Merge tag 'drm-intel-next-fixes-2019-11-22' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
- Reverts a patch to avoid spinning forever when context's timeline
is active but has no requests
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20191122155523.GA20167@jlahtine-desk.ger.corp.intel.com
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 58f6b3b..8ff69a5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -3110,6 +3110,9 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
DRM_INFO("amdgpu: finishing device.\n");
adev->shutdown = true;
+
+ flush_delayed_work(&adev->delayed_init_work);
+
/* disable all interrupts */
amdgpu_irq_disable_all(adev);
if (adev->mode_info.mode_config_initialized){
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index bbe9ac7..1f044c8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -567,7 +567,9 @@ static int psp_xgmi_initialize(struct psp_context *psp)
struct ta_xgmi_shared_memory *xgmi_cmd;
int ret;
- if (!psp->adev->psp.ta_fw)
+ if (!psp->adev->psp.ta_fw ||
+ !psp->adev->psp.ta_xgmi_ucode_size ||
+ !psp->adev->psp.ta_xgmi_start_addr)
return -ENOENT;
if (!psp->xgmi_context.initialized) {
@@ -777,6 +779,12 @@ static int psp_ras_initialize(struct psp_context *psp)
{
int ret;
+ if (!psp->adev->psp.ta_ras_ucode_size ||
+ !psp->adev->psp.ta_ras_start_addr) {
+ dev_warn(psp->adev->dev, "RAS: ras ta ucode is not available\n");
+ return 0;
+ }
+
if (!psp->ras.ras_initialized) {
ret = psp_ras_init_shared_buf(psp);
if (ret)
@@ -866,6 +874,12 @@ static int psp_hdcp_initialize(struct psp_context *psp)
{
int ret;
+ if (!psp->adev->psp.ta_hdcp_ucode_size ||
+ !psp->adev->psp.ta_hdcp_start_addr) {
+ dev_warn(psp->adev->dev, "HDCP: hdcp ta ucode is not available\n");
+ return 0;
+ }
+
if (!psp->hdcp_context.hdcp_initialized) {
ret = psp_hdcp_init_shared_buf(psp);
if (ret)
@@ -1039,6 +1053,12 @@ static int psp_dtm_initialize(struct psp_context *psp)
{
int ret;
+ if (!psp->adev->psp.ta_dtm_ucode_size ||
+ !psp->adev->psp.ta_dtm_start_addr) {
+ dev_warn(psp->adev->dev, "DTM: dtm ta ucode is not available\n");
+ return 0;
+ }
+
if (!psp->dtm_context.dtm_initialized) {
ret = psp_dtm_init_shared_buf(psp);
if (ret)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index b4dd89a..e324bfe 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -299,6 +299,7 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
{
int i, j;
+ cancel_delayed_work_sync(&adev->uvd.idle_work);
drm_sched_entity_destroy(&adev->uvd.entity);
for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index 703677f..46b590a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -216,6 +216,7 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev)
if (adev->vce.vcpu_bo == NULL)
return 0;
+ cancel_delayed_work_sync(&adev->vce.idle_work);
drm_sched_entity_destroy(&adev->vce.entity);
amdgpu_bo_free_kernel(&adev->vce.vcpu_bo, &adev->vce.gpu_addr,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 3199e4a..9d87044 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -193,6 +193,8 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev)
{
int i, j;
+ cancel_delayed_work_sync(&adev->vcn.idle_work);
+
if (adev->vcn.indirect_sram) {
amdgpu_bo_free_kernel(&adev->vcn.dpg_sram_bo,
&adev->vcn.dpg_sram_gpu_addr,
diff --git a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
index 1e2da4d..69243a8 100644
--- a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
+++ b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
@@ -1068,10 +1068,6 @@ static int smu_smc_table_hw_init(struct smu_context *smu,
return ret;
if (adev->asic_type != CHIP_ARCTURUS) {
- ret = smu_override_pcie_parameters(smu);
- if (ret)
- return ret;
-
ret = smu_notify_display_change(smu);
if (ret)
return ret;
@@ -1100,6 +1096,12 @@ static int smu_smc_table_hw_init(struct smu_context *smu,
return ret;
}
+ if (adev->asic_type != CHIP_ARCTURUS) {
+ ret = smu_override_pcie_parameters(smu);
+ if (ret)
+ return ret;
+ }
+
ret = smu_set_default_od_settings(smu, initialize);
if (ret)
return ret;
@@ -1109,7 +1111,7 @@ static int smu_smc_table_hw_init(struct smu_context *smu,
if (ret)
return ret;
- ret = smu_get_power_limit(smu, &smu->default_power_limit, true, false);
+ ret = smu_get_power_limit(smu, &smu->default_power_limit, false, false);
if (ret)
return ret;
}
@@ -2511,3 +2513,13 @@ int smu_get_dpm_clock_table(struct smu_context *smu,
return ret;
}
+
+uint32_t smu_get_pptable_power_limit(struct smu_context *smu)
+{
+ uint32_t ret = 0;
+
+ if (smu->ppt_funcs->get_pptable_power_limit)
+ ret = smu->ppt_funcs->get_pptable_power_limit(smu);
+
+ return ret;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c
index 3099ac2..58c7c4a 100644
--- a/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/arcturus_ppt.c
@@ -1261,15 +1261,14 @@ arcturus_get_profiling_clk_mask(struct smu_context *smu,
static int arcturus_get_power_limit(struct smu_context *smu,
uint32_t *limit,
- bool asic_default)
+ bool cap)
{
PPTable_t *pptable = smu->smu_table.driver_pptable;
uint32_t asic_default_power_limit = 0;
int ret = 0;
int power_src;
- if (!smu->default_power_limit ||
- !smu->power_limit) {
+ if (!smu->power_limit) {
if (smu_feature_is_enabled(smu, SMU_FEATURE_PPT_BIT)) {
power_src = smu_power_get_index(smu, SMU_POWER_SOURCE_AC);
if (power_src < 0)
@@ -1292,17 +1291,11 @@ static int arcturus_get_power_limit(struct smu_context *smu,
pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
}
- if (smu->od_enabled) {
- asic_default_power_limit *= (100 + smu->smu_table.TDPODLimit);
- asic_default_power_limit /= 100;
- }
-
- smu->default_power_limit = asic_default_power_limit;
smu->power_limit = asic_default_power_limit;
}
- if (asic_default)
- *limit = smu->default_power_limit;
+ if (cap)
+ *limit = smu_v11_0_get_max_power_limit(smu);
else
*limit = smu->power_limit;
@@ -2070,6 +2063,13 @@ static void arcturus_i2c_eeprom_control_fini(struct i2c_adapter *control)
i2c_del_adapter(control);
}
+static uint32_t arcturus_get_pptable_power_limit(struct smu_context *smu)
+{
+ PPTable_t *pptable = smu->smu_table.driver_pptable;
+
+ return pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
+}
+
static const struct pptable_funcs arcturus_ppt_funcs = {
/* translate smu index into arcturus specific index */
.get_smu_msg_index = arcturus_get_smu_msg_index,
@@ -2160,6 +2160,7 @@ static const struct pptable_funcs arcturus_ppt_funcs = {
.get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq,
.set_soft_freq_limited_range = smu_v11_0_set_soft_freq_limited_range,
.override_pcie_parameters = smu_v11_0_override_pcie_parameters,
+ .get_pptable_power_limit = arcturus_get_pptable_power_limit,
};
void arcturus_set_ppt_funcs(struct smu_context *smu)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index c805c6f..390e752 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -3969,6 +3969,13 @@ static int smu7_set_power_state_tasks(struct pp_hwmgr *hwmgr, const void *input)
"Failed to populate and upload SCLK MCLK DPM levels!",
result = tmp_result);
+ /*
+ * If a custom pp table is loaded, set DPMTABLE_OD_UPDATE_VDDC flag.
+ * That effectively disables AVFS feature.
+ */
+ if (hwmgr->hardcode_pp_table != NULL)
+ data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC;
+
tmp_result = smu7_update_avfs(hwmgr);
PP_ASSERT_WITH_CODE((0 == tmp_result),
"Failed to update avfs voltages!",
diff --git a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
index 8120e75..999445c 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
@@ -261,7 +261,6 @@ struct smu_table_context
struct smu_table *tables;
struct smu_table memory_pool;
uint8_t thermal_controller_type;
- uint16_t TDPODLimit;
void *overdrive_table;
};
@@ -548,6 +547,7 @@ struct pptable_funcs {
int (*get_dpm_ultimate_freq)(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t *min, uint32_t *max);
int (*set_soft_freq_limited_range)(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t min, uint32_t max);
int (*override_pcie_parameters)(struct smu_context *smu);
+ uint32_t (*get_pptable_power_limit)(struct smu_context *smu);
};
int smu_load_microcode(struct smu_context *smu);
@@ -717,4 +717,6 @@ int smu_get_uclk_dpm_states(struct smu_context *smu,
int smu_get_dpm_clock_table(struct smu_context *smu,
struct dpm_clocks *clock_table);
+uint32_t smu_get_pptable_power_limit(struct smu_context *smu);
+
#endif
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h
index fd6ec90..6061490 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h
@@ -48,6 +48,8 @@
#define SMU11_TOOL_SIZE 0x19000
+#define MAX_PCIE_CONF 2
+
#define CLK_MAP(clk, index) \
[SMU_##clk] = {1, (index)}
@@ -88,6 +90,11 @@ struct smu_11_0_dpm_table {
uint32_t max; /* MHz */
};
+struct smu_11_0_pcie_table {
+ uint8_t pcie_gen[MAX_PCIE_CONF];
+ uint8_t pcie_lane[MAX_PCIE_CONF];
+};
+
struct smu_11_0_dpm_tables {
struct smu_11_0_dpm_table soc_table;
struct smu_11_0_dpm_table gfx_table;
@@ -100,6 +107,7 @@ struct smu_11_0_dpm_tables {
struct smu_11_0_dpm_table display_table;
struct smu_11_0_dpm_table phy_table;
struct smu_11_0_dpm_table fclk_table;
+ struct smu_11_0_pcie_table pcie_table;
};
struct smu_11_0_dpm_context {
@@ -250,4 +258,8 @@ int smu_v11_0_set_soft_freq_limited_range(struct smu_context *smu, enum smu_clk_
int smu_v11_0_override_pcie_parameters(struct smu_context *smu);
+int smu_v11_0_set_default_od_settings(struct smu_context *smu, bool initialize, size_t overdrive_table_size);
+
+uint32_t smu_v11_0_get_max_power_limit(struct smu_context *smu);
+
#endif
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h
index 86cdc33..b2f96a1 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0_pptable.h
@@ -141,7 +141,9 @@ struct smu_11_0_powerplay_table
struct smu_11_0_power_saving_clock_table power_saving_clock;
struct smu_11_0_overdrive_table overdrive_table;
+#ifndef SMU_11_0_PARTIAL_PPTABLE
PPTable_t smc_pptable; //PPTable_t in smu11_driver_if.h
+#endif
} __attribute__((packed));
#endif
diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
index 354f709..14be350 100644
--- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
@@ -36,6 +36,7 @@
#include "navi10_ppt.h"
#include "smu_v11_0_pptable.h"
#include "smu_v11_0_ppsmc.h"
+#include "nbio/nbio_7_4_sh_mask.h"
#include "asic_reg/mp/mp_11_0_sh_mask.h"
@@ -599,6 +600,7 @@ static int navi10_set_default_dpm_table(struct smu_context *smu)
struct smu_table_context *table_context = &smu->smu_table;
struct smu_11_0_dpm_context *dpm_context = smu_dpm->dpm_context;
PPTable_t *driver_ppt = NULL;
+ int i;
driver_ppt = table_context->driver_pptable;
@@ -629,6 +631,11 @@ static int navi10_set_default_dpm_table(struct smu_context *smu)
dpm_context->dpm_tables.phy_table.min = driver_ppt->FreqTablePhyclk[0];
dpm_context->dpm_tables.phy_table.max = driver_ppt->FreqTablePhyclk[NUM_PHYCLK_DPM_LEVELS - 1];
+ for (i = 0; i < MAX_PCIE_CONF; i++) {
+ dpm_context->dpm_tables.pcie_table.pcie_gen[i] = driver_ppt->PcieGenSpeed[i];
+ dpm_context->dpm_tables.pcie_table.pcie_lane[i] = driver_ppt->PcieLaneCount[i];
+ }
+
return 0;
}
@@ -691,13 +698,29 @@ static bool navi10_is_support_fine_grained_dpm(struct smu_context *smu, enum smu
return dpm_desc->SnapToDiscrete == 0 ? true : false;
}
+static inline bool navi10_od_feature_is_supported(struct smu_11_0_overdrive_table *od_table, enum SMU_11_0_ODFEATURE_ID feature)
+{
+ return od_table->cap[feature];
+}
+
+
static int navi10_print_clk_levels(struct smu_context *smu,
enum smu_clk_type clk_type, char *buf)
{
+ uint16_t *curve_settings;
int i, size = 0, ret = 0;
uint32_t cur_value = 0, value = 0, count = 0;
uint32_t freq_values[3] = {0};
uint32_t mark_index = 0;
+ struct smu_table_context *table_context = &smu->smu_table;
+ uint32_t gen_speed, lane_width;
+ struct smu_dpm_context *smu_dpm = &smu->smu_dpm;
+ struct smu_11_0_dpm_context *dpm_context = smu_dpm->dpm_context;
+ struct amdgpu_device *adev = smu->adev;
+ PPTable_t *pptable = (PPTable_t *)table_context->driver_pptable;
+ OverDriveTable_t *od_table =
+ (OverDriveTable_t *)table_context->overdrive_table;
+ struct smu_11_0_overdrive_table *od_settings = smu->od_settings;
switch (clk_type) {
case SMU_GFXCLK:
@@ -748,6 +771,69 @@ static int navi10_print_clk_levels(struct smu_context *smu,
}
break;
+ case SMU_PCIE:
+ gen_speed = (RREG32_PCIE(smnPCIE_LC_SPEED_CNTL) &
+ PSWUSP0_PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE_MASK)
+ >> PSWUSP0_PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE__SHIFT;
+ lane_width = (RREG32_PCIE(smnPCIE_LC_LINK_WIDTH_CNTL) &
+ PCIE_LC_LINK_WIDTH_CNTL__LC_LINK_WIDTH_RD_MASK)
+ >> PCIE_LC_LINK_WIDTH_CNTL__LC_LINK_WIDTH_RD__SHIFT;
+ for (i = 0; i < NUM_LINK_LEVELS; i++)
+ size += sprintf(buf + size, "%d: %s %s %dMhz %s\n", i,
+ (dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 0) ? "2.5GT/s," :
+ (dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 1) ? "5.0GT/s," :
+ (dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 2) ? "8.0GT/s," :
+ (dpm_context->dpm_tables.pcie_table.pcie_gen[i] == 3) ? "16.0GT/s," : "",
+ (dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 1) ? "x1" :
+ (dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 2) ? "x2" :
+ (dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 3) ? "x4" :
+ (dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 4) ? "x8" :
+ (dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 5) ? "x12" :
+ (dpm_context->dpm_tables.pcie_table.pcie_lane[i] == 6) ? "x16" : "",
+ pptable->LclkFreq[i],
+ (gen_speed == dpm_context->dpm_tables.pcie_table.pcie_gen[i]) &&
+ (lane_width == dpm_context->dpm_tables.pcie_table.pcie_lane[i]) ?
+ "*" : "");
+ break;
+ case SMU_OD_SCLK:
+ if (!smu->od_enabled || !od_table || !od_settings)
+ break;
+ if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_LIMITS))
+ break;
+ size += sprintf(buf + size, "OD_SCLK:\n");
+ size += sprintf(buf + size, "0: %uMhz\n1: %uMhz\n", od_table->GfxclkFmin, od_table->GfxclkFmax);
+ break;
+ case SMU_OD_MCLK:
+ if (!smu->od_enabled || !od_table || !od_settings)
+ break;
+ if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_UCLK_MAX))
+ break;
+ size += sprintf(buf + size, "OD_MCLK:\n");
+ size += sprintf(buf + size, "0: %uMHz\n", od_table->UclkFmax);
+ break;
+ case SMU_OD_VDDC_CURVE:
+ if (!smu->od_enabled || !od_table || !od_settings)
+ break;
+ if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_CURVE))
+ break;
+ size += sprintf(buf + size, "OD_VDDC_CURVE:\n");
+ for (i = 0; i < 3; i++) {
+ switch (i) {
+ case 0:
+ curve_settings = &od_table->GfxclkFreq1;
+ break;
+ case 1:
+ curve_settings = &od_table->GfxclkFreq2;
+ break;
+ case 2:
+ curve_settings = &od_table->GfxclkFreq3;
+ break;
+ default:
+ break;
+ }
+ size += sprintf(buf + size, "%d: %uMHz @ %umV\n", i, curve_settings[0], curve_settings[1] / NAVI10_VOLTAGE_SCALE);
+ }
+ break;
default:
break;
}
@@ -1582,17 +1668,22 @@ static int navi10_display_disable_memory_clock_switch(struct smu_context *smu,
return ret;
}
+static uint32_t navi10_get_pptable_power_limit(struct smu_context *smu)
+{
+ PPTable_t *pptable = smu->smu_table.driver_pptable;
+ return pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
+}
+
static int navi10_get_power_limit(struct smu_context *smu,
uint32_t *limit,
- bool asic_default)
+ bool cap)
{
PPTable_t *pptable = smu->smu_table.driver_pptable;
uint32_t asic_default_power_limit = 0;
int ret = 0;
int power_src;
- if (!smu->default_power_limit ||
- !smu->power_limit) {
+ if (!smu->power_limit) {
if (smu_feature_is_enabled(smu, SMU_FEATURE_PPT_BIT)) {
power_src = smu_power_get_index(smu, SMU_POWER_SOURCE_AC);
if (power_src < 0)
@@ -1615,17 +1706,11 @@ static int navi10_get_power_limit(struct smu_context *smu,
pptable->SocketPowerLimitAc[PPT_THROTTLER_PPT0];
}
- if (smu->od_enabled) {
- asic_default_power_limit *= (100 + smu->smu_table.TDPODLimit);
- asic_default_power_limit /= 100;
- }
-
- smu->default_power_limit = asic_default_power_limit;
smu->power_limit = asic_default_power_limit;
}
- if (asic_default)
- *limit = smu->default_power_limit;
+ if (cap)
+ *limit = smu_v11_0_get_max_power_limit(smu);
else
*limit = smu->power_limit;
@@ -1640,6 +1725,9 @@ static int navi10_update_pcie_parameters(struct smu_context *smu,
int ret, i;
uint32_t smu_pcie_arg;
+ struct smu_dpm_context *smu_dpm = &smu->smu_dpm;
+ struct smu_11_0_dpm_context *dpm_context = smu_dpm->dpm_context;
+
for (i = 0; i < NUM_LINK_LEVELS; i++) {
smu_pcie_arg = (i << 16) |
((pptable->PcieGenSpeed[i] <= pcie_gen_cap) ? (pptable->PcieGenSpeed[i] << 8) :
@@ -1648,11 +1736,249 @@ static int navi10_update_pcie_parameters(struct smu_context *smu,
ret = smu_send_smc_msg_with_param(smu,
SMU_MSG_OverridePcieParameters,
smu_pcie_arg);
+
+ if (ret)
+ return ret;
+
+ if (pptable->PcieGenSpeed[i] > pcie_gen_cap)
+ dpm_context->dpm_tables.pcie_table.pcie_gen[i] = pcie_gen_cap;
+ if (pptable->PcieLaneCount[i] > pcie_width_cap)
+ dpm_context->dpm_tables.pcie_table.pcie_lane[i] = pcie_width_cap;
+ }
+
+ return 0;
+}
+
+static inline void navi10_dump_od_table(OverDriveTable_t *od_table) {
+ pr_debug("OD: Gfxclk: (%d, %d)\n", od_table->GfxclkFmin, od_table->GfxclkFmax);
+ pr_debug("OD: Gfx1: (%d, %d)\n", od_table->GfxclkFreq1, od_table->GfxclkVolt1);
+ pr_debug("OD: Gfx2: (%d, %d)\n", od_table->GfxclkFreq2, od_table->GfxclkVolt2);
+ pr_debug("OD: Gfx3: (%d, %d)\n", od_table->GfxclkFreq3, od_table->GfxclkVolt3);
+ pr_debug("OD: UclkFmax: %d\n", od_table->UclkFmax);
+ pr_debug("OD: OverDrivePct: %d\n", od_table->OverDrivePct);
+}
+
+static int navi10_od_setting_check_range(struct smu_11_0_overdrive_table *od_table, enum SMU_11_0_ODSETTING_ID setting, uint32_t value)
+{
+ if (value < od_table->min[setting]) {
+ pr_warn("OD setting (%d, %d) is less than the minimum allowed (%d)\n", setting, value, od_table->min[setting]);
+ return -EINVAL;
+ }
+ if (value > od_table->max[setting]) {
+ pr_warn("OD setting (%d, %d) is greater than the maximum allowed (%d)\n", setting, value, od_table->max[setting]);
+ return -EINVAL;
+ }
+ return 0;
+}
+
+static int navi10_setup_od_limits(struct smu_context *smu) {
+ struct smu_11_0_overdrive_table *overdrive_table = NULL;
+ struct smu_11_0_powerplay_table *powerplay_table = NULL;
+
+ if (!smu->smu_table.power_play_table) {
+ pr_err("powerplay table uninitialized!\n");
+ return -ENOENT;
+ }
+ powerplay_table = (struct smu_11_0_powerplay_table *)smu->smu_table.power_play_table;
+ overdrive_table = &powerplay_table->overdrive_table;
+ if (!smu->od_settings) {
+ smu->od_settings = kmemdup(overdrive_table, sizeof(struct smu_11_0_overdrive_table), GFP_KERNEL);
+ } else {
+ memcpy(smu->od_settings, overdrive_table, sizeof(struct smu_11_0_overdrive_table));
+ }
+ return 0;
+}
+
+static int navi10_set_default_od_settings(struct smu_context *smu, bool initialize) {
+ OverDriveTable_t *od_table;
+ int ret = 0;
+
+ ret = smu_v11_0_set_default_od_settings(smu, initialize, sizeof(OverDriveTable_t));
+ if (ret)
+ return ret;
+
+ if (initialize) {
+ ret = navi10_setup_od_limits(smu);
+ if (ret) {
+ pr_err("Failed to retrieve board OD limits\n");
+ return ret;
+ }
+
+ }
+
+ od_table = (OverDriveTable_t *)smu->smu_table.overdrive_table;
+ if (od_table) {
+ navi10_dump_od_table(od_table);
}
return ret;
}
+static int navi10_od_edit_dpm_table(struct smu_context *smu, enum PP_OD_DPM_TABLE_COMMAND type, long input[], uint32_t size) {
+ int i;
+ int ret = 0;
+ struct smu_table_context *table_context = &smu->smu_table;
+ OverDriveTable_t *od_table;
+ struct smu_11_0_overdrive_table *od_settings;
+ enum SMU_11_0_ODSETTING_ID freq_setting, voltage_setting;
+ uint16_t *freq_ptr, *voltage_ptr;
+ od_table = (OverDriveTable_t *)table_context->overdrive_table;
+
+ if (!smu->od_enabled) {
+ pr_warn("OverDrive is not enabled!\n");
+ return -EINVAL;
+ }
+
+ if (!smu->od_settings) {
+ pr_err("OD board limits are not set!\n");
+ return -ENOENT;
+ }
+
+ od_settings = smu->od_settings;
+
+ switch (type) {
+ case PP_OD_EDIT_SCLK_VDDC_TABLE:
+ if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_LIMITS)) {
+ pr_warn("GFXCLK_LIMITS not supported!\n");
+ return -ENOTSUPP;
+ }
+ if (!table_context->overdrive_table) {
+ pr_err("Overdrive is not initialized\n");
+ return -EINVAL;
+ }
+ for (i = 0; i < size; i += 2) {
+ if (i + 2 > size) {
+ pr_info("invalid number of input parameters %d\n", size);
+ return -EINVAL;
+ }
+ switch (input[i]) {
+ case 0:
+ freq_setting = SMU_11_0_ODSETTING_GFXCLKFMIN;
+ freq_ptr = &od_table->GfxclkFmin;
+ if (input[i + 1] > od_table->GfxclkFmax) {
+ pr_info("GfxclkFmin (%ld) must be <= GfxclkFmax (%u)!\n",
+ input[i + 1],
+ od_table->GfxclkFmin);
+ return -EINVAL;
+ }
+ break;
+ case 1:
+ freq_setting = SMU_11_0_ODSETTING_GFXCLKFMAX;
+ freq_ptr = &od_table->GfxclkFmax;
+ if (input[i + 1] < od_table->GfxclkFmin) {
+ pr_info("GfxclkFmax (%ld) must be >= GfxclkFmin (%u)!\n",
+ input[i + 1],
+ od_table->GfxclkFmax);
+ return -EINVAL;
+ }
+ break;
+ default:
+ pr_info("Invalid SCLK_VDDC_TABLE index: %ld\n", input[i]);
+ pr_info("Supported indices: [0:min,1:max]\n");
+ return -EINVAL;
+ }
+ ret = navi10_od_setting_check_range(od_settings, freq_setting, input[i + 1]);
+ if (ret)
+ return ret;
+ *freq_ptr = input[i + 1];
+ }
+ break;
+ case PP_OD_EDIT_MCLK_VDDC_TABLE:
+ if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_UCLK_MAX)) {
+ pr_warn("UCLK_MAX not supported!\n");
+ return -ENOTSUPP;
+ }
+ if (size < 2) {
+ pr_info("invalid number of parameters: %d\n", size);
+ return -EINVAL;
+ }
+ if (input[0] != 1) {
+ pr_info("Invalid MCLK_VDDC_TABLE index: %ld\n", input[0]);
+ pr_info("Supported indices: [1:max]\n");
+ return -EINVAL;
+ }
+ ret = navi10_od_setting_check_range(od_settings, SMU_11_0_ODSETTING_UCLKFMAX, input[1]);
+ if (ret)
+ return ret;
+ od_table->UclkFmax = input[1];
+ break;
+ case PP_OD_COMMIT_DPM_TABLE:
+ navi10_dump_od_table(od_table);
+ ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)od_table, true);
+ if (ret) {
+ pr_err("Failed to import overdrive table!\n");
+ return ret;
+ }
+ // no lock needed because smu_od_edit_dpm_table has it
+ ret = smu_handle_task(smu, smu->smu_dpm.dpm_level,
+ AMD_PP_TASK_READJUST_POWER_STATE,
+ false);
+ if (ret) {
+ return ret;
+ }
+ break;
+ case PP_OD_EDIT_VDDC_CURVE:
+ if (!navi10_od_feature_is_supported(od_settings, SMU_11_0_ODFEATURE_GFXCLK_CURVE)) {
+ pr_warn("GFXCLK_CURVE not supported!\n");
+ return -ENOTSUPP;
+ }
+ if (size < 3) {
+ pr_info("invalid number of parameters: %d\n", size);
+ return -EINVAL;
+ }
+ if (!od_table) {
+ pr_info("Overdrive is not initialized\n");
+ return -EINVAL;
+ }
+
+ switch (input[0]) {
+ case 0:
+ freq_setting = SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P1;
+ voltage_setting = SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P1;
+ freq_ptr = &od_table->GfxclkFreq1;
+ voltage_ptr = &od_table->GfxclkVolt1;
+ break;
+ case 1:
+ freq_setting = SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P2;
+ voltage_setting = SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P2;
+ freq_ptr = &od_table->GfxclkFreq2;
+ voltage_ptr = &od_table->GfxclkVolt2;
+ break;
+ case 2:
+ freq_setting = SMU_11_0_ODSETTING_VDDGFXCURVEFREQ_P3;
+ voltage_setting = SMU_11_0_ODSETTING_VDDGFXCURVEVOLTAGE_P3;
+ freq_ptr = &od_table->GfxclkFreq3;
+ voltage_ptr = &od_table->GfxclkVolt3;
+ break;
+ default:
+ pr_info("Invalid VDDC_CURVE index: %ld\n", input[0]);
+ pr_info("Supported indices: [0, 1, 2]\n");
+ return -EINVAL;
+ }
+ ret = navi10_od_setting_check_range(od_settings, freq_setting, input[1]);
+ if (ret)
+ return ret;
+ // Allow setting zero to disable the OverDrive VDDC curve
+ if (input[2] != 0) {
+ ret = navi10_od_setting_check_range(od_settings, voltage_setting, input[2]);
+ if (ret)
+ return ret;
+ *freq_ptr = input[1];
+ *voltage_ptr = ((uint16_t)input[2]) * NAVI10_VOLTAGE_SCALE;
+ pr_debug("OD: set curve %ld: (%d, %d)\n", input[0], *freq_ptr, *voltage_ptr);
+ } else {
+ // If setting 0, disable all voltage curve settings
+ od_table->GfxclkVolt1 = 0;
+ od_table->GfxclkVolt2 = 0;
+ od_table->GfxclkVolt3 = 0;
+ }
+ navi10_dump_od_table(od_table);
+ break;
+ default:
+ return -ENOSYS;
+ }
+ return ret;
+}
static const struct pptable_funcs navi10_ppt_funcs = {
.tables_init = navi10_tables_init,
@@ -1742,6 +2068,9 @@ static const struct pptable_funcs navi10_ppt_funcs = {
.get_dpm_ultimate_freq = smu_v11_0_get_dpm_ultimate_freq,
.set_soft_freq_limited_range = smu_v11_0_set_soft_freq_limited_range,
.override_pcie_parameters = smu_v11_0_override_pcie_parameters,
+ .set_default_od_settings = navi10_set_default_od_settings,
+ .od_edit_dpm_table = navi10_od_edit_dpm_table,
+ .get_pptable_power_limit = navi10_get_pptable_power_limit,
};
void navi10_set_ppt_funcs(struct smu_context *smu)
diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.h b/drivers/gpu/drm/amd/powerplay/navi10_ppt.h
index a37e37c..ec03c79 100644
--- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.h
+++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.h
@@ -33,6 +33,11 @@
#define NAVI14_UMD_PSTATE_PEAK_XTX_GFXCLK (1717)
#define NAVI14_UMD_PSTATE_PEAK_XL_GFXCLK (1448)
+#define NAVI10_VOLTAGE_SCALE (4)
+
+#define smnPCIE_LC_SPEED_CNTL 0x11140290
+#define smnPCIE_LC_LINK_WIDTH_CNTL 0x11140288
+
extern void navi10_set_ppt_funcs(struct smu_context *smu);
#endif
diff --git a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c
index e859bb1..fc9679e 100644
--- a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c
+++ b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c
@@ -24,6 +24,8 @@
#include <linux/module.h>
#include <linux/pci.h>
+#define SMU_11_0_PARTIAL_PPTABLE
+
#include "pp_debug.h"
#include "amdgpu.h"
#include "amdgpu_smu.h"
@@ -31,6 +33,7 @@
#include "atomfirmware.h"
#include "amdgpu_atomfirmware.h"
#include "smu_v11_0.h"
+#include "smu_v11_0_pptable.h"
#include "soc15_common.h"
#include "atom.h"
#include "amd_pcie.h"
@@ -1045,13 +1048,44 @@ int smu_v11_0_init_max_sustainable_clocks(struct smu_context *smu)
return 0;
}
+uint32_t smu_v11_0_get_max_power_limit(struct smu_context *smu) {
+ uint32_t od_limit, max_power_limit;
+ struct smu_11_0_powerplay_table *powerplay_table = NULL;
+ struct smu_table_context *table_context = &smu->smu_table;
+ powerplay_table = table_context->power_play_table;
+
+ max_power_limit = smu_get_pptable_power_limit(smu);
+
+ if (!max_power_limit) {
+ // If we couldn't get the table limit, fall back on first-read value
+ if (!smu->default_power_limit)
+ smu->default_power_limit = smu->power_limit;
+ max_power_limit = smu->default_power_limit;
+ }
+
+ if (smu->od_enabled) {
+ od_limit = le32_to_cpu(powerplay_table->overdrive_table.max[SMU_11_0_ODSETTING_POWERPERCENTAGE]);
+
+ pr_debug("ODSETTING_POWERPERCENTAGE: %d (default: %d)\n", od_limit, smu->default_power_limit);
+
+ max_power_limit *= (100 + od_limit);
+ max_power_limit /= 100;
+ }
+
+ return max_power_limit;
+}
+
int smu_v11_0_set_power_limit(struct smu_context *smu, uint32_t n)
{
int ret = 0;
+ uint32_t max_power_limit;
- if (n > smu->default_power_limit) {
- pr_err("New power limit is over the max allowed %d\n",
- smu->default_power_limit);
+ max_power_limit = smu_v11_0_get_max_power_limit(smu);
+
+ if (n > max_power_limit) {
+ pr_err("New power limit (%d) is over the max allowed %d\n",
+ n,
+ max_power_limit);
return -EINVAL;
}
@@ -1779,3 +1813,30 @@ int smu_v11_0_override_pcie_parameters(struct smu_context *smu)
return ret;
}
+
+int smu_v11_0_set_default_od_settings(struct smu_context *smu, bool initialize, size_t overdrive_table_size)
+{
+ struct smu_table_context *table_context = &smu->smu_table;
+ int ret = 0;
+
+ if (initialize) {
+ if (table_context->overdrive_table) {
+ return -EINVAL;
+ }
+ table_context->overdrive_table = kzalloc(overdrive_table_size, GFP_KERNEL);
+ if (!table_context->overdrive_table) {
+ return -ENOMEM;
+ }
+ ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, false);
+ if (ret) {
+ pr_err("Failed to export overdrive table!\n");
+ return ret;
+ }
+ }
+ ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, true);
+ if (ret) {
+ pr_err("Failed to import overdrive table!\n");
+ return ret;
+ }
+ return ret;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c
index 5b21386..0b48928 100644
--- a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c
@@ -466,7 +466,6 @@ static int vega20_store_powerplay_table(struct smu_context *smu)
sizeof(PPTable_t));
table_context->thermal_controller_type = powerplay_table->ucThermalControllerType;
- table_context->TDPODLimit = le32_to_cpu(powerplay_table->OverDrive8Table.ODSettingsMax[ATOM_VEGA20_ODSETTING_POWERPERCENTAGE]);
return 0;
}
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index 2f2b889..000fa4a 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -1105,29 +1105,6 @@ int drm_gem_mmap_obj(struct drm_gem_object *obj, unsigned long obj_size,
if (obj_size < vma->vm_end - vma->vm_start)
return -EINVAL;
- if (obj->funcs && obj->funcs->mmap) {
- /* Remove the fake offset */
- vma->vm_pgoff -= drm_vma_node_start(&obj->vma_node);
-
- ret = obj->funcs->mmap(obj, vma);
- if (ret)
- return ret;
- WARN_ON(!(vma->vm_flags & VM_DONTEXPAND));
- } else {
- if (obj->funcs && obj->funcs->vm_ops)
- vma->vm_ops = obj->funcs->vm_ops;
- else if (dev->driver->gem_vm_ops)
- vma->vm_ops = dev->driver->gem_vm_ops;
- else
- return -EINVAL;
-
- vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP;
- vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
- vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
- }
-
- vma->vm_private_data = obj;
-
/* Take a ref for this mapping of the object, so that the fault
* handler can dereference the mmap offset's pointer to the object.
* This reference is cleaned up by the corresponding vm_close
@@ -1136,6 +1113,33 @@ int drm_gem_mmap_obj(struct drm_gem_object *obj, unsigned long obj_size,
*/
drm_gem_object_get(obj);
+ if (obj->funcs && obj->funcs->mmap) {
+ /* Remove the fake offset */
+ vma->vm_pgoff -= drm_vma_node_start(&obj->vma_node);
+
+ ret = obj->funcs->mmap(obj, vma);
+ if (ret) {
+ drm_gem_object_put_unlocked(obj);
+ return ret;
+ }
+ WARN_ON(!(vma->vm_flags & VM_DONTEXPAND));
+ } else {
+ if (obj->funcs && obj->funcs->vm_ops)
+ vma->vm_ops = obj->funcs->vm_ops;
+ else if (dev->driver->gem_vm_ops)
+ vma->vm_ops = dev->driver->gem_vm_ops;
+ else {
+ drm_gem_object_put_unlocked(obj);
+ return -EINVAL;
+ }
+
+ vma->vm_flags |= VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP;
+ vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+ vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
+ }
+
+ vma->vm_private_data = obj;
+
return 0;
}
EXPORT_SYMBOL(drm_gem_mmap_obj);
diff --git a/drivers/gpu/drm/drm_gem_ttm_helper.c b/drivers/gpu/drm/drm_gem_ttm_helper.c
index 7412bfc..605a8a3 100644
--- a/drivers/gpu/drm/drm_gem_ttm_helper.c
+++ b/drivers/gpu/drm/drm_gem_ttm_helper.c
@@ -64,8 +64,19 @@ int drm_gem_ttm_mmap(struct drm_gem_object *gem,
struct vm_area_struct *vma)
{
struct ttm_buffer_object *bo = drm_gem_ttm_of_gem(gem);
+ int ret;
- return ttm_bo_mmap_obj(vma, bo);
+ ret = ttm_bo_mmap_obj(vma, bo);
+ if (ret < 0)
+ return ret;
+
+ /*
+ * ttm has its own object refcounting, so drop gem reference
+ * to avoid double accounting counting.
+ */
+ drm_gem_object_put_unlocked(gem);
+
+ return 0;
}
EXPORT_SYMBOL(drm_gem_ttm_mmap);
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index 41f0832..fd74e26 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -379,10 +379,6 @@ radeon_pci_remove(struct pci_dev *pdev)
static void
radeon_pci_shutdown(struct pci_dev *pdev)
{
-#ifdef CONFIG_PPC64
- struct drm_device *ddev = pci_get_drvdata(pdev);
-#endif
-
/* if we are running in a VM, make sure the device
* torn down properly on reboot/shutdown
*/
@@ -390,13 +386,14 @@ radeon_pci_shutdown(struct pci_dev *pdev)
radeon_pci_remove(pdev);
#ifdef CONFIG_PPC64
- /* Some adapters need to be suspended before a
+ /*
+ * Some adapters need to be suspended before a
* shutdown occurs in order to prevent an error
* during kexec.
* Make this power specific becauase it breaks
* some non-power boards.
*/
- radeon_suspend_kms(ddev, true, true, false);
+ radeon_suspend_kms(pci_get_drvdata(pdev), true, true, false);
#endif
}