{0, 1, 400, 800, 6042, 36042, 6584, 36584}, /* DDR3-800 SC */
};
-static const struct cxsr_latency *intel_get_cxsr_latency(int is_desktop,
- int is_ddr3,
+static const struct cxsr_latency *intel_get_cxsr_latency(bool is_desktop,
+ bool is_ddr3,
int fsb,
int mem)
{
struct drm_device *dev = &dev_priv->drm;
u32 val;
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
I915_WRITE(FW_BLC_SELF_VLV, enable ? FW_CSPWRDWNEN : 0);
POSTING_READ(FW_BLC_SELF_VLV);
dev_priv->wm.vlv.cxsr = enable;
- } else if (IS_G4X(dev) || IS_CRESTLINE(dev)) {
+ } else if (IS_G4X(dev_priv) || IS_CRESTLINE(dev_priv)) {
I915_WRITE(FW_BLC_SELF, enable ? FW_BLC_SELF_EN : 0);
POSTING_READ(FW_BLC_SELF);
} else if (IS_PINEVIEW(dev)) {
val |= enable ? PINEVIEW_SELF_REFRESH_EN : 0;
I915_WRITE(DSPFW3, val);
POSTING_READ(DSPFW3);
- } else if (IS_I945G(dev) || IS_I945GM(dev)) {
+ } else if (IS_I945G(dev_priv) || IS_I945GM(dev_priv)) {
val = enable ? _MASKED_BIT_ENABLE(FW_BLC_SELF_EN) :
_MASKED_BIT_DISABLE(FW_BLC_SELF_EN);
I915_WRITE(FW_BLC_SELF, val);
POSTING_READ(FW_BLC_SELF);
- } else if (IS_I915GM(dev)) {
+ } else if (IS_I915GM(dev_priv)) {
+ /*
+ * FIXME can't find a bit like this for 915G, and
+ * and yet it does have the related watermark in
+ * FW_BLC_SELF. What's going on?
+ */
val = enable ? _MASKED_BIT_ENABLE(INSTPM_SELF_EN) :
_MASKED_BIT_DISABLE(INSTPM_SELF_EN);
I915_WRITE(INSTPM, val);
u32 reg;
unsigned long wm;
- latency = intel_get_cxsr_latency(IS_PINEVIEW_G(dev), dev_priv->is_ddr3,
- dev_priv->fsb_freq, dev_priv->mem_freq);
+ latency = intel_get_cxsr_latency(IS_PINEVIEW_G(dev_priv),
+ dev_priv->is_ddr3,
+ dev_priv->fsb_freq,
+ dev_priv->mem_freq);
if (!latency) {
DRM_DEBUG_KMS("Unknown FSB/MEM found, disable CxSR\n");
intel_set_memory_cxsr(dev_priv, false);
display_wm, cursor_wm);
if (display_wm > display->max_wm) {
- DRM_DEBUG_KMS("display watermark is too large(%d/%ld), disabling\n",
+ DRM_DEBUG_KMS("display watermark is too large(%d/%u), disabling\n",
display_wm, display->max_wm);
return false;
}
if (cursor_wm > cursor->max_wm) {
- DRM_DEBUG_KMS("cursor watermark is too large(%d/%ld), disabling\n",
+ DRM_DEBUG_KMS("cursor watermark is too large(%d/%u), disabling\n",
cursor_wm, cursor->max_wm);
return false;
}
if (dev_priv->wm.pri_latency[level] == 0)
return USHRT_MAX;
- if (!state->visible)
+ if (!state->base.visible)
return 0;
cpp = drm_format_plane_cpp(state->base.fb->pixel_format, 0);
if (plane->base.type == DRM_PLANE_TYPE_CURSOR)
continue;
- if (state->visible) {
+ if (state->base.visible) {
wm_state->num_active_planes++;
total_rate += drm_format_plane_cpp(state->base.fb->pixel_format, 0);
}
continue;
}
- if (!state->visible) {
+ if (!state->base.visible) {
plane->wm.fifo_size = 0;
continue;
}
struct intel_plane_state *state =
to_intel_plane_state(plane->base.state);
- if (!state->visible)
+ if (!state->base.visible)
continue;
/* normal watermarks */
if (IS_I945GM(dev))
wm_info = &i945_wm_info;
- else if (!IS_GEN2(dev))
+ else if (!IS_GEN2(dev_priv))
wm_info = &i915_wm_info;
else
wm_info = &i830_a_wm_info;
if (intel_crtc_active(crtc)) {
const struct drm_display_mode *adjusted_mode;
int cpp = drm_format_plane_cpp(crtc->primary->state->fb->pixel_format, 0);
- if (IS_GEN2(dev))
+ if (IS_GEN2(dev_priv))
cpp = 4;
adjusted_mode = &to_intel_crtc(crtc)->config->base.adjusted_mode;
planea_wm = wm_info->max_wm;
}
- if (IS_GEN2(dev))
+ if (IS_GEN2(dev_priv))
wm_info = &i830_bc_wm_info;
fifo_size = dev_priv->display.get_fifo_size(dev, 1);
if (intel_crtc_active(crtc)) {
const struct drm_display_mode *adjusted_mode;
int cpp = drm_format_plane_cpp(crtc->primary->state->fb->pixel_format, 0);
- if (IS_GEN2(dev))
+ if (IS_GEN2(dev_priv))
cpp = 4;
adjusted_mode = &to_intel_crtc(crtc)->config->base.adjusted_mode;
DRM_DEBUG_KMS("FIFO watermarks - A: %d, B: %d\n", planea_wm, planeb_wm);
- if (IS_I915GM(dev) && enabled) {
+ if (IS_I915GM(dev_priv) && enabled) {
struct drm_i915_gem_object *obj;
obj = intel_fb_obj(enabled->primary->state->fb);
/* self-refresh seems busted with untiled */
- if (obj->tiling_mode == I915_TILING_NONE)
+ if (!i915_gem_object_is_tiled(obj))
enabled = NULL;
}
unsigned long line_time_us;
int entries;
+ if (IS_I915GM(dev_priv) || IS_I945GM(dev_priv))
+ cpp = 4;
+
line_time_us = max(htotal * 1000 / clock, 1);
/* Use ns/us then divide to preserve precision */
if (srwm < 0)
srwm = 1;
- if (IS_I945G(dev) || IS_I945GM(dev))
+ if (IS_I945G(dev_priv) || IS_I945GM(dev_priv))
I915_WRITE(FW_BLC_SELF,
FW_BLC_SELF_FIFO_MASK | (srwm & 0xff));
- else if (IS_I915GM(dev))
+ else
I915_WRITE(FW_BLC_SELF, srwm & 0x3f);
}
drm_format_plane_cpp(pstate->base.fb->pixel_format, 0) : 0;
uint32_t method1, method2;
- if (!cstate->base.active || !pstate->visible)
+ if (!cstate->base.active || !pstate->base.visible)
return 0;
method1 = ilk_wm_method1(ilk_pipe_pixel_rate(cstate), cpp, mem_value);
method2 = ilk_wm_method2(ilk_pipe_pixel_rate(cstate),
cstate->base.adjusted_mode.crtc_htotal,
- drm_rect_width(&pstate->dst),
+ drm_rect_width(&pstate->base.dst),
cpp, mem_value);
return min(method1, method2);
drm_format_plane_cpp(pstate->base.fb->pixel_format, 0) : 0;
uint32_t method1, method2;
- if (!cstate->base.active || !pstate->visible)
+ if (!cstate->base.active || !pstate->base.visible)
return 0;
method1 = ilk_wm_method1(ilk_pipe_pixel_rate(cstate), cpp, mem_value);
method2 = ilk_wm_method2(ilk_pipe_pixel_rate(cstate),
cstate->base.adjusted_mode.crtc_htotal,
- drm_rect_width(&pstate->dst),
+ drm_rect_width(&pstate->base.dst),
cpp, mem_value);
return min(method1, method2);
}
* this is necessary to avoid flickering.
*/
int cpp = 4;
- int width = pstate->visible ? pstate->base.crtc_w : 64;
+ int width = pstate->base.visible ? pstate->base.crtc_w : 64;
if (!cstate->base.active)
return 0;
int cpp = pstate->base.fb ?
drm_format_plane_cpp(pstate->base.fb->pixel_format, 0) : 0;
- if (!cstate->base.active || !pstate->visible)
+ if (!cstate->base.active || !pstate->base.visible)
return 0;
- return ilk_wm_fbc(pri_val, drm_rect_width(&pstate->dst), cpp);
+ return ilk_wm_fbc(pri_val, drm_rect_width(&pstate->base.dst), cpp);
}
static unsigned int ilk_display_fifo_size(const struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
- if (IS_GEN9(dev)) {
+ if (IS_GEN9(dev_priv)) {
uint32_t val;
int ret, i;
- int level, max_level = ilk_wm_max_level(dev);
+ int level, max_level = ilk_wm_max_level(dev_priv);
/* read the first set of memory latencies[0:3] */
val = 0; /* data0 to be programmed to 0 for first set */
wm[7] = (val >> GEN9_MEM_LATENCY_LEVEL_3_7_SHIFT) &
GEN9_MEM_LATENCY_LEVEL_MASK;
+ /*
+ * If a level n (n > 1) has a 0us latency, all levels m (m >= n)
+ * need to be disabled. We make sure to sanitize the values out
+ * of the punit to satisfy this requirement.
+ */
+ for (level = 1; level <= max_level; level++) {
+ if (wm[level] == 0) {
+ for (i = level + 1; i <= max_level; i++)
+ wm[i] = 0;
+ break;
+ }
+ }
+
/*
* WaWmMemoryReadLatency:skl
*
* punit doesn't take into account the read latency so we need
- * to add 2us to the various latency levels we retrieve from
- * the punit.
- * - W0 is a bit special in that it's the only level that
- * can't be disabled if we want to have display working, so
- * we always add 2us there.
- * - For levels >=1, punit returns 0us latency when they are
- * disabled, so we respect that and don't add 2us then
- *
- * Additionally, if a level n (n > 1) has a 0us latency, all
- * levels m (m >= n) need to be disabled. We make sure to
- * sanitize the values out of the punit to satisfy this
- * requirement.
+ * to add 2us to the various latency levels we retrieve from the
+ * punit when level 0 response data us 0us.
*/
- wm[0] += 2;
- for (level = 1; level <= max_level; level++)
- if (wm[level] != 0)
+ if (wm[0] == 0) {
+ wm[0] += 2;
+ for (level = 1; level <= max_level; level++) {
+ if (wm[level] == 0)
+ break;
wm[level] += 2;
- else {
- for (i = level + 1; i <= max_level; i++)
- wm[i] = 0;
-
- break;
}
- } else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+ }
+
+ } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
uint64_t sskpd = I915_READ64(MCH_SSKPD);
wm[0] = (sskpd >> 56) & 0xFF;
}
}
-static void intel_fixup_spr_wm_latency(struct drm_device *dev, uint16_t wm[5])
+static void intel_fixup_spr_wm_latency(struct drm_i915_private *dev_priv,
+ uint16_t wm[5])
{
/* ILK sprite LP0 latency is 1300 ns */
- if (IS_GEN5(dev))
+ if (IS_GEN5(dev_priv))
wm[0] = 13;
}
-static void intel_fixup_cur_wm_latency(struct drm_device *dev, uint16_t wm[5])
+static void intel_fixup_cur_wm_latency(struct drm_i915_private *dev_priv,
+ uint16_t wm[5])
{
/* ILK cursor LP0 latency is 1300 ns */
- if (IS_GEN5(dev))
+ if (IS_GEN5(dev_priv))
wm[0] = 13;
/* WaDoubleCursorLP3Latency:ivb */
- if (IS_IVYBRIDGE(dev))
+ if (IS_IVYBRIDGE(dev_priv))
wm[3] *= 2;
}
-int ilk_wm_max_level(const struct drm_device *dev)
+int ilk_wm_max_level(const struct drm_i915_private *dev_priv)
{
/* how many WM levels are we expecting */
- if (INTEL_INFO(dev)->gen >= 9)
+ if (INTEL_GEN(dev_priv) >= 9)
return 7;
- else if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+ else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
return 4;
- else if (INTEL_INFO(dev)->gen >= 6)
+ else if (INTEL_GEN(dev_priv) >= 6)
return 3;
else
return 2;
}
-static void intel_print_wm_latency(struct drm_device *dev,
+static void intel_print_wm_latency(struct drm_i915_private *dev_priv,
const char *name,
const uint16_t wm[8])
{
- int level, max_level = ilk_wm_max_level(dev);
+ int level, max_level = ilk_wm_max_level(dev_priv);
for (level = 0; level <= max_level; level++) {
unsigned int latency = wm[level];
* - latencies are in us on gen9.
* - before then, WM1+ latency values are in 0.5us units
*/
- if (IS_GEN9(dev))
+ if (IS_GEN9(dev_priv))
latency *= 10;
else if (level > 0)
latency *= 5;
static bool ilk_increase_wm_latency(struct drm_i915_private *dev_priv,
uint16_t wm[5], uint16_t min)
{
- int level, max_level = ilk_wm_max_level(&dev_priv->drm);
+ int level, max_level = ilk_wm_max_level(dev_priv);
if (wm[0] >= min)
return false;
return;
DRM_DEBUG_KMS("WM latency values increased to avoid potential underruns\n");
- intel_print_wm_latency(dev, "Primary", dev_priv->wm.pri_latency);
- intel_print_wm_latency(dev, "Sprite", dev_priv->wm.spr_latency);
- intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency);
+ intel_print_wm_latency(dev_priv, "Primary", dev_priv->wm.pri_latency);
+ intel_print_wm_latency(dev_priv, "Sprite", dev_priv->wm.spr_latency);
+ intel_print_wm_latency(dev_priv, "Cursor", dev_priv->wm.cur_latency);
}
static void ilk_setup_wm_latency(struct drm_device *dev)
memcpy(dev_priv->wm.cur_latency, dev_priv->wm.pri_latency,
sizeof(dev_priv->wm.pri_latency));
- intel_fixup_spr_wm_latency(dev, dev_priv->wm.spr_latency);
- intel_fixup_cur_wm_latency(dev, dev_priv->wm.cur_latency);
+ intel_fixup_spr_wm_latency(dev_priv, dev_priv->wm.spr_latency);
+ intel_fixup_cur_wm_latency(dev_priv, dev_priv->wm.cur_latency);
- intel_print_wm_latency(dev, "Primary", dev_priv->wm.pri_latency);
- intel_print_wm_latency(dev, "Sprite", dev_priv->wm.spr_latency);
- intel_print_wm_latency(dev, "Cursor", dev_priv->wm.cur_latency);
+ intel_print_wm_latency(dev_priv, "Primary", dev_priv->wm.pri_latency);
+ intel_print_wm_latency(dev_priv, "Sprite", dev_priv->wm.spr_latency);
+ intel_print_wm_latency(dev_priv, "Cursor", dev_priv->wm.cur_latency);
- if (IS_GEN6(dev))
+ if (IS_GEN6(dev_priv))
snb_wm_latency_quirk(dev);
}
struct drm_i915_private *dev_priv = to_i915(dev);
intel_read_wm_latency(dev, dev_priv->wm.skl_latency);
- intel_print_wm_latency(dev, "Gen9 Plane", dev_priv->wm.skl_latency);
+ intel_print_wm_latency(dev_priv, "Gen9 Plane", dev_priv->wm.skl_latency);
}
static bool ilk_validate_pipe_wm(struct drm_device *dev,
struct intel_plane_state *pristate = NULL;
struct intel_plane_state *sprstate = NULL;
struct intel_plane_state *curstate = NULL;
- int level, max_level = ilk_wm_max_level(dev), usable_level;
+ int level, max_level = ilk_wm_max_level(dev_priv), usable_level;
struct ilk_wm_maximums max;
pipe_wm = &cstate->wm.ilk.optimal;
pipe_wm->pipe_enabled = cstate->base.active;
if (sprstate) {
- pipe_wm->sprites_enabled = sprstate->visible;
- pipe_wm->sprites_scaled = sprstate->visible &&
- (drm_rect_width(&sprstate->dst) != drm_rect_width(&sprstate->src) >> 16 ||
- drm_rect_height(&sprstate->dst) != drm_rect_height(&sprstate->src) >> 16);
+ pipe_wm->sprites_enabled = sprstate->base.visible;
+ pipe_wm->sprites_scaled = sprstate->base.visible &&
+ (drm_rect_width(&sprstate->base.dst) != drm_rect_width(&sprstate->base.src) >> 16 ||
+ drm_rect_height(&sprstate->base.dst) != drm_rect_height(&sprstate->base.src) >> 16);
}
usable_level = max_level;
memset(&pipe_wm->wm, 0, sizeof(pipe_wm->wm));
pipe_wm->wm[0] = pipe_wm->raw_wm[0];
- if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
pipe_wm->linetime = hsw_compute_linetime_wm(cstate);
if (!ilk_validate_pipe_wm(dev, pipe_wm))
{
struct intel_pipe_wm *a = &newstate->wm.ilk.intermediate;
struct intel_pipe_wm *b = &intel_crtc->wm.active.ilk;
- int level, max_level = ilk_wm_max_level(dev);
+ int level, max_level = ilk_wm_max_level(to_i915(dev));
/*
* Start with the final, target watermarks, then combine with the
struct intel_pipe_wm *merged)
{
struct drm_i915_private *dev_priv = to_i915(dev);
- int level, max_level = ilk_wm_max_level(dev);
+ int level, max_level = ilk_wm_max_level(dev_priv);
int last_enabled_level = max_level;
/* ILK/SNB/IVB: LP1+ watermarks only w/ single pipe */
- if ((INTEL_INFO(dev)->gen <= 6 || IS_IVYBRIDGE(dev)) &&
+ if ((INTEL_GEN(dev_priv) <= 6 || IS_IVYBRIDGE(dev_priv)) &&
config->num_pipes_active > 1)
last_enabled_level = 0;
* What we should check here is whether FBC can be
* enabled sometime later.
*/
- if (IS_GEN5(dev) && !merged->fbc_wm_enabled &&
+ if (IS_GEN5(dev_priv) && !merged->fbc_wm_enabled &&
intel_fbc_is_active(dev_priv)) {
for (level = 2; level <= max_level; level++) {
struct intel_wm_level *wm = &merged->wm[level];
{
struct drm_i915_private *dev_priv = to_i915(dev);
- if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
return 2 * level;
else
return dev_priv->wm.pri_latency[level];
struct intel_pipe_wm *r1,
struct intel_pipe_wm *r2)
{
- int level, max_level = ilk_wm_max_level(dev);
+ int level, max_level = ilk_wm_max_level(to_i915(dev));
int level1 = 0, level2 = 0;
for (level = 1; level <= max_level; level++) {
I915_WRITE(PIPE_WM_LINETIME(PIPE_C), results->wm_linetime[2]);
if (dirty & WM_DIRTY_DDB) {
- if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
val = I915_READ(WM_MISC);
if (results->partitioning == INTEL_DDB_PART_1_2)
val &= ~WM_MISC_DATA_PARTITION_5_6;
return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL);
}
-/*
- * On gen9, we need to allocate Display Data Buffer (DDB) portions to the
- * different active planes.
- */
-
-#define SKL_DDB_SIZE 896 /* in blocks */
-#define BXT_DDB_SIZE 512
#define SKL_SAGV_BLOCK_TIME 30 /* µs */
/*
}
}
+/*
+ * FIXME: We still don't have the proper code detect if we need to apply the WA,
+ * so assume we'll always need it in order to avoid underruns.
+ */
+static bool skl_needs_memory_bw_wa(struct intel_atomic_state *state)
+{
+ struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+
+ if (IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv) ||
+ IS_KABYLAKE(dev_priv))
+ return true;
+
+ return false;
+}
+
+static bool
+intel_has_sagv(struct drm_i915_private *dev_priv)
+{
+ if (IS_KABYLAKE(dev_priv))
+ return true;
+
+ if (IS_SKYLAKE(dev_priv) &&
+ dev_priv->sagv_status != I915_SAGV_NOT_CONTROLLED)
+ return true;
+
+ return false;
+}
+
/*
* SAGV dynamically adjusts the system agent voltage and clock frequencies
* depending on power and performance requirements. The display engine access
* - We're not using an interlaced display configuration
*/
int
-skl_enable_sagv(struct drm_i915_private *dev_priv)
+intel_enable_sagv(struct drm_i915_private *dev_priv)
{
int ret;
- if (dev_priv->skl_sagv_status == I915_SKL_SAGV_NOT_CONTROLLED ||
- dev_priv->skl_sagv_status == I915_SKL_SAGV_ENABLED)
+ if (!intel_has_sagv(dev_priv))
+ return 0;
+
+ if (dev_priv->sagv_status == I915_SAGV_ENABLED)
return 0;
DRM_DEBUG_KMS("Enabling the SAGV\n");
* Some skl systems, pre-release machines in particular,
* don't actually have an SAGV.
*/
- if (ret == -ENXIO) {
+ if (IS_SKYLAKE(dev_priv) && ret == -ENXIO) {
DRM_DEBUG_DRIVER("No SAGV found on system, ignoring\n");
- dev_priv->skl_sagv_status = I915_SKL_SAGV_NOT_CONTROLLED;
+ dev_priv->sagv_status = I915_SAGV_NOT_CONTROLLED;
return 0;
} else if (ret < 0) {
DRM_ERROR("Failed to enable the SAGV\n");
return ret;
}
- dev_priv->skl_sagv_status = I915_SKL_SAGV_ENABLED;
+ dev_priv->sagv_status = I915_SAGV_ENABLED;
return 0;
}
static int
-skl_do_sagv_disable(struct drm_i915_private *dev_priv)
+intel_do_sagv_disable(struct drm_i915_private *dev_priv)
{
int ret;
uint32_t temp = GEN9_SAGV_DISABLE;
}
int
-skl_disable_sagv(struct drm_i915_private *dev_priv)
+intel_disable_sagv(struct drm_i915_private *dev_priv)
{
int ret, result;
- if (dev_priv->skl_sagv_status == I915_SKL_SAGV_NOT_CONTROLLED ||
- dev_priv->skl_sagv_status == I915_SKL_SAGV_DISABLED)
+ if (!intel_has_sagv(dev_priv))
+ return 0;
+
+ if (dev_priv->sagv_status == I915_SAGV_DISABLED)
return 0;
DRM_DEBUG_KMS("Disabling the SAGV\n");
mutex_lock(&dev_priv->rps.hw_lock);
/* bspec says to keep retrying for at least 1 ms */
- ret = wait_for(result = skl_do_sagv_disable(dev_priv), 1);
+ ret = wait_for(result = intel_do_sagv_disable(dev_priv), 1);
mutex_unlock(&dev_priv->rps.hw_lock);
if (ret == -ETIMEDOUT) {
* Some skl systems, pre-release machines in particular,
* don't actually have an SAGV.
*/
- if (result == -ENXIO) {
+ if (IS_SKYLAKE(dev_priv) && result == -ENXIO) {
DRM_DEBUG_DRIVER("No SAGV found on system, ignoring\n");
- dev_priv->skl_sagv_status = I915_SKL_SAGV_NOT_CONTROLLED;
+ dev_priv->sagv_status = I915_SAGV_NOT_CONTROLLED;
return 0;
} else if (result < 0) {
DRM_ERROR("Failed to disable the SAGV\n");
return result;
}
- dev_priv->skl_sagv_status = I915_SKL_SAGV_DISABLED;
+ dev_priv->sagv_status = I915_SAGV_DISABLED;
return 0;
}
-bool skl_can_enable_sagv(struct drm_atomic_state *state)
+bool intel_can_enable_sagv(struct drm_atomic_state *state)
{
struct drm_device *dev = state->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
- struct drm_crtc *crtc;
+ struct intel_crtc *crtc;
+ struct intel_plane *plane;
+ struct intel_crtc_state *cstate;
+ struct skl_plane_wm *wm;
enum pipe pipe;
- int level, plane;
+ int level, latency;
+
+ if (!intel_has_sagv(dev_priv))
+ return false;
/*
* SKL workaround: bspec recommends we disable the SAGV when we have
/* Since we're now guaranteed to only have one active CRTC... */
pipe = ffs(intel_state->active_crtcs) - 1;
- crtc = dev_priv->pipe_to_crtc_mapping[pipe];
+ crtc = to_intel_crtc(dev_priv->pipe_to_crtc_mapping[pipe]);
+ cstate = to_intel_crtc_state(crtc->base.state);
- if (crtc->state->mode.flags & DRM_MODE_FLAG_INTERLACE)
+ if (crtc->base.state->adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
return false;
- for_each_plane(dev_priv, pipe, plane) {
+ for_each_intel_plane_on_crtc(dev, crtc, plane) {
+ wm = &cstate->wm.skl.optimal.planes[skl_wm_plane_id(plane)];
+
/* Skip this plane if it's not enabled */
- if (intel_state->wm_results.plane[pipe][plane][0] == 0)
+ if (!wm->wm[0].plane_en)
continue;
/* Find the highest enabled wm level for this plane */
- for (level = ilk_wm_max_level(dev);
- intel_state->wm_results.plane[pipe][plane][level] == 0; --level)
+ for (level = ilk_wm_max_level(dev_priv);
+ !wm->wm[level].plane_en; --level)
{ }
+ latency = dev_priv->wm.skl_latency[level];
+
+ if (skl_needs_memory_bw_wa(intel_state) &&
+ plane->base.state->fb->modifier[0] ==
+ I915_FORMAT_MOD_X_TILED)
+ latency += 15;
+
/*
* If any of the planes on this pipe don't enable wm levels
* that incur memory latencies higher then 30µs we can't enable
* the SAGV
*/
- if (dev_priv->wm.skl_latency[level] < SKL_SAGV_BLOCK_TIME)
+ if (latency < SKL_SAGV_BLOCK_TIME)
return false;
}
struct drm_crtc *for_crtc = cstate->base.crtc;
unsigned int pipe_size, ddb_size;
int nth_active_pipe;
- int pipe = to_intel_crtc(for_crtc)->pipe;
if (WARN_ON(!state) || !cstate->base.active) {
alloc->start = 0;
else
*num_active = hweight32(dev_priv->active_crtcs);
- if (IS_BROXTON(dev))
- ddb_size = BXT_DDB_SIZE;
- else
- ddb_size = SKL_DDB_SIZE;
+ ddb_size = INTEL_INFO(dev_priv)->ddb_size;
+ WARN_ON(ddb_size == 0);
ddb_size -= 4; /* 4 blocks for bypass path allocation */
* we currently hold.
*/
if (!intel_state->active_pipe_changes) {
- *alloc = dev_priv->wm.skl_hw.ddb.pipe[pipe];
+ *alloc = to_intel_crtc(for_crtc)->hw_ddb;
return;
}
uint32_t downscale_h, downscale_w;
uint32_t src_w, src_h, dst_w, dst_h;
- if (WARN_ON(!pstate->visible))
+ if (WARN_ON(!pstate->base.visible))
return DRM_PLANE_HELPER_NO_SCALING;
/* n.b., src is 16.16 fixed point, dst is whole integer */
- src_w = drm_rect_width(&pstate->src);
- src_h = drm_rect_height(&pstate->src);
- dst_w = drm_rect_width(&pstate->dst);
- dst_h = drm_rect_height(&pstate->dst);
- if (intel_rotation_90_or_270(pstate->base.rotation))
+ src_w = drm_rect_width(&pstate->base.src);
+ src_h = drm_rect_height(&pstate->base.src);
+ dst_w = drm_rect_width(&pstate->base.dst);
+ dst_h = drm_rect_height(&pstate->base.dst);
+ if (drm_rotation_90_or_270(pstate->base.rotation))
swap(dst_w, dst_h);
downscale_h = max(src_h / dst_h, (uint32_t)DRM_PLANE_HELPER_NO_SCALING);
uint32_t width = 0, height = 0;
unsigned format = fb ? fb->pixel_format : DRM_FORMAT_XRGB8888;
- if (!intel_pstate->visible)
+ if (!intel_pstate->base.visible)
return 0;
if (pstate->plane->type == DRM_PLANE_TYPE_CURSOR)
return 0;
if (y && format != DRM_FORMAT_NV12)
return 0;
- width = drm_rect_width(&intel_pstate->src) >> 16;
- height = drm_rect_height(&intel_pstate->src) >> 16;
+ width = drm_rect_width(&intel_pstate->base.src) >> 16;
+ height = drm_rect_height(&intel_pstate->base.src) >> 16;
- if (intel_rotation_90_or_270(pstate->rotation))
+ if (drm_rotation_90_or_270(pstate->rotation))
swap(width, height);
/* for planar format */
fb->modifier[0] != I915_FORMAT_MOD_Yf_TILED)
return 8;
- src_w = drm_rect_width(&intel_pstate->src) >> 16;
- src_h = drm_rect_height(&intel_pstate->src) >> 16;
+ src_w = drm_rect_width(&intel_pstate->base.src) >> 16;
+ src_h = drm_rect_height(&intel_pstate->base.src) >> 16;
- if (intel_rotation_90_or_270(pstate->rotation))
+ if (drm_rotation_90_or_270(pstate->rotation))
swap(src_w, src_h);
/* Halve UV plane width and height for NV12 */
else
plane_bpp = drm_format_plane_cpp(fb->pixel_format, 0);
- if (intel_rotation_90_or_270(pstate->rotation)) {
+ if (drm_rotation_90_or_270(pstate->rotation)) {
switch (plane_bpp) {
case 1:
min_scanlines = 32;
struct drm_plane *plane;
struct drm_plane_state *pstate;
enum pipe pipe = intel_crtc->pipe;
- struct skl_ddb_entry *alloc = &ddb->pipe[pipe];
+ struct skl_ddb_entry *alloc = &cstate->wm.skl.ddb;
uint16_t alloc_size, start, cursor_blocks;
uint16_t *minimum = cstate->wm.skl.minimum_blocks;
uint16_t *y_minimum = cstate->wm.skl.minimum_y_blocks;
int num_active;
int id, i;
+ /* Clear the partitioning for disabled planes. */
+ memset(ddb->plane[pipe], 0, sizeof(ddb->plane[pipe]));
+ memset(ddb->y_plane[pipe], 0, sizeof(ddb->y_plane[pipe]));
+
if (WARN_ON(!state))
return 0;
if (!cstate->base.active) {
- ddb->pipe[pipe].start = ddb->pipe[pipe].end = 0;
- memset(ddb->plane[pipe], 0, sizeof(ddb->plane[pipe]));
- memset(ddb->y_plane[pipe], 0, sizeof(ddb->y_plane[pipe]));
+ alloc->start = alloc->end = 0;
return 0;
}
if (intel_plane->pipe != pipe)
continue;
- if (!to_intel_plane_state(pstate)->visible) {
+ if (!to_intel_plane_state(pstate)->base.visible) {
minimum[id] = 0;
y_minimum[id] = 0;
continue;
return 0;
}
-static uint32_t skl_pipe_pixel_rate(const struct intel_crtc_state *config)
-{
- /* TODO: Take into account the scalers once we support them */
- return config->base.adjusted_mode.crtc_clock;
-}
-
/*
* The max latency should be 257 (max the punit can code is 255 and we add 2us
* for the read latency) and cpp should always be <= 8, so that
}
static uint32_t skl_wm_method2(uint32_t pixel_rate, uint32_t pipe_htotal,
- uint32_t horiz_pixels, uint8_t cpp,
- uint64_t tiling, uint32_t latency)
+ uint32_t latency, uint32_t plane_blocks_per_line)
{
uint32_t ret;
- uint32_t plane_bytes_per_line, plane_blocks_per_line;
uint32_t wm_intermediate_val;
if (latency == 0)
return UINT_MAX;
- plane_bytes_per_line = horiz_pixels * cpp;
-
- if (tiling == I915_FORMAT_MOD_Y_TILED ||
- tiling == I915_FORMAT_MOD_Yf_TILED) {
- plane_bytes_per_line *= 4;
- plane_blocks_per_line = DIV_ROUND_UP(plane_bytes_per_line, 512);
- plane_blocks_per_line /= 4;
- } else if (tiling == DRM_FORMAT_MOD_NONE) {
- plane_blocks_per_line = DIV_ROUND_UP(plane_bytes_per_line, 512) + 1;
- } else {
- plane_blocks_per_line = DIV_ROUND_UP(plane_bytes_per_line, 512);
- }
-
wm_intermediate_val = latency * pixel_rate;
ret = DIV_ROUND_UP(wm_intermediate_val, pipe_htotal * 1000) *
plane_blocks_per_line;
uint64_t pixel_rate;
/* Shouldn't reach here on disabled planes... */
- if (WARN_ON(!pstate->visible))
+ if (WARN_ON(!pstate->base.visible))
return 0;
/*
* Adjusted plane pixel rate is just the pipe's adjusted pixel rate
* with additional adjustments for plane-specific scaling.
*/
- adjusted_pixel_rate = skl_pipe_pixel_rate(cstate);
+ adjusted_pixel_rate = ilk_pipe_pixel_rate(cstate);
downscale_amount = skl_plane_downscale_amount(pstate);
pixel_rate = adjusted_pixel_rate * downscale_amount >> 16;
uint8_t cpp;
uint32_t width = 0, height = 0;
uint32_t plane_pixel_rate;
+ uint32_t y_tile_minimum, y_min_scanlines;
+ struct intel_atomic_state *state =
+ to_intel_atomic_state(cstate->base.state);
+ bool apply_memory_bw_wa = skl_needs_memory_bw_wa(state);
- if (latency == 0 || !cstate->base.active || !intel_pstate->visible) {
+ if (latency == 0 || !cstate->base.active || !intel_pstate->base.visible) {
*enabled = false;
return 0;
}
- width = drm_rect_width(&intel_pstate->src) >> 16;
- height = drm_rect_height(&intel_pstate->src) >> 16;
+ if (apply_memory_bw_wa && fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
+ latency += 15;
- if (intel_rotation_90_or_270(pstate->rotation))
+ width = drm_rect_width(&intel_pstate->base.src) >> 16;
+ height = drm_rect_height(&intel_pstate->base.src) >> 16;
+
+ if (drm_rotation_90_or_270(pstate->rotation))
swap(width, height);
cpp = drm_format_plane_cpp(fb->pixel_format, 0);
plane_pixel_rate = skl_adjusted_plane_pixel_rate(cstate, intel_pstate);
+ if (drm_rotation_90_or_270(pstate->rotation)) {
+ int cpp = (fb->pixel_format == DRM_FORMAT_NV12) ?
+ drm_format_plane_cpp(fb->pixel_format, 1) :
+ drm_format_plane_cpp(fb->pixel_format, 0);
+
+ switch (cpp) {
+ case 1:
+ y_min_scanlines = 16;
+ break;
+ case 2:
+ y_min_scanlines = 8;
+ break;
+ case 4:
+ y_min_scanlines = 4;
+ break;
+ default:
+ MISSING_CASE(cpp);
+ return -EINVAL;
+ }
+ } else {
+ y_min_scanlines = 4;
+ }
+
+ plane_bytes_per_line = width * cpp;
+ if (fb->modifier[0] == I915_FORMAT_MOD_Y_TILED ||
+ fb->modifier[0] == I915_FORMAT_MOD_Yf_TILED) {
+ plane_blocks_per_line =
+ DIV_ROUND_UP(plane_bytes_per_line * y_min_scanlines, 512);
+ plane_blocks_per_line /= y_min_scanlines;
+ } else if (fb->modifier[0] == DRM_FORMAT_MOD_NONE) {
+ plane_blocks_per_line = DIV_ROUND_UP(plane_bytes_per_line, 512)
+ + 1;
+ } else {
+ plane_blocks_per_line = DIV_ROUND_UP(plane_bytes_per_line, 512);
+ }
+
method1 = skl_wm_method1(plane_pixel_rate, cpp, latency);
method2 = skl_wm_method2(plane_pixel_rate,
cstate->base.adjusted_mode.crtc_htotal,
- width,
- cpp,
- fb->modifier[0],
- latency);
+ latency,
+ plane_blocks_per_line);
- plane_bytes_per_line = width * cpp;
- plane_blocks_per_line = DIV_ROUND_UP(plane_bytes_per_line, 512);
+ y_tile_minimum = plane_blocks_per_line * y_min_scanlines;
+ if (apply_memory_bw_wa)
+ y_tile_minimum *= 2;
if (fb->modifier[0] == I915_FORMAT_MOD_Y_TILED ||
fb->modifier[0] == I915_FORMAT_MOD_Yf_TILED) {
- uint32_t min_scanlines = 4;
- uint32_t y_tile_minimum;
- if (intel_rotation_90_or_270(pstate->rotation)) {
- int cpp = (fb->pixel_format == DRM_FORMAT_NV12) ?
- drm_format_plane_cpp(fb->pixel_format, 1) :
- drm_format_plane_cpp(fb->pixel_format, 0);
-
- switch (cpp) {
- case 1:
- min_scanlines = 16;
- break;
- case 2:
- min_scanlines = 8;
- break;
- case 8:
- WARN(1, "Unsupported pixel depth for rotation");
- }
- }
- y_tile_minimum = plane_blocks_per_line * min_scanlines;
selected_result = max(method2, y_tile_minimum);
} else {
- if ((ddb_allocation / plane_blocks_per_line) >= 1)
+ if ((cpp * cstate->base.adjusted_mode.crtc_htotal / 512 < 1) &&
+ (plane_bytes_per_line / 512 < 1))
+ selected_result = method2;
+ else if ((ddb_allocation / plane_blocks_per_line) >= 1)
selected_result = min(method1, method2);
else
selected_result = method1;
if (level >= 1 && level <= 7) {
if (fb->modifier[0] == I915_FORMAT_MOD_Y_TILED ||
- fb->modifier[0] == I915_FORMAT_MOD_Yf_TILED)
- res_lines += 4;
- else
+ fb->modifier[0] == I915_FORMAT_MOD_Yf_TILED) {
+ res_blocks += y_tile_minimum;
+ res_lines += y_min_scanlines;
+ } else {
res_blocks++;
+ }
}
if (res_blocks >= ddb_allocation || res_lines > 31) {
skl_compute_wm_level(const struct drm_i915_private *dev_priv,
struct skl_ddb_allocation *ddb,
struct intel_crtc_state *cstate,
+ struct intel_plane *intel_plane,
int level,
struct skl_wm_level *result)
{
struct drm_atomic_state *state = cstate->base.state;
struct intel_crtc *intel_crtc = to_intel_crtc(cstate->base.crtc);
- struct drm_plane *plane;
- struct intel_plane *intel_plane;
- struct intel_plane_state *intel_pstate;
+ struct drm_plane *plane = &intel_plane->base;
+ struct intel_plane_state *intel_pstate = NULL;
uint16_t ddb_blocks;
enum pipe pipe = intel_crtc->pipe;
int ret;
+ int i = skl_wm_plane_id(intel_plane);
+
+ if (state)
+ intel_pstate =
+ intel_atomic_get_existing_plane_state(state,
+ intel_plane);
/*
- * We'll only calculate watermarks for planes that are actually
- * enabled, so make sure all other planes are set as disabled.
+ * Note: If we start supporting multiple pending atomic commits against
+ * the same planes/CRTC's in the future, plane->state will no longer be
+ * the correct pre-state to use for the calculations here and we'll
+ * need to change where we get the 'unchanged' plane data from.
+ *
+ * For now this is fine because we only allow one queued commit against
+ * a CRTC. Even if the plane isn't modified by this transaction and we
+ * don't have a plane lock, we still have the CRTC's lock, so we know
+ * that no other transactions are racing with us to update it.
*/
- memset(result, 0, sizeof(*result));
-
- for_each_intel_plane_mask(&dev_priv->drm,
- intel_plane,
- cstate->base.plane_mask) {
- int i = skl_wm_plane_id(intel_plane);
+ if (!intel_pstate)
+ intel_pstate = to_intel_plane_state(plane->state);
- plane = &intel_plane->base;
- intel_pstate = NULL;
- if (state)
- intel_pstate =
- intel_atomic_get_existing_plane_state(state,
- intel_plane);
+ WARN_ON(!intel_pstate->base.fb);
- /*
- * Note: If we start supporting multiple pending atomic commits
- * against the same planes/CRTC's in the future, plane->state
- * will no longer be the correct pre-state to use for the
- * calculations here and we'll need to change where we get the
- * 'unchanged' plane data from.
- *
- * For now this is fine because we only allow one queued commit
- * against a CRTC. Even if the plane isn't modified by this
- * transaction and we don't have a plane lock, we still have
- * the CRTC's lock, so we know that no other transactions are
- * racing with us to update it.
- */
- if (!intel_pstate)
- intel_pstate = to_intel_plane_state(plane->state);
-
- WARN_ON(!intel_pstate->base.fb);
+ ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][i]);
- ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][i]);
-
- ret = skl_compute_plane_wm(dev_priv,
- cstate,
- intel_pstate,
- ddb_blocks,
- level,
- &result->plane_res_b[i],
- &result->plane_res_l[i],
- &result->plane_en[i]);
- if (ret)
- return ret;
- }
+ ret = skl_compute_plane_wm(dev_priv,
+ cstate,
+ intel_pstate,
+ ddb_blocks,
+ level,
+ &result->plane_res_b,
+ &result->plane_res_l,
+ &result->plane_en);
+ if (ret)
+ return ret;
return 0;
}
static uint32_t
skl_compute_linetime_wm(struct intel_crtc_state *cstate)
{
+ uint32_t pixel_rate;
+
if (!cstate->base.active)
return 0;
- if (WARN_ON(skl_pipe_pixel_rate(cstate) == 0))
+ pixel_rate = ilk_pipe_pixel_rate(cstate);
+
+ if (WARN_ON(pixel_rate == 0))
return 0;
return DIV_ROUND_UP(8 * cstate->base.adjusted_mode.crtc_htotal * 1000,
- skl_pipe_pixel_rate(cstate));
+ pixel_rate);
}
static void skl_compute_transition_wm(struct intel_crtc_state *cstate,
struct skl_wm_level *trans_wm /* out */)
{
- struct drm_crtc *crtc = cstate->base.crtc;
- struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_plane *intel_plane;
-
if (!cstate->base.active)
return;
/* Until we know more, just disable transition WMs */
- for_each_intel_plane_on_crtc(crtc->dev, intel_crtc, intel_plane) {
- int i = skl_wm_plane_id(intel_plane);
-
- trans_wm->plane_en[i] = false;
- }
+ trans_wm->plane_en = false;
}
static int skl_build_pipe_wm(struct intel_crtc_state *cstate,
{
struct drm_device *dev = cstate->base.crtc->dev;
const struct drm_i915_private *dev_priv = to_i915(dev);
- int level, max_level = ilk_wm_max_level(dev);
+ struct intel_plane *intel_plane;
+ struct skl_plane_wm *wm;
+ int level, max_level = ilk_wm_max_level(dev_priv);
int ret;
- for (level = 0; level <= max_level; level++) {
- ret = skl_compute_wm_level(dev_priv, ddb, cstate,
- level, &pipe_wm->wm[level]);
- if (ret)
- return ret;
- }
- pipe_wm->linetime = skl_compute_linetime_wm(cstate);
-
- skl_compute_transition_wm(cstate, &pipe_wm->trans_wm);
-
- return 0;
-}
-
-static void skl_compute_wm_results(struct drm_device *dev,
- struct skl_pipe_wm *p_wm,
- struct skl_wm_values *r,
- struct intel_crtc *intel_crtc)
-{
- int level, max_level = ilk_wm_max_level(dev);
- enum pipe pipe = intel_crtc->pipe;
- uint32_t temp;
- int i;
-
- for (level = 0; level <= max_level; level++) {
- for (i = 0; i < intel_num_planes(intel_crtc); i++) {
- temp = 0;
+ /*
+ * We'll only calculate watermarks for planes that are actually
+ * enabled, so make sure all other planes are set as disabled.
+ */
+ memset(pipe_wm->planes, 0, sizeof(pipe_wm->planes));
- temp |= p_wm->wm[level].plane_res_l[i] <<
- PLANE_WM_LINES_SHIFT;
- temp |= p_wm->wm[level].plane_res_b[i];
- if (p_wm->wm[level].plane_en[i])
- temp |= PLANE_WM_EN;
+ for_each_intel_plane_mask(&dev_priv->drm,
+ intel_plane,
+ cstate->base.plane_mask) {
+ wm = &pipe_wm->planes[skl_wm_plane_id(intel_plane)];
- r->plane[pipe][i][level] = temp;
+ for (level = 0; level <= max_level; level++) {
+ ret = skl_compute_wm_level(dev_priv, ddb, cstate,
+ intel_plane, level,
+ &wm->wm[level]);
+ if (ret)
+ return ret;
}
-
- temp = 0;
-
- temp |= p_wm->wm[level].plane_res_l[PLANE_CURSOR] << PLANE_WM_LINES_SHIFT;
- temp |= p_wm->wm[level].plane_res_b[PLANE_CURSOR];
-
- if (p_wm->wm[level].plane_en[PLANE_CURSOR])
- temp |= PLANE_WM_EN;
-
- r->plane[pipe][PLANE_CURSOR][level] = temp;
-
+ skl_compute_transition_wm(cstate, &wm->trans_wm);
}
+ pipe_wm->linetime = skl_compute_linetime_wm(cstate);
- /* transition WMs */
- for (i = 0; i < intel_num_planes(intel_crtc); i++) {
- temp = 0;
- temp |= p_wm->trans_wm.plane_res_l[i] << PLANE_WM_LINES_SHIFT;
- temp |= p_wm->trans_wm.plane_res_b[i];
- if (p_wm->trans_wm.plane_en[i])
- temp |= PLANE_WM_EN;
-
- r->plane_trans[pipe][i] = temp;
- }
-
- temp = 0;
- temp |= p_wm->trans_wm.plane_res_l[PLANE_CURSOR] << PLANE_WM_LINES_SHIFT;
- temp |= p_wm->trans_wm.plane_res_b[PLANE_CURSOR];
- if (p_wm->trans_wm.plane_en[PLANE_CURSOR])
- temp |= PLANE_WM_EN;
-
- r->plane_trans[pipe][PLANE_CURSOR] = temp;
-
- r->wm_linetime[pipe] = p_wm->linetime;
+ return 0;
}
static void skl_ddb_entry_write(struct drm_i915_private *dev_priv,
I915_WRITE(reg, 0);
}
-static void skl_write_wm_values(struct drm_i915_private *dev_priv,
- const struct skl_wm_values *new)
+static void skl_write_wm_level(struct drm_i915_private *dev_priv,
+ i915_reg_t reg,
+ const struct skl_wm_level *level)
{
- struct drm_device *dev = &dev_priv->drm;
- struct intel_crtc *crtc;
-
- for_each_intel_crtc(dev, crtc) {
- int i, level, max_level = ilk_wm_max_level(dev);
- enum pipe pipe = crtc->pipe;
+ uint32_t val = 0;
- if ((new->dirty_pipes & drm_crtc_mask(&crtc->base)) == 0)
- continue;
- if (!crtc->active)
- continue;
-
- I915_WRITE(PIPE_WM_LINETIME(pipe), new->wm_linetime[pipe]);
-
- for (level = 0; level <= max_level; level++) {
- for (i = 0; i < intel_num_planes(crtc); i++)
- I915_WRITE(PLANE_WM(pipe, i, level),
- new->plane[pipe][i][level]);
- I915_WRITE(CUR_WM(pipe, level),
- new->plane[pipe][PLANE_CURSOR][level]);
- }
- for (i = 0; i < intel_num_planes(crtc); i++)
- I915_WRITE(PLANE_WM_TRANS(pipe, i),
- new->plane_trans[pipe][i]);
- I915_WRITE(CUR_WM_TRANS(pipe),
- new->plane_trans[pipe][PLANE_CURSOR]);
-
- for (i = 0; i < intel_num_planes(crtc); i++) {
- skl_ddb_entry_write(dev_priv,
- PLANE_BUF_CFG(pipe, i),
- &new->ddb.plane[pipe][i]);
- skl_ddb_entry_write(dev_priv,
- PLANE_NV12_BUF_CFG(pipe, i),
- &new->ddb.y_plane[pipe][i]);
- }
-
- skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe),
- &new->ddb.plane[pipe][PLANE_CURSOR]);
+ if (level->plane_en) {
+ val |= PLANE_WM_EN;
+ val |= level->plane_res_b;
+ val |= level->plane_res_l << PLANE_WM_LINES_SHIFT;
}
-}
-/*
- * When setting up a new DDB allocation arrangement, we need to correctly
- * sequence the times at which the new allocations for the pipes are taken into
- * account or we'll have pipes fetching from space previously allocated to
- * another pipe.
- *
- * Roughly the sequence looks like:
- * 1. re-allocate the pipe(s) with the allocation being reduced and not
- * overlapping with a previous light-up pipe (another way to put it is:
- * pipes with their new allocation strickly included into their old ones).
- * 2. re-allocate the other pipes that get their allocation reduced
- * 3. allocate the pipes having their allocation increased
- *
- * Steps 1. and 2. are here to take care of the following case:
- * - Initially DDB looks like this:
- * | B | C |
- * - enable pipe A.
- * - pipe B has a reduced DDB allocation that overlaps with the old pipe C
- * allocation
- * | A | B | C |
- *
- * We need to sequence the re-allocation: C, B, A (and not B, C, A).
- */
+ I915_WRITE(reg, val);
+}
-static void
-skl_wm_flush_pipe(struct drm_i915_private *dev_priv, enum pipe pipe, int pass)
+void skl_write_plane_wm(struct intel_crtc *intel_crtc,
+ const struct skl_plane_wm *wm,
+ const struct skl_ddb_allocation *ddb,
+ int plane)
{
- int plane;
-
- DRM_DEBUG_KMS("flush pipe %c (pass %d)\n", pipe_name(pipe), pass);
+ struct drm_crtc *crtc = &intel_crtc->base;
+ struct drm_device *dev = crtc->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ int level, max_level = ilk_wm_max_level(dev_priv);
+ enum pipe pipe = intel_crtc->pipe;
- for_each_plane(dev_priv, pipe, plane) {
- I915_WRITE(PLANE_SURF(pipe, plane),
- I915_READ(PLANE_SURF(pipe, plane)));
+ for (level = 0; level <= max_level; level++) {
+ skl_write_wm_level(dev_priv, PLANE_WM(pipe, plane, level),
+ &wm->wm[level]);
}
- I915_WRITE(CURBASE(pipe), I915_READ(CURBASE(pipe)));
+ skl_write_wm_level(dev_priv, PLANE_WM_TRANS(pipe, plane),
+ &wm->trans_wm);
+
+ skl_ddb_entry_write(dev_priv, PLANE_BUF_CFG(pipe, plane),
+ &ddb->plane[pipe][plane]);
+ skl_ddb_entry_write(dev_priv, PLANE_NV12_BUF_CFG(pipe, plane),
+ &ddb->y_plane[pipe][plane]);
}
-static bool
-skl_ddb_allocation_included(const struct skl_ddb_allocation *old,
- const struct skl_ddb_allocation *new,
- enum pipe pipe)
+void skl_write_cursor_wm(struct intel_crtc *intel_crtc,
+ const struct skl_plane_wm *wm,
+ const struct skl_ddb_allocation *ddb)
{
- uint16_t old_size, new_size;
+ struct drm_crtc *crtc = &intel_crtc->base;
+ struct drm_device *dev = crtc->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ int level, max_level = ilk_wm_max_level(dev_priv);
+ enum pipe pipe = intel_crtc->pipe;
- old_size = skl_ddb_entry_size(&old->pipe[pipe]);
- new_size = skl_ddb_entry_size(&new->pipe[pipe]);
+ for (level = 0; level <= max_level; level++) {
+ skl_write_wm_level(dev_priv, CUR_WM(pipe, level),
+ &wm->wm[level]);
+ }
+ skl_write_wm_level(dev_priv, CUR_WM_TRANS(pipe), &wm->trans_wm);
- return old_size != new_size &&
- new->pipe[pipe].start >= old->pipe[pipe].start &&
- new->pipe[pipe].end <= old->pipe[pipe].end;
+ skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe),
+ &ddb->plane[pipe][PLANE_CURSOR]);
}
-static void skl_flush_wm_values(struct drm_i915_private *dev_priv,
- struct skl_wm_values *new_values)
+bool skl_wm_level_equals(const struct skl_wm_level *l1,
+ const struct skl_wm_level *l2)
{
- struct drm_device *dev = &dev_priv->drm;
- struct skl_ddb_allocation *cur_ddb, *new_ddb;
- bool reallocated[I915_MAX_PIPES] = {};
- struct intel_crtc *crtc;
- enum pipe pipe;
-
- new_ddb = &new_values->ddb;
- cur_ddb = &dev_priv->wm.skl_hw.ddb;
-
- /*
- * First pass: flush the pipes with the new allocation contained into
- * the old space.
- *
- * We'll wait for the vblank on those pipes to ensure we can safely
- * re-allocate the freed space without this pipe fetching from it.
- */
- for_each_intel_crtc(dev, crtc) {
- if (!crtc->active)
- continue;
-
- pipe = crtc->pipe;
-
- if (!skl_ddb_allocation_included(cur_ddb, new_ddb, pipe))
- continue;
+ if (l1->plane_en != l2->plane_en)
+ return false;
- skl_wm_flush_pipe(dev_priv, pipe, 1);
- intel_wait_for_vblank(dev, pipe);
+ /* If both planes aren't enabled, the rest shouldn't matter */
+ if (!l1->plane_en)
+ return true;
- reallocated[pipe] = true;
- }
+ return (l1->plane_res_l == l2->plane_res_l &&
+ l1->plane_res_b == l2->plane_res_b);
+}
+static inline bool skl_ddb_entries_overlap(const struct skl_ddb_entry *a,
+ const struct skl_ddb_entry *b)
+{
+ return a->start < b->end && b->start < a->end;
+}
- /*
- * Second pass: flush the pipes that are having their allocation
- * reduced, but overlapping with a previous allocation.
- *
- * Here as well we need to wait for the vblank to make sure the freed
- * space is not used anymore.
- */
- for_each_intel_crtc(dev, crtc) {
- if (!crtc->active)
- continue;
+bool skl_ddb_allocation_overlaps(struct drm_atomic_state *state,
+ struct intel_crtc *intel_crtc)
+{
+ struct drm_crtc *other_crtc;
+ struct drm_crtc_state *other_cstate;
+ struct intel_crtc *other_intel_crtc;
+ const struct skl_ddb_entry *ddb =
+ &to_intel_crtc_state(intel_crtc->base.state)->wm.skl.ddb;
+ int i;
- pipe = crtc->pipe;
+ for_each_crtc_in_state(state, other_crtc, other_cstate, i) {
+ other_intel_crtc = to_intel_crtc(other_crtc);
- if (reallocated[pipe])
+ if (other_intel_crtc == intel_crtc)
continue;
- if (skl_ddb_entry_size(&new_ddb->pipe[pipe]) <
- skl_ddb_entry_size(&cur_ddb->pipe[pipe])) {
- skl_wm_flush_pipe(dev_priv, pipe, 2);
- intel_wait_for_vblank(dev, pipe);
- reallocated[pipe] = true;
- }
+ if (skl_ddb_entries_overlap(ddb, &other_intel_crtc->hw_ddb))
+ return true;
}
- /*
- * Third pass: flush the pipes that got more space allocated.
- *
- * We don't need to actively wait for the update here, next vblank
- * will just get more DDB space with the correct WM values.
- */
- for_each_intel_crtc(dev, crtc) {
- if (!crtc->active)
- continue;
-
- pipe = crtc->pipe;
-
- /*
- * At this point, only the pipes more space than before are
- * left to re-allocate.
- */
- if (reallocated[pipe])
- continue;
-
- skl_wm_flush_pipe(dev_priv, pipe, 3);
- }
+ return false;
}
static int skl_update_pipe_wm(struct drm_crtc_state *cstate,
return ret;
}
+static int
+skl_ddb_add_affected_planes(struct intel_crtc_state *cstate)
+{
+ struct drm_atomic_state *state = cstate->base.state;
+ struct drm_device *dev = state->dev;
+ struct drm_crtc *crtc = cstate->base.crtc;
+ struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
+ struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb;
+ struct skl_ddb_allocation *cur_ddb = &dev_priv->wm.skl_hw.ddb;
+ struct drm_plane_state *plane_state;
+ struct drm_plane *plane;
+ enum pipe pipe = intel_crtc->pipe;
+ int id;
+
+ WARN_ON(!drm_atomic_get_existing_crtc_state(state, crtc));
+
+ drm_for_each_plane_mask(plane, dev, crtc->state->plane_mask) {
+ id = skl_wm_plane_id(to_intel_plane(plane));
+
+ if (skl_ddb_entry_equal(&cur_ddb->plane[pipe][id],
+ &new_ddb->plane[pipe][id]) &&
+ skl_ddb_entry_equal(&cur_ddb->y_plane[pipe][id],
+ &new_ddb->y_plane[pipe][id]))
+ continue;
+
+ plane_state = drm_atomic_get_plane_state(state, plane);
+ if (IS_ERR(plane_state))
+ return PTR_ERR(plane_state);
+ }
+
+ return 0;
+}
+
static int
skl_compute_ddb(struct drm_atomic_state *state)
{
intel_state->wm_results.dirty_pipes = ~0;
}
+ /*
+ * We're not recomputing for the pipes not included in the commit, so
+ * make sure we start with the current state.
+ */
+ memcpy(ddb, &dev_priv->wm.skl_hw.ddb, sizeof(*ddb));
+
for_each_intel_crtc_mask(dev, intel_crtc, realloc_pipes) {
struct intel_crtc_state *cstate;
if (ret)
return ret;
- ret = drm_atomic_add_affected_planes(state, &intel_crtc->base);
+ ret = skl_ddb_add_affected_planes(cstate);
if (ret)
return ret;
}
struct skl_wm_values *src,
enum pipe pipe)
{
- dst->wm_linetime[pipe] = src->wm_linetime[pipe];
- memcpy(dst->plane[pipe], src->plane[pipe],
- sizeof(dst->plane[pipe]));
- memcpy(dst->plane_trans[pipe], src->plane_trans[pipe],
- sizeof(dst->plane_trans[pipe]));
-
- dst->ddb.pipe[pipe] = src->ddb.pipe[pipe];
memcpy(dst->ddb.y_plane[pipe], src->ddb.y_plane[pipe],
sizeof(dst->ddb.y_plane[pipe]));
memcpy(dst->ddb.plane[pipe], src->ddb.plane[pipe],
sizeof(dst->ddb.plane[pipe]));
}
+static void
+skl_print_wm_changes(const struct drm_atomic_state *state)
+{
+ const struct drm_device *dev = state->dev;
+ const struct drm_i915_private *dev_priv = to_i915(dev);
+ const struct intel_atomic_state *intel_state =
+ to_intel_atomic_state(state);
+ const struct drm_crtc *crtc;
+ const struct drm_crtc_state *cstate;
+ const struct drm_plane *plane;
+ const struct intel_plane *intel_plane;
+ const struct drm_plane_state *pstate;
+ const struct skl_ddb_allocation *old_ddb = &dev_priv->wm.skl_hw.ddb;
+ const struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb;
+ enum pipe pipe;
+ int id;
+ int i, j;
+
+ for_each_crtc_in_state(state, crtc, cstate, i) {
+ pipe = to_intel_crtc(crtc)->pipe;
+
+ for_each_plane_in_state(state, plane, pstate, j) {
+ const struct skl_ddb_entry *old, *new;
+
+ intel_plane = to_intel_plane(plane);
+ id = skl_wm_plane_id(intel_plane);
+ old = &old_ddb->plane[pipe][id];
+ new = &new_ddb->plane[pipe][id];
+
+ if (intel_plane->pipe != pipe)
+ continue;
+
+ if (skl_ddb_entry_equal(old, new))
+ continue;
+
+ if (id != PLANE_CURSOR) {
+ DRM_DEBUG_ATOMIC("[PLANE:%d:plane %d%c] ddb (%d - %d) -> (%d - %d)\n",
+ plane->base.id, id + 1,
+ pipe_name(pipe),
+ old->start, old->end,
+ new->start, new->end);
+ } else {
+ DRM_DEBUG_ATOMIC("[PLANE:%d:cursor %c] ddb (%d - %d) -> (%d - %d)\n",
+ plane->base.id,
+ pipe_name(pipe),
+ old->start, old->end,
+ new->start, new->end);
+ }
+ }
+ }
+}
+
static int
skl_compute_wm(struct drm_atomic_state *state)
{
* no suitable watermark values can be found.
*/
for_each_crtc_in_state(state, crtc, cstate, i) {
- struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
struct intel_crtc_state *intel_cstate =
to_intel_crtc_state(cstate);
continue;
intel_cstate->update_wm_pre = true;
- skl_compute_wm_results(crtc->dev, pipe_wm, results, intel_crtc);
}
+ skl_print_wm_changes(state);
+
return 0;
}
struct skl_wm_values *hw_vals = &dev_priv->wm.skl_hw;
struct intel_crtc_state *cstate = to_intel_crtc_state(crtc->state);
struct skl_pipe_wm *pipe_wm = &cstate->wm.skl.optimal;
- int pipe;
+ enum pipe pipe = intel_crtc->pipe;
if ((results->dirty_pipes & drm_crtc_mask(crtc)) == 0)
return;
mutex_lock(&dev_priv->wm.wm_mutex);
- skl_write_wm_values(dev_priv, results);
- skl_flush_wm_values(dev_priv, results);
-
/*
- * Store the new configuration (but only for the pipes that have
- * changed; the other values weren't recomputed).
+ * If this pipe isn't active already, we're going to be enabling it
+ * very soon. Since it's safe to update a pipe's ddb allocation while
+ * the pipe's shut off, just do so here. Already active pipes will have
+ * their watermarks updated once we update their planes.
*/
- for_each_pipe_masked(dev_priv, pipe, results->dirty_pipes)
- skl_copy_wm_for_pipe(hw_vals, results, pipe);
+ if (crtc->state->active_changed) {
+ int plane;
+
+ for (plane = 0; plane < intel_num_planes(intel_crtc); plane++)
+ skl_write_plane_wm(intel_crtc, &pipe_wm->planes[plane],
+ &results->ddb, plane);
+
+ skl_write_cursor_wm(intel_crtc, &pipe_wm->planes[PLANE_CURSOR],
+ &results->ddb);
+ }
+
+ skl_copy_wm_for_pipe(hw_vals, results, pipe);
+
+ intel_crtc->hw_ddb = cstate->wm.skl.ddb;
mutex_unlock(&dev_priv->wm.wm_mutex);
}
mutex_unlock(&dev_priv->wm.wm_mutex);
}
-static void skl_pipe_wm_active_state(uint32_t val,
- struct skl_pipe_wm *active,
- bool is_transwm,
- bool is_cursor,
- int i,
- int level)
+static inline void skl_wm_level_from_reg_val(uint32_t val,
+ struct skl_wm_level *level)
{
- bool is_enabled = (val & PLANE_WM_EN) != 0;
-
- if (!is_transwm) {
- if (!is_cursor) {
- active->wm[level].plane_en[i] = is_enabled;
- active->wm[level].plane_res_b[i] =
- val & PLANE_WM_BLOCKS_MASK;
- active->wm[level].plane_res_l[i] =
- (val >> PLANE_WM_LINES_SHIFT) &
- PLANE_WM_LINES_MASK;
- } else {
- active->wm[level].plane_en[PLANE_CURSOR] = is_enabled;
- active->wm[level].plane_res_b[PLANE_CURSOR] =
- val & PLANE_WM_BLOCKS_MASK;
- active->wm[level].plane_res_l[PLANE_CURSOR] =
- (val >> PLANE_WM_LINES_SHIFT) &
- PLANE_WM_LINES_MASK;
- }
- } else {
- if (!is_cursor) {
- active->trans_wm.plane_en[i] = is_enabled;
- active->trans_wm.plane_res_b[i] =
- val & PLANE_WM_BLOCKS_MASK;
- active->trans_wm.plane_res_l[i] =
- (val >> PLANE_WM_LINES_SHIFT) &
- PLANE_WM_LINES_MASK;
- } else {
- active->trans_wm.plane_en[PLANE_CURSOR] = is_enabled;
- active->trans_wm.plane_res_b[PLANE_CURSOR] =
- val & PLANE_WM_BLOCKS_MASK;
- active->trans_wm.plane_res_l[PLANE_CURSOR] =
- (val >> PLANE_WM_LINES_SHIFT) &
- PLANE_WM_LINES_MASK;
- }
- }
+ level->plane_en = val & PLANE_WM_EN;
+ level->plane_res_b = val & PLANE_WM_BLOCKS_MASK;
+ level->plane_res_l = (val >> PLANE_WM_LINES_SHIFT) &
+ PLANE_WM_LINES_MASK;
}
-static void skl_pipe_wm_get_hw_state(struct drm_crtc *crtc)
+void skl_pipe_wm_get_hw_state(struct drm_crtc *crtc,
+ struct skl_pipe_wm *out)
{
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
- struct skl_wm_values *hw = &dev_priv->wm.skl_hw;
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_crtc_state *cstate = to_intel_crtc_state(crtc->state);
- struct skl_pipe_wm *active = &cstate->wm.skl.optimal;
+ struct intel_plane *intel_plane;
+ struct skl_plane_wm *wm;
enum pipe pipe = intel_crtc->pipe;
- int level, i, max_level;
- uint32_t temp;
-
- max_level = ilk_wm_max_level(dev);
-
- hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe));
-
- for (level = 0; level <= max_level; level++) {
- for (i = 0; i < intel_num_planes(intel_crtc); i++)
- hw->plane[pipe][i][level] =
- I915_READ(PLANE_WM(pipe, i, level));
- hw->plane[pipe][PLANE_CURSOR][level] = I915_READ(CUR_WM(pipe, level));
- }
-
- for (i = 0; i < intel_num_planes(intel_crtc); i++)
- hw->plane_trans[pipe][i] = I915_READ(PLANE_WM_TRANS(pipe, i));
- hw->plane_trans[pipe][PLANE_CURSOR] = I915_READ(CUR_WM_TRANS(pipe));
+ int level, id, max_level;
+ uint32_t val;
- if (!intel_crtc->active)
- return;
+ max_level = ilk_wm_max_level(dev_priv);
- hw->dirty_pipes |= drm_crtc_mask(crtc);
+ for_each_intel_plane_on_crtc(dev, intel_crtc, intel_plane) {
+ id = skl_wm_plane_id(intel_plane);
+ wm = &out->planes[id];
- active->linetime = hw->wm_linetime[pipe];
+ for (level = 0; level <= max_level; level++) {
+ if (id != PLANE_CURSOR)
+ val = I915_READ(PLANE_WM(pipe, id, level));
+ else
+ val = I915_READ(CUR_WM(pipe, level));
- for (level = 0; level <= max_level; level++) {
- for (i = 0; i < intel_num_planes(intel_crtc); i++) {
- temp = hw->plane[pipe][i][level];
- skl_pipe_wm_active_state(temp, active, false,
- false, i, level);
+ skl_wm_level_from_reg_val(val, &wm->wm[level]);
}
- temp = hw->plane[pipe][PLANE_CURSOR][level];
- skl_pipe_wm_active_state(temp, active, false, true, i, level);
- }
- for (i = 0; i < intel_num_planes(intel_crtc); i++) {
- temp = hw->plane_trans[pipe][i];
- skl_pipe_wm_active_state(temp, active, true, false, i, 0);
+ if (id != PLANE_CURSOR)
+ val = I915_READ(PLANE_WM_TRANS(pipe, id));
+ else
+ val = I915_READ(CUR_WM_TRANS(pipe));
+
+ skl_wm_level_from_reg_val(val, &wm->trans_wm);
}
- temp = hw->plane_trans[pipe][PLANE_CURSOR];
- skl_pipe_wm_active_state(temp, active, true, true, i, 0);
+ if (!intel_crtc->active)
+ return;
- intel_crtc->wm.active.skl = *active;
+ out->linetime = I915_READ(PIPE_WM_LINETIME(pipe));
}
void skl_wm_get_hw_state(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct skl_wm_values *hw = &dev_priv->wm.skl_hw;
struct skl_ddb_allocation *ddb = &dev_priv->wm.skl_hw.ddb;
struct drm_crtc *crtc;
+ struct intel_crtc *intel_crtc;
+ struct intel_crtc_state *cstate;
skl_ddb_get_hw_state(dev_priv, ddb);
- list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
- skl_pipe_wm_get_hw_state(crtc);
+ list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
+ intel_crtc = to_intel_crtc(crtc);
+ cstate = to_intel_crtc_state(crtc->state);
+
+ skl_pipe_wm_get_hw_state(crtc, &cstate->wm.skl.optimal);
+
+ if (intel_crtc->active) {
+ hw->dirty_pipes |= drm_crtc_mask(crtc);
+ intel_crtc->wm.active.skl = cstate->wm.skl.optimal;
+ }
+ }
if (dev_priv->active_crtcs) {
/* Fully recompute DDB on first atomic commit */
};
hw->wm_pipe[pipe] = I915_READ(wm0_pipe_reg[pipe]);
- if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
hw->wm_linetime[pipe] = I915_READ(PIPE_WM_LINETIME(pipe));
memset(active, 0, sizeof(*active));
active->wm[0].cur_val = tmp & WM0_PIPE_CURSOR_MASK;
active->linetime = hw->wm_linetime[pipe];
} else {
- int level, max_level = ilk_wm_max_level(dev);
+ int level, max_level = ilk_wm_max_level(dev_priv);
/*
* For inactive pipes, all watermark levels
hw->wm_lp_spr[2] = I915_READ(WM3S_LP_IVB);
}
- if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
hw->partitioning = (I915_READ(WM_MISC) & WM_MISC_DATA_PARTITION_5_6) ?
INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
- else if (IS_IVYBRIDGE(dev))
+ else if (IS_IVYBRIDGE(dev_priv))
hw->partitioning = (I915_READ(DISP_ARB_CTL2) & DISP_DATA_PARTITION_5_6) ?
INTEL_DDB_PART_5_6 : INTEL_DDB_PART_1_2;
*/
if (!(dev_priv->gt.awake &&
dev_priv->rps.enabled &&
- dev_priv->rps.cur_freq < dev_priv->rps.max_freq_softlimit))
+ dev_priv->rps.cur_freq < dev_priv->rps.boost_freq))
return;
/* Force a RPS boost (and don't count it against the client) if
static void gen6_init_rps_frequencies(struct drm_i915_private *dev_priv)
{
- uint32_t rp_state_cap;
- u32 ddcc_status = 0;
- int ret;
-
/* All of these values are in units of 50MHz */
- dev_priv->rps.cur_freq = 0;
+
/* static values from HW: RP0 > RP1 > RPn (min_freq) */
if (IS_BROXTON(dev_priv)) {
- rp_state_cap = I915_READ(BXT_RP_STATE_CAP);
+ u32 rp_state_cap = I915_READ(BXT_RP_STATE_CAP);
dev_priv->rps.rp0_freq = (rp_state_cap >> 16) & 0xff;
dev_priv->rps.rp1_freq = (rp_state_cap >> 8) & 0xff;
dev_priv->rps.min_freq = (rp_state_cap >> 0) & 0xff;
} else {
- rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+ u32 rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
dev_priv->rps.rp0_freq = (rp_state_cap >> 0) & 0xff;
dev_priv->rps.rp1_freq = (rp_state_cap >> 8) & 0xff;
dev_priv->rps.min_freq = (rp_state_cap >> 16) & 0xff;
}
-
/* hw_max = RP0 until we check for overclocking */
- dev_priv->rps.max_freq = dev_priv->rps.rp0_freq;
+ dev_priv->rps.max_freq = dev_priv->rps.rp0_freq;
dev_priv->rps.efficient_freq = dev_priv->rps.rp1_freq;
if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv) ||
IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) {
- ret = sandybridge_pcode_read(dev_priv,
- HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL,
- &ddcc_status);
- if (0 == ret)
+ u32 ddcc_status = 0;
+
+ if (sandybridge_pcode_read(dev_priv,
+ HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL,
+ &ddcc_status) == 0)
dev_priv->rps.efficient_freq =
clamp_t(u8,
((ddcc_status >> 8) & 0xff),
if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) {
/* Store the frequency values in 16.66 MHZ units, which is
- the natural hardware unit for SKL */
+ * the natural hardware unit for SKL
+ */
dev_priv->rps.rp0_freq *= GEN9_FREQ_SCALER;
dev_priv->rps.rp1_freq *= GEN9_FREQ_SCALER;
dev_priv->rps.min_freq *= GEN9_FREQ_SCALER;
dev_priv->rps.max_freq *= GEN9_FREQ_SCALER;
dev_priv->rps.efficient_freq *= GEN9_FREQ_SCALER;
}
+}
- dev_priv->rps.idle_freq = dev_priv->rps.min_freq;
+static void reset_rps(struct drm_i915_private *dev_priv,
+ void (*set)(struct drm_i915_private *, u8))
+{
+ u8 freq = dev_priv->rps.cur_freq;
- /* Preserve min/max settings in case of re-init */
- if (dev_priv->rps.max_freq_softlimit == 0)
- dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
+ /* force a reset */
+ dev_priv->rps.power = -1;
+ dev_priv->rps.cur_freq = -1;
- if (dev_priv->rps.min_freq_softlimit == 0) {
- if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
- dev_priv->rps.min_freq_softlimit =
- max_t(int, dev_priv->rps.efficient_freq,
- intel_freq_opcode(dev_priv, 450));
- else
- dev_priv->rps.min_freq_softlimit =
- dev_priv->rps.min_freq;
- }
+ set(dev_priv, freq);
}
/* See the Gen9_GT_PM_Programming_Guide doc for the below */
{
intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
- gen6_init_rps_frequencies(dev_priv);
-
/* WaGsvDisableTurbo: Workaround to disable turbo on BXT A* */
if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) {
/*
/* Leaning on the below call to gen6_set_rps to program/setup the
* Up/Down EI & threshold registers, as well as the RP_CONTROL,
* RP_INTERRUPT_LIMITS & RPNSWREQ registers */
- dev_priv->rps.power = HIGH_POWER; /* force a reset */
- gen6_set_rps(dev_priv, dev_priv->rps.idle_freq);
+ reset_rps(dev_priv, gen6_set_rps);
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
}
static void gen9_enable_rc6(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
+ enum intel_engine_id id;
uint32_t rc6_mask = 0;
/* 1a: Software RC state - RC0 */
I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 54 << 16);
I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
- for_each_engine(engine, dev_priv)
+ for_each_engine(engine, dev_priv, id)
I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
if (HAS_GUC(dev_priv))
if (intel_enable_rc6() & INTEL_RC6_ENABLE)
rc6_mask = GEN6_RC_CTL_RC6_ENABLE;
DRM_INFO("RC6 %s\n", onoff(rc6_mask & GEN6_RC_CTL_RC6_ENABLE));
- /* WaRsUseTimeoutMode */
- if (IS_SKL_REVID(dev_priv, 0, SKL_REVID_D0) ||
- IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) {
+ /* WaRsUseTimeoutMode:bxt */
+ if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) {
I915_WRITE(GEN6_RC6_THRESHOLD, 625); /* 800us */
I915_WRITE(GEN6_RC_CONTROL, GEN6_RC_CTL_HW_ENABLE |
GEN7_RC_CTL_TO_MODE |
static void gen8_enable_rps(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
+ enum intel_engine_id id;
uint32_t rc6_mask = 0;
/* 1a: Software RC state - RC0 */
/* 2a: Disable RC states. */
I915_WRITE(GEN6_RC_CONTROL, 0);
- /* Initialize rps frequencies */
- gen6_init_rps_frequencies(dev_priv);
-
/* 2b: Program RC6 thresholds.*/
I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16);
I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
- for_each_engine(engine, dev_priv)
+ for_each_engine(engine, dev_priv, id)
I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
I915_WRITE(GEN6_RC_SLEEP, 0);
if (IS_BROADWELL(dev_priv))
/* 6: Ring frequency + overclocking (our driver does this later */
- dev_priv->rps.power = HIGH_POWER; /* force a reset */
- gen6_set_rps(dev_priv, dev_priv->rps.idle_freq);
+ reset_rps(dev_priv, gen6_set_rps);
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
}
static void gen6_enable_rps(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
- u32 rc6vids, pcu_mbox = 0, rc6_mask = 0;
+ enum intel_engine_id id;
+ u32 rc6vids, rc6_mask = 0;
u32 gtfifodbg;
int rc6_mode;
int ret;
intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
- /* Initialize rps frequencies */
- gen6_init_rps_frequencies(dev_priv);
-
/* disable the counters and set deterministic thresholds */
I915_WRITE(GEN6_RC_CONTROL, 0);
I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000);
I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25);
- for_each_engine(engine, dev_priv)
+ for_each_engine(engine, dev_priv, id)
I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
I915_WRITE(GEN6_RC_SLEEP, 0);
I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 50000);
I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
- ret = sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_MIN_FREQ_TABLE, 0);
- if (ret)
- DRM_DEBUG_DRIVER("Failed to set the min frequency\n");
-
- ret = sandybridge_pcode_read(dev_priv, GEN6_READ_OC_PARAMS, &pcu_mbox);
- if (!ret && (pcu_mbox & (1<<31))) { /* OC supported */
- DRM_DEBUG_DRIVER("Overclocking supported. Max: %dMHz, Overclock max: %dMHz\n",
- (dev_priv->rps.max_freq_softlimit & 0xff) * 50,
- (pcu_mbox & 0xff) * 50);
- dev_priv->rps.max_freq = pcu_mbox & 0xff;
- }
-
- dev_priv->rps.power = HIGH_POWER; /* force a reset */
- gen6_set_rps(dev_priv, dev_priv->rps.idle_freq);
+ reset_rps(dev_priv, gen6_set_rps);
rc6vids = 0;
ret = sandybridge_pcode_read(dev_priv, GEN6_PCODE_READ_RC6VIDS, &rc6vids);
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
}
-static void __gen6_update_ring_freq(struct drm_i915_private *dev_priv)
+static void gen6_update_ring_freq(struct drm_i915_private *dev_priv)
{
int min_freq = 15;
unsigned int gpu_freq;
}
}
-void gen6_update_ring_freq(struct drm_i915_private *dev_priv)
-{
- if (!HAS_CORE_RING_FREQ(dev_priv))
- return;
-
- mutex_lock(&dev_priv->rps.hw_lock);
- __gen6_update_ring_freq(dev_priv);
- mutex_unlock(&dev_priv->rps.hw_lock);
-}
-
static int cherryview_rps_max_freq(struct drm_i915_private *dev_priv)
{
u32 val, rp0;
val = vlv_punit_read(dev_priv, FB_GFX_FMAX_AT_VMAX_FUSE);
- switch (INTEL_INFO(dev_priv)->eu_total) {
+ switch (INTEL_INFO(dev_priv)->sseu.eu_total) {
case 8:
/* (2 * 4) config */
rp0 = (val >> FB_GFX_FMAX_AT_VMAX_2SS4EU_FUSE_SHIFT);
u32 pcbr;
int pctx_size = 24*1024;
- mutex_lock(&dev_priv->drm.struct_mutex);
-
pcbr = I915_READ(VLV_PCBR);
if (pcbr) {
/* BIOS set it up already, grab the pre-alloc'd space */
out:
DRM_DEBUG_DRIVER("PCBR: 0x%08x\n", I915_READ(VLV_PCBR));
dev_priv->vlv_pctx = pctx;
- mutex_unlock(&dev_priv->drm.struct_mutex);
}
static void valleyview_cleanup_pctx(struct drm_i915_private *dev_priv)
if (WARN_ON(!dev_priv->vlv_pctx))
return;
- drm_gem_object_unreference_unlocked(&dev_priv->vlv_pctx->base);
+ i915_gem_object_put_unlocked(dev_priv->vlv_pctx);
dev_priv->vlv_pctx = NULL;
}
vlv_init_gpll_ref_freq(dev_priv);
- mutex_lock(&dev_priv->rps.hw_lock);
-
val = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
switch ((val >> 6) & 3) {
case 0:
DRM_DEBUG_DRIVER("min GPU freq: %d MHz (%u)\n",
intel_gpu_freq(dev_priv, dev_priv->rps.min_freq),
dev_priv->rps.min_freq);
-
- dev_priv->rps.idle_freq = dev_priv->rps.min_freq;
-
- /* Preserve min/max settings in case of re-init */
- if (dev_priv->rps.max_freq_softlimit == 0)
- dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
-
- if (dev_priv->rps.min_freq_softlimit == 0)
- dev_priv->rps.min_freq_softlimit = dev_priv->rps.min_freq;
-
- mutex_unlock(&dev_priv->rps.hw_lock);
}
static void cherryview_init_gt_powersave(struct drm_i915_private *dev_priv)
vlv_init_gpll_ref_freq(dev_priv);
- mutex_lock(&dev_priv->rps.hw_lock);
-
mutex_lock(&dev_priv->sb_lock);
val = vlv_cck_read(dev_priv, CCK_FUSE_REG);
mutex_unlock(&dev_priv->sb_lock);
dev_priv->rps.rp1_freq |
dev_priv->rps.min_freq) & 1,
"Odd GPU freq values\n");
-
- dev_priv->rps.idle_freq = dev_priv->rps.min_freq;
-
- /* Preserve min/max settings in case of re-init */
- if (dev_priv->rps.max_freq_softlimit == 0)
- dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
-
- if (dev_priv->rps.min_freq_softlimit == 0)
- dev_priv->rps.min_freq_softlimit = dev_priv->rps.min_freq;
-
- mutex_unlock(&dev_priv->rps.hw_lock);
}
static void valleyview_cleanup_gt_powersave(struct drm_i915_private *dev_priv)
static void cherryview_enable_rps(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
+ enum intel_engine_id id;
u32 gtfifodbg, val, rc6_mode = 0, pcbr;
WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000); /* 12500 * 1280ns */
I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25); /* 25 * 1280ns */
- for_each_engine(engine, dev_priv)
+ for_each_engine(engine, dev_priv, id)
I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
I915_WRITE(GEN6_RC_SLEEP, 0);
DRM_DEBUG_DRIVER("GPLL enabled? %s\n", yesno(val & GPLLENABLE));
DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
- dev_priv->rps.cur_freq = (val >> 8) & 0xff;
- DRM_DEBUG_DRIVER("current GPU freq: %d MHz (%u)\n",
- intel_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
- dev_priv->rps.cur_freq);
-
- DRM_DEBUG_DRIVER("setting GPU freq to %d MHz (%u)\n",
- intel_gpu_freq(dev_priv, dev_priv->rps.idle_freq),
- dev_priv->rps.idle_freq);
-
- valleyview_set_rps(dev_priv, dev_priv->rps.idle_freq);
+ reset_rps(dev_priv, valleyview_set_rps);
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
}
static void valleyview_enable_rps(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
+ enum intel_engine_id id;
u32 gtfifodbg, val, rc6_mode = 0;
WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000);
I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25);
- for_each_engine(engine, dev_priv)
+ for_each_engine(engine, dev_priv, id)
I915_WRITE(RING_MAX_IDLE(engine->mmio_base), 10);
I915_WRITE(GEN6_RC6_THRESHOLD, 0x557);
DRM_DEBUG_DRIVER("GPLL enabled? %s\n", yesno(val & GPLLENABLE));
DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
- dev_priv->rps.cur_freq = (val >> 8) & 0xff;
- DRM_DEBUG_DRIVER("current GPU freq: %d MHz (%u)\n",
- intel_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
- dev_priv->rps.cur_freq);
-
- DRM_DEBUG_DRIVER("setting GPU freq to %d MHz (%u)\n",
- intel_gpu_freq(dev_priv, dev_priv->rps.idle_freq),
- dev_priv->rps.idle_freq);
-
- valleyview_set_rps(dev_priv, dev_priv->rps.idle_freq);
+ reset_rps(dev_priv, valleyview_set_rps);
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
}
*/
bool i915_gpu_busy(void)
{
- struct drm_i915_private *dev_priv;
- struct intel_engine_cs *engine;
bool ret = false;
spin_lock_irq(&mchdev_lock);
- if (!i915_mch_dev)
- goto out_unlock;
- dev_priv = i915_mch_dev;
-
- for_each_engine(engine, dev_priv)
- ret |= !list_empty(&engine->request_list);
-
-out_unlock:
+ if (i915_mch_dev)
+ ret = i915_mch_dev->gt.awake;
spin_unlock_irq(&mchdev_lock);
return ret;
intel_runtime_pm_get(dev_priv);
}
+ mutex_lock(&dev_priv->drm.struct_mutex);
+ mutex_lock(&dev_priv->rps.hw_lock);
+
+ /* Initialize RPS limits (for userspace) */
if (IS_CHERRYVIEW(dev_priv))
cherryview_init_gt_powersave(dev_priv);
else if (IS_VALLEYVIEW(dev_priv))
valleyview_init_gt_powersave(dev_priv);
+ else if (INTEL_GEN(dev_priv) >= 6)
+ gen6_init_rps_frequencies(dev_priv);
+
+ /* Derive initial user preferences/limits from the hardware limits */
+ dev_priv->rps.idle_freq = dev_priv->rps.min_freq;
+ dev_priv->rps.cur_freq = dev_priv->rps.idle_freq;
+
+ dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq;
+ dev_priv->rps.min_freq_softlimit = dev_priv->rps.min_freq;
+
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
+ dev_priv->rps.min_freq_softlimit =
+ max_t(int,
+ dev_priv->rps.efficient_freq,
+ intel_freq_opcode(dev_priv, 450));
+
+ /* After setting max-softlimit, find the overclock max freq */
+ if (IS_GEN6(dev_priv) ||
+ IS_IVYBRIDGE(dev_priv) || IS_HASWELL(dev_priv)) {
+ u32 params = 0;
+
+ sandybridge_pcode_read(dev_priv, GEN6_READ_OC_PARAMS, ¶ms);
+ if (params & BIT(31)) { /* OC supported */
+ DRM_DEBUG_DRIVER("Overclocking supported, max: %dMHz, overclock: %dMHz\n",
+ (dev_priv->rps.max_freq & 0xff) * 50,
+ (params & 0xff) * 50);
+ dev_priv->rps.max_freq = params & 0xff;
+ }
+ }
+
+ /* Finally allow us to boost to max by default */
+ dev_priv->rps.boost_freq = dev_priv->rps.max_freq;
+
+ mutex_unlock(&dev_priv->rps.hw_lock);
+ mutex_unlock(&dev_priv->drm.struct_mutex);
+
+ intel_autoenable_gt_powersave(dev_priv);
}
void intel_cleanup_gt_powersave(struct drm_i915_private *dev_priv)
intel_runtime_pm_put(dev_priv);
}
-static void gen6_suspend_rps(struct drm_i915_private *dev_priv)
-{
- flush_delayed_work(&dev_priv->rps.delayed_resume_work);
-
- gen6_disable_rps_interrupts(dev_priv);
-}
-
/**
* intel_suspend_gt_powersave - suspend PM work and helper threads
* @dev_priv: i915 device
if (INTEL_GEN(dev_priv) < 6)
return;
- gen6_suspend_rps(dev_priv);
+ if (cancel_delayed_work_sync(&dev_priv->rps.autoenable_work))
+ intel_runtime_pm_put(dev_priv);
+
+ /* gen6_rps_idle() will be called later to disable interrupts */
+}
- /* Force GPU to min freq during suspend */
- gen6_rps_idle(dev_priv);
+void intel_sanitize_gt_powersave(struct drm_i915_private *dev_priv)
+{
+ dev_priv->rps.enabled = true; /* force disabling */
+ intel_disable_gt_powersave(dev_priv);
+
+ gen6_reset_rps_interrupts(dev_priv);
}
void intel_disable_gt_powersave(struct drm_i915_private *dev_priv)
{
- if (IS_IRONLAKE_M(dev_priv)) {
- ironlake_disable_drps(dev_priv);
- } else if (INTEL_INFO(dev_priv)->gen >= 6) {
- intel_suspend_gt_powersave(dev_priv);
+ if (!READ_ONCE(dev_priv->rps.enabled))
+ return;
- mutex_lock(&dev_priv->rps.hw_lock);
- if (INTEL_INFO(dev_priv)->gen >= 9) {
- gen9_disable_rc6(dev_priv);
- gen9_disable_rps(dev_priv);
- } else if (IS_CHERRYVIEW(dev_priv))
- cherryview_disable_rps(dev_priv);
- else if (IS_VALLEYVIEW(dev_priv))
- valleyview_disable_rps(dev_priv);
- else
- gen6_disable_rps(dev_priv);
+ mutex_lock(&dev_priv->rps.hw_lock);
- dev_priv->rps.enabled = false;
- mutex_unlock(&dev_priv->rps.hw_lock);
+ if (INTEL_GEN(dev_priv) >= 9) {
+ gen9_disable_rc6(dev_priv);
+ gen9_disable_rps(dev_priv);
+ } else if (IS_CHERRYVIEW(dev_priv)) {
+ cherryview_disable_rps(dev_priv);
+ } else if (IS_VALLEYVIEW(dev_priv)) {
+ valleyview_disable_rps(dev_priv);
+ } else if (INTEL_GEN(dev_priv) >= 6) {
+ gen6_disable_rps(dev_priv);
+ } else if (IS_IRONLAKE_M(dev_priv)) {
+ ironlake_disable_drps(dev_priv);
}
+
+ dev_priv->rps.enabled = false;
+ mutex_unlock(&dev_priv->rps.hw_lock);
}
-static void intel_gen6_powersave_work(struct work_struct *work)
+void intel_enable_gt_powersave(struct drm_i915_private *dev_priv)
{
- struct drm_i915_private *dev_priv =
- container_of(work, struct drm_i915_private,
- rps.delayed_resume_work.work);
+ /* We shouldn't be disabling as we submit, so this should be less
+ * racy than it appears!
+ */
+ if (READ_ONCE(dev_priv->rps.enabled))
+ return;
- mutex_lock(&dev_priv->rps.hw_lock);
+ /* Powersaving is controlled by the host when inside a VM */
+ if (intel_vgpu_active(dev_priv))
+ return;
- gen6_reset_rps_interrupts(dev_priv);
+ mutex_lock(&dev_priv->rps.hw_lock);
if (IS_CHERRYVIEW(dev_priv)) {
cherryview_enable_rps(dev_priv);
} else if (IS_VALLEYVIEW(dev_priv)) {
valleyview_enable_rps(dev_priv);
- } else if (INTEL_INFO(dev_priv)->gen >= 9) {
+ } else if (INTEL_GEN(dev_priv) >= 9) {
gen9_enable_rc6(dev_priv);
gen9_enable_rps(dev_priv);
if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))
- __gen6_update_ring_freq(dev_priv);
+ gen6_update_ring_freq(dev_priv);
} else if (IS_BROADWELL(dev_priv)) {
gen8_enable_rps(dev_priv);
- __gen6_update_ring_freq(dev_priv);
- } else {
+ gen6_update_ring_freq(dev_priv);
+ } else if (INTEL_GEN(dev_priv) >= 6) {
gen6_enable_rps(dev_priv);
- __gen6_update_ring_freq(dev_priv);
+ gen6_update_ring_freq(dev_priv);
+ } else if (IS_IRONLAKE_M(dev_priv)) {
+ ironlake_enable_drps(dev_priv);
+ intel_init_emon(dev_priv);
}
WARN_ON(dev_priv->rps.max_freq < dev_priv->rps.min_freq);
WARN_ON(dev_priv->rps.efficient_freq > dev_priv->rps.max_freq);
dev_priv->rps.enabled = true;
+ mutex_unlock(&dev_priv->rps.hw_lock);
+}
- gen6_enable_rps_interrupts(dev_priv);
+static void __intel_autoenable_gt_powersave(struct work_struct *work)
+{
+ struct drm_i915_private *dev_priv =
+ container_of(work, typeof(*dev_priv), rps.autoenable_work.work);
+ struct intel_engine_cs *rcs;
+ struct drm_i915_gem_request *req;
- mutex_unlock(&dev_priv->rps.hw_lock);
+ if (READ_ONCE(dev_priv->rps.enabled))
+ goto out;
+
+ rcs = dev_priv->engine[RCS];
+ if (rcs->last_context)
+ goto out;
+ if (!rcs->init_context)
+ goto out;
+
+ mutex_lock(&dev_priv->drm.struct_mutex);
+
+ req = i915_gem_request_alloc(rcs, dev_priv->kernel_context);
+ if (IS_ERR(req))
+ goto unlock;
+
+ if (!i915.enable_execlists && i915_switch_context(req) == 0)
+ rcs->init_context(req);
+
+ /* Mark the device busy, calling intel_enable_gt_powersave() */
+ i915_add_request_no_flush(req);
+
+unlock:
+ mutex_unlock(&dev_priv->drm.struct_mutex);
+out:
intel_runtime_pm_put(dev_priv);
}
-void intel_enable_gt_powersave(struct drm_i915_private *dev_priv)
+void intel_autoenable_gt_powersave(struct drm_i915_private *dev_priv)
{
- /* Powersaving is controlled by the host when inside a VM */
- if (intel_vgpu_active(dev_priv))
+ if (READ_ONCE(dev_priv->rps.enabled))
return;
if (IS_IRONLAKE_M(dev_priv)) {
ironlake_enable_drps(dev_priv);
- mutex_lock(&dev_priv->drm.struct_mutex);
intel_init_emon(dev_priv);
- mutex_unlock(&dev_priv->drm.struct_mutex);
} else if (INTEL_INFO(dev_priv)->gen >= 6) {
/*
* PCU communication is slow and this doesn't need to be
* paths, so the _noresume version is enough (and in case of
* runtime resume it's necessary).
*/
- if (schedule_delayed_work(&dev_priv->rps.delayed_resume_work,
- round_jiffies_up_relative(HZ)))
+ if (queue_delayed_work(dev_priv->wq,
+ &dev_priv->rps.autoenable_work,
+ round_jiffies_up_relative(HZ)))
intel_runtime_pm_get_noresume(dev_priv);
}
}
-void intel_reset_gt_powersave(struct drm_i915_private *dev_priv)
-{
- if (INTEL_INFO(dev_priv)->gen < 6)
- return;
-
- gen6_suspend_rps(dev_priv);
- dev_priv->rps.enabled = false;
-}
-
static void ibx_init_clock_gating(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
* The bit 22 of 0x42004
* The bit 7,8,9 of 0x42020.
*/
- if (IS_IRONLAKE_M(dev)) {
+ if (IS_IRONLAKE_M(dev_priv)) {
/* WaFbcAsynchFlipDisableFbcQueue:ilk */
I915_WRITE(ILK_DISPLAY_CHICKEN1,
I915_READ(ILK_DISPLAY_CHICKEN1) |
* TODO: this bit should only be enabled when really needed, then
* disabled when not needed anymore in order to save power.
*/
- if (HAS_PCH_LPT_LP(dev))
+ if (HAS_PCH_LPT_LP(dev_priv))
I915_WRITE(SOUTH_DSPCLK_GATE_D,
I915_READ(SOUTH_DSPCLK_GATE_D) |
PCH_LP_PARTITION_LEVEL_DISABLE);
{
struct drm_i915_private *dev_priv = to_i915(dev);
- if (HAS_PCH_LPT_LP(dev)) {
+ if (HAS_PCH_LPT_LP(dev_priv)) {
uint32_t val = I915_READ(SOUTH_DSPCLK_GATE_D);
val &= ~PCH_LP_PARTITION_LEVEL_DISABLE;
CHICKEN3_DGMG_DONE_FIX_DISABLE);
/* WaDisablePSDDualDispatchEnable:ivb */
- if (IS_IVB_GT1(dev))
+ if (IS_IVB_GT1(dev_priv))
I915_WRITE(GEN7_HALF_SLICE_CHICKEN1,
_MASKED_BIT_ENABLE(GEN7_PSD_SINGLE_PORT_DISPATCH_ENABLE));
GEN7_WA_FOR_GEN7_L3_CONTROL);
I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER,
GEN7_WA_L3_CHICKEN_MODE);
- if (IS_IVB_GT1(dev))
+ if (IS_IVB_GT1(dev_priv))
I915_WRITE(GEN7_ROW_CHICKEN2,
_MASKED_BIT_ENABLE(DOP_CLOCK_GATING_DISABLE));
else {
snpcr |= GEN6_MBC_SNPCR_MED;
I915_WRITE(GEN6_MBCUNIT_SNPCR, snpcr);
- if (!HAS_PCH_NOP(dev))
+ if (!HAS_PCH_NOP(dev_priv))
cpt_init_clock_gating(dev);
gen6_check_mch_setup(dev);
dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE |
OVRUNIT_CLOCK_GATE_DISABLE |
OVCUNIT_CLOCK_GATE_DISABLE;
- if (IS_GM45(dev))
+ if (IS_GM45(dev_priv))
dspclk_gate |= DSSUNIT_CLOCK_GATE_DISABLE;
I915_WRITE(DSPCLK_GATE_D, dspclk_gate);
void intel_suspend_hw(struct drm_device *dev)
{
- if (HAS_PCH_LPT(dev))
+ if (HAS_PCH_LPT(to_i915(dev)))
lpt_suspend_hw(dev);
}
/* For cxsr */
if (IS_PINEVIEW(dev))
i915_pineview_get_mem_freq(dev);
- else if (IS_GEN5(dev))
+ else if (IS_GEN5(dev_priv))
i915_ironlake_get_mem_freq(dev);
/* For FIFO watermark updates */
skl_setup_wm_latency(dev);
dev_priv->display.update_wm = skl_update_wm;
dev_priv->display.compute_global_watermarks = skl_compute_wm;
- } else if (HAS_PCH_SPLIT(dev)) {
+ } else if (HAS_PCH_SPLIT(dev_priv)) {
ilk_setup_wm_latency(dev);
- if ((IS_GEN5(dev) && dev_priv->wm.pri_latency[1] &&
+ if ((IS_GEN5(dev_priv) && dev_priv->wm.pri_latency[1] &&
dev_priv->wm.spr_latency[1] && dev_priv->wm.cur_latency[1]) ||
- (!IS_GEN5(dev) && dev_priv->wm.pri_latency[0] &&
+ (!IS_GEN5(dev_priv) && dev_priv->wm.pri_latency[0] &&
dev_priv->wm.spr_latency[0] && dev_priv->wm.cur_latency[0])) {
dev_priv->display.compute_pipe_wm = ilk_compute_pipe_wm;
dev_priv->display.compute_intermediate_wm =
DRM_DEBUG_KMS("Failed to read display plane latency. "
"Disable CxSR\n");
}
- } else if (IS_CHERRYVIEW(dev)) {
+ } else if (IS_CHERRYVIEW(dev_priv)) {
vlv_setup_wm_latency(dev);
dev_priv->display.update_wm = vlv_update_wm;
- } else if (IS_VALLEYVIEW(dev)) {
+ } else if (IS_VALLEYVIEW(dev_priv)) {
vlv_setup_wm_latency(dev);
dev_priv->display.update_wm = vlv_update_wm;
} else if (IS_PINEVIEW(dev)) {
- if (!intel_get_cxsr_latency(IS_PINEVIEW_G(dev),
+ if (!intel_get_cxsr_latency(IS_PINEVIEW_G(dev_priv),
dev_priv->is_ddr3,
dev_priv->fsb_freq,
dev_priv->mem_freq)) {
dev_priv->display.update_wm = NULL;
} else
dev_priv->display.update_wm = pineview_update_wm;
- } else if (IS_G4X(dev)) {
+ } else if (IS_G4X(dev_priv)) {
dev_priv->display.update_wm = g4x_update_wm;
- } else if (IS_GEN4(dev)) {
+ } else if (IS_GEN4(dev_priv)) {
dev_priv->display.update_wm = i965_update_wm;
- } else if (IS_GEN3(dev)) {
+ } else if (IS_GEN3(dev_priv)) {
dev_priv->display.update_wm = i9xx_update_wm;
dev_priv->display.get_fifo_size = i9xx_get_fifo_size;
- } else if (IS_GEN2(dev)) {
+ } else if (IS_GEN2(dev_priv)) {
if (INTEL_INFO(dev)->num_pipes == 1) {
dev_priv->display.update_wm = i845_update_wm;
dev_priv->display.get_fifo_size = i845_get_fifo_size;
if (!i915_gem_request_completed(req))
gen6_rps_boost(req->i915, NULL, req->emitted_jiffies);
- i915_gem_request_unreference(req);
+ i915_gem_request_put(req);
kfree(boost);
}
if (boost == NULL)
return;
- i915_gem_request_reference(req);
- boost->req = req;
+ boost->req = i915_gem_request_get(req);
INIT_WORK(&boost->work, __intel_rps_boost_work);
queue_work(req->i915->wq, &boost->work);
mutex_init(&dev_priv->rps.hw_lock);
spin_lock_init(&dev_priv->rps.client_lock);
- INIT_DELAYED_WORK(&dev_priv->rps.delayed_resume_work,
- intel_gen6_powersave_work);
+ INIT_DELAYED_WORK(&dev_priv->rps.autoenable_work,
+ __intel_autoenable_gt_powersave);
INIT_LIST_HEAD(&dev_priv->rps.clients);
- INIT_LIST_HEAD(&dev_priv->rps.semaphores.link);
- INIT_LIST_HEAD(&dev_priv->rps.mmioflips.link);
dev_priv->pm.suspended = false;
atomic_set(&dev_priv->pm.wakeref_count, 0);