_SYS_CTRL_LT_ENABLE		BIT(0)
> @@ -101,6 +103,9 @@
>  
>  #define QCOM_PCIE_2_1_0_MAX_SUPPLY	3
>  #define QCOM_PCIE_2_1_0_MAX_CLOCKS	5
> +
> +#define QCOM_PCIE_CRC8_POLYNOMIAL (BIT(2) | BIT(1) | BIT(0))
> +
>  struct qcom_pcie_resources_2_1_0 {
>  	struct clk_bulk_data clks[QCOM_PCIE_2_1_0_MAX_CLOCKS];
>  	struct reset_control *pci_reset;
> @@ -183,6 +188,16 @@ struct qcom_pcie_ops {
>  	void (*deinit)(struct qcom_pcie *pcie);
>  	void (*post_deinit)(struct qcom_pcie *pcie);
>  	void (*ltssm_enable)(struct qcom_pcie *pcie);
> +	int (*config_sid)(struct qcom_pcie *pcie);
> +};
> +
> +/* sid info structure */
> +struct qcom_pcie_sid_info_t {

why _t postfix? Maybe qcom_pcie_sid ?

SID - Stream ID ?

> +	u16 bdf;
> +	u8 pcie_sid;
> +	u8 hash;
> +	u32 smmu_sid;
> +	u32 value;
>  };
>  
>  struct qcom_pcie {
> @@ -193,6 +208,8 @@ struct qcom_pcie {
>  	struct phy *phy;
>  	struct gpio_desc *reset;
>  	const struct qcom_pcie_ops *ops;
> +	struct qcom_pcie_sid_info_t *sid_info;
> +	u32 sid_info_len;
>  	int gen;
>  };
>  
> @@ -1257,6 +1274,120 @@ static int qcom_pcie_link_up(struct dw_pcie *pci)
>  	return !!(val & PCI_EXP_LNKSTA_DLLLA);
>  }
>  
> +static int qcom_pcie_get_iommu_map(struct qcom_pcie *pcie)
> +{
> +	/* iommu map structure */
> +	struct {
> +		u32 bdf;
> +		u32 phandle;
> +		u32 smmu_sid;
> +		u32 smmu_sid_len;
> +	} *map;
> +	struct device *dev = pcie->pci->dev;
> +	int i, size = 0;
> +	u32 smmu_sid_base;
> +
> +	of_get_property(dev->of_node, "iommu-map", &size);
> +	if (!size)
> +		return 0;
> +
> +	map = kzalloc(size, GFP_KERNEL);
> +	if (!map)
> +		return -ENOMEM;
> +
> +	of_property_read_u32_array(dev->of_node,
> +		"iommu-map", (u32 *)map, size / sizeof(u32));

iommu-map is a standard DT property why we have to parse it manually?

> +
> +	pcie->sid_info_len = size / (sizeof(*map));
> +	pcie->sid_info = devm_kcalloc(dev, pcie->sid_info_len,
> +				sizeof(*pcie->sid_info), GFP_KERNEL);
> +	if (!pcie->sid_info) {
> +		kfree(map);
> +		return -ENOMEM;
> +	}
> +
> +	/* Extract the SMMU SID base from the first entry of iommu-map */
> +	smmu_sid_base = map[0].smmu_sid;
> +	for (i = 0; i < pcie->sid_info_len; i++) {
> +		pcie->sid_info[i].bdf = map[i].bdf;
> +		pcie->sid_info[i].smmu_sid = map[i].smmu_sid;
> +		pcie->sid_info[i].pcie_sid =
> +				pcie->sid_info[i].smmu_sid - smmu_sid_base;
> +	}
> +
> +	kfree(map);
> +
> +	return 0;
> +}
> +
> +static int qcom_pcie_config_sid_sm8250(struct qcom_pcie *pcie)
> +{
> +	void __iomem *bdf_to_sid_base = pcie->parf +
> +		PCIE20_PARF_BDF_TO_SID_TABLE_N;
> +	u8 qcom_pcie_crc8_table[CRC8_TABLE_SIZE];
> +	int ret, i;
> +
> +	ret = qcom_pcie_get_iommu_map(pcie);
> +	if (ret)
> +		return ret;
> +
> +	if (!pcie->sid_info)
> +		return 0;
> +
> +	crc8_populate_msb(qcom_pcie_crc8_table, QCOM_PCIE_CRC8_POLYNOMIAL);
> +
> +	/* Registers need to be zero out first */
> +	memset_io(bdf_to_sid_base, 0, CRC8_TABLE_SIZE * sizeof(u32));
> +
> +	/* Initial setup for boot */

Could you elaborate more what the code below is trying to achieve. Is
that connected to bootloaders?

> +	for (i = 0; i < pcie->sid_info_len; i++) {
> +		struct qcom_pcie_sid_info_t *sid_info = &pcie->sid_info[i];
> +		u16 bdf_be = cpu_to_be16(sid_info->bdf);
> +		u32 val;
> +		u8 hash;
> +
> +		hash = crc8(qcom_pcie_crc8_table, (u8 *)&bdf_be, sizeof(bdf_be),
> +			0);
> +
> +		val = readl(bdf_to_sid_base + hash * sizeof(u32));
> +
> +		/* If there is a collision, look for next available entry */
> +		while (val) {
> +			u8 current_hash = hash++;
> +			u8 next_mask = 0xff;
> +
> +			/* If NEXT field is NULL then update it with next hash */
> +			if (!(val & next_mask)) {
> +				int j;
> +
> +				val |= (u32)hash;
> +				writel(val, bdf_to_sid_base +
> +					current_hash * sizeof(u32));
> +
> +				/* Look for sid_info of current hash and update it */
> +				for (j = 0; j < pcie->sid_info_len; j++) {
> +					if (pcie->sid_info[j].hash !=
> +						current_hash)
> +						continue;
> +
> +					pcie->sid_info[j].value = val;
> +					break;
> +				}
> +			}
> +
> +			val = readl(bdf_to_sid_base + hash * sizeof(u32));
> +		}
> +
> +		val = sid_info->bdf << 16 | sid_info->pcie_sid << 8 | 0;
> +		writel(val, bdf_to_sid_base + hash * sizeof(u32));
> +
> +		sid_info->hash = hash;
> +		sid_info->value = val;
> +	}
> +
> +	return 0;
> +}
> +
>  static int qcom_pcie_host_init(struct pcie_port *pp)
>  {
>  	struct dw_pcie *pci = to_dw_pcie_from_pp(pp);
> @@ -1290,6 +1421,12 @@ static int qcom_pcie_host_init(struct pcie_port *pp)
>  	if (ret)
>  		goto err;
>  
> +	if (pcie->ops->config_sid) {
> +		ret = pcie->ops->config_sid(pcie);
> +		if (ret)
> +			goto err;
> +	}
> +
>  	return 0;
>  err:
>  	qcom_ep_reset_assert(pcie);
> @@ -1367,6 +1504,7 @@ static const struct qcom_pcie_ops ops_sm8250 = {
>  	.ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
>  	.post_init = qcom_pcie_post_init_2_7_0,
>  	.post_deinit = qcom_pcie_post_deinit_2_7_0,
> +	.config_sid = qcom_pcie_config_sid_sm8250,
>  };
>  
>  static const struct dw_pcie_ops dw_pcie_ops = {
> 

-- 
regards,
Stan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 4/5] PCI: qcom: Add SM8250 SoC support
  2020-09-30 15:09 ` [PATCH v2 4/5] PCI: qcom: Add SM8250 SoC support Manivannan Sadhasivam
@ 2020-09-30 21:56   ` Stanimir Varbanov
  2020-10-01  5:34     ` Manivannan Sadhasivam
  0 siblings, 1 reply; 14+ messages in thread
From: Stanimir Varbanov @ 2020-09-30 21:56 UTC (permalink / raw)
  To: Manivannan Sadhasivam, agross, bjorn.andersson, kishon, vkoul, robh
  Cc: bhelgaas, lorenzo.pieralisi, linux-arm-msm, linux-pci,
	linux-kernel, mgautam, devicetree

Hi Mani,

On 9/30/20 6:09 PM, Manivannan Sadhasivam wrote:
> The PCIe IP on SM8250 SoC is similar to the one used on SDM845. Hence
> the support is added reusing the members of ops_2_7_0. The key
> difference between ops_2_7_0 and ops_sm8250 is the config_sid callback,
> which will be added in successive commit.
> 
> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> ---
>  drivers/pci/controller/dwc/pcie-qcom.c | 11 +++++++++++
>  1 file changed, 11 insertions(+)
> 
> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
> index 3aac77a295ba..44db91861b47 100644
> --- a/drivers/pci/controller/dwc/pcie-qcom.c
> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
> @@ -1359,6 +1359,16 @@ static const struct qcom_pcie_ops ops_2_7_0 = {
>  	.post_deinit = qcom_pcie_post_deinit_2_7_0,
>  };
>  
> +/* Qcom IP rev.: 1.9.0 */
> +static const struct qcom_pcie_ops ops_sm8250 = {

This breaks the policy compatible -> ops_X_Y_Z. Could you introduce new
method config_sid and check into for compatible qcom,pcie-sm8250 string
there?

> +	.get_resources = qcom_pcie_get_resources_2_7_0,
> +	.init = qcom_pcie_init_2_7_0,
> +	.deinit = qcom_pcie_deinit_2_7_0,
> +	.ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
> +	.post_init = qcom_pcie_post_init_2_7_0,
> +	.post_deinit = qcom_pcie_post_deinit_2_7_0,
> +};
> +
>  static const struct dw_pcie_ops dw_pcie_ops = {
>  	.link_up = qcom_pcie_link_up,
>  };
> @@ -1476,6 +1486,7 @@ static const struct of_device_id qcom_pcie_match[] = {
>  	{ .compatible = "qcom,pcie-ipq4019", .data = &ops_2_4_0 },
>  	{ .compatible = "qcom,pcie-qcs404", .data = &ops_2_4_0 },
>  	{ .compatible = "qcom,pcie-sdm845", .data = &ops_2_7_0 },
> +	{ .compatible = "qcom,pcie-sm8250", .data = &ops_sm8250 },
>  	{ }
>  };
>  
> 

-- 
regards,
Stan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 4/5] PCI: qcom: Add SM8250 SoC support
  2020-09-30 21:56   ` Stanimir Varbanov
@ 2020-10-01  5:34     ` Manivannan Sadhasivam
  0 siblings, 0 replies; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-10-01  5:34 UTC (permalink / raw)
  To: Stanimir Varbanov
  Cc: agross, bjorn.andersson, kishon, vkoul, robh, bhelgaas,
	lorenzo.pieralisi, linux-arm-msm, linux-pci, linux-kernel,
	mgautam, devicetree

Hi Stan,

On Thu, Oct 01, 2020 at 12:56:28AM +0300, Stanimir Varbanov wrote:
> Hi Mani,
> 
> On 9/30/20 6:09 PM, Manivannan Sadhasivam wrote:
> > The PCIe IP on SM8250 SoC is similar to the one used on SDM845. Hence
> > the support is added reusing the members of ops_2_7_0. The key
> > difference between ops_2_7_0 and ops_sm8250 is the config_sid callback,
> > which will be added in successive commit.
> > 
> > Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> > ---
> >  drivers/pci/controller/dwc/pcie-qcom.c | 11 +++++++++++
> >  1 file changed, 11 insertions(+)
> > 
> > diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
> > index 3aac77a295ba..44db91861b47 100644
> > --- a/drivers/pci/controller/dwc/pcie-qcom.c
> > +++ b/drivers/pci/controller/dwc/pcie-qcom.c
> > @@ -1359,6 +1359,16 @@ static const struct qcom_pcie_ops ops_2_7_0 = {
> >  	.post_deinit = qcom_pcie_post_deinit_2_7_0,
> >  };
> >  
> > +/* Qcom IP rev.: 1.9.0 */
> > +static const struct qcom_pcie_ops ops_sm8250 = {
> 
> This breaks the policy compatible -> ops_X_Y_Z. Could you introduce new
> method config_sid and check into for compatible qcom,pcie-sm8250 string
> there?
> 

I thought about it but during previous submission review Bjorn mentioned that
this config_sid got introduced in SM8150 and there might be chances that future
SoCs could also use it. That's why I was inclined to introduce a new ops instead
of checking for the compatible.

And the reason to use "sm8250" instead of IP revision is that I can't find the
Synopsys IP revision for this. But if you strongly prefer IP revision then I can
just use "ops_1_9_0"!

Thanks,
Mani

> > +	.get_resources = qcom_pcie_get_resources_2_7_0,
> > +	.init = qcom_pcie_init_2_7_0,
> > +	.deinit = qcom_pcie_deinit_2_7_0,
> > +	.ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
> > +	.post_init = qcom_pcie_post_init_2_7_0,
> > +	.post_deinit = qcom_pcie_post_deinit_2_7_0,
> > +};
> > +
> >  static const struct dw_pcie_ops dw_pcie_ops = {
> >  	.link_up = qcom_pcie_link_up,
> >  };
> > @@ -1476,6 +1486,7 @@ static const struct of_device_id qcom_pcie_match[] = {
> >  	{ .compatible = "qcom,pcie-ipq4019", .data = &ops_2_4_0 },
> >  	{ .compatible = "qcom,pcie-qcs404", .data = &ops_2_4_0 },
> >  	{ .compatible = "qcom,pcie-sdm845", .data = &ops_2_7_0 },
> > +	{ .compatible = "qcom,pcie-sm8250", .data = &ops_sm8250 },
> >  	{ }
> >  };
> >  
> > 
> 
> -- 
> regards,
> Stan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250
  2020-09-30 21:46   ` Stanimir Varbanov
@ 2020-10-01  5:57     ` Manivannan Sadhasivam
  2020-10-01 10:57       ` Stanimir Varbanov
  0 siblings, 1 reply; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-10-01  5:57 UTC (permalink / raw)
  To: Stanimir Varbanov
  Cc: agross, bjorn.andersson, kishon, vkoul, robh, bhelgaas,
	lorenzo.pieralisi, linux-arm-msm, linux-pci, linux-kernel,
	mgautam, devicetree

Hi Stan,

On Thu, Oct 01, 2020 at 12:46:46AM +0300, Stanimir Varbanov wrote:
> Hi Mani,
> 
> On 9/30/20 6:09 PM, Manivannan Sadhasivam wrote:
> > For SM8250, we need to write the BDF to SID mapping in PCIe controller
> > register space for proper working. This is accomplished by extracting
> > the BDF and SID values from "iommu-map" property in DT and writing those
> > in the register address calculated from the hash value of BDF. In case
> > of collisions, the index of the next entry will also be written.
> 
> This describes what the patch is doing. But why? Is that done in the
> other DWC low-level drivers or this is qcom specialty?
> 

AFAIK, only some NXP SoCs deal with similar kind of mapping but right now
this is a Qcom only stuff.

> > 
> > For the sake of it, let's introduce a "config_sid" callback and do it
> > conditionally for SM8250.
> > 
> > Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> > ---
> >  drivers/pci/controller/dwc/Kconfig     |   1 +
> >  drivers/pci/controller/dwc/pcie-qcom.c | 138 +++++++++++++++++++++++++
> >  2 files changed, 139 insertions(+)
> > 

[...]

> > +
> > +/* sid info structure */
> > +struct qcom_pcie_sid_info_t {
> 
> why _t postfix? Maybe qcom_pcie_sid ?
> 

Just to differentiate the struct name and its variable. But I can
remove the _t suffix.

> SID - Stream ID ?
> 

Yes! Will expand in commit message also.

> > +	u16 bdf;
> > +	u8 pcie_sid;
> > +	u8 hash;
> > +	u32 smmu_sid;
> > +	u32 value;
> >  };
> >  
> >  struct qcom_pcie {
> > @@ -193,6 +208,8 @@ struct qcom_pcie {
> >  	struct phy *phy;
> >  	struct gpio_desc *reset;
> >  	const struct qcom_pcie_ops *ops;
> > +	struct qcom_pcie_sid_info_t *sid_info;
> > +	u32 sid_info_len;
> >  	int gen;
> >  };
> >  
> > @@ -1257,6 +1274,120 @@ static int qcom_pcie_link_up(struct dw_pcie *pci)
> >  	return !!(val & PCI_EXP_LNKSTA_DLLLA);
> >  }
> >  
> > +static int qcom_pcie_get_iommu_map(struct qcom_pcie *pcie)
> > +{
> > +	/* iommu map structure */
> > +	struct {
> > +		u32 bdf;
> > +		u32 phandle;
> > +		u32 smmu_sid;
> > +		u32 smmu_sid_len;
> > +	} *map;
> > +	struct device *dev = pcie->pci->dev;
> > +	int i, size = 0;
> > +	u32 smmu_sid_base;
> > +
> > +	of_get_property(dev->of_node, "iommu-map", &size);
> > +	if (!size)
> > +		return 0;
> > +
> > +	map = kzalloc(size, GFP_KERNEL);
> > +	if (!map)
> > +		return -ENOMEM;
> > +
> > +	of_property_read_u32_array(dev->of_node,
> > +		"iommu-map", (u32 *)map, size / sizeof(u32));
> 
> iommu-map is a standard DT property why we have to parse it manually?
> 

So right now we don't have a way to pass this information from DT. And there
is no IOMMU API to parse the fields also. We need to extract this information
to program the hash tables (BDF, SID) as the mapping between BDF and SID is not
1:1 in SM8250.

Perhaps I can add this information in commit message.

> > +
> > +	pcie->sid_info_len = size / (sizeof(*map));
> > +	pcie->sid_info = devm_kcalloc(dev, pcie->sid_info_len,
> > +				sizeof(*pcie->sid_info), GFP_KERNEL);
> > +	if (!pcie->sid_info) {
> > +		kfree(map);
> > +		return -ENOMEM;
> > +	}
> > +
> > +	/* Extract the SMMU SID base from the first entry of iommu-map */
> > +	smmu_sid_base = map[0].smmu_sid;
> > +	for (i = 0; i < pcie->sid_info_len; i++) {
> > +		pcie->sid_info[i].bdf = map[i].bdf;
> > +		pcie->sid_info[i].smmu_sid = map[i].smmu_sid;
> > +		pcie->sid_info[i].pcie_sid =
> > +				pcie->sid_info[i].smmu_sid - smmu_sid_base;
> > +	}
> > +
> > +	kfree(map);
> > +
> > +	return 0;
> > +}
> > +
> > +static int qcom_pcie_config_sid_sm8250(struct qcom_pcie *pcie)
> > +{
> > +	void __iomem *bdf_to_sid_base = pcie->parf +
> > +		PCIE20_PARF_BDF_TO_SID_TABLE_N;
> > +	u8 qcom_pcie_crc8_table[CRC8_TABLE_SIZE];
> > +	int ret, i;
> > +
> > +	ret = qcom_pcie_get_iommu_map(pcie);
> > +	if (ret)
> > +		return ret;
> > +
> > +	if (!pcie->sid_info)
> > +		return 0;
> > +
> > +	crc8_populate_msb(qcom_pcie_crc8_table, QCOM_PCIE_CRC8_POLYNOMIAL);
> > +
> > +	/* Registers need to be zero out first */
> > +	memset_io(bdf_to_sid_base, 0, CRC8_TABLE_SIZE * sizeof(u32));
> > +
> > +	/* Initial setup for boot */
> 
> Could you elaborate more what the code below is trying to achieve. Is
> that connected to bootloaders?
> 

No. This is trying to program the hash tables for initial boot but I think this
doesn't make sense here as it will be done all the time. I'll just remove this
comment.

Thanks,
Mani

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250
  2020-10-01  5:57     ` Manivannan Sadhasivam
@ 2020-10-01 10:57       ` Stanimir Varbanov
  2020-10-01 11:46         ` Manivannan Sadhasivam
  2020-10-06  8:15         ` Marc Gonzalez
  0 siblings, 2 replies; 14+ messages in thread
From: Stanimir Varbanov @ 2020-10-01 10:57 UTC (permalink / raw)
  To: Manivannan Sadhasivam
  Cc: agross, bjorn.andersson, kishon, vkoul, robh, bhelgaas,
	lorenzo.pieralisi, linux-arm-msm, linux-pci, linux-kernel,
	mgautam, devicetree, Marc Gonzalez

Hi Mani,

On 10/1/20 8:57 AM, Manivannan Sadhasivam wrote:
> Hi Stan,
> 
> On Thu, Oct 01, 2020 at 12:46:46AM +0300, Stanimir Varbanov wrote:
>> Hi Mani,
>>
>> On 9/30/20 6:09 PM, Manivannan Sadhasivam wrote:
>>> For SM8250, we need to write the BDF to SID mapping in PCIe controller
>>> register space for proper working. This is accomplished by extracting
>>> the BDF and SID values from "iommu-map" property in DT and writing those
>>> in the register address calculated from the hash value of BDF. In case
>>> of collisions, the index of the next entry will also be written.
>>
>> This describes what the patch is doing. But why? Is that done in the
>> other DWC low-level drivers or this is qcom specialty?
>>
> 
> AFAIK, only some NXP SoCs deal with similar kind of mapping but right now
> this is a Qcom only stuff.
> 
>>>
>>> For the sake of it, let's introduce a "config_sid" callback and do it
>>> conditionally for SM8250.
>>>
>>> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
>>> ---
>>>  drivers/pci/controller/dwc/Kconfig     |   1 +
>>>  drivers/pci/controller/dwc/pcie-qcom.c | 138 +++++++++++++++++++++++++
>>>  2 files changed, 139 insertions(+)

<snip>

>>>  
>>> +static int qcom_pcie_get_iommu_map(struct qcom_pcie *pcie)
>>> +{
>>> +	/* iommu map structure */
>>> +	struct {
>>> +		u32 bdf;
>>> +		u32 phandle;
>>> +		u32 smmu_sid;
>>> +		u32 smmu_sid_len;
>>> +	} *map;
>>> +	struct device *dev = pcie->pci->dev;
>>> +	int i, size = 0;
>>> +	u32 smmu_sid_base;
>>> +
>>> +	of_get_property(dev->of_node, "iommu-map", &size);
>>> +	if (!size)
>>> +		return 0;
>>> +
>>> +	map = kzalloc(size, GFP_KERNEL);
>>> +	if (!map)
>>> +		return -ENOMEM;
>>> +
>>> +	of_property_read_u32_array(dev->of_node,
>>> +		"iommu-map", (u32 *)map, size / sizeof(u32));
>>
>> iommu-map is a standard DT property why we have to parse it manually?
>>
> 
> So right now we don't have a way to pass this information from DT. And there
> is no IOMMU API to parse the fields also. We need to extract this information
> to program the hash tables (BDF, SID) as the mapping between BDF and SID is not
> 1:1 in SM8250.

We used iommu-map for msm8998 see this commit:

b84dfd175c09888751f501e471fdca346f582e06
("arm64: dts: qcom: msm8998: Add PCIe PHY and RC nodes")

I also Cc-ed Marc if he knows something more.

> 
> Perhaps I can add this information in commit message.


-- 
regards,
Stan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250
  2020-10-01 10:57       ` Stanimir Varbanov
@ 2020-10-01 11:46         ` Manivannan Sadhasivam
  2020-10-06  8:15         ` Marc Gonzalez
  1 sibling, 0 replies; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-10-01 11:46 UTC (permalink / raw)
  To: Stanimir Varbanov
  Cc: agross, bjorn.andersson, kishon, vkoul, robh, bhelgaas,
	lorenzo.pieralisi, linux-arm-msm, linux-pci, linux-kernel,
	mgautam, devicetree, Marc Gonzalez

Hi Stan,

On Thu, Oct 01, 2020 at 01:57:59PM +0300, Stanimir Varbanov wrote:
> Hi Mani,
> 
> On 10/1/20 8:57 AM, Manivannan Sadhasivam wrote:
> > Hi Stan,
> > 
> > On Thu, Oct 01, 2020 at 12:46:46AM +0300, Stanimir Varbanov wrote:
> >> Hi Mani,
> >>
> >> On 9/30/20 6:09 PM, Manivannan Sadhasivam wrote:
> >>> For SM8250, we need to write the BDF to SID mapping in PCIe controller
> >>> register space for proper working. This is accomplished by extracting
> >>> the BDF and SID values from "iommu-map" property in DT and writing those
> >>> in the register address calculated from the hash value of BDF. In case
> >>> of collisions, the index of the next entry will also be written.
> >>
> >> This describes what the patch is doing. But why? Is that done in the
> >> other DWC low-level drivers or this is qcom specialty?
> >>
> > 
> > AFAIK, only some NXP SoCs deal with similar kind of mapping but right now
> > this is a Qcom only stuff.
> > 
> >>>
> >>> For the sake of it, let's introduce a "config_sid" callback and do it
> >>> conditionally for SM8250.
> >>>
> >>> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> >>> ---
> >>>  drivers/pci/controller/dwc/Kconfig     |   1 +
> >>>  drivers/pci/controller/dwc/pcie-qcom.c | 138 +++++++++++++++++++++++++
> >>>  2 files changed, 139 insertions(+)
> 
> <snip>
> 
> >>>  
> >>> +static int qcom_pcie_get_iommu_map(struct qcom_pcie *pcie)
> >>> +{
> >>> +	/* iommu map structure */
> >>> +	struct {
> >>> +		u32 bdf;
> >>> +		u32 phandle;
> >>> +		u32 smmu_sid;
> >>> +		u32 smmu_sid_len;
> >>> +	} *map;
> >>> +	struct device *dev = pcie->pci->dev;
> >>> +	int i, size = 0;
> >>> +	u32 smmu_sid_base;
> >>> +
> >>> +	of_get_property(dev->of_node, "iommu-map", &size);
> >>> +	if (!size)
> >>> +		return 0;
> >>> +
> >>> +	map = kzalloc(size, GFP_KERNEL);
> >>> +	if (!map)
> >>> +		return -ENOMEM;
> >>> +
> >>> +	of_property_read_u32_array(dev->of_node,
> >>> +		"iommu-map", (u32 *)map, size / sizeof(u32));
> >>
> >> iommu-map is a standard DT property why we have to parse it manually?
> >>
> > 
> > So right now we don't have a way to pass this information from DT. And there
> > is no IOMMU API to parse the fields also. We need to extract this information
> > to program the hash tables (BDF, SID) as the mapping between BDF and SID is not
> > 1:1 in SM8250.
> 
> We used iommu-map for msm8998 see this commit:
> 
> b84dfd175c09888751f501e471fdca346f582e06
> ("arm64: dts: qcom: msm8998: Add PCIe PHY and RC nodes")
> 

The iommu-map property will be present in most of the SoCs as needed for the
SMMU translation of DMA transfers. But in the case of SM8250, we need to map the
BDF (Bus Device Function) identifier (aka RID) to the SMMU Stream ID (SID)
in the PCIe controller, and we get that information from existing iommu-map
property.

Thanks,
Mani

> I also Cc-ed Marc if he knows something more.
> 
> > 
> > Perhaps I can add this information in commit message.
> 
> 
> -- 
> regards,
> Stan

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250
  2020-10-01 10:57       ` Stanimir Varbanov
  2020-10-01 11:46         ` Manivannan Sadhasivam
@ 2020-10-06  8:15         ` Marc Gonzalez
  2020-10-06 10:12           ` Manivannan Sadhasivam
  1 sibling, 1 reply; 14+ messages in thread
From: Marc Gonzalez @ 2020-10-06  8:15 UTC (permalink / raw)
  To: Stanimir Varbanov, Manivannan Sadhasivam, Robin Murphy
  Cc: agross, bjorn.andersson, kishon, vkoul, robh, bhelgaas,
	lorenzo.pieralisi, linux-arm-msm, linux-pci, mgautam

On 01/10/2020 12:57, Stanimir Varbanov wrote:

> On 10/1/20 8:57 AM, Manivannan Sadhasivam wrote:
>
>> On Thu, Oct 01, 2020 at 12:46:46AM +0300, Stanimir Varbanov wrote:
>>
>>> On 9/30/20 6:09 PM, Manivannan Sadhasivam wrote:
>>>
>>>> For SM8250, we need to write the BDF to SID mapping in PCIe controller
>>>> register space for proper working. This is accomplished by extracting
>>>> the BDF and SID values from "iommu-map" property in DT and writing those
>>>> in the register address calculated from the hash value of BDF. In case
>>>> of collisions, the index of the next entry will also be written.
>>>
>>> This describes what the patch is doing. But why? Is that done in the
>>> other DWC low-level drivers or this is qcom specialty?
>>
>> AFAIK, only some NXP SoCs deal with similar kind of mapping but right now
>> this is a Qcom only stuff.
>>
>>>> For the sake of it, let's introduce a "config_sid" callback and do it
>>>> conditionally for SM8250.
>>>>
>>>> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
>>>> ---
>>>>  drivers/pci/controller/dwc/Kconfig     |   1 +
>>>>  drivers/pci/controller/dwc/pcie-qcom.c | 138 +++++++++++++++++++++++++
>>>>  2 files changed, 139 insertions(+)
> 
> <snip>
> 
>>>>  
>>>> +static int qcom_pcie_get_iommu_map(struct qcom_pcie *pcie)
>>>> +{
>>>> +	/* iommu map structure */
>>>> +	struct {
>>>> +		u32 bdf;
>>>> +		u32 phandle;
>>>> +		u32 smmu_sid;
>>>> +		u32 smmu_sid_len;
>>>> +	} *map;
>>>> +	struct device *dev = pcie->pci->dev;
>>>> +	int i, size = 0;
>>>> +	u32 smmu_sid_base;
>>>> +
>>>> +	of_get_property(dev->of_node, "iommu-map", &size);
>>>> +	if (!size)
>>>> +		return 0;
>>>> +
>>>> +	map = kzalloc(size, GFP_KERNEL);
>>>> +	if (!map)
>>>> +		return -ENOMEM;
>>>> +
>>>> +	of_property_read_u32_array(dev->of_node,
>>>> +		"iommu-map", (u32 *)map, size / sizeof(u32));
>>>
>>> iommu-map is a standard DT property why we have to parse it manually?
>>>
>>
>> So right now we don't have a way to pass this information from DT. And there
>> is no IOMMU API to parse the fields also. We need to extract this information
>> to program the hash tables (BDF, SID) as the mapping between BDF and SID is not
>> 1:1 in SM8250.
> 
> We used iommu-map for msm8998 see this commit:
> 
> b84dfd175c09888751f501e471fdca346f582e06
> ("arm64: dts: qcom: msm8998: Add PCIe PHY and RC nodes")
> 
> I also Cc-ed Marc if he knows something more.

My memory is hazy.

I remember an odd quirk in the downstream kernel:

[v1,1/3] PCI: qcom: Setup PCIE20_PARF_BDF_TRANSLATE_N
http://patchwork.ozlabs.org/project/linux-pci/patch/958ae127-3aa2-6824-c875-e3012644ed3d@free.fr/

Manivannan, are you trying to deal with PCIE20_PARF_BDF_TRANSLATE_N
or some equivalent register?

+Robin, he's the one who helped me figure this stuff out (iommu-map).
It was in reply to patch 2:
http://patchwork.ozlabs.org/project/linux-pci/patch/82ab78ee-4a38-4eee-f064-272b6f964f17@free.fr/

In the end, I dropped patch 1 because... everything seemed to work
without it (?!) (Makes one wonder what it actually does. But qcom
refused to provide any register documentation, which is idiotic
because this is DW IP, and they are open-source friendly, IIUC.)

Regards.

^ permalink raw reply	[flat|nested] 14+ messages in thread

* Re: [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250
  2020-10-06  8:15         ` Marc Gonzalez
@ 2020-10-06 10:12           ` Manivannan Sadhasivam
  0 siblings, 0 replies; 14+ messages in thread
From: Manivannan Sadhasivam @ 2020-10-06 10:12 UTC (permalink / raw)
  To: Marc Gonzalez
  Cc: Stanimir Varbanov, Robin Murphy, agross, bjorn.andersson, kishon,
	vkoul, robh, bhelgaas, lorenzo.pieralisi, linux-arm-msm,
	linux-pci, mgautam

Hi,

On Tue, Oct 06, 2020 at 10:15:08AM +0200, Marc Gonzalez wrote:
> On 01/10/2020 12:57, Stanimir Varbanov wrote:
> 
> > On 10/1/20 8:57 AM, Manivannan Sadhasivam wrote:
> >
> >> On Thu, Oct 01, 2020 at 12:46:46AM +0300, Stanimir Varbanov wrote:
> >>

[...]

> >>> iommu-map is a standard DT property why we have to parse it manually?
> >>>
> >>
> >> So right now we don't have a way to pass this information from DT. And there
> >> is no IOMMU API to parse the fields also. We need to extract this information
> >> to program the hash tables (BDF, SID) as the mapping between BDF and SID is not
> >> 1:1 in SM8250.
> > 
> > We used iommu-map for msm8998 see this commit:
> > 
> > b84dfd175c09888751f501e471fdca346f582e06
> > ("arm64: dts: qcom: msm8998: Add PCIe PHY and RC nodes")
> > 
> > I also Cc-ed Marc if he knows something more.
> 
> My memory is hazy.
> 
> I remember an odd quirk in the downstream kernel:
> 
> [v1,1/3] PCI: qcom: Setup PCIE20_PARF_BDF_TRANSLATE_N
> http://patchwork.ozlabs.org/project/linux-pci/patch/958ae127-3aa2-6824-c875-e3012644ed3d@free.fr/
> 
> Manivannan, are you trying to deal with PCIE20_PARF_BDF_TRANSLATE_N
> or some equivalent register?
> 

I'm dealing with PCIE20_PARF_BDF_TO_SID_TABLE_N but I'm not sure how it
relates to PCIE20_PARF_BDF_TRANSLATE_N.

> +Robin, he's the one who helped me figure this stuff out (iommu-map).
> It was in reply to patch 2:
> http://patchwork.ozlabs.org/project/linux-pci/patch/82ab78ee-4a38-4eee-f064-272b6f964f17@free.fr/
> 
> In the end, I dropped patch 1 because... everything seemed to work
> without it (?!) (Makes one wonder what it actually does. But qcom
> refused to provide any register documentation, which is idiotic
> because this is DW IP, and they are open-source friendly, IIUC.)
> 

Without this patch, PCIe will work on SM8250 but during successive reboots
I started seeing issues which gets fixed by this patch.

Thanks,
Mani

> Regards.

^ permalink raw reply	[flat|nested] 14+ messages in thread

end of thread, other threads:[~2020-10-06 10:13 UTC | newest]

Thread overview: 14+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-09-30 15:09 [PATCH v2 0/5] Add PCIe support for SM8250 SoC Manivannan Sadhasivam
2020-09-30 15:09 ` [PATCH v2 1/5] dt-bindings: phy: qcom,qmp: Add SM8250 PCIe PHY bindings Manivannan Sadhasivam
2020-09-30 15:09 ` [PATCH v2 2/5] phy: qcom-qmp: Add SM8250 PCIe QMP PHYs Manivannan Sadhasivam
2020-09-30 15:09 ` [PATCH v2 3/5] dt-bindings: pci: qcom: Document PCIe bindings for SM8250 SoC Manivannan Sadhasivam
2020-09-30 15:09 ` [PATCH v2 4/5] PCI: qcom: Add SM8250 SoC support Manivannan Sadhasivam
2020-09-30 21:56   ` Stanimir Varbanov
2020-10-01  5:34     ` Manivannan Sadhasivam
2020-09-30 15:09 ` [PATCH v2 5/5] PCI: qcom: Add support for configuring BDF to SID mapping for SM8250 Manivannan Sadhasivam
2020-09-30 21:46   ` Stanimir Varbanov
2020-10-01  5:57     ` Manivannan Sadhasivam
2020-10-01 10:57       ` Stanimir Varbanov
2020-10-01 11:46         ` Manivannan Sadhasivam
2020-10-06  8:15         ` Marc Gonzalez
2020-10-06 10:12           ` Manivannan Sadhasivam

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).