drm/radeon: implement pcie gen2/3 support for SI

If both the motherboard and GPU support pcie gen2 or 3,
enable it.  PCIE gen2 and 3 offer more bandwidth than
pcie gen1.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

2 files changed, 212 insertions, 0 deletions

diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 882509ab166..b073b2caf54 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -66,6 +66,7 @@ MODULE_FIRMWARE("radeon/HAINAN_ce.bin");
 MODULE_FIRMWARE("radeon/HAINAN_mc.bin");
 MODULE_FIRMWARE("radeon/HAINAN_rlc.bin");
 
+static void si_pcie_gen3_enable(struct radeon_device *rdev);
 extern int r600_ih_ring_alloc(struct radeon_device *rdev);
 extern void r600_ih_ring_fini(struct radeon_device *rdev);
 extern void evergreen_fix_pci_max_read_req_size(struct radeon_device *rdev);
@@ -5316,6 +5317,9 @@ static int si_startup(struct radeon_device *rdev)
 	struct radeon_ring *ring;
 	int r;
 
+	/* enable pcie gen2/3 link */
+	si_pcie_gen3_enable(rdev);
+
 	if (!rdev->me_fw || !rdev->pfp_fw || !rdev->ce_fw ||
 	    !rdev->rlc_fw || !rdev->mc_fw) {
 		r = si_init_microcode(rdev);
@@ -5781,3 +5785,161 @@ int si_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk)
 
 	return 0;
 }
+
+static void si_pcie_gen3_enable(struct radeon_device *rdev)
+{
+	struct pci_dev *root = rdev->pdev->bus->self;
+	int bridge_pos, gpu_pos;
+	u32 speed_cntl, mask, current_data_rate;
+	int ret, i;
+	u16 tmp16;
+
+	if (radeon_pcie_gen2 == 0)
+		return;
+
+	if (rdev->flags & RADEON_IS_IGP)
+		return;
+
+	if (!(rdev->flags & RADEON_IS_PCIE))
+		return;
+
+	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
+	if (ret != 0)
+		return;
+
+	if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80)))
+		return;
+
+	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
+	current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >>
+		LC_CURRENT_DATA_RATE_SHIFT;
+	if (mask & DRM_PCIE_SPEED_80) {
+		if (current_data_rate == 2) {
+			DRM_INFO("PCIE gen 3 link speeds already enabled\n");
+			return;
+		}
+		DRM_INFO("enabling PCIE gen 3 link speeds, disable with radeon.pcie_gen2=0\n");
+	} else if (mask & DRM_PCIE_SPEED_50) {
+		if (current_data_rate == 1) {
+			DRM_INFO("PCIE gen 2 link speeds already enabled\n");
+			return;
+		}
+		DRM_INFO("enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0\n");
+	}
+
+	bridge_pos = pci_pcie_cap(root);
+	if (!bridge_pos)
+		return;
+
+	gpu_pos = pci_pcie_cap(rdev->pdev);
+	if (!gpu_pos)
+		return;
+
+	if (mask & DRM_PCIE_SPEED_80) {
+		/* re-try equalization if gen3 is not already enabled */
+		if (current_data_rate != 2) {
+			u16 bridge_cfg, gpu_cfg;
+			u16 bridge_cfg2, gpu_cfg2;
+			u32 max_lw, current_lw, tmp;
+
+			pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
+			pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+
+			tmp16 = bridge_cfg | PCI_EXP_LNKCTL_HAWD;
+			pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+
+			tmp16 = gpu_cfg | PCI_EXP_LNKCTL_HAWD;
+			pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+
+			tmp = RREG32_PCIE(PCIE_LC_STATUS1);
+			max_lw = (tmp & LC_DETECTED_LINK_WIDTH_MASK) >> LC_DETECTED_LINK_WIDTH_SHIFT;
+			current_lw = (tmp & LC_OPERATING_LINK_WIDTH_MASK) >> LC_OPERATING_LINK_WIDTH_SHIFT;
+
+			if (current_lw < max_lw) {
+				tmp = RREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL);
+				if (tmp & LC_RENEGOTIATION_SUPPORT) {
+					tmp &= ~(LC_LINK_WIDTH_MASK | LC_UPCONFIGURE_DIS);
+					tmp |= (max_lw << LC_LINK_WIDTH_SHIFT);
+					tmp |= LC_UPCONFIGURE_SUPPORT | LC_RENEGOTIATE_EN | LC_RECONFIG_NOW;
+					WREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL, tmp);
+				}
+			}
+
+			for (i = 0; i < 10; i++) {
+				/* check status */
+				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_DEVSTA, &tmp16);
+				if (tmp16 & PCI_EXP_DEVSTA_TRPND)
+					break;
+
+				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
+				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+
+				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &bridge_cfg2);
+				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &gpu_cfg2);
+
+				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
+				tmp |= LC_SET_QUIESCE;
+				WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp);
+
+				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
+				tmp |= LC_REDO_EQ;
+				WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp);
+
+				mdelay(100);
+
+				/* linkctl */
+				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16);
+				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
+				tmp16 |= (bridge_cfg & PCI_EXP_LNKCTL_HAWD);
+				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+
+				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, &tmp16);
+				tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
+				tmp16 |= (gpu_cfg & PCI_EXP_LNKCTL_HAWD);
+				pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+
+				/* linkctl2 */
+				pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &tmp16);
+				tmp16 &= ~((1 << 4) | (7 << 9));
+				tmp16 |= (bridge_cfg2 & ((1 << 4) | (7 << 9)));
+				pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, tmp16);
+
+				pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
+				tmp16 &= ~((1 << 4) | (7 << 9));
+				tmp16 |= (gpu_cfg2 & ((1 << 4) | (7 << 9)));
+				pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+
+				tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
+				tmp &= ~LC_SET_QUIESCE;
+				WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp);
+			}
+		}
+	}
+
+	/* set the link speed */
+	speed_cntl |= LC_FORCE_EN_SW_SPEED_CHANGE | LC_FORCE_DIS_HW_SPEED_CHANGE;
+	speed_cntl &= ~LC_FORCE_DIS_SW_SPEED_CHANGE;
+	WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl);
+
+	pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
+	tmp16 &= ~0xf;
+	if (mask & DRM_PCIE_SPEED_80)
+		tmp16 |= 3; /* gen3 */
+	else if (mask & DRM_PCIE_SPEED_50)
+		tmp16 |= 2; /* gen2 */
+	else
+		tmp16 |= 1; /* gen1 */
+	pci_write_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+
+	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
+	speed_cntl |= LC_INITIATE_LINK_SPEED_CHANGE;
+	WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+		speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
+		if ((speed_cntl & LC_INITIATE_LINK_SPEED_CHANGE) == 0)
+			break;
+		udelay(1);
+	}
+}
+
diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h
index 8f2d7d4f9b2..6d4bdbc797a 100644
--- a/drivers/gpu/drm/radeon/sid.h
+++ b/drivers/gpu/drm/radeon/sid.h
@@ -829,6 +829,56 @@
 #       define THREAD_TRACE_FLUSH                       (54 << 0)
 #       define THREAD_TRACE_FINISH                      (55 << 0)
 
+/* PCIE registers idx/data 0x30/0x34 */
+#define PCIE_LC_STATUS1                                   0x28 /* PCIE */
+#       define LC_OPERATING_LINK_WIDTH_MASK               (0x7 << 2)
+#       define LC_OPERATING_LINK_WIDTH_SHIFT              2
+#       define LC_DETECTED_LINK_WIDTH_MASK                (0x7 << 5)
+#       define LC_DETECTED_LINK_WIDTH_SHIFT               5
+
+/* PCIE PORT registers idx/data 0x38/0x3c */
+#define PCIE_LC_LINK_WIDTH_CNTL                           0xa2 /* PCIE_P */
+#       define LC_LINK_WIDTH_SHIFT                        0
+#       define LC_LINK_WIDTH_MASK                         0x7
+#       define LC_LINK_WIDTH_X0                           0
+#       define LC_LINK_WIDTH_X1                           1
+#       define LC_LINK_WIDTH_X2                           2
+#       define LC_LINK_WIDTH_X4                           3
+#       define LC_LINK_WIDTH_X8                           4
+#       define LC_LINK_WIDTH_X16                          6
+#       define LC_LINK_WIDTH_RD_SHIFT                     4
+#       define LC_LINK_WIDTH_RD_MASK                      0x70
+#       define LC_RECONFIG_ARC_MISSING_ESCAPE             (1 << 7)
+#       define LC_RECONFIG_NOW                            (1 << 8)
+#       define LC_RENEGOTIATION_SUPPORT                   (1 << 9)
+#       define LC_RENEGOTIATE_EN                          (1 << 10)
+#       define LC_SHORT_RECONFIG_EN                       (1 << 11)
+#       define LC_UPCONFIGURE_SUPPORT                     (1 << 12)
+#       define LC_UPCONFIGURE_DIS                         (1 << 13)
+#define PCIE_LC_SPEED_CNTL                                0xa4 /* PCIE_P */
+#       define LC_GEN2_EN_STRAP                           (1 << 0)
+#       define LC_GEN3_EN_STRAP                           (1 << 1)
+#       define LC_TARGET_LINK_SPEED_OVERRIDE_EN           (1 << 2)
+#       define LC_TARGET_LINK_SPEED_OVERRIDE_MASK         (0x3 << 3)
+#       define LC_TARGET_LINK_SPEED_OVERRIDE_SHIFT        3
+#       define LC_FORCE_EN_SW_SPEED_CHANGE                (1 << 5)
+#       define LC_FORCE_DIS_SW_SPEED_CHANGE               (1 << 6)
+#       define LC_FORCE_EN_HW_SPEED_CHANGE                (1 << 7)
+#       define LC_FORCE_DIS_HW_SPEED_CHANGE               (1 << 8)
+#       define LC_INITIATE_LINK_SPEED_CHANGE              (1 << 9)
+#       define LC_SPEED_CHANGE_ATTEMPTS_ALLOWED_MASK      (0x3 << 10)
+#       define LC_SPEED_CHANGE_ATTEMPTS_ALLOWED_SHIFT     10
+#       define LC_CURRENT_DATA_RATE_MASK                  (0x3 << 13) /* 0/1/2 = gen1/2/3 */
+#       define LC_CURRENT_DATA_RATE_SHIFT                 13
+#       define LC_CLR_FAILED_SPD_CHANGE_CNT               (1 << 16)
+#       define LC_OTHER_SIDE_EVER_SENT_GEN2               (1 << 18)
+#       define LC_OTHER_SIDE_SUPPORTS_GEN2                (1 << 19)
+#       define LC_OTHER_SIDE_EVER_SENT_GEN3               (1 << 20)
+#       define LC_OTHER_SIDE_SUPPORTS_GEN3                (1 << 21)
+#define PCIE_LC_CNTL4                                     0xb6 /* PCIE_P */
+#       define LC_REDO_EQ                                 (1 << 5)
+#       define LC_SET_QUIESCE                             (1 << 13)
+
 /*
  * UVD
  */