<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40"> <head> <meta http-equiv="Content-Type" content="text/html; charset=gb2312"> <meta name="Generator" content="Microsoft Word 15 (filtered medium)"> <style></style> </head> <body lang="EN-US" link="blue" vlink="purple"> <div class="WordSection1"> Hi Rex,<o:p></o:p> <o:p> </o:p> Yes, per discussed, we may need another patch to consider how to set the PHM_PlatformCaps_UMDPState flag.<o:p></o:p> But for now, I will keep the patch as it is.<o:p></o:p> <o:p> </o:p> Regards,<o:p></o:p> Evan<o:p></o:p> <div style="border:none;border-left:solid blue 1.5pt;padding:0in 0in 0in 4.0pt"> <div> <div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0in 0in 0in"> From: Zhu, Rex Sent: Tuesday, June 19, 2018 11:43 PM To: Quan, Evan <Evan.Quan@amd.com>; Alex Deucher <alexdeucher@gmail.com> Cc: amd-gfx list <amd-gfx@lists.freedesktop.org> Subject: Re: [PATCH 10/13] drm/amd/powerplay: apply clocks adjust rules on power state change<o:p></o:p> </div> </div> <o:p> </o:p> <div> <div> Hi Evan,<o:p></o:p> </div> <div> did we need to check the following flags on vega12？will driver set those flags when user select the umd_pstate?<o:p></o:p> </div> <div> PHM_PlatformCaps_UMDPState/PHM_PlatformCaps_PState.<o:p></o:p> </div> <div> Best Regards<o:p></o:p> </div> <div> Rex <o:p></o:p> </div> <div> <div> 获取 <a href="https://aka.ms/ghei36">Outlook for Android</a><o:p></o:p> </div> <o:p> </o:p> </div> <div class="MsoNormal" align="center" style="text-align:center"> <hr size="2" width="98%" align="center"> </div> <div id="x_divRplyFwdMsg"> From: amd-gfx <<a href="mailto:amd-gfx-bounces@lists.freedesktop.org">amd-gfx-bounces@lists.freedesktop.org</a>> on behalf of Alex Deucher <<a href="mailto:alexdeucher@gmail.com">alexdeucher@gmail.com</a>> Sent: Tuesday, June 19, 2018 11:16:44 PM To: Quan, Evan Cc: amd-gfx list Subject: Re: [PATCH 10/13] drm/amd/powerplay: apply clocks adjust rules on power state change<o:p></o:p> <div> <o:p></o:p> </div> </div> </div> <div> On Tue, Jun 19, 2018 at 3:39 AM, Evan Quan <<a href="mailto:evan.quan@amd.com">evan.quan@amd.com</a>> wrote: > The clocks hard/soft min/max clock levels will be adjusted > correspondingly. Also note that this add the apply_clocks_adjust_rules callback which is used to validate the clock settings on a power state change. One other comment below. > > Change-Id: I2c4b6cd6756d40a28933f0c26b9e1a3d5078bab8 > Signed-off-by: Evan Quan <<a href="mailto:evan.quan@amd.com">evan.quan@amd.com</a>> > --- > drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c | 162 +++++++++++++++++++++ > drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h | 2 + > 2 files changed, 164 insertions(+) > > diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c > index a227ace..26bdfff 100644 > --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c > +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c > @@ -1950,6 +1950,166 @@ static int vega12_print_clock_levels(struct pp_hwmgr *hwmgr, > return size; > } > > +static int vega12_apply_clocks_adjust_rules(struct pp_hwmgr *hwmgr) > +{ > + struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); > + struct vega12_single_dpm_table *dpm_table; > + bool vblank_too_short = false; > + bool disable_mclk_switching; > + uint32_t i, latency; > + > + disable_mclk_switching = ((1 < hwmgr->display_config->num_display) && > + !hwmgr->display_config->multi_monitor_in_sync) || > + vblank_too_short; > + latency = hwmgr->display_config->dce_tolerable_mclk_in_active_latency; > + > + /* gfxclk */ > + dpm_table = &(data->dpm_table.gfx_table); > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + > + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { > + if (VEGA12_UMD_PSTATE_GFXCLK_LEVEL < dpm_table->count) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_GFXCLK_LEVEL].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_GFXCLK_LEVEL].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + } > + } > + > + /* memclk */ > + dpm_table = &(data->dpm_table.mem_table); > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + > + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { > + if (VEGA12_UMD_PSTATE_MCLK_LEVEL < dpm_table->count) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_MCLK_LEVEL].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_MCLK_LEVEL].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + } > + } > + > + /* honour DAL's UCLK Hardmin */ > + if (dpm_table->dpm_state.hard_min_level < (hwmgr->display_config->min_mem_set_clock / 100)) > + dpm_table->dpm_state.hard_min_level = hwmgr->display_config->min_mem_set_clock / 100; > + Didn't you just remove the uclk hard min setting in a previous patch? > + /* Hardmin is dependent on displayconfig */ > + if (disable_mclk_switching) { > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + for (i = 0; i < data->mclk_latency_table.count - 1; i++) { > + if (data->mclk_latency_table.entries[i].latency <= latency) { > + if (dpm_table->dpm_levels[i].value >= (hwmgr->display_config->min_mem_set_clock / 100)) { > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[i].value; > + break; > + } > + } > + } > + } > + > + if (hwmgr->display_config->nb_pstate_switch_disable) > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + > + /* vclk */ > + dpm_table = &(data->dpm_table.vclk_table); > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + > + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { > + if (VEGA12_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + } > + } > + > + /* dclk */ > + dpm_table = &(data->dpm_table.dclk_table); > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + > + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { > + if (VEGA12_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + } > + } > + > + /* socclk */ > + dpm_table = &(data->dpm_table.soc_table); > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + > + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { > + if (VEGA12_UMD_PSTATE_SOCCLK_LEVEL < dpm_table->count) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_SOCCLK_LEVEL].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_SOCCLK_LEVEL].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + } > + } > + > + /* eclk */ > + dpm_table = &(data->dpm_table.eclk_table); > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value; > + dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + > + if (PP_CAP(PHM_PlatformCaps_UMDPState)) { > + if (VEGA12_UMD_PSTATE_VCEMCLK_LEVEL < dpm_table->count) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_VCEMCLK_LEVEL].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_VCEMCLK_LEVEL].value; > + } > + > + if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) { > + dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value; > + } > + } > + > + return 0; > +} > + > static int vega12_display_configuration_changed_task(struct pp_hwmgr *hwmgr) > { > struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend); > @@ -2196,6 +2356,8 @@ static const struct pp_hwmgr_func vega12_hwmgr_funcs = { > .display_clock_voltage_request = vega12_display_clock_voltage_request, > .force_clock_level = vega12_force_clock_level, > .print_clock_levels = vega12_print_clock_levels, > + .apply_clocks_adjust_rules = > + vega12_apply_clocks_adjust_rules, > .display_config_changed = vega12_display_configuration_changed_task, > .powergate_uvd = vega12_power_gate_uvd, > .powergate_vce = vega12_power_gate_vce, > diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h > index e18c083..e17237c 100644 > --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h > +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h > @@ -443,6 +443,8 @@ struct vega12_hwmgr { > #define VEGA12_UMD_PSTATE_GFXCLK_LEVEL 0x3 > #define VEGA12_UMD_PSTATE_SOCCLK_LEVEL 0x3 > #define VEGA12_UMD_PSTATE_MCLK_LEVEL 0x2 > +#define VEGA12_UMD_PSTATE_UVDCLK_LEVEL 0x3 > +#define VEGA12_UMD_PSTATE_VCEMCLK_LEVEL 0x3 > > int vega12_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable); > > -- > 2.7.4 > > _______________________________________________ > amd-gfx mailing list > <a href="mailto:amd-gfx@lists.freedesktop.org">amd-gfx@lists.freedesktop.org</a> > <a href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a> _______________________________________________ amd-gfx mailing list <a href="mailto:amd-gfx@lists.freedesktop.org">amd-gfx@lists.freedesktop.org</a> <a href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a><o:p></o:p> </div> </div> </div> </body> </html>