[PATCH 2/3] drm/amd/powerplay: force sclk limit for peak profile
Quan, Evan
Evan.Quan at amd.com
Thu Jul 18 10:32:08 UTC 2019
1. In navi10_force_clk_levels, i think you need to compare the max level user requested with the peak limit and set the smaller one.
2. can you help me to understand why the change in apply_clock_rules is needed?
________________________________
发件人: amd-gfx <amd-gfx-bounces at lists.freedesktop.org> 代表 Chengming Gui <Jack.Gui at amd.com>
发送时间: Thursday, July 18, 2019 6:02:17 PM
收件人: amd-gfx at lists.freedesktop.org <amd-gfx at lists.freedesktop.org>
抄送: Gui, Jack <Jack.Gui at amd.com>
主题: [PATCH 2/3] drm/amd/powerplay: force sclk limit for peak profile
force different GFX clocks with different SKUs for navi10:
XL (other rev_id): 1625MHz
XT (F1/C1): 1755MHz
XTX (F0/C0): 1830MHz
Signed-off-by: Chengming Gui <Jack.Gui at amd.com>
---
drivers/gpu/drm/amd/powerplay/amdgpu_smu.c | 2 +
drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h | 2 +
drivers/gpu/drm/amd/powerplay/navi10_ppt.c | 66 +++++++++++++++++++++++++-
3 files changed, 68 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
index 122985c..693414f 100644
--- a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
+++ b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c
@@ -664,6 +664,8 @@ static int smu_sw_init(void *handle)
smu->watermarks_bitmap = 0;
smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT;
+ smu->smu_dpm.default_sclk_limit = 0;
+ smu->smu_dpm.peak_sclk_limit = 0;
smu->workload_mask = 1 << smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT];
smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT] = 0;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
index 135a323..acb522b 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h
@@ -441,6 +441,8 @@ struct smu_dpm_context {
void *dpm_context;
void *golden_dpm_context;
bool enable_umd_pstate;
+ uint32_t default_sclk_limit;
+ uint32_t peak_sclk_limit;
enum amd_dpm_forced_level dpm_level;
enum amd_dpm_forced_level saved_dpm_level;
enum amd_dpm_forced_level requested_dpm_level;
diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
index 895a4e5..b4deb9e 100644
--- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
+++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c
@@ -37,6 +37,15 @@
#include "asic_reg/mp/mp_11_0_sh_mask.h"
+#define NV_NV10_F0 0xF0
+#define NV_NV10_C0 0xC0
+#define NV_NV10_F1 0xF1
+#define NV_NV10_C1 0xC1
+
+#define NV_NV10_PEAK_SCLK_XTX 1830
+#define NV_NV10_PEAK_SCLK_XT 1755
+#define NV_NV10_PEAK_SCLK_XL 1625
+
#define FEATURE_MASK(feature) (1ULL << feature)
#define SMC_DPM_FEATURE ( \
FEATURE_MASK(FEATURE_DPM_PREFETCHER_BIT) | \
@@ -675,6 +684,7 @@ static int navi10_force_clk_levels(struct smu_context *smu,
int ret = 0, size = 0;
uint32_t soft_min_level = 0, soft_max_level = 0, min_freq = 0, max_freq = 0;
+ struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
soft_min_level = mask ? (ffs(mask) - 1) : 0;
soft_max_level = mask ? (fls(mask) - 1) : 0;
@@ -682,6 +692,23 @@ static int navi10_force_clk_levels(struct smu_context *smu,
switch (clk_type) {
case SMU_GFXCLK:
case SMU_SCLK:
+ if (smu_dpm_ctx->peak_sclk_limit) {
+ max_freq = smu_dpm_ctx->peak_sclk_limit;
+ ret = smu_get_dpm_freq_by_index(smu, clk_type, soft_min_level, &min_freq);
+ if (ret)
+ return size;
+ } else {
+ ret = smu_get_dpm_freq_by_index(smu, clk_type, soft_min_level, &min_freq);
+ if (ret)
+ return size;
+ ret = smu_get_dpm_freq_by_index(smu, clk_type, soft_max_level, &max_freq);
+ if (ret)
+ return size;
+ }
+ ret = smu_set_soft_freq_range(smu, clk_type, min_freq, max_freq);
+ if (ret)
+ return size;
+ break;
case SMU_SOCCLK:
case SMU_MCLK:
case SMU_UCLK:
@@ -690,11 +717,9 @@ static int navi10_force_clk_levels(struct smu_context *smu,
ret = smu_get_dpm_freq_by_index(smu, clk_type, soft_min_level, &min_freq);
if (ret)
return size;
-
ret = smu_get_dpm_freq_by_index(smu, clk_type, soft_max_level, &max_freq);
if (ret)
return size;
-
ret = smu_set_soft_freq_range(smu, clk_type, min_freq, max_freq);
if (ret)
return size;
@@ -838,6 +863,7 @@ static int navi10_unforce_dpm_levels(struct smu_context *smu)
int ret = 0, i = 0;
uint32_t min_freq, max_freq;
enum smu_clk_type clk_type;
+ struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
enum smu_clk_type clks[] = {
SMU_GFXCLK,
@@ -851,10 +877,18 @@ static int navi10_unforce_dpm_levels(struct smu_context *smu)
if (ret)
return ret;
+ if (clk_type == SMU_GFXCLK && smu_dpm_ctx->default_sclk_limit != 0) {
+ max_freq = smu_dpm_ctx->default_sclk_limit;
+ ret = smu_get_dpm_freq_range(smu, SMU_GFXCLK,
+ &min_freq, NULL);
+ if (ret)
+ return ret;
+ }
ret = smu_set_soft_freq_range(smu, clk_type, min_freq, max_freq);
if (ret)
return ret;
}
+ smu_dpm_ctx->peak_sclk_limit = 0;
return ret;
}
@@ -1531,6 +1565,33 @@ static int navi10_set_ppfeature_status(struct smu_context *smu,
return 0;
}
+static int navi10_apply_clocks_adjust_rules(struct smu_context *smu)
+{
+ int ret = 0;
+ struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm);
+ struct amdgpu_device *adev = smu->adev;
+
+ if (smu_dpm_ctx->default_sclk_limit == 0) {
+ ret = smu_get_dpm_freq_range(smu, SMU_SCLK, NULL,
+ &smu_dpm_ctx->default_sclk_limit);
+ return ret;
+ }
+
+ if (smu_dpm_ctx->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK &&
+ smu_dpm_ctx->dpm_level != smu_dpm_ctx->saved_dpm_level) {
+ if (adev->rev_id == NV_NV10_F0 || adev->rev_id == NV_NV10_C0)
+ smu_dpm_ctx->peak_sclk_limit = NV_NV10_PEAK_SCLK_XTX;
+ else if (adev->rev_id == NV_NV10_F1 || adev->rev_id == NV_NV10_C1)
+ smu_dpm_ctx->peak_sclk_limit = NV_NV10_PEAK_SCLK_XT;
+ else
+ smu_dpm_ctx->peak_sclk_limit = NV_NV10_PEAK_SCLK_XL;
+ } else if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_PROFILE_PEAK && smu_dpm_ctx->peak_sclk_limit != 0) {
+ smu_dpm_ctx->peak_sclk_limit = 0;
+ }
+
+ return 0;
+}
+
static const struct pptable_funcs navi10_ppt_funcs = {
.tables_init = navi10_tables_init,
.alloc_dpm_context = navi10_allocate_dpm_context,
@@ -1566,6 +1627,7 @@ static const struct pptable_funcs navi10_ppt_funcs = {
.get_uclk_dpm_states = navi10_get_uclk_dpm_states,
.get_ppfeature_status = navi10_get_ppfeature_status,
.set_ppfeature_status = navi10_set_ppfeature_status,
+ .apply_clocks_adjust_rules = navi10_apply_clocks_adjust_rules,
};
void navi10_set_ppt_funcs(struct smu_context *smu)
--
2.7.4
_______________________________________________
amd-gfx mailing list
amd-gfx at lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/amd-gfx/attachments/20190718/d69fd083/attachment-0001.html>
More information about the amd-gfx
mailing list