Merge branch 'drm-fixes-3.11' of git://people.freedesktop.org/~agd5f/linux

more DPM fixes for radeon. * 'drm-fixes-3.11' of git://people.freedesktop.org/~agd5f/linux: drm/radeon/dpm: add debugfs support for RS780/RS880 (v3) drm/radeon/dpm/atom: fix broken gcc harder drm/radeon/dpm/atom: restructure logic to work around a compiler bug drm/radeon/dpm: fix atom vram table parsing drm/radeon: fix an endian bug in atom table parsing drm/radeon: add a module parameter to disable aspm
2013-07-18 10:19:46 +10:00 · 2013-07-18 10:19:46 +10:00 · fb328d7365
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@ -5515,6 +5515,9 @@ void evergreen_program_aspm(struct radeon_device *rdev)
 	 */
 	bool fusion_platform = false;

+	if (radeon_aspm == 0)
+		return;
+
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		return;

--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@ -97,6 +97,7 @@ extern int radeon_msi;
 extern int radeon_lockup_timeout;
 extern int radeon_fastfb;
 extern int radeon_dpm;
+extern int radeon_aspm;

 /*
 * Copy from radeon_drv.h so we don't have to include both and have conflicting
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@ -1270,6 +1270,7 @@ static struct radeon_asic rs780_asic = {
 		.get_sclk = &rs780_dpm_get_sclk,
 		.get_mclk = &rs780_dpm_get_mclk,
 		.print_power_state = &rs780_dpm_print_power_state,
+		.debugfs_print_current_performance_level = &rs780_dpm_debugfs_print_current_performance_level,
 	},
 	.pflip = {
 		.pre_page_flip = &rs600_pre_page_flip,
--- a/drivers/gpu/drm/radeon/radeon_asic.h
+++ b/drivers/gpu/drm/radeon/radeon_asic.h
@ -433,6 +433,8 @@ u32 rs780_dpm_get_sclk(struct radeon_device *rdev, bool low);
 u32 rs780_dpm_get_mclk(struct radeon_device *rdev, bool low);
 void rs780_dpm_print_power_state(struct radeon_device *rdev,
 				 struct radeon_ps *ps);
+void rs780_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
+						       struct seq_file *m);

 /* uvd */
 int r600_uvd_init(struct radeon_device *rdev);
--- a/drivers/gpu/drm/radeon/radeon_atombios.c
+++ b/drivers/gpu/drm/radeon/radeon_atombios.c
@ -3513,7 +3513,6 @@ int radeon_atom_get_memory_info(struct radeon_device *rdev,
 	u8 frev, crev, i;
 	u16 data_offset, size;
 	union vram_info *vram_info;
-	u8 *p;

 	memset(mem_info, 0, sizeof(struct atom_memory_info));

@ -3529,13 +3528,12 @@ int radeon_atom_get_memory_info(struct radeon_device *rdev,
 				if (module_index < vram_info->v1_3.ucNumOfVRAMModule) {
 					ATOM_VRAM_MODULE_V3 *vram_module =
 						(ATOM_VRAM_MODULE_V3 *)vram_info->v1_3.aVramInfo;
-					p = (u8 *)vram_info->v1_3.aVramInfo;

 					for (i = 0; i < module_index; i++) {
-						vram_module = (ATOM_VRAM_MODULE_V3 *)p;
 						if (le16_to_cpu(vram_module->usSize) == 0)
 							return -EINVAL;
-						p += le16_to_cpu(vram_module->usSize);
+						vram_module = (ATOM_VRAM_MODULE_V3 *)
+							((u8 *)vram_module + le16_to_cpu(vram_module->usSize));
 					}
 					mem_info->mem_vendor = vram_module->asMemory.ucMemoryVenderID & 0xf;
 					mem_info->mem_type = vram_module->asMemory.ucMemoryType & 0xf0;
@ -3547,13 +3545,12 @@ int radeon_atom_get_memory_info(struct radeon_device *rdev,
 				if (module_index < vram_info->v1_4.ucNumOfVRAMModule) {
 					ATOM_VRAM_MODULE_V4 *vram_module =
 						(ATOM_VRAM_MODULE_V4 *)vram_info->v1_4.aVramInfo;
-					p = (u8 *)vram_info->v1_4.aVramInfo;

 					for (i = 0; i < module_index; i++) {
-						vram_module = (ATOM_VRAM_MODULE_V4 *)p;
 						if (le16_to_cpu(vram_module->usModuleSize) == 0)
 							return -EINVAL;
-						p += le16_to_cpu(vram_module->usModuleSize);
+						vram_module = (ATOM_VRAM_MODULE_V4 *)
+							((u8 *)vram_module + le16_to_cpu(vram_module->usModuleSize));
 					}
 					mem_info->mem_vendor = vram_module->ucMemoryVenderID & 0xf;
 					mem_info->mem_type = vram_module->ucMemoryType & 0xf0;
@ -3572,13 +3569,12 @@ int radeon_atom_get_memory_info(struct radeon_device *rdev,
 				if (module_index < vram_info->v2_1.ucNumOfVRAMModule) {
 					ATOM_VRAM_MODULE_V7 *vram_module =
 						(ATOM_VRAM_MODULE_V7 *)vram_info->v2_1.aVramInfo;
-					p = (u8 *)vram_info->v2_1.aVramInfo;

 					for (i = 0; i < module_index; i++) {
-						vram_module = (ATOM_VRAM_MODULE_V7 *)p;
 						if (le16_to_cpu(vram_module->usModuleSize) == 0)
 							return -EINVAL;
-						p += le16_to_cpu(vram_module->usModuleSize);
+						vram_module = (ATOM_VRAM_MODULE_V7 *)
+							((u8 *)vram_module + le16_to_cpu(vram_module->usModuleSize));
 					}
 					mem_info->mem_vendor = vram_module->ucMemoryVenderID & 0xf;
 					mem_info->mem_type = vram_module->ucMemoryType & 0xf0;
@ -3628,21 +3624,19 @@ int radeon_atom_get_mclk_range_table(struct radeon_device *rdev,
 				if (module_index < vram_info->v1_4.ucNumOfVRAMModule) {
 					ATOM_VRAM_MODULE_V4 *vram_module =
 						(ATOM_VRAM_MODULE_V4 *)vram_info->v1_4.aVramInfo;
-					ATOM_MEMORY_TIMING_FORMAT *format;
-					p = (u8 *)vram_info->v1_4.aVramInfo;

 					for (i = 0; i < module_index; i++) {
-						vram_module = (ATOM_VRAM_MODULE_V4 *)p;
 						if (le16_to_cpu(vram_module->usModuleSize) == 0)
 							return -EINVAL;
-						p += le16_to_cpu(vram_module->usModuleSize);
+						vram_module = (ATOM_VRAM_MODULE_V4 *)
+							((u8 *)vram_module + le16_to_cpu(vram_module->usModuleSize));
 					}
 					mclk_range_table->num_entries = (u8)
-						((vram_module->usModuleSize - offsetof(ATOM_VRAM_MODULE_V4, asMemTiming)) /
+						((le16_to_cpu(vram_module->usModuleSize) - offsetof(ATOM_VRAM_MODULE_V4, asMemTiming)) /
 						 mem_timing_size);
-					p = (u8 *)vram_module->asMemTiming;
+					p = (u8 *)&vram_module->asMemTiming[0];
 					for (i = 0; i < mclk_range_table->num_entries; i++) {
-						format = (ATOM_MEMORY_TIMING_FORMAT *)p;
+						ATOM_MEMORY_TIMING_FORMAT *format = (ATOM_MEMORY_TIMING_FORMAT *)p;
 						mclk_range_table->mclk[i] = le32_to_cpu(format->ulClkRange);
 						p += mem_timing_size;
 					}
@ -3705,17 +3699,21 @@ int radeon_atom_init_mc_reg_table(struct radeon_device *rdev,
 						(ATOM_MEMORY_SETTING_DATA_BLOCK *)
 						((u8 *)reg_block + (2 * sizeof(u16)) +
 						 le16_to_cpu(reg_block->usRegIndexTblSize));
+					ATOM_INIT_REG_INDEX_FORMAT *format = &reg_block->asRegIndexBuf[0];
 					num_entries = (u8)((le16_to_cpu(reg_block->usRegIndexTblSize)) /
 							   sizeof(ATOM_INIT_REG_INDEX_FORMAT)) - 1;
 					if (num_entries > VBIOS_MC_REGISTER_ARRAY_SIZE)
 						return -EINVAL;
-					while (!(reg_block->asRegIndexBuf[i].ucPreRegDataLength & ACCESS_PLACEHOLDER) &&
-					      (i < num_entries)) {
+					while (i < num_entries) {
+						if (format->ucPreRegDataLength & ACCESS_PLACEHOLDER)
+							break;
 						reg_table->mc_reg_address[i].s1 =
-							(u16)(le16_to_cpu(reg_block->asRegIndexBuf[i].usRegIndex));
+							(u16)(le16_to_cpu(format->usRegIndex));
 						reg_table->mc_reg_address[i].pre_reg_data =
-							(u8)(reg_block->asRegIndexBuf[i].ucPreRegDataLength);
+							(u8)(format->ucPreRegDataLength);
 						i++;
+						format = (ATOM_INIT_REG_INDEX_FORMAT *)
+							((u8 *)format + sizeof(ATOM_INIT_REG_INDEX_FORMAT));
 					}
 					reg_table->last = i;
 					while ((*(u32 *)reg_data != END_OF_REG_DATA_BLOCK) &&
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@ -167,6 +167,7 @@ int radeon_msi = -1;
 int radeon_lockup_timeout = 10000;
 int radeon_fastfb = 0;
 int radeon_dpm = -1;
+int radeon_aspm = -1;

 MODULE_PARM_DESC(no_wb, "Disable AGP writeback for scratch registers");
 module_param_named(no_wb, radeon_no_wb, int, 0444);
@ -225,6 +226,9 @@ module_param_named(fastfb, radeon_fastfb, int, 0444);
 MODULE_PARM_DESC(dpm, "DPM support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(dpm, radeon_dpm, int, 0444);

+MODULE_PARM_DESC(aspm, "ASPM support (1 = enable, 0 = disable, -1 = auto)");
+module_param_named(aspm, radeon_aspm, int, 0444);
+
 static struct pci_device_id pciidlist[] = {
 	radeon_PCI_IDS
 };
--- a/drivers/gpu/drm/radeon/rs780_dpm.c
+++ b/drivers/gpu/drm/radeon/rs780_dpm.c
@ -28,6 +28,7 @@
 #include "r600_dpm.h"
 #include "rs780_dpm.h"
 #include "atom.h"
+#include <linux/seq_file.h>

 static struct igp_ps *rs780_get_ps(struct radeon_ps *rps)
 {
@ -961,3 +962,27 @@ u32 rs780_dpm_get_mclk(struct radeon_device *rdev, bool low)

 	return pi->bootup_uma_clk;
 }
+
+void rs780_dpm_debugfs_print_current_performance_level(struct radeon_device *rdev,
+						       struct seq_file *m)
+{
+	struct radeon_ps *rps = rdev->pm.dpm.current_ps;
+	struct igp_ps *ps = rs780_get_ps(rps);
+	u32 current_fb_div = RREG32(FVTHROT_STATUS_REG0) & CURRENT_FEEDBACK_DIV_MASK;
+	u32 func_cntl = RREG32(CG_SPLL_FUNC_CNTL);
+	u32 ref_div = ((func_cntl & SPLL_REF_DIV_MASK) >> SPLL_REF_DIV_SHIFT) + 1;
+	u32 post_div = ((func_cntl & SPLL_SW_HILEN_MASK) >> SPLL_SW_HILEN_SHIFT) + 1 +
+		((func_cntl & SPLL_SW_LOLEN_MASK) >> SPLL_SW_LOLEN_SHIFT) + 1;
+	u32 sclk = (rdev->clock.spll.reference_freq * current_fb_div) /
+		(post_div * ref_div);
+
+	seq_printf(m, "uvd    vclk: %d dclk: %d\n", rps->vclk, rps->dclk);
+
+	/* guess based on the current sclk */
+	if (sclk < (ps->sclk_low + 500))
+		seq_printf(m, "power level 0    sclk: %u vddc_index: %d\n",
+			   ps->sclk_low, ps->min_voltage);
+	else
+		seq_printf(m, "power level 1    sclk: %u vddc_index: %d\n",
+			   ps->sclk_high, ps->max_voltage);
+}
--- a/drivers/gpu/drm/radeon/rs780d.h
+++ b/drivers/gpu/drm/radeon/rs780d.h
@ -28,6 +28,7 @@
 #       define SPLL_SLEEP                                (1 << 1)
 #       define SPLL_REF_DIV(x)                           ((x) << 2)
 #       define SPLL_REF_DIV_MASK                         (7 << 2)
+#       define SPLL_REF_DIV_SHIFT                        2
 #       define SPLL_FB_DIV(x)                            ((x) << 5)
 #       define SPLL_FB_DIV_MASK                          (0xff << 2)
 #       define SPLL_FB_DIV_SHIFT                         2
@ -36,8 +37,10 @@
 #       define SPLL_PULSENUM_MASK                        (3 << 14)
 #       define SPLL_SW_HILEN(x)                          ((x) << 16)
 #       define SPLL_SW_HILEN_MASK                        (0xf << 16)
+#       define SPLL_SW_HILEN_SHIFT                       16
 #       define SPLL_SW_LOLEN(x)                          ((x) << 20)
 #       define SPLL_SW_LOLEN_MASK                        (0xf << 20)
+#       define SPLL_SW_LOLEN_SHIFT                       20
 #       define SPLL_DIVEN                                (1 << 24)
 #       define SPLL_BYPASS_EN                            (1 << 25)
 #       define SPLL_CHG_STATUS                           (1 << 29)
--- a/drivers/gpu/drm/radeon/rv6xx_dpm.c
+++ b/drivers/gpu/drm/radeon/rv6xx_dpm.c
@ -1763,12 +1763,14 @@ void rv6xx_setup_asic(struct radeon_device *rdev)
 {
 	r600_enable_acpi_pm(rdev);

+	if (radeon_aspm != 0) {
 		if (rdev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_ASPM_L0s)
 			rv6xx_enable_l0s(rdev);
 		if (rdev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_ASPM_L1)
 			rv6xx_enable_l1(rdev);
 		if (rdev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_TURNOFFPLL_ASPML1)
 			rv6xx_enable_pll_sleep_in_l1(rdev);
+	}
 }

 void rv6xx_dpm_display_configuration_changed(struct radeon_device *rdev)
--- a/drivers/gpu/drm/radeon/rv770_dpm.c
+++ b/drivers/gpu/drm/radeon/rv770_dpm.c
@ -2099,12 +2099,14 @@ void rv770_dpm_setup_asic(struct radeon_device *rdev)

 	rv770_enable_acpi_pm(rdev);

+	if (radeon_aspm != 0) {
 		if (rdev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_ASPM_L0s)
 			rv770_enable_l0s(rdev);
 		if (rdev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_ASPM_L1)
 			rv770_enable_l1(rdev);
 		if (rdev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_TURNOFFPLL_ASPML1)
 			rv770_enable_pll_sleep_in_l1(rdev);
+	}
 }

 void rv770_dpm_display_configuration_changed(struct radeon_device *rdev)
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@ -7053,6 +7053,9 @@ static void si_program_aspm(struct radeon_device *rdev)
 	bool disable_l0s = false, disable_l1 = false, disable_plloff_in_l1 = false;
 	bool disable_clkreq = false;

+	if (radeon_aspm == 0)
+		return;
+
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		return;