linux_dsm_epyc7002/drivers/gpu/drm/i915/intel_pm.c

/*
 * Copyright © 2012 Intel Corporation
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice (including the next
 * paragraph) shall be included in all copies or substantial portions of the
 * Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 *
 * Authors:
 *    Eugeni Dodonov <eugeni.dodonov@intel.com>
 *
 */

#include <linux/cpufreq.h>
#include "i915_drv.h"
#include "intel_drv.h"

/* FBC, or Frame Buffer Compression, is a technique employed to compress the
 * framebuffer contents in-memory, aiming at reducing the required bandwidth
 * during in-memory transfers and, therefore, reduce the power packet.
 *
 * The benefits of FBC are mostly visible with solid backgrounds and
 * variation-less patterns.
 *
 * FBC-related functionality can be enabled by the means of the
 * i915.i915_enable_fbc parameter
 */

void i8xx_disable_fbc(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 fbc_ctl;

	/* Disable compression */
	fbc_ctl = I915_READ(FBC_CONTROL);
	if ((fbc_ctl & FBC_CTL_EN) == 0)
		return;

	fbc_ctl &= ~FBC_CTL_EN;
	I915_WRITE(FBC_CONTROL, fbc_ctl);

	/* Wait for compressing bit to clear */
	if (wait_for((I915_READ(FBC_STATUS) & FBC_STAT_COMPRESSING) == 0, 10)) {
		DRM_DEBUG_KMS("FBC idle timed out\n");
		return;
	}

	DRM_DEBUG_KMS("disabled FBC\n");
}

void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
{
	struct drm_device *dev = crtc->dev;
	struct drm_i915_private *dev_priv = dev->dev_private;
	struct drm_framebuffer *fb = crtc->fb;
	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
	struct drm_i915_gem_object *obj = intel_fb->obj;
	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
	int cfb_pitch;
	int plane, i;
	u32 fbc_ctl, fbc_ctl2;

	cfb_pitch = dev_priv->cfb_size / FBC_LL_SIZE;
	if (fb->pitches[0] < cfb_pitch)
		cfb_pitch = fb->pitches[0];

	/* FBC_CTL wants 64B units */
	cfb_pitch = (cfb_pitch / 64) - 1;
	plane = intel_crtc->plane == 0 ? FBC_CTL_PLANEA : FBC_CTL_PLANEB;

	/* Clear old tags */
	for (i = 0; i < (FBC_LL_SIZE / 32) + 1; i++)
		I915_WRITE(FBC_TAG + (i * 4), 0);

	/* Set it up... */
	fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
	fbc_ctl2 |= plane;
	I915_WRITE(FBC_CONTROL2, fbc_ctl2);
	I915_WRITE(FBC_FENCE_OFF, crtc->y);

	/* enable it... */
	fbc_ctl = FBC_CTL_EN | FBC_CTL_PERIODIC;
	if (IS_I945GM(dev))
		fbc_ctl |= FBC_CTL_C3_IDLE; /* 945 needs special SR handling */
	fbc_ctl |= (cfb_pitch & 0xff) << FBC_CTL_STRIDE_SHIFT;
	fbc_ctl |= (interval & 0x2fff) << FBC_CTL_INTERVAL_SHIFT;
	fbc_ctl |= obj->fence_reg;
	I915_WRITE(FBC_CONTROL, fbc_ctl);

	DRM_DEBUG_KMS("enabled FBC, pitch %d, yoff %d, plane %d, ",
		      cfb_pitch, crtc->y, intel_crtc->plane);
}

bool i8xx_fbc_enabled(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	return I915_READ(FBC_CONTROL) & FBC_CTL_EN;
}

void g4x_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
{
	struct drm_device *dev = crtc->dev;
	struct drm_i915_private *dev_priv = dev->dev_private;
	struct drm_framebuffer *fb = crtc->fb;
	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
	struct drm_i915_gem_object *obj = intel_fb->obj;
	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
	int plane = intel_crtc->plane == 0 ? DPFC_CTL_PLANEA : DPFC_CTL_PLANEB;
	unsigned long stall_watermark = 200;
	u32 dpfc_ctl;

	dpfc_ctl = plane | DPFC_SR_EN | DPFC_CTL_LIMIT_1X;
	dpfc_ctl |= DPFC_CTL_FENCE_EN | obj->fence_reg;
	I915_WRITE(DPFC_CHICKEN, DPFC_HT_MODIFY);

	I915_WRITE(DPFC_RECOMP_CTL, DPFC_RECOMP_STALL_EN |
		   (stall_watermark << DPFC_RECOMP_STALL_WM_SHIFT) |
		   (interval << DPFC_RECOMP_TIMER_COUNT_SHIFT));
	I915_WRITE(DPFC_FENCE_YOFF, crtc->y);

	/* enable it... */
	I915_WRITE(DPFC_CONTROL, I915_READ(DPFC_CONTROL) | DPFC_CTL_EN);

	DRM_DEBUG_KMS("enabled fbc on plane %d\n", intel_crtc->plane);
}

void g4x_disable_fbc(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 dpfc_ctl;

	/* Disable compression */
	dpfc_ctl = I915_READ(DPFC_CONTROL);
	if (dpfc_ctl & DPFC_CTL_EN) {
		dpfc_ctl &= ~DPFC_CTL_EN;
		I915_WRITE(DPFC_CONTROL, dpfc_ctl);

		DRM_DEBUG_KMS("disabled FBC\n");
	}
}

bool g4x_fbc_enabled(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	return I915_READ(DPFC_CONTROL) & DPFC_CTL_EN;
}

static void sandybridge_blit_fbc_update(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 blt_ecoskpd;

	/* Make sure blitter notifies FBC of writes */
	gen6_gt_force_wake_get(dev_priv);
	blt_ecoskpd = I915_READ(GEN6_BLITTER_ECOSKPD);
	blt_ecoskpd |= GEN6_BLITTER_FBC_NOTIFY <<
		GEN6_BLITTER_LOCK_SHIFT;
	I915_WRITE(GEN6_BLITTER_ECOSKPD, blt_ecoskpd);
	blt_ecoskpd |= GEN6_BLITTER_FBC_NOTIFY;
	I915_WRITE(GEN6_BLITTER_ECOSKPD, blt_ecoskpd);
	blt_ecoskpd &= ~(GEN6_BLITTER_FBC_NOTIFY <<
			 GEN6_BLITTER_LOCK_SHIFT);
	I915_WRITE(GEN6_BLITTER_ECOSKPD, blt_ecoskpd);
	POSTING_READ(GEN6_BLITTER_ECOSKPD);
	gen6_gt_force_wake_put(dev_priv);
}

void ironlake_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
{
	struct drm_device *dev = crtc->dev;
	struct drm_i915_private *dev_priv = dev->dev_private;
	struct drm_framebuffer *fb = crtc->fb;
	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
	struct drm_i915_gem_object *obj = intel_fb->obj;
	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
	int plane = intel_crtc->plane == 0 ? DPFC_CTL_PLANEA : DPFC_CTL_PLANEB;
	unsigned long stall_watermark = 200;
	u32 dpfc_ctl;

	dpfc_ctl = I915_READ(ILK_DPFC_CONTROL);
	dpfc_ctl &= DPFC_RESERVED;
	dpfc_ctl |= (plane | DPFC_CTL_LIMIT_1X);
	/* Set persistent mode for front-buffer rendering, ala X. */
	dpfc_ctl |= DPFC_CTL_PERSISTENT_MODE;
	dpfc_ctl |= (DPFC_CTL_FENCE_EN | obj->fence_reg);
	I915_WRITE(ILK_DPFC_CHICKEN, DPFC_HT_MODIFY);

	I915_WRITE(ILK_DPFC_RECOMP_CTL, DPFC_RECOMP_STALL_EN |
		   (stall_watermark << DPFC_RECOMP_STALL_WM_SHIFT) |
		   (interval << DPFC_RECOMP_TIMER_COUNT_SHIFT));
	I915_WRITE(ILK_DPFC_FENCE_YOFF, crtc->y);
	I915_WRITE(ILK_FBC_RT_BASE, obj->gtt_offset | ILK_FBC_RT_VALID);
	/* enable it... */
	I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);

	if (IS_GEN6(dev)) {
		I915_WRITE(SNB_DPFC_CTL_SA,
			   SNB_CPU_FENCE_ENABLE | obj->fence_reg);
		I915_WRITE(DPFC_CPU_FENCE_OFFSET, crtc->y);
		sandybridge_blit_fbc_update(dev);
	}

	DRM_DEBUG_KMS("enabled fbc on plane %d\n", intel_crtc->plane);
}

void ironlake_disable_fbc(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 dpfc_ctl;

	/* Disable compression */
	dpfc_ctl = I915_READ(ILK_DPFC_CONTROL);
	if (dpfc_ctl & DPFC_CTL_EN) {
		dpfc_ctl &= ~DPFC_CTL_EN;
		I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl);

		DRM_DEBUG_KMS("disabled FBC\n");
	}
}

bool ironlake_fbc_enabled(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	return I915_READ(ILK_DPFC_CONTROL) & DPFC_CTL_EN;
}

bool intel_fbc_enabled(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	if (!dev_priv->display.fbc_enabled)
		return false;

	return dev_priv->display.fbc_enabled(dev);
}

static void intel_fbc_work_fn(struct work_struct *__work)
{
	struct intel_fbc_work *work =
		container_of(to_delayed_work(__work),
			     struct intel_fbc_work, work);
	struct drm_device *dev = work->crtc->dev;
	struct drm_i915_private *dev_priv = dev->dev_private;

	mutex_lock(&dev->struct_mutex);
	if (work == dev_priv->fbc_work) {
		/* Double check that we haven't switched fb without cancelling
		 * the prior work.
		 */
		if (work->crtc->fb == work->fb) {
			dev_priv->display.enable_fbc(work->crtc,
						     work->interval);

			dev_priv->cfb_plane = to_intel_crtc(work->crtc)->plane;
			dev_priv->cfb_fb = work->crtc->fb->base.id;
			dev_priv->cfb_y = work->crtc->y;
		}

		dev_priv->fbc_work = NULL;
	}
	mutex_unlock(&dev->struct_mutex);

	kfree(work);
}

static void intel_cancel_fbc_work(struct drm_i915_private *dev_priv)
{
	if (dev_priv->fbc_work == NULL)
		return;

	DRM_DEBUG_KMS("cancelling pending FBC enable\n");

	/* Synchronisation is provided by struct_mutex and checking of
	 * dev_priv->fbc_work, so we can perform the cancellation
	 * entirely asynchronously.
	 */
	if (cancel_delayed_work(&dev_priv->fbc_work->work))
		/* tasklet was killed before being run, clean up */
		kfree(dev_priv->fbc_work);

	/* Mark the work as no longer wanted so that if it does
	 * wake-up (because the work was already running and waiting
	 * for our mutex), it will discover that is no longer
	 * necessary to run.
	 */
	dev_priv->fbc_work = NULL;
}

void intel_enable_fbc(struct drm_crtc *crtc, unsigned long interval)
{
	struct intel_fbc_work *work;
	struct drm_device *dev = crtc->dev;
	struct drm_i915_private *dev_priv = dev->dev_private;

	if (!dev_priv->display.enable_fbc)
		return;

	intel_cancel_fbc_work(dev_priv);

	work = kzalloc(sizeof *work, GFP_KERNEL);
	if (work == NULL) {
		dev_priv->display.enable_fbc(crtc, interval);
		return;
	}

	work->crtc = crtc;
	work->fb = crtc->fb;
	work->interval = interval;
	INIT_DELAYED_WORK(&work->work, intel_fbc_work_fn);

	dev_priv->fbc_work = work;

	DRM_DEBUG_KMS("scheduling delayed FBC enable\n");

	/* Delay the actual enabling to let pageflipping cease and the
	 * display to settle before starting the compression. Note that
	 * this delay also serves a second purpose: it allows for a
	 * vblank to pass after disabling the FBC before we attempt
	 * to modify the control registers.
	 *
	 * A more complicated solution would involve tracking vblanks
	 * following the termination of the page-flipping sequence
	 * and indeed performing the enable as a co-routine and not
	 * waiting synchronously upon the vblank.
	 */
	schedule_delayed_work(&work->work, msecs_to_jiffies(50));
}

void intel_disable_fbc(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	intel_cancel_fbc_work(dev_priv);

	if (!dev_priv->display.disable_fbc)
		return;

	dev_priv->display.disable_fbc(dev);
	dev_priv->cfb_plane = -1;
}

/**
 * intel_update_fbc - enable/disable FBC as needed
 * @dev: the drm_device
 *
 * Set up the framebuffer compression hardware at mode set time.  We
 * enable it if possible:
 *   - plane A only (on pre-965)
 *   - no pixel mulitply/line duplication
 *   - no alpha buffer discard
 *   - no dual wide
 *   - framebuffer <= 2048 in width, 1536 in height
 *
 * We can't assume that any compression will take place (worst case),
 * so the compressed buffer has to be the same size as the uncompressed
 * one.  It also must reside (along with the line length buffer) in
 * stolen memory.
 *
 * We need to enable/disable FBC on a global basis.
 */
void intel_update_fbc(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	struct drm_crtc *crtc = NULL, *tmp_crtc;
	struct intel_crtc *intel_crtc;
	struct drm_framebuffer *fb;
	struct intel_framebuffer *intel_fb;
	struct drm_i915_gem_object *obj;
	int enable_fbc;

	DRM_DEBUG_KMS("\n");

	if (!i915_powersave)
		return;

	if (!I915_HAS_FBC(dev))
		return;

	/*
	 * If FBC is already on, we just have to verify that we can
	 * keep it that way...
	 * Need to disable if:
	 *   - more than one pipe is active
	 *   - changing FBC params (stride, fence, mode)
	 *   - new fb is too large to fit in compressed buffer
	 *   - going to an unsupported config (interlace, pixel multiply, etc.)
	 */
	list_for_each_entry(tmp_crtc, &dev->mode_config.crtc_list, head) {
		if (tmp_crtc->enabled && tmp_crtc->fb) {
			if (crtc) {
				DRM_DEBUG_KMS("more than one pipe active, disabling compression\n");
				dev_priv->no_fbc_reason = FBC_MULTIPLE_PIPES;
				goto out_disable;
			}
			crtc = tmp_crtc;
		}
	}

	if (!crtc || crtc->fb == NULL) {
		DRM_DEBUG_KMS("no output, disabling\n");
		dev_priv->no_fbc_reason = FBC_NO_OUTPUT;
		goto out_disable;
	}

	intel_crtc = to_intel_crtc(crtc);
	fb = crtc->fb;
	intel_fb = to_intel_framebuffer(fb);
	obj = intel_fb->obj;

	enable_fbc = i915_enable_fbc;
	if (enable_fbc < 0) {
		DRM_DEBUG_KMS("fbc set to per-chip default\n");
		enable_fbc = 1;
		if (INTEL_INFO(dev)->gen <= 6)
			enable_fbc = 0;
	}
	if (!enable_fbc) {
		DRM_DEBUG_KMS("fbc disabled per module param\n");
		dev_priv->no_fbc_reason = FBC_MODULE_PARAM;
		goto out_disable;
	}
	if (intel_fb->obj->base.size > dev_priv->cfb_size) {
		DRM_DEBUG_KMS("framebuffer too large, disabling "
			      "compression\n");
		dev_priv->no_fbc_reason = FBC_STOLEN_TOO_SMALL;
		goto out_disable;
	}
	if ((crtc->mode.flags & DRM_MODE_FLAG_INTERLACE) ||
	    (crtc->mode.flags & DRM_MODE_FLAG_DBLSCAN)) {
		DRM_DEBUG_KMS("mode incompatible with compression, "
			      "disabling\n");
		dev_priv->no_fbc_reason = FBC_UNSUPPORTED_MODE;
		goto out_disable;
	}
	if ((crtc->mode.hdisplay > 2048) ||
	    (crtc->mode.vdisplay > 1536)) {
		DRM_DEBUG_KMS("mode too large for compression, disabling\n");
		dev_priv->no_fbc_reason = FBC_MODE_TOO_LARGE;
		goto out_disable;
	}
	if ((IS_I915GM(dev) || IS_I945GM(dev)) && intel_crtc->plane != 0) {
		DRM_DEBUG_KMS("plane not 0, disabling compression\n");
		dev_priv->no_fbc_reason = FBC_BAD_PLANE;
		goto out_disable;
	}

	/* The use of a CPU fence is mandatory in order to detect writes
	 * by the CPU to the scanout and trigger updates to the FBC.
	 */
	if (obj->tiling_mode != I915_TILING_X ||
	    obj->fence_reg == I915_FENCE_REG_NONE) {
		DRM_DEBUG_KMS("framebuffer not tiled or fenced, disabling compression\n");
		dev_priv->no_fbc_reason = FBC_NOT_TILED;
		goto out_disable;
	}

	/* If the kernel debugger is active, always disable compression */
	if (in_dbg_master())
		goto out_disable;

	/* If the scanout has not changed, don't modify the FBC settings.
	 * Note that we make the fundamental assumption that the fb->obj
	 * cannot be unpinned (and have its GTT offset and fence revoked)
	 * without first being decoupled from the scanout and FBC disabled.
	 */
	if (dev_priv->cfb_plane == intel_crtc->plane &&
	    dev_priv->cfb_fb == fb->base.id &&
	    dev_priv->cfb_y == crtc->y)
		return;

	if (intel_fbc_enabled(dev)) {
		/* We update FBC along two paths, after changing fb/crtc
		 * configuration (modeswitching) and after page-flipping
		 * finishes. For the latter, we know that not only did
		 * we disable the FBC at the start of the page-flip
		 * sequence, but also more than one vblank has passed.
		 *
		 * For the former case of modeswitching, it is possible
		 * to switch between two FBC valid configurations
		 * instantaneously so we do need to disable the FBC
		 * before we can modify its control registers. We also
		 * have to wait for the next vblank for that to take
		 * effect. However, since we delay enabling FBC we can
		 * assume that a vblank has passed since disabling and
		 * that we can safely alter the registers in the deferred
		 * callback.
		 *
		 * In the scenario that we go from a valid to invalid
		 * and then back to valid FBC configuration we have
		 * no strict enforcement that a vblank occurred since
		 * disabling the FBC. However, along all current pipe
		 * disabling paths we do need to wait for a vblank at
		 * some point. And we wait before enabling FBC anyway.
		 */
		DRM_DEBUG_KMS("disabling active FBC for update\n");
		intel_disable_fbc(dev);
	}

	intel_enable_fbc(crtc, 500);
	return;

out_disable:
	/* Multiple disables should be harmless */
	if (intel_fbc_enabled(dev)) {
		DRM_DEBUG_KMS("unsupported config, disabling FBC\n");
		intel_disable_fbc(dev);
	}
}

static const struct cxsr_latency cxsr_latency_table[] = {
	{1, 0, 800, 400, 3382, 33382, 3983, 33983},    /* DDR2-400 SC */
	{1, 0, 800, 667, 3354, 33354, 3807, 33807},    /* DDR2-667 SC */
	{1, 0, 800, 800, 3347, 33347, 3763, 33763},    /* DDR2-800 SC */
	{1, 1, 800, 667, 6420, 36420, 6873, 36873},    /* DDR3-667 SC */
	{1, 1, 800, 800, 5902, 35902, 6318, 36318},    /* DDR3-800 SC */

	{1, 0, 667, 400, 3400, 33400, 4021, 34021},    /* DDR2-400 SC */
	{1, 0, 667, 667, 3372, 33372, 3845, 33845},    /* DDR2-667 SC */
	{1, 0, 667, 800, 3386, 33386, 3822, 33822},    /* DDR2-800 SC */
	{1, 1, 667, 667, 6438, 36438, 6911, 36911},    /* DDR3-667 SC */
	{1, 1, 667, 800, 5941, 35941, 6377, 36377},    /* DDR3-800 SC */

	{1, 0, 400, 400, 3472, 33472, 4173, 34173},    /* DDR2-400 SC */
	{1, 0, 400, 667, 3443, 33443, 3996, 33996},    /* DDR2-667 SC */
	{1, 0, 400, 800, 3430, 33430, 3946, 33946},    /* DDR2-800 SC */
	{1, 1, 400, 667, 6509, 36509, 7062, 37062},    /* DDR3-667 SC */
	{1, 1, 400, 800, 5985, 35985, 6501, 36501},    /* DDR3-800 SC */

	{0, 0, 800, 400, 3438, 33438, 4065, 34065},    /* DDR2-400 SC */
	{0, 0, 800, 667, 3410, 33410, 3889, 33889},    /* DDR2-667 SC */
	{0, 0, 800, 800, 3403, 33403, 3845, 33845},    /* DDR2-800 SC */
	{0, 1, 800, 667, 6476, 36476, 6955, 36955},    /* DDR3-667 SC */
	{0, 1, 800, 800, 5958, 35958, 6400, 36400},    /* DDR3-800 SC */

	{0, 0, 667, 400, 3456, 33456, 4103, 34106},    /* DDR2-400 SC */
	{0, 0, 667, 667, 3428, 33428, 3927, 33927},    /* DDR2-667 SC */
	{0, 0, 667, 800, 3443, 33443, 3905, 33905},    /* DDR2-800 SC */
	{0, 1, 667, 667, 6494, 36494, 6993, 36993},    /* DDR3-667 SC */
	{0, 1, 667, 800, 5998, 35998, 6460, 36460},    /* DDR3-800 SC */

	{0, 0, 400, 400, 3528, 33528, 4255, 34255},    /* DDR2-400 SC */
	{0, 0, 400, 667, 3500, 33500, 4079, 34079},    /* DDR2-667 SC */
	{0, 0, 400, 800, 3487, 33487, 4029, 34029},    /* DDR2-800 SC */
	{0, 1, 400, 667, 6566, 36566, 7145, 37145},    /* DDR3-667 SC */
	{0, 1, 400, 800, 6042, 36042, 6584, 36584},    /* DDR3-800 SC */
};

const struct cxsr_latency *intel_get_cxsr_latency(int is_desktop,
							 int is_ddr3,
							 int fsb,
							 int mem)
{
	const struct cxsr_latency *latency;
	int i;

	if (fsb == 0 || mem == 0)
		return NULL;

	for (i = 0; i < ARRAY_SIZE(cxsr_latency_table); i++) {
		latency = &cxsr_latency_table[i];
		if (is_desktop == latency->is_desktop &&
		    is_ddr3 == latency->is_ddr3 &&
		    fsb == latency->fsb_freq && mem == latency->mem_freq)
			return latency;
	}

	DRM_DEBUG_KMS("Unknown FSB/MEM found, disable CxSR\n");

	return NULL;
}

void pineview_disable_cxsr(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	/* deactivate cxsr */
	I915_WRITE(DSPFW3, I915_READ(DSPFW3) & ~PINEVIEW_SELF_REFRESH_EN);
}

/*
 * Latency for FIFO fetches is dependent on several factors:
 *   - memory configuration (speed, channels)
 *   - chipset
 *   - current MCH state
 * It can be fairly high in some situations, so here we assume a fairly
 * pessimal value.  It's a tradeoff between extra memory fetches (if we
 * set this value too high, the FIFO will fetch frequently to stay full)
 * and power consumption (set it too low to save power and we might see
 * FIFO underruns and display "flicker").
 *
 * A value of 5us seems to be a good balance; safe for very low end
 * platforms but not overly aggressive on lower latency configs.
 */
static const int latency_ns = 5000;

int i9xx_get_fifo_size(struct drm_device *dev, int plane)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	uint32_t dsparb = I915_READ(DSPARB);
	int size;

	size = dsparb & 0x7f;
	if (plane)
		size = ((dsparb >> DSPARB_CSTART_SHIFT) & 0x7f) - size;

	DRM_DEBUG_KMS("FIFO size - (0x%08x) %s: %d\n", dsparb,
		      plane ? "B" : "A", size);

	return size;
}

int i85x_get_fifo_size(struct drm_device *dev, int plane)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	uint32_t dsparb = I915_READ(DSPARB);
	int size;

	size = dsparb & 0x1ff;
	if (plane)
		size = ((dsparb >> DSPARB_BEND_SHIFT) & 0x1ff) - size;
	size >>= 1; /* Convert to cachelines */

	DRM_DEBUG_KMS("FIFO size - (0x%08x) %s: %d\n", dsparb,
		      plane ? "B" : "A", size);

	return size;
}

int i845_get_fifo_size(struct drm_device *dev, int plane)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	uint32_t dsparb = I915_READ(DSPARB);
	int size;

	size = dsparb & 0x7f;
	size >>= 2; /* Convert to cachelines */

	DRM_DEBUG_KMS("FIFO size - (0x%08x) %s: %d\n", dsparb,
		      plane ? "B" : "A",
		      size);

	return size;
}

int i830_get_fifo_size(struct drm_device *dev, int plane)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	uint32_t dsparb = I915_READ(DSPARB);
	int size;

	size = dsparb & 0x7f;
	size >>= 1; /* Convert to cachelines */

	DRM_DEBUG_KMS("FIFO size - (0x%08x) %s: %d\n", dsparb,
		      plane ? "B" : "A", size);

	return size;
}

/* Pineview has different values for various configs */
static const struct intel_watermark_params pineview_display_wm = {
	PINEVIEW_DISPLAY_FIFO,
	PINEVIEW_MAX_WM,
	PINEVIEW_DFT_WM,
	PINEVIEW_GUARD_WM,
	PINEVIEW_FIFO_LINE_SIZE
};
static const struct intel_watermark_params pineview_display_hplloff_wm = {
	PINEVIEW_DISPLAY_FIFO,
	PINEVIEW_MAX_WM,
	PINEVIEW_DFT_HPLLOFF_WM,
	PINEVIEW_GUARD_WM,
	PINEVIEW_FIFO_LINE_SIZE
};
static const struct intel_watermark_params pineview_cursor_wm = {
	PINEVIEW_CURSOR_FIFO,
	PINEVIEW_CURSOR_MAX_WM,
	PINEVIEW_CURSOR_DFT_WM,
	PINEVIEW_CURSOR_GUARD_WM,
	PINEVIEW_FIFO_LINE_SIZE,
};
static const struct intel_watermark_params pineview_cursor_hplloff_wm = {
	PINEVIEW_CURSOR_FIFO,
	PINEVIEW_CURSOR_MAX_WM,
	PINEVIEW_CURSOR_DFT_WM,
	PINEVIEW_CURSOR_GUARD_WM,
	PINEVIEW_FIFO_LINE_SIZE
};
static const struct intel_watermark_params g4x_wm_info = {
	G4X_FIFO_SIZE,
	G4X_MAX_WM,
	G4X_MAX_WM,
	2,
	G4X_FIFO_LINE_SIZE,
};
static const struct intel_watermark_params g4x_cursor_wm_info = {
	I965_CURSOR_FIFO,
	I965_CURSOR_MAX_WM,
	I965_CURSOR_DFT_WM,
	2,
	G4X_FIFO_LINE_SIZE,
};
static const struct intel_watermark_params valleyview_wm_info = {
	VALLEYVIEW_FIFO_SIZE,
	VALLEYVIEW_MAX_WM,
	VALLEYVIEW_MAX_WM,
	2,
	G4X_FIFO_LINE_SIZE,
};
static const struct intel_watermark_params valleyview_cursor_wm_info = {
	I965_CURSOR_FIFO,
	VALLEYVIEW_CURSOR_MAX_WM,
	I965_CURSOR_DFT_WM,
	2,
	G4X_FIFO_LINE_SIZE,
};
static const struct intel_watermark_params i965_cursor_wm_info = {
	I965_CURSOR_FIFO,
	I965_CURSOR_MAX_WM,
	I965_CURSOR_DFT_WM,
	2,
	I915_FIFO_LINE_SIZE,
};
static const struct intel_watermark_params i945_wm_info = {
	I945_FIFO_SIZE,
	I915_MAX_WM,
	1,
	2,
	I915_FIFO_LINE_SIZE
};
static const struct intel_watermark_params i915_wm_info = {
	I915_FIFO_SIZE,
	I915_MAX_WM,
	1,
	2,
	I915_FIFO_LINE_SIZE
};
static const struct intel_watermark_params i855_wm_info = {
	I855GM_FIFO_SIZE,
	I915_MAX_WM,
	1,
	2,
	I830_FIFO_LINE_SIZE
};
static const struct intel_watermark_params i830_wm_info = {
	I830_FIFO_SIZE,
	I915_MAX_WM,
	1,
	2,
	I830_FIFO_LINE_SIZE
};

static const struct intel_watermark_params ironlake_display_wm_info = {
	ILK_DISPLAY_FIFO,
	ILK_DISPLAY_MAXWM,
	ILK_DISPLAY_DFTWM,
	2,
	ILK_FIFO_LINE_SIZE
};
static const struct intel_watermark_params ironlake_cursor_wm_info = {
	ILK_CURSOR_FIFO,
	ILK_CURSOR_MAXWM,
	ILK_CURSOR_DFTWM,
	2,
	ILK_FIFO_LINE_SIZE
};
static const struct intel_watermark_params ironlake_display_srwm_info = {
	ILK_DISPLAY_SR_FIFO,
	ILK_DISPLAY_MAX_SRWM,
	ILK_DISPLAY_DFT_SRWM,
	2,
	ILK_FIFO_LINE_SIZE
};
static const struct intel_watermark_params ironlake_cursor_srwm_info = {
	ILK_CURSOR_SR_FIFO,
	ILK_CURSOR_MAX_SRWM,
	ILK_CURSOR_DFT_SRWM,
	2,
	ILK_FIFO_LINE_SIZE
};

static const struct intel_watermark_params sandybridge_display_wm_info = {
	SNB_DISPLAY_FIFO,
	SNB_DISPLAY_MAXWM,
	SNB_DISPLAY_DFTWM,
	2,
	SNB_FIFO_LINE_SIZE
};
static const struct intel_watermark_params sandybridge_cursor_wm_info = {
	SNB_CURSOR_FIFO,
	SNB_CURSOR_MAXWM,
	SNB_CURSOR_DFTWM,
	2,
	SNB_FIFO_LINE_SIZE
};
static const struct intel_watermark_params sandybridge_display_srwm_info = {
	SNB_DISPLAY_SR_FIFO,
	SNB_DISPLAY_MAX_SRWM,
	SNB_DISPLAY_DFT_SRWM,
	2,
	SNB_FIFO_LINE_SIZE
};
static const struct intel_watermark_params sandybridge_cursor_srwm_info = {
	SNB_CURSOR_SR_FIFO,
	SNB_CURSOR_MAX_SRWM,
	SNB_CURSOR_DFT_SRWM,
	2,
	SNB_FIFO_LINE_SIZE
};


/**
 * intel_calculate_wm - calculate watermark level
 * @clock_in_khz: pixel clock
 * @wm: chip FIFO params
 * @pixel_size: display pixel size
 * @latency_ns: memory latency for the platform
 *
 * Calculate the watermark level (the level at which the display plane will
 * start fetching from memory again).  Each chip has a different display
 * FIFO size and allocation, so the caller needs to figure that out and pass
 * in the correct intel_watermark_params structure.
 *
 * As the pixel clock runs, the FIFO will be drained at a rate that depends
 * on the pixel size.  When it reaches the watermark level, it'll start
 * fetching FIFO line sized based chunks from memory until the FIFO fills
 * past the watermark point.  If the FIFO drains completely, a FIFO underrun
 * will occur, and a display engine hang could result.
 */
static unsigned long intel_calculate_wm(unsigned long clock_in_khz,
					const struct intel_watermark_params *wm,
					int fifo_size,
					int pixel_size,
					unsigned long latency_ns)
{
	long entries_required, wm_size;

	/*
	 * Note: we need to make sure we don't overflow for various clock &
	 * latency values.
	 * clocks go from a few thousand to several hundred thousand.
	 * latency is usually a few thousand
	 */
	entries_required = ((clock_in_khz / 1000) * pixel_size * latency_ns) /
		1000;
	entries_required = DIV_ROUND_UP(entries_required, wm->cacheline_size);

	DRM_DEBUG_KMS("FIFO entries required for mode: %ld\n", entries_required);

	wm_size = fifo_size - (entries_required + wm->guard_size);

	DRM_DEBUG_KMS("FIFO watermark level: %ld\n", wm_size);

	/* Don't promote wm_size to unsigned... */
	if (wm_size > (long)wm->max_wm)
		wm_size = wm->max_wm;
	if (wm_size <= 0)
		wm_size = wm->default_wm;
	return wm_size;
}

static struct drm_crtc *single_enabled_crtc(struct drm_device *dev)
{
	struct drm_crtc *crtc, *enabled = NULL;

	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
		if (crtc->enabled && crtc->fb) {
			if (enabled)
				return NULL;
			enabled = crtc;
		}
	}

	return enabled;
}

void pineview_update_wm(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	struct drm_crtc *crtc;
	const struct cxsr_latency *latency;
	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);
	if (!latency) {
		DRM_DEBUG_KMS("Unknown FSB/MEM found, disable CxSR\n");
		pineview_disable_cxsr(dev);
		return;
	}

	crtc = single_enabled_crtc(dev);
	if (crtc) {
		int clock = crtc->mode.clock;
		int pixel_size = crtc->fb->bits_per_pixel / 8;

		/* Display SR */
		wm = intel_calculate_wm(clock, &pineview_display_wm,
					pineview_display_wm.fifo_size,
					pixel_size, latency->display_sr);
		reg = I915_READ(DSPFW1);
		reg &= ~DSPFW_SR_MASK;
		reg |= wm << DSPFW_SR_SHIFT;
		I915_WRITE(DSPFW1, reg);
		DRM_DEBUG_KMS("DSPFW1 register is %x\n", reg);

		/* cursor SR */
		wm = intel_calculate_wm(clock, &pineview_cursor_wm,
					pineview_display_wm.fifo_size,
					pixel_size, latency->cursor_sr);
		reg = I915_READ(DSPFW3);
		reg &= ~DSPFW_CURSOR_SR_MASK;
		reg |= (wm & 0x3f) << DSPFW_CURSOR_SR_SHIFT;
		I915_WRITE(DSPFW3, reg);

		/* Display HPLL off SR */
		wm = intel_calculate_wm(clock, &pineview_display_hplloff_wm,
					pineview_display_hplloff_wm.fifo_size,
					pixel_size, latency->display_hpll_disable);
		reg = I915_READ(DSPFW3);
		reg &= ~DSPFW_HPLL_SR_MASK;
		reg |= wm & DSPFW_HPLL_SR_MASK;
		I915_WRITE(DSPFW3, reg);

		/* cursor HPLL off SR */
		wm = intel_calculate_wm(clock, &pineview_cursor_hplloff_wm,
					pineview_display_hplloff_wm.fifo_size,
					pixel_size, latency->cursor_hpll_disable);
		reg = I915_READ(DSPFW3);
		reg &= ~DSPFW_HPLL_CURSOR_MASK;
		reg |= (wm & 0x3f) << DSPFW_HPLL_CURSOR_SHIFT;
		I915_WRITE(DSPFW3, reg);
		DRM_DEBUG_KMS("DSPFW3 register is %x\n", reg);

		/* activate cxsr */
		I915_WRITE(DSPFW3,
			   I915_READ(DSPFW3) | PINEVIEW_SELF_REFRESH_EN);
		DRM_DEBUG_KMS("Self-refresh is enabled\n");
	} else {
		pineview_disable_cxsr(dev);
		DRM_DEBUG_KMS("Self-refresh is disabled\n");
	}
}

static bool g4x_compute_wm0(struct drm_device *dev,
			    int plane,
			    const struct intel_watermark_params *display,
			    int display_latency_ns,
			    const struct intel_watermark_params *cursor,
			    int cursor_latency_ns,
			    int *plane_wm,
			    int *cursor_wm)
{
	struct drm_crtc *crtc;
	int htotal, hdisplay, clock, pixel_size;
	int line_time_us, line_count;
	int entries, tlb_miss;

	crtc = intel_get_crtc_for_plane(dev, plane);
	if (crtc->fb == NULL || !crtc->enabled) {
		*cursor_wm = cursor->guard_size;
		*plane_wm = display->guard_size;
		return false;
	}

	htotal = crtc->mode.htotal;
	hdisplay = crtc->mode.hdisplay;
	clock = crtc->mode.clock;
	pixel_size = crtc->fb->bits_per_pixel / 8;

	/* Use the small buffer method to calculate plane watermark */
	entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000;
	tlb_miss = display->fifo_size*display->cacheline_size - hdisplay * 8;
	if (tlb_miss > 0)
		entries += tlb_miss;
	entries = DIV_ROUND_UP(entries, display->cacheline_size);
	*plane_wm = entries + display->guard_size;
	if (*plane_wm > (int)display->max_wm)
		*plane_wm = display->max_wm;

	/* Use the large buffer method to calculate cursor watermark */
	line_time_us = ((htotal * 1000) / clock);
	line_count = (cursor_latency_ns / line_time_us + 1000) / 1000;
	entries = line_count * 64 * pixel_size;
	tlb_miss = cursor->fifo_size*cursor->cacheline_size - hdisplay * 8;
	if (tlb_miss > 0)
		entries += tlb_miss;
	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
	*cursor_wm = entries + cursor->guard_size;
	if (*cursor_wm > (int)cursor->max_wm)
		*cursor_wm = (int)cursor->max_wm;

	return true;
}

/*
 * Check the wm result.
 *
 * If any calculated watermark values is larger than the maximum value that
 * can be programmed into the associated watermark register, that watermark
 * must be disabled.
 */
static bool g4x_check_srwm(struct drm_device *dev,
			   int display_wm, int cursor_wm,
			   const struct intel_watermark_params *display,
			   const struct intel_watermark_params *cursor)
{
	DRM_DEBUG_KMS("SR watermark: display plane %d, cursor %d\n",
		      display_wm, cursor_wm);

	if (display_wm > display->max_wm) {
		DRM_DEBUG_KMS("display watermark is too large(%d/%ld), 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",
			      cursor_wm, cursor->max_wm);
		return false;
	}

	if (!(display_wm || cursor_wm)) {
		DRM_DEBUG_KMS("SR latency is 0, disabling\n");
		return false;
	}

	return true;
}

static bool g4x_compute_srwm(struct drm_device *dev,
			     int plane,
			     int latency_ns,
			     const struct intel_watermark_params *display,
			     const struct intel_watermark_params *cursor,
			     int *display_wm, int *cursor_wm)
{
	struct drm_crtc *crtc;
	int hdisplay, htotal, pixel_size, clock;
	unsigned long line_time_us;
	int line_count, line_size;
	int small, large;
	int entries;

	if (!latency_ns) {
		*display_wm = *cursor_wm = 0;
		return false;
	}

	crtc = intel_get_crtc_for_plane(dev, plane);
	hdisplay = crtc->mode.hdisplay;
	htotal = crtc->mode.htotal;
	clock = crtc->mode.clock;
	pixel_size = crtc->fb->bits_per_pixel / 8;

	line_time_us = (htotal * 1000) / clock;
	line_count = (latency_ns / line_time_us + 1000) / 1000;
	line_size = hdisplay * pixel_size;

	/* Use the minimum of the small and large buffer method for primary */
	small = ((clock * pixel_size / 1000) * latency_ns) / 1000;
	large = line_count * line_size;

	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
	*display_wm = entries + display->guard_size;

	/* calculate the self-refresh watermark for display cursor */
	entries = line_count * pixel_size * 64;
	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
	*cursor_wm = entries + cursor->guard_size;

	return g4x_check_srwm(dev,
			      *display_wm, *cursor_wm,
			      display, cursor);
}

static bool vlv_compute_drain_latency(struct drm_device *dev,
				     int plane,
				     int *plane_prec_mult,
				     int *plane_dl,
				     int *cursor_prec_mult,
				     int *cursor_dl)
{
	struct drm_crtc *crtc;
	int clock, pixel_size;
	int entries;

	crtc = intel_get_crtc_for_plane(dev, plane);
	if (crtc->fb == NULL || !crtc->enabled)
		return false;

	clock = crtc->mode.clock;	/* VESA DOT Clock */
	pixel_size = crtc->fb->bits_per_pixel / 8;	/* BPP */

	entries = (clock / 1000) * pixel_size;
	*plane_prec_mult = (entries > 256) ?
		DRAIN_LATENCY_PRECISION_32 : DRAIN_LATENCY_PRECISION_16;
	*plane_dl = (64 * (*plane_prec_mult) * 4) / ((clock / 1000) *
						     pixel_size);

	entries = (clock / 1000) * 4;	/* BPP is always 4 for cursor */
	*cursor_prec_mult = (entries > 256) ?
		DRAIN_LATENCY_PRECISION_32 : DRAIN_LATENCY_PRECISION_16;
	*cursor_dl = (64 * (*cursor_prec_mult) * 4) / ((clock / 1000) * 4);

	return true;
}

/*
 * Update drain latency registers of memory arbiter
 *
 * Valleyview SoC has a new memory arbiter and needs drain latency registers
 * to be programmed. Each plane has a drain latency multiplier and a drain
 * latency value.
 */

static void vlv_update_drain_latency(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int planea_prec, planea_dl, planeb_prec, planeb_dl;
	int cursora_prec, cursora_dl, cursorb_prec, cursorb_dl;
	int plane_prec_mult, cursor_prec_mult; /* Precision multiplier is
							either 16 or 32 */

	/* For plane A, Cursor A */
	if (vlv_compute_drain_latency(dev, 0, &plane_prec_mult, &planea_dl,
				      &cursor_prec_mult, &cursora_dl)) {
		cursora_prec = (cursor_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
			DDL_CURSORA_PRECISION_32 : DDL_CURSORA_PRECISION_16;
		planea_prec = (plane_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
			DDL_PLANEA_PRECISION_32 : DDL_PLANEA_PRECISION_16;

		I915_WRITE(VLV_DDL1, cursora_prec |
				(cursora_dl << DDL_CURSORA_SHIFT) |
				planea_prec | planea_dl);
	}

	/* For plane B, Cursor B */
	if (vlv_compute_drain_latency(dev, 1, &plane_prec_mult, &planeb_dl,
				      &cursor_prec_mult, &cursorb_dl)) {
		cursorb_prec = (cursor_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
			DDL_CURSORB_PRECISION_32 : DDL_CURSORB_PRECISION_16;
		planeb_prec = (plane_prec_mult == DRAIN_LATENCY_PRECISION_32) ?
			DDL_PLANEB_PRECISION_32 : DDL_PLANEB_PRECISION_16;

		I915_WRITE(VLV_DDL2, cursorb_prec |
				(cursorb_dl << DDL_CURSORB_SHIFT) |
				planeb_prec | planeb_dl);
	}
}

#define single_plane_enabled(mask) is_power_of_2(mask)

void valleyview_update_wm(struct drm_device *dev)
{
	static const int sr_latency_ns = 12000;
	struct drm_i915_private *dev_priv = dev->dev_private;
	int planea_wm, planeb_wm, cursora_wm, cursorb_wm;
	int plane_sr, cursor_sr;
	unsigned int enabled = 0;

	vlv_update_drain_latency(dev);

	if (g4x_compute_wm0(dev, 0,
			    &valleyview_wm_info, latency_ns,
			    &valleyview_cursor_wm_info, latency_ns,
			    &planea_wm, &cursora_wm))
		enabled |= 1;

	if (g4x_compute_wm0(dev, 1,
			    &valleyview_wm_info, latency_ns,
			    &valleyview_cursor_wm_info, latency_ns,
			    &planeb_wm, &cursorb_wm))
		enabled |= 2;

	plane_sr = cursor_sr = 0;
	if (single_plane_enabled(enabled) &&
	    g4x_compute_srwm(dev, ffs(enabled) - 1,
			     sr_latency_ns,
			     &valleyview_wm_info,
			     &valleyview_cursor_wm_info,
			     &plane_sr, &cursor_sr))
		I915_WRITE(FW_BLC_SELF_VLV, FW_CSPWRDWNEN);
	else
		I915_WRITE(FW_BLC_SELF_VLV,
			   I915_READ(FW_BLC_SELF_VLV) & ~FW_CSPWRDWNEN);

	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
		      planea_wm, cursora_wm,
		      planeb_wm, cursorb_wm,
		      plane_sr, cursor_sr);

	I915_WRITE(DSPFW1,
		   (plane_sr << DSPFW_SR_SHIFT) |
		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
		   planea_wm);
	I915_WRITE(DSPFW2,
		   (I915_READ(DSPFW2) & DSPFW_CURSORA_MASK) |
		   (cursora_wm << DSPFW_CURSORA_SHIFT));
	I915_WRITE(DSPFW3,
		   (I915_READ(DSPFW3) | (cursor_sr << DSPFW_CURSOR_SR_SHIFT)));
}

void g4x_update_wm(struct drm_device *dev)
{
	static const int sr_latency_ns = 12000;
	struct drm_i915_private *dev_priv = dev->dev_private;
	int planea_wm, planeb_wm, cursora_wm, cursorb_wm;
	int plane_sr, cursor_sr;
	unsigned int enabled = 0;

	if (g4x_compute_wm0(dev, 0,
			    &g4x_wm_info, latency_ns,
			    &g4x_cursor_wm_info, latency_ns,
			    &planea_wm, &cursora_wm))
		enabled |= 1;

	if (g4x_compute_wm0(dev, 1,
			    &g4x_wm_info, latency_ns,
			    &g4x_cursor_wm_info, latency_ns,
			    &planeb_wm, &cursorb_wm))
		enabled |= 2;

	plane_sr = cursor_sr = 0;
	if (single_plane_enabled(enabled) &&
	    g4x_compute_srwm(dev, ffs(enabled) - 1,
			     sr_latency_ns,
			     &g4x_wm_info,
			     &g4x_cursor_wm_info,
			     &plane_sr, &cursor_sr))
		I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN);
	else
		I915_WRITE(FW_BLC_SELF,
			   I915_READ(FW_BLC_SELF) & ~FW_BLC_SELF_EN);

	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
		      planea_wm, cursora_wm,
		      planeb_wm, cursorb_wm,
		      plane_sr, cursor_sr);

	I915_WRITE(DSPFW1,
		   (plane_sr << DSPFW_SR_SHIFT) |
		   (cursorb_wm << DSPFW_CURSORB_SHIFT) |
		   (planeb_wm << DSPFW_PLANEB_SHIFT) |
		   planea_wm);
	I915_WRITE(DSPFW2,
		   (I915_READ(DSPFW2) & DSPFW_CURSORA_MASK) |
		   (cursora_wm << DSPFW_CURSORA_SHIFT));
	/* HPLL off in SR has some issues on G4x... disable it */
	I915_WRITE(DSPFW3,
		   (I915_READ(DSPFW3) & ~DSPFW_HPLL_SR_EN) |
		   (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
}

void i965_update_wm(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	struct drm_crtc *crtc;
	int srwm = 1;
	int cursor_sr = 16;

	/* Calc sr entries for one plane configs */
	crtc = single_enabled_crtc(dev);
	if (crtc) {
		/* self-refresh has much higher latency */
		static const int sr_latency_ns = 12000;
		int clock = crtc->mode.clock;
		int htotal = crtc->mode.htotal;
		int hdisplay = crtc->mode.hdisplay;
		int pixel_size = crtc->fb->bits_per_pixel / 8;
		unsigned long line_time_us;
		int entries;

		line_time_us = ((htotal * 1000) / clock);

		/* Use ns/us then divide to preserve precision */
		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
			pixel_size * hdisplay;
		entries = DIV_ROUND_UP(entries, I915_FIFO_LINE_SIZE);
		srwm = I965_FIFO_SIZE - entries;
		if (srwm < 0)
			srwm = 1;
		srwm &= 0x1ff;
		DRM_DEBUG_KMS("self-refresh entries: %d, wm: %d\n",
			      entries, srwm);

		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
			pixel_size * 64;
		entries = DIV_ROUND_UP(entries,
					  i965_cursor_wm_info.cacheline_size);
		cursor_sr = i965_cursor_wm_info.fifo_size -
			(entries + i965_cursor_wm_info.guard_size);

		if (cursor_sr > i965_cursor_wm_info.max_wm)
			cursor_sr = i965_cursor_wm_info.max_wm;

		DRM_DEBUG_KMS("self-refresh watermark: display plane %d "
			      "cursor %d\n", srwm, cursor_sr);

		if (IS_CRESTLINE(dev))
			I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN);
	} else {
		/* Turn off self refresh if both pipes are enabled */
		if (IS_CRESTLINE(dev))
			I915_WRITE(FW_BLC_SELF, I915_READ(FW_BLC_SELF)
				   & ~FW_BLC_SELF_EN);
	}

	DRM_DEBUG_KMS("Setting FIFO watermarks - A: 8, B: 8, C: 8, SR %d\n",
		      srwm);

	/* 965 has limitations... */
	I915_WRITE(DSPFW1, (srwm << DSPFW_SR_SHIFT) |
		   (8 << 16) | (8 << 8) | (8 << 0));
	I915_WRITE(DSPFW2, (8 << 8) | (8 << 0));
	/* update cursor SR watermark */
	I915_WRITE(DSPFW3, (cursor_sr << DSPFW_CURSOR_SR_SHIFT));
}

void i9xx_update_wm(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	const struct intel_watermark_params *wm_info;
	uint32_t fwater_lo;
	uint32_t fwater_hi;
	int cwm, srwm = 1;
	int fifo_size;
	int planea_wm, planeb_wm;
	struct drm_crtc *crtc, *enabled = NULL;

	if (IS_I945GM(dev))
		wm_info = &i945_wm_info;
	else if (!IS_GEN2(dev))
		wm_info = &i915_wm_info;
	else
		wm_info = &i855_wm_info;

	fifo_size = dev_priv->display.get_fifo_size(dev, 0);
	crtc = intel_get_crtc_for_plane(dev, 0);
	if (crtc->enabled && crtc->fb) {
		planea_wm = intel_calculate_wm(crtc->mode.clock,
					       wm_info, fifo_size,
					       crtc->fb->bits_per_pixel / 8,
					       latency_ns);
		enabled = crtc;
	} else
		planea_wm = fifo_size - wm_info->guard_size;

	fifo_size = dev_priv->display.get_fifo_size(dev, 1);
	crtc = intel_get_crtc_for_plane(dev, 1);
	if (crtc->enabled && crtc->fb) {
		planeb_wm = intel_calculate_wm(crtc->mode.clock,
					       wm_info, fifo_size,
					       crtc->fb->bits_per_pixel / 8,
					       latency_ns);
		if (enabled == NULL)
			enabled = crtc;
		else
			enabled = NULL;
	} else
		planeb_wm = fifo_size - wm_info->guard_size;

	DRM_DEBUG_KMS("FIFO watermarks - A: %d, B: %d\n", planea_wm, planeb_wm);

	/*
	 * Overlay gets an aggressive default since video jitter is bad.
	 */
	cwm = 2;

	/* Play safe and disable self-refresh before adjusting watermarks. */
	if (IS_I945G(dev) || IS_I945GM(dev))
		I915_WRITE(FW_BLC_SELF, FW_BLC_SELF_EN_MASK | 0);
	else if (IS_I915GM(dev))
		I915_WRITE(INSTPM, I915_READ(INSTPM) & ~INSTPM_SELF_EN);

	/* Calc sr entries for one plane configs */
	if (HAS_FW_BLC(dev) && enabled) {
		/* self-refresh has much higher latency */
		static const int sr_latency_ns = 6000;
		int clock = enabled->mode.clock;
		int htotal = enabled->mode.htotal;
		int hdisplay = enabled->mode.hdisplay;
		int pixel_size = enabled->fb->bits_per_pixel / 8;
		unsigned long line_time_us;
		int entries;

		line_time_us = (htotal * 1000) / clock;

		/* Use ns/us then divide to preserve precision */
		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
			pixel_size * hdisplay;
		entries = DIV_ROUND_UP(entries, wm_info->cacheline_size);
		DRM_DEBUG_KMS("self-refresh entries: %d\n", entries);
		srwm = wm_info->fifo_size - entries;
		if (srwm < 0)
			srwm = 1;

		if (IS_I945G(dev) || IS_I945GM(dev))
			I915_WRITE(FW_BLC_SELF,
				   FW_BLC_SELF_FIFO_MASK | (srwm & 0xff));
		else if (IS_I915GM(dev))
			I915_WRITE(FW_BLC_SELF, srwm & 0x3f);
	}

	DRM_DEBUG_KMS("Setting FIFO watermarks - A: %d, B: %d, C: %d, SR %d\n",
		      planea_wm, planeb_wm, cwm, srwm);

	fwater_lo = ((planeb_wm & 0x3f) << 16) | (planea_wm & 0x3f);
	fwater_hi = (cwm & 0x1f);

	/* Set request length to 8 cachelines per fetch */
	fwater_lo = fwater_lo | (1 << 24) | (1 << 8);
	fwater_hi = fwater_hi | (1 << 8);

	I915_WRITE(FW_BLC, fwater_lo);
	I915_WRITE(FW_BLC2, fwater_hi);

	if (HAS_FW_BLC(dev)) {
		if (enabled) {
			if (IS_I945G(dev) || IS_I945GM(dev))
				I915_WRITE(FW_BLC_SELF,
					   FW_BLC_SELF_EN_MASK | FW_BLC_SELF_EN);
			else if (IS_I915GM(dev))
				I915_WRITE(INSTPM, I915_READ(INSTPM) | INSTPM_SELF_EN);
			DRM_DEBUG_KMS("memory self refresh enabled\n");
		} else
			DRM_DEBUG_KMS("memory self refresh disabled\n");
	}
}

void i830_update_wm(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	struct drm_crtc *crtc;
	uint32_t fwater_lo;
	int planea_wm;

	crtc = single_enabled_crtc(dev);
	if (crtc == NULL)
		return;

	planea_wm = intel_calculate_wm(crtc->mode.clock, &i830_wm_info,
				       dev_priv->display.get_fifo_size(dev, 0),
				       crtc->fb->bits_per_pixel / 8,
				       latency_ns);
	fwater_lo = I915_READ(FW_BLC) & ~0xfff;
	fwater_lo |= (3<<8) | planea_wm;

	DRM_DEBUG_KMS("Setting FIFO watermarks - A: %d\n", planea_wm);

	I915_WRITE(FW_BLC, fwater_lo);
}

#define ILK_LP0_PLANE_LATENCY		700
#define ILK_LP0_CURSOR_LATENCY		1300

/*
 * Check the wm result.
 *
 * If any calculated watermark values is larger than the maximum value that
 * can be programmed into the associated watermark register, that watermark
 * must be disabled.
 */
static bool ironlake_check_srwm(struct drm_device *dev, int level,
				int fbc_wm, int display_wm, int cursor_wm,
				const struct intel_watermark_params *display,
				const struct intel_watermark_params *cursor)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	DRM_DEBUG_KMS("watermark %d: display plane %d, fbc lines %d,"
		      " cursor %d\n", level, display_wm, fbc_wm, cursor_wm);

	if (fbc_wm > SNB_FBC_MAX_SRWM) {
		DRM_DEBUG_KMS("fbc watermark(%d) is too large(%d), disabling wm%d+\n",
			      fbc_wm, SNB_FBC_MAX_SRWM, level);

		/* fbc has it's own way to disable FBC WM */
		I915_WRITE(DISP_ARB_CTL,
			   I915_READ(DISP_ARB_CTL) | DISP_FBC_WM_DIS);
		return false;
	}

	if (display_wm > display->max_wm) {
		DRM_DEBUG_KMS("display watermark(%d) is too large(%d), disabling wm%d+\n",
			      display_wm, SNB_DISPLAY_MAX_SRWM, level);
		return false;
	}

	if (cursor_wm > cursor->max_wm) {
		DRM_DEBUG_KMS("cursor watermark(%d) is too large(%d), disabling wm%d+\n",
			      cursor_wm, SNB_CURSOR_MAX_SRWM, level);
		return false;
	}

	if (!(fbc_wm || display_wm || cursor_wm)) {
		DRM_DEBUG_KMS("latency %d is 0, disabling wm%d+\n", level, level);
		return false;
	}

	return true;
}

/*
 * Compute watermark values of WM[1-3],
 */
static bool ironlake_compute_srwm(struct drm_device *dev, int level, int plane,
				  int latency_ns,
				  const struct intel_watermark_params *display,
				  const struct intel_watermark_params *cursor,
				  int *fbc_wm, int *display_wm, int *cursor_wm)
{
	struct drm_crtc *crtc;
	unsigned long line_time_us;
	int hdisplay, htotal, pixel_size, clock;
	int line_count, line_size;
	int small, large;
	int entries;

	if (!latency_ns) {
		*fbc_wm = *display_wm = *cursor_wm = 0;
		return false;
	}

	crtc = intel_get_crtc_for_plane(dev, plane);
	hdisplay = crtc->mode.hdisplay;
	htotal = crtc->mode.htotal;
	clock = crtc->mode.clock;
	pixel_size = crtc->fb->bits_per_pixel / 8;

	line_time_us = (htotal * 1000) / clock;
	line_count = (latency_ns / line_time_us + 1000) / 1000;
	line_size = hdisplay * pixel_size;

	/* Use the minimum of the small and large buffer method for primary */
	small = ((clock * pixel_size / 1000) * latency_ns) / 1000;
	large = line_count * line_size;

	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
	*display_wm = entries + display->guard_size;

	/*
	 * Spec says:
	 * FBC WM = ((Final Primary WM * 64) / number of bytes per line) + 2
	 */
	*fbc_wm = DIV_ROUND_UP(*display_wm * 64, line_size) + 2;

	/* calculate the self-refresh watermark for display cursor */
	entries = line_count * pixel_size * 64;
	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
	*cursor_wm = entries + cursor->guard_size;

	return ironlake_check_srwm(dev, level,
				   *fbc_wm, *display_wm, *cursor_wm,
				   display, cursor);
}

void ironlake_update_wm(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int fbc_wm, plane_wm, cursor_wm;
	unsigned int enabled;

	enabled = 0;
	if (g4x_compute_wm0(dev, 0,
			    &ironlake_display_wm_info,
			    ILK_LP0_PLANE_LATENCY,
			    &ironlake_cursor_wm_info,
			    ILK_LP0_CURSOR_LATENCY,
			    &plane_wm, &cursor_wm)) {
		I915_WRITE(WM0_PIPEA_ILK,
			   (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm);
		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
			      " plane %d, " "cursor: %d\n",
			      plane_wm, cursor_wm);
		enabled |= 1;
	}

	if (g4x_compute_wm0(dev, 1,
			    &ironlake_display_wm_info,
			    ILK_LP0_PLANE_LATENCY,
			    &ironlake_cursor_wm_info,
			    ILK_LP0_CURSOR_LATENCY,
			    &plane_wm, &cursor_wm)) {
		I915_WRITE(WM0_PIPEB_ILK,
			   (plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm);
		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
			      " plane %d, cursor: %d\n",
			      plane_wm, cursor_wm);
		enabled |= 2;
	}

	/*
	 * Calculate and update the self-refresh watermark only when one
	 * display plane is used.
	 */
	I915_WRITE(WM3_LP_ILK, 0);
	I915_WRITE(WM2_LP_ILK, 0);
	I915_WRITE(WM1_LP_ILK, 0);

	if (!single_plane_enabled(enabled))
		return;
	enabled = ffs(enabled) - 1;

	/* WM1 */
	if (!ironlake_compute_srwm(dev, 1, enabled,
				   ILK_READ_WM1_LATENCY() * 500,
				   &ironlake_display_srwm_info,
				   &ironlake_cursor_srwm_info,
				   &fbc_wm, &plane_wm, &cursor_wm))
		return;

	I915_WRITE(WM1_LP_ILK,
		   WM1_LP_SR_EN |
		   (ILK_READ_WM1_LATENCY() << WM1_LP_LATENCY_SHIFT) |
		   (fbc_wm << WM1_LP_FBC_SHIFT) |
		   (plane_wm << WM1_LP_SR_SHIFT) |
		   cursor_wm);

	/* WM2 */
	if (!ironlake_compute_srwm(dev, 2, enabled,
				   ILK_READ_WM2_LATENCY() * 500,
				   &ironlake_display_srwm_info,
				   &ironlake_cursor_srwm_info,
				   &fbc_wm, &plane_wm, &cursor_wm))
		return;

	I915_WRITE(WM2_LP_ILK,
		   WM2_LP_EN |
		   (ILK_READ_WM2_LATENCY() << WM1_LP_LATENCY_SHIFT) |
		   (fbc_wm << WM1_LP_FBC_SHIFT) |
		   (plane_wm << WM1_LP_SR_SHIFT) |
		   cursor_wm);

	/*
	 * WM3 is unsupported on ILK, probably because we don't have latency
	 * data for that power state
	 */
}

void sandybridge_update_wm(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int latency = SNB_READ_WM0_LATENCY() * 100;	/* In unit 0.1us */
	u32 val;
	int fbc_wm, plane_wm, cursor_wm;
	unsigned int enabled;

	enabled = 0;
	if (g4x_compute_wm0(dev, 0,
			    &sandybridge_display_wm_info, latency,
			    &sandybridge_cursor_wm_info, latency,
			    &plane_wm, &cursor_wm)) {
		val = I915_READ(WM0_PIPEA_ILK);
		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
		I915_WRITE(WM0_PIPEA_ILK, val |
			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
		DRM_DEBUG_KMS("FIFO watermarks For pipe A -"
			      " plane %d, " "cursor: %d\n",
			      plane_wm, cursor_wm);
		enabled |= 1;
	}

	if (g4x_compute_wm0(dev, 1,
			    &sandybridge_display_wm_info, latency,
			    &sandybridge_cursor_wm_info, latency,
			    &plane_wm, &cursor_wm)) {
		val = I915_READ(WM0_PIPEB_ILK);
		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
		I915_WRITE(WM0_PIPEB_ILK, val |
			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
		DRM_DEBUG_KMS("FIFO watermarks For pipe B -"
			      " plane %d, cursor: %d\n",
			      plane_wm, cursor_wm);
		enabled |= 2;
	}

	/* IVB has 3 pipes */
	if (IS_IVYBRIDGE(dev) &&
	    g4x_compute_wm0(dev, 2,
			    &sandybridge_display_wm_info, latency,
			    &sandybridge_cursor_wm_info, latency,
			    &plane_wm, &cursor_wm)) {
		val = I915_READ(WM0_PIPEC_IVB);
		val &= ~(WM0_PIPE_PLANE_MASK | WM0_PIPE_CURSOR_MASK);
		I915_WRITE(WM0_PIPEC_IVB, val |
			   ((plane_wm << WM0_PIPE_PLANE_SHIFT) | cursor_wm));
		DRM_DEBUG_KMS("FIFO watermarks For pipe C -"
			      " plane %d, cursor: %d\n",
			      plane_wm, cursor_wm);
		enabled |= 3;
	}

	/*
	 * Calculate and update the self-refresh watermark only when one
	 * display plane is used.
	 *
	 * SNB support 3 levels of watermark.
	 *
	 * WM1/WM2/WM2 watermarks have to be enabled in the ascending order,
	 * and disabled in the descending order
	 *
	 */
	I915_WRITE(WM3_LP_ILK, 0);
	I915_WRITE(WM2_LP_ILK, 0);
	I915_WRITE(WM1_LP_ILK, 0);

	if (!single_plane_enabled(enabled) ||
	    dev_priv->sprite_scaling_enabled)
		return;
	enabled = ffs(enabled) - 1;

	/* WM1 */
	if (!ironlake_compute_srwm(dev, 1, enabled,
				   SNB_READ_WM1_LATENCY() * 500,
				   &sandybridge_display_srwm_info,
				   &sandybridge_cursor_srwm_info,
				   &fbc_wm, &plane_wm, &cursor_wm))
		return;

	I915_WRITE(WM1_LP_ILK,
		   WM1_LP_SR_EN |
		   (SNB_READ_WM1_LATENCY() << WM1_LP_LATENCY_SHIFT) |
		   (fbc_wm << WM1_LP_FBC_SHIFT) |
		   (plane_wm << WM1_LP_SR_SHIFT) |
		   cursor_wm);

	/* WM2 */
	if (!ironlake_compute_srwm(dev, 2, enabled,
				   SNB_READ_WM2_LATENCY() * 500,
				   &sandybridge_display_srwm_info,
				   &sandybridge_cursor_srwm_info,
				   &fbc_wm, &plane_wm, &cursor_wm))
		return;

	I915_WRITE(WM2_LP_ILK,
		   WM2_LP_EN |
		   (SNB_READ_WM2_LATENCY() << WM1_LP_LATENCY_SHIFT) |
		   (fbc_wm << WM1_LP_FBC_SHIFT) |
		   (plane_wm << WM1_LP_SR_SHIFT) |
		   cursor_wm);

	/* WM3 */
	if (!ironlake_compute_srwm(dev, 3, enabled,
				   SNB_READ_WM3_LATENCY() * 500,
				   &sandybridge_display_srwm_info,
				   &sandybridge_cursor_srwm_info,
				   &fbc_wm, &plane_wm, &cursor_wm))
		return;

	I915_WRITE(WM3_LP_ILK,
		   WM3_LP_EN |
		   (SNB_READ_WM3_LATENCY() << WM1_LP_LATENCY_SHIFT) |
		   (fbc_wm << WM1_LP_FBC_SHIFT) |
		   (plane_wm << WM1_LP_SR_SHIFT) |
		   cursor_wm);
}

static bool
sandybridge_compute_sprite_wm(struct drm_device *dev, int plane,
			      uint32_t sprite_width, int pixel_size,
			      const struct intel_watermark_params *display,
			      int display_latency_ns, int *sprite_wm)
{
	struct drm_crtc *crtc;
	int clock;
	int entries, tlb_miss;

	crtc = intel_get_crtc_for_plane(dev, plane);
	if (crtc->fb == NULL || !crtc->enabled) {
		*sprite_wm = display->guard_size;
		return false;
	}

	clock = crtc->mode.clock;

	/* Use the small buffer method to calculate the sprite watermark */
	entries = ((clock * pixel_size / 1000) * display_latency_ns) / 1000;
	tlb_miss = display->fifo_size*display->cacheline_size -
		sprite_width * 8;
	if (tlb_miss > 0)
		entries += tlb_miss;
	entries = DIV_ROUND_UP(entries, display->cacheline_size);
	*sprite_wm = entries + display->guard_size;
	if (*sprite_wm > (int)display->max_wm)
		*sprite_wm = display->max_wm;

	return true;
}

static bool
sandybridge_compute_sprite_srwm(struct drm_device *dev, int plane,
				uint32_t sprite_width, int pixel_size,
				const struct intel_watermark_params *display,
				int latency_ns, int *sprite_wm)
{
	struct drm_crtc *crtc;
	unsigned long line_time_us;
	int clock;
	int line_count, line_size;
	int small, large;
	int entries;

	if (!latency_ns) {
		*sprite_wm = 0;
		return false;
	}

	crtc = intel_get_crtc_for_plane(dev, plane);
	clock = crtc->mode.clock;
	if (!clock) {
		*sprite_wm = 0;
		return false;
	}

	line_time_us = (sprite_width * 1000) / clock;
	if (!line_time_us) {
		*sprite_wm = 0;
		return false;
	}

	line_count = (latency_ns / line_time_us + 1000) / 1000;
	line_size = sprite_width * pixel_size;

	/* Use the minimum of the small and large buffer method for primary */
	small = ((clock * pixel_size / 1000) * latency_ns) / 1000;
	large = line_count * line_size;

	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
	*sprite_wm = entries + display->guard_size;

	return *sprite_wm > 0x3ff ? false : true;
}

void sandybridge_update_sprite_wm(struct drm_device *dev, int pipe,
					 uint32_t sprite_width, int pixel_size)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int latency = SNB_READ_WM0_LATENCY() * 100;	/* In unit 0.1us */
	u32 val;
	int sprite_wm, reg;
	int ret;

	switch (pipe) {
	case 0:
		reg = WM0_PIPEA_ILK;
		break;
	case 1:
		reg = WM0_PIPEB_ILK;
		break;
	case 2:
		reg = WM0_PIPEC_IVB;
		break;
	default:
		return; /* bad pipe */
	}

	ret = sandybridge_compute_sprite_wm(dev, pipe, sprite_width, pixel_size,
					    &sandybridge_display_wm_info,
					    latency, &sprite_wm);
	if (!ret) {
		DRM_DEBUG_KMS("failed to compute sprite wm for pipe %d\n",
			      pipe);
		return;
	}

	val = I915_READ(reg);
	val &= ~WM0_PIPE_SPRITE_MASK;
	I915_WRITE(reg, val | (sprite_wm << WM0_PIPE_SPRITE_SHIFT));
	DRM_DEBUG_KMS("sprite watermarks For pipe %d - %d\n", pipe, sprite_wm);


	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
					      pixel_size,
					      &sandybridge_display_srwm_info,
					      SNB_READ_WM1_LATENCY() * 500,
					      &sprite_wm);
	if (!ret) {
		DRM_DEBUG_KMS("failed to compute sprite lp1 wm on pipe %d\n",
			      pipe);
		return;
	}
	I915_WRITE(WM1S_LP_ILK, sprite_wm);

	/* Only IVB has two more LP watermarks for sprite */
	if (!IS_IVYBRIDGE(dev))
		return;

	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
					      pixel_size,
					      &sandybridge_display_srwm_info,
					      SNB_READ_WM2_LATENCY() * 500,
					      &sprite_wm);
	if (!ret) {
		DRM_DEBUG_KMS("failed to compute sprite lp2 wm on pipe %d\n",
			      pipe);
		return;
	}
	I915_WRITE(WM2S_LP_IVB, sprite_wm);

	ret = sandybridge_compute_sprite_srwm(dev, pipe, sprite_width,
					      pixel_size,
					      &sandybridge_display_srwm_info,
					      SNB_READ_WM3_LATENCY() * 500,
					      &sprite_wm);
	if (!ret) {
		DRM_DEBUG_KMS("failed to compute sprite lp3 wm on pipe %d\n",
			      pipe);
		return;
	}
	I915_WRITE(WM3S_LP_IVB, sprite_wm);
}

/**
 * intel_update_watermarks - update FIFO watermark values based on current modes
 *
 * Calculate watermark values for the various WM regs based on current mode
 * and plane configuration.
 *
 * There are several cases to deal with here:
 *   - normal (i.e. non-self-refresh)
 *   - self-refresh (SR) mode
 *   - lines are large relative to FIFO size (buffer can hold up to 2)
 *   - lines are small relative to FIFO size (buffer can hold more than 2
 *     lines), so need to account for TLB latency
 *
 *   The normal calculation is:
 *     watermark = dotclock * bytes per pixel * latency
 *   where latency is platform & configuration dependent (we assume pessimal
 *   values here).
 *
 *   The SR calculation is:
 *     watermark = (trunc(latency/line time)+1) * surface width *
 *       bytes per pixel
 *   where
 *     line time = htotal / dotclock
 *     surface width = hdisplay for normal plane and 64 for cursor
 *   and latency is assumed to be high, as above.
 *
 * The final value programmed to the register should always be rounded up,
 * and include an extra 2 entries to account for clock crossings.
 *
 * We don't use the sprite, so we can ignore that.  And on Crestline we have
 * to set the non-SR watermarks to 8.
 */
void intel_update_watermarks(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	if (dev_priv->display.update_wm)
		dev_priv->display.update_wm(dev);
}

void intel_update_sprite_watermarks(struct drm_device *dev, int pipe,
				    uint32_t sprite_width, int pixel_size)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	if (dev_priv->display.update_sprite_wm)
		dev_priv->display.update_sprite_wm(dev, pipe, sprite_width,
						   pixel_size);
}

static struct drm_i915_gem_object *
intel_alloc_context_page(struct drm_device *dev)
{
	struct drm_i915_gem_object *ctx;
	int ret;

	WARN_ON(!mutex_is_locked(&dev->struct_mutex));

	ctx = i915_gem_alloc_object(dev, 4096);
	if (!ctx) {
		DRM_DEBUG("failed to alloc power context, RC6 disabled\n");
		return NULL;
	}

	ret = i915_gem_object_pin(ctx, 4096, true);
	if (ret) {
		DRM_ERROR("failed to pin power context: %d\n", ret);
		goto err_unref;
	}

	ret = i915_gem_object_set_to_gtt_domain(ctx, 1);
	if (ret) {
		DRM_ERROR("failed to set-domain on power context: %d\n", ret);
		goto err_unpin;
	}

	return ctx;

err_unpin:
	i915_gem_object_unpin(ctx);
err_unref:
	drm_gem_object_unreference(&ctx->base);
	mutex_unlock(&dev->struct_mutex);
	return NULL;
}

bool ironlake_set_drps(struct drm_device *dev, u8 val)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u16 rgvswctl;

	rgvswctl = I915_READ16(MEMSWCTL);
	if (rgvswctl & MEMCTL_CMD_STS) {
		DRM_DEBUG("gpu busy, RCS change rejected\n");
		return false; /* still busy with another command */
	}

	rgvswctl = (MEMCTL_CMD_CHFREQ << MEMCTL_CMD_SHIFT) |
		(val << MEMCTL_FREQ_SHIFT) | MEMCTL_SFCAVM;
	I915_WRITE16(MEMSWCTL, rgvswctl);
	POSTING_READ16(MEMSWCTL);

	rgvswctl |= MEMCTL_CMD_STS;
	I915_WRITE16(MEMSWCTL, rgvswctl);

	return true;
}

void ironlake_enable_drps(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 rgvmodectl = I915_READ(MEMMODECTL);
	u8 fmax, fmin, fstart, vstart;

	/* Enable temp reporting */
	I915_WRITE16(PMMISC, I915_READ(PMMISC) | MCPPCE_EN);
	I915_WRITE16(TSC1, I915_READ(TSC1) | TSE);

	/* 100ms RC evaluation intervals */
	I915_WRITE(RCUPEI, 100000);
	I915_WRITE(RCDNEI, 100000);

	/* Set max/min thresholds to 90ms and 80ms respectively */
	I915_WRITE(RCBMAXAVG, 90000);
	I915_WRITE(RCBMINAVG, 80000);

	I915_WRITE(MEMIHYST, 1);

	/* Set up min, max, and cur for interrupt handling */
	fmax = (rgvmodectl & MEMMODE_FMAX_MASK) >> MEMMODE_FMAX_SHIFT;
	fmin = (rgvmodectl & MEMMODE_FMIN_MASK);
	fstart = (rgvmodectl & MEMMODE_FSTART_MASK) >>
		MEMMODE_FSTART_SHIFT;

	vstart = (I915_READ(PXVFREQ_BASE + (fstart * 4)) & PXVFREQ_PX_MASK) >>
		PXVFREQ_PX_SHIFT;

	dev_priv->fmax = fmax; /* IPS callback will increase this */
	dev_priv->fstart = fstart;

	dev_priv->max_delay = fstart;
	dev_priv->min_delay = fmin;
	dev_priv->cur_delay = fstart;

	DRM_DEBUG_DRIVER("fmax: %d, fmin: %d, fstart: %d\n",
			 fmax, fmin, fstart);

	I915_WRITE(MEMINTREN, MEMINT_CX_SUPR_EN | MEMINT_EVAL_CHG_EN);

	/*
	 * Interrupts will be enabled in ironlake_irq_postinstall
	 */

	I915_WRITE(VIDSTART, vstart);
	POSTING_READ(VIDSTART);

	rgvmodectl |= MEMMODE_SWMODE_EN;
	I915_WRITE(MEMMODECTL, rgvmodectl);

	if (wait_for((I915_READ(MEMSWCTL) & MEMCTL_CMD_STS) == 0, 10))
		DRM_ERROR("stuck trying to change perf mode\n");
	msleep(1);

	ironlake_set_drps(dev, fstart);

	dev_priv->last_count1 = I915_READ(0x112e4) + I915_READ(0x112e8) +
		I915_READ(0x112e0);
	dev_priv->last_time1 = jiffies_to_msecs(jiffies);
	dev_priv->last_count2 = I915_READ(0x112f4);
	getrawmonotonic(&dev_priv->last_time2);
}

void ironlake_disable_drps(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u16 rgvswctl = I915_READ16(MEMSWCTL);

	/* Ack interrupts, disable EFC interrupt */
	I915_WRITE(MEMINTREN, I915_READ(MEMINTREN) & ~MEMINT_EVAL_CHG_EN);
	I915_WRITE(MEMINTRSTS, MEMINT_EVAL_CHG);
	I915_WRITE(DEIER, I915_READ(DEIER) & ~DE_PCU_EVENT);
	I915_WRITE(DEIIR, DE_PCU_EVENT);
	I915_WRITE(DEIMR, I915_READ(DEIMR) | DE_PCU_EVENT);

	/* Go back to the starting frequency */
	ironlake_set_drps(dev, dev_priv->fstart);
	msleep(1);
	rgvswctl |= MEMCTL_CMD_STS;
	I915_WRITE(MEMSWCTL, rgvswctl);
	msleep(1);

}

void gen6_set_rps(struct drm_device *dev, u8 val)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 swreq;

	swreq = (val & 0x3ff) << 25;
	I915_WRITE(GEN6_RPNSWREQ, swreq);
}

void gen6_disable_rps(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	I915_WRITE(GEN6_RPNSWREQ, 1 << 31);
	I915_WRITE(GEN6_PMINTRMSK, 0xffffffff);
	I915_WRITE(GEN6_PMIER, 0);
	/* Complete PM interrupt masking here doesn't race with the rps work
	 * item again unmasking PM interrupts because that is using a different
	 * register (PMIMR) to mask PM interrupts. The only risk is in leaving
	 * stale bits in PMIIR and PMIMR which gen6_enable_rps will clean up. */

	spin_lock_irq(&dev_priv->rps_lock);
	dev_priv->pm_iir = 0;
	spin_unlock_irq(&dev_priv->rps_lock);

	I915_WRITE(GEN6_PMIIR, I915_READ(GEN6_PMIIR));
}

int intel_enable_rc6(const struct drm_device *dev)
{
	/*
	 * Respect the kernel parameter if it is set
	 */
	if (i915_enable_rc6 >= 0)
		return i915_enable_rc6;

	/*
	 * Disable RC6 on Ironlake
	 */
	if (INTEL_INFO(dev)->gen == 5)
		return 0;

	/* Sorry Haswell, no RC6 for you for now. */
	if (IS_HASWELL(dev))
		return 0;

	/*
	 * Disable rc6 on Sandybridge
	 */
	if (INTEL_INFO(dev)->gen == 6) {
		DRM_DEBUG_DRIVER("Sandybridge: deep RC6 disabled\n");
		return INTEL_RC6_ENABLE;
	}
	DRM_DEBUG_DRIVER("RC6 and deep RC6 enabled\n");
	return (INTEL_RC6_ENABLE | INTEL_RC6p_ENABLE);
}

void gen6_enable_rps(struct drm_i915_private *dev_priv)
{
	u32 rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
	u32 gt_perf_status = I915_READ(GEN6_GT_PERF_STATUS);
	u32 pcu_mbox, rc6_mask = 0;
	u32 gtfifodbg;
	int cur_freq, min_freq, max_freq;
	int rc6_mode;
	int i;

	/* Here begins a magic sequence of register writes to enable
	 * auto-downclocking.
	 *
	 * Perhaps there might be some value in exposing these to
	 * userspace...
	 */
	I915_WRITE(GEN6_RC_STATE, 0);
	mutex_lock(&dev_priv->dev->struct_mutex);

	/* Clear the DBG now so we don't confuse earlier errors */
	if ((gtfifodbg = I915_READ(GTFIFODBG))) {
		DRM_ERROR("GT fifo had a previous error %x\n", gtfifodbg);
		I915_WRITE(GTFIFODBG, gtfifodbg);
	}

	gen6_gt_force_wake_get(dev_priv);

	/* disable the counters and set deterministic thresholds */
	I915_WRITE(GEN6_RC_CONTROL, 0);

	I915_WRITE(GEN6_RC1_WAKE_RATE_LIMIT, 1000 << 16);
	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 40 << 16 | 30);
	I915_WRITE(GEN6_RC6pp_WAKE_RATE_LIMIT, 30);
	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000);
	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25);

	for (i = 0; i < I915_NUM_RINGS; i++)
		I915_WRITE(RING_MAX_IDLE(dev_priv->ring[i].mmio_base), 10);

	I915_WRITE(GEN6_RC_SLEEP, 0);
	I915_WRITE(GEN6_RC1e_THRESHOLD, 1000);
	I915_WRITE(GEN6_RC6_THRESHOLD, 50000);
	I915_WRITE(GEN6_RC6p_THRESHOLD, 100000);
	I915_WRITE(GEN6_RC6pp_THRESHOLD, 64000); /* unused */

	rc6_mode = intel_enable_rc6(dev_priv->dev);
	if (rc6_mode & INTEL_RC6_ENABLE)
		rc6_mask |= GEN6_RC_CTL_RC6_ENABLE;

	if (rc6_mode & INTEL_RC6p_ENABLE)
		rc6_mask |= GEN6_RC_CTL_RC6p_ENABLE;

	if (rc6_mode & INTEL_RC6pp_ENABLE)
		rc6_mask |= GEN6_RC_CTL_RC6pp_ENABLE;

	DRM_INFO("Enabling RC6 states: RC6 %s, RC6p %s, RC6pp %s\n",
			(rc6_mode & INTEL_RC6_ENABLE) ? "on" : "off",
			(rc6_mode & INTEL_RC6p_ENABLE) ? "on" : "off",
			(rc6_mode & INTEL_RC6pp_ENABLE) ? "on" : "off");

	I915_WRITE(GEN6_RC_CONTROL,
		   rc6_mask |
		   GEN6_RC_CTL_EI_MODE(1) |
		   GEN6_RC_CTL_HW_ENABLE);

	I915_WRITE(GEN6_RPNSWREQ,
		   GEN6_FREQUENCY(10) |
		   GEN6_OFFSET(0) |
		   GEN6_AGGRESSIVE_TURBO);
	I915_WRITE(GEN6_RC_VIDEO_FREQ,
		   GEN6_FREQUENCY(12));

	I915_WRITE(GEN6_RP_DOWN_TIMEOUT, 1000000);
	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS,
		   18 << 24 |
		   6 << 16);
	I915_WRITE(GEN6_RP_UP_THRESHOLD, 10000);
	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 1000000);
	I915_WRITE(GEN6_RP_UP_EI, 100000);
	I915_WRITE(GEN6_RP_DOWN_EI, 5000000);
	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
	I915_WRITE(GEN6_RP_CONTROL,
		   GEN6_RP_MEDIA_TURBO |
		   GEN6_RP_MEDIA_HW_MODE |
		   GEN6_RP_MEDIA_IS_GFX |
		   GEN6_RP_ENABLE |
		   GEN6_RP_UP_BUSY_AVG |
		   GEN6_RP_DOWN_IDLE_CONT);

	if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0,
		     500))
		DRM_ERROR("timeout waiting for pcode mailbox to become idle\n");

	I915_WRITE(GEN6_PCODE_DATA, 0);
	I915_WRITE(GEN6_PCODE_MAILBOX,
		   GEN6_PCODE_READY |
		   GEN6_PCODE_WRITE_MIN_FREQ_TABLE);
	if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0,
		     500))
		DRM_ERROR("timeout waiting for pcode mailbox to finish\n");

	min_freq = (rp_state_cap & 0xff0000) >> 16;
	max_freq = rp_state_cap & 0xff;
	cur_freq = (gt_perf_status & 0xff00) >> 8;

	/* Check for overclock support */
	if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0,
		     500))
		DRM_ERROR("timeout waiting for pcode mailbox to become idle\n");
	I915_WRITE(GEN6_PCODE_MAILBOX, GEN6_READ_OC_PARAMS);
	pcu_mbox = I915_READ(GEN6_PCODE_DATA);
	if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) & GEN6_PCODE_READY) == 0,
		     500))
		DRM_ERROR("timeout waiting for pcode mailbox to finish\n");
	if (pcu_mbox & (1<<31)) { /* OC supported */
		max_freq = pcu_mbox & 0xff;
		DRM_DEBUG_DRIVER("overclocking supported, adjusting frequency max to %dMHz\n", pcu_mbox * 50);
	}

	/* In units of 100MHz */
	dev_priv->max_delay = max_freq;
	dev_priv->min_delay = min_freq;
	dev_priv->cur_delay = cur_freq;

	/* requires MSI enabled */
	I915_WRITE(GEN6_PMIER,
		   GEN6_PM_MBOX_EVENT |
		   GEN6_PM_THERMAL_EVENT |
		   GEN6_PM_RP_DOWN_TIMEOUT |
		   GEN6_PM_RP_UP_THRESHOLD |
		   GEN6_PM_RP_DOWN_THRESHOLD |
		   GEN6_PM_RP_UP_EI_EXPIRED |
		   GEN6_PM_RP_DOWN_EI_EXPIRED);
	spin_lock_irq(&dev_priv->rps_lock);
	WARN_ON(dev_priv->pm_iir != 0);
	I915_WRITE(GEN6_PMIMR, 0);
	spin_unlock_irq(&dev_priv->rps_lock);
	/* enable all PM interrupts */
	I915_WRITE(GEN6_PMINTRMSK, 0);

	gen6_gt_force_wake_put(dev_priv);
	mutex_unlock(&dev_priv->dev->struct_mutex);
}

void gen6_update_ring_freq(struct drm_i915_private *dev_priv)
{
	int min_freq = 15;
	int gpu_freq, ia_freq, max_ia_freq;
	int scaling_factor = 180;

	max_ia_freq = cpufreq_quick_get_max(0);
	/*
	 * Default to measured freq if none found, PCU will ensure we don't go
	 * over
	 */
	if (!max_ia_freq)
		max_ia_freq = tsc_khz;

	/* Convert from kHz to MHz */
	max_ia_freq /= 1000;

	mutex_lock(&dev_priv->dev->struct_mutex);

	/*
	 * For each potential GPU frequency, load a ring frequency we'd like
	 * to use for memory access.  We do this by specifying the IA frequency
	 * the PCU should use as a reference to determine the ring frequency.
	 */
	for (gpu_freq = dev_priv->max_delay; gpu_freq >= dev_priv->min_delay;
	     gpu_freq--) {
		int diff = dev_priv->max_delay - gpu_freq;

		/*
		 * For GPU frequencies less than 750MHz, just use the lowest
		 * ring freq.
		 */
		if (gpu_freq < min_freq)
			ia_freq = 800;
		else
			ia_freq = max_ia_freq - ((diff * scaling_factor) / 2);
		ia_freq = DIV_ROUND_CLOSEST(ia_freq, 100);

		I915_WRITE(GEN6_PCODE_DATA,
			   (ia_freq << GEN6_PCODE_FREQ_IA_RATIO_SHIFT) |
			   gpu_freq);
		I915_WRITE(GEN6_PCODE_MAILBOX, GEN6_PCODE_READY |
			   GEN6_PCODE_WRITE_MIN_FREQ_TABLE);
		if (wait_for((I915_READ(GEN6_PCODE_MAILBOX) &
			      GEN6_PCODE_READY) == 0, 10)) {
			DRM_ERROR("pcode write of freq table timed out\n");
			continue;
		}
	}

	mutex_unlock(&dev_priv->dev->struct_mutex);
}

static void ironlake_teardown_rc6(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	if (dev_priv->renderctx) {
		i915_gem_object_unpin(dev_priv->renderctx);
		drm_gem_object_unreference(&dev_priv->renderctx->base);
		dev_priv->renderctx = NULL;
	}

	if (dev_priv->pwrctx) {
		i915_gem_object_unpin(dev_priv->pwrctx);
		drm_gem_object_unreference(&dev_priv->pwrctx->base);
		dev_priv->pwrctx = NULL;
	}
}

void ironlake_disable_rc6(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	if (I915_READ(PWRCTXA)) {
		/* Wake the GPU, prevent RC6, then restore RSTDBYCTL */
		I915_WRITE(RSTDBYCTL, I915_READ(RSTDBYCTL) | RCX_SW_EXIT);
		wait_for(((I915_READ(RSTDBYCTL) & RSX_STATUS_MASK) == RSX_STATUS_ON),
			 50);

		I915_WRITE(PWRCTXA, 0);
		POSTING_READ(PWRCTXA);

		I915_WRITE(RSTDBYCTL, I915_READ(RSTDBYCTL) & ~RCX_SW_EXIT);
		POSTING_READ(RSTDBYCTL);
	}

	ironlake_teardown_rc6(dev);
}

static int ironlake_setup_rc6(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	if (dev_priv->renderctx == NULL)
		dev_priv->renderctx = intel_alloc_context_page(dev);
	if (!dev_priv->renderctx)
		return -ENOMEM;

	if (dev_priv->pwrctx == NULL)
		dev_priv->pwrctx = intel_alloc_context_page(dev);
	if (!dev_priv->pwrctx) {
		ironlake_teardown_rc6(dev);
		return -ENOMEM;
	}

	return 0;
}

void ironlake_enable_rc6(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int ret;

	/* rc6 disabled by default due to repeated reports of hanging during
	 * boot and resume.
	 */
	if (!intel_enable_rc6(dev))
		return;

	mutex_lock(&dev->struct_mutex);
	ret = ironlake_setup_rc6(dev);
	if (ret) {
		mutex_unlock(&dev->struct_mutex);
		return;
	}

	/*
	 * GPU can automatically power down the render unit if given a page
	 * to save state.
	 */
	ret = BEGIN_LP_RING(6);
	if (ret) {
		ironlake_teardown_rc6(dev);
		mutex_unlock(&dev->struct_mutex);
		return;
	}

	OUT_RING(MI_SUSPEND_FLUSH | MI_SUSPEND_FLUSH_EN);
	OUT_RING(MI_SET_CONTEXT);
	OUT_RING(dev_priv->renderctx->gtt_offset |
		 MI_MM_SPACE_GTT |
		 MI_SAVE_EXT_STATE_EN |
		 MI_RESTORE_EXT_STATE_EN |
		 MI_RESTORE_INHIBIT);
	OUT_RING(MI_SUSPEND_FLUSH);
	OUT_RING(MI_NOOP);
	OUT_RING(MI_FLUSH);
	ADVANCE_LP_RING();

	/*
	 * Wait for the command parser to advance past MI_SET_CONTEXT. The HW
	 * does an implicit flush, combined with MI_FLUSH above, it should be
	 * safe to assume that renderctx is valid
	 */
	ret = intel_wait_ring_idle(LP_RING(dev_priv));
	if (ret) {
		DRM_ERROR("failed to enable ironlake power power savings\n");
		ironlake_teardown_rc6(dev);
		mutex_unlock(&dev->struct_mutex);
		return;
	}

	I915_WRITE(PWRCTXA, dev_priv->pwrctx->gtt_offset | PWRCTX_EN);
	I915_WRITE(RSTDBYCTL, I915_READ(RSTDBYCTL) & ~RCX_SW_EXIT);
	mutex_unlock(&dev->struct_mutex);
}

static unsigned long intel_pxfreq(u32 vidfreq)
{
	unsigned long freq;
	int div = (vidfreq & 0x3f0000) >> 16;
	int post = (vidfreq & 0x3000) >> 12;
	int pre = (vidfreq & 0x7);

	if (!pre)
		return 0;

	freq = ((div * 133333) / ((1<<post) * pre));

	return freq;
}

void intel_init_emon(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 lcfuse;
	u8 pxw[16];
	int i;

	/* Disable to program */
	I915_WRITE(ECR, 0);
	POSTING_READ(ECR);

	/* Program energy weights for various events */
	I915_WRITE(SDEW, 0x15040d00);
	I915_WRITE(CSIEW0, 0x007f0000);
	I915_WRITE(CSIEW1, 0x1e220004);
	I915_WRITE(CSIEW2, 0x04000004);

	for (i = 0; i < 5; i++)
		I915_WRITE(PEW + (i * 4), 0);
	for (i = 0; i < 3; i++)
		I915_WRITE(DEW + (i * 4), 0);

	/* Program P-state weights to account for frequency power adjustment */
	for (i = 0; i < 16; i++) {
		u32 pxvidfreq = I915_READ(PXVFREQ_BASE + (i * 4));
		unsigned long freq = intel_pxfreq(pxvidfreq);
		unsigned long vid = (pxvidfreq & PXVFREQ_PX_MASK) >>
			PXVFREQ_PX_SHIFT;
		unsigned long val;

		val = vid * vid;
		val *= (freq / 1000);
		val *= 255;
		val /= (127*127*900);
		if (val > 0xff)
			DRM_ERROR("bad pxval: %ld\n", val);
		pxw[i] = val;
	}
	/* Render standby states get 0 weight */
	pxw[14] = 0;
	pxw[15] = 0;

	for (i = 0; i < 4; i++) {
		u32 val = (pxw[i*4] << 24) | (pxw[(i*4)+1] << 16) |
			(pxw[(i*4)+2] << 8) | (pxw[(i*4)+3]);
		I915_WRITE(PXW + (i * 4), val);
	}

	/* Adjust magic regs to magic values (more experimental results) */
	I915_WRITE(OGW0, 0);
	I915_WRITE(OGW1, 0);
	I915_WRITE(EG0, 0x00007f00);
	I915_WRITE(EG1, 0x0000000e);
	I915_WRITE(EG2, 0x000e0000);
	I915_WRITE(EG3, 0x68000300);
	I915_WRITE(EG4, 0x42000000);
	I915_WRITE(EG5, 0x00140031);
	I915_WRITE(EG6, 0);
	I915_WRITE(EG7, 0);

	for (i = 0; i < 8; i++)
		I915_WRITE(PXWL + (i * 4), 0);

	/* Enable PMON + select events */
	I915_WRITE(ECR, 0x80000019);

	lcfuse = I915_READ(LCFUSE02);

	dev_priv->corr = (lcfuse & LCFUSE_HIV_MASK);
}

void ironlake_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE;

	/* Required for FBC */
	dspclk_gate |= DPFCUNIT_CLOCK_GATE_DISABLE |
		DPFCRUNIT_CLOCK_GATE_DISABLE |
		DPFDUNIT_CLOCK_GATE_DISABLE;
	/* Required for CxSR */
	dspclk_gate |= DPARBUNIT_CLOCK_GATE_DISABLE;

	I915_WRITE(PCH_3DCGDIS0,
		   MARIUNIT_CLOCK_GATE_DISABLE |
		   SVSMUNIT_CLOCK_GATE_DISABLE);
	I915_WRITE(PCH_3DCGDIS1,
		   VFMUNIT_CLOCK_GATE_DISABLE);

	I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate);

	/*
	 * According to the spec the following bits should be set in
	 * order to enable memory self-refresh
	 * The bit 22/21 of 0x42004
	 * The bit 5 of 0x42020
	 * The bit 15 of 0x45000
	 */
	I915_WRITE(ILK_DISPLAY_CHICKEN2,
		   (I915_READ(ILK_DISPLAY_CHICKEN2) |
		    ILK_DPARB_GATE | ILK_VSDPFD_FULL));
	I915_WRITE(ILK_DSPCLK_GATE,
		   (I915_READ(ILK_DSPCLK_GATE) |
		    ILK_DPARB_CLK_GATE));
	I915_WRITE(DISP_ARB_CTL,
		   (I915_READ(DISP_ARB_CTL) |
		    DISP_FBC_WM_DIS));
	I915_WRITE(WM3_LP_ILK, 0);
	I915_WRITE(WM2_LP_ILK, 0);
	I915_WRITE(WM1_LP_ILK, 0);

	/*
	 * Based on the document from hardware guys the following bits
	 * should be set unconditionally in order to enable FBC.
	 * The bit 22 of 0x42000
	 * The bit 22 of 0x42004
	 * The bit 7,8,9 of 0x42020.
	 */
	if (IS_IRONLAKE_M(dev)) {
		I915_WRITE(ILK_DISPLAY_CHICKEN1,
			   I915_READ(ILK_DISPLAY_CHICKEN1) |
			   ILK_FBCQ_DIS);
		I915_WRITE(ILK_DISPLAY_CHICKEN2,
			   I915_READ(ILK_DISPLAY_CHICKEN2) |
			   ILK_DPARB_GATE);
		I915_WRITE(ILK_DSPCLK_GATE,
			   I915_READ(ILK_DSPCLK_GATE) |
			   ILK_DPFC_DIS1 |
			   ILK_DPFC_DIS2 |
			   ILK_CLK_FBC);
	}

	I915_WRITE(ILK_DISPLAY_CHICKEN2,
		   I915_READ(ILK_DISPLAY_CHICKEN2) |
		   ILK_ELPIN_409_SELECT);
	I915_WRITE(_3D_CHICKEN2,
		   _3D_CHICKEN2_WM_READ_PIPELINED << 16 |
		   _3D_CHICKEN2_WM_READ_PIPELINED);
}

void gen6_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int pipe;
	uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE;

	I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate);

	I915_WRITE(ILK_DISPLAY_CHICKEN2,
		   I915_READ(ILK_DISPLAY_CHICKEN2) |
		   ILK_ELPIN_409_SELECT);

	I915_WRITE(WM3_LP_ILK, 0);
	I915_WRITE(WM2_LP_ILK, 0);
	I915_WRITE(WM1_LP_ILK, 0);

	/* clear masked bit */
	I915_WRITE(CACHE_MODE_0,
		   CM0_STC_EVICT_DISABLE_LRA_SNB << CM0_MASK_SHIFT);

	I915_WRITE(GEN6_UCGCTL1,
		   I915_READ(GEN6_UCGCTL1) |
		   GEN6_BLBUNIT_CLOCK_GATE_DISABLE |
		   GEN6_CSUNIT_CLOCK_GATE_DISABLE);

	/* According to the BSpec vol1g, bit 12 (RCPBUNIT) clock
	 * gating disable must be set.  Failure to set it results in
	 * flickering pixels due to Z write ordering failures after
	 * some amount of runtime in the Mesa "fire" demo, and Unigine
	 * Sanctuary and Tropics, and apparently anything else with
	 * alpha test or pixel discard.
	 *
	 * According to the spec, bit 11 (RCCUNIT) must also be set,
	 * but we didn't debug actual testcases to find it out.
	 */
	I915_WRITE(GEN6_UCGCTL2,
		   GEN6_RCPBUNIT_CLOCK_GATE_DISABLE |
		   GEN6_RCCUNIT_CLOCK_GATE_DISABLE);

	/* Bspec says we need to always set all mask bits. */
	I915_WRITE(_3D_CHICKEN, (0xFFFF << 16) |
		   _3D_CHICKEN_SF_DISABLE_FASTCLIP_CULL);

	/*
	 * According to the spec the following bits should be
	 * set in order to enable memory self-refresh and fbc:
	 * The bit21 and bit22 of 0x42000
	 * The bit21 and bit22 of 0x42004
	 * The bit5 and bit7 of 0x42020
	 * The bit14 of 0x70180
	 * The bit14 of 0x71180
	 */
	I915_WRITE(ILK_DISPLAY_CHICKEN1,
		   I915_READ(ILK_DISPLAY_CHICKEN1) |
		   ILK_FBCQ_DIS | ILK_PABSTRETCH_DIS);
	I915_WRITE(ILK_DISPLAY_CHICKEN2,
		   I915_READ(ILK_DISPLAY_CHICKEN2) |
		   ILK_DPARB_GATE | ILK_VSDPFD_FULL);
	I915_WRITE(ILK_DSPCLK_GATE,
		   I915_READ(ILK_DSPCLK_GATE) |
		   ILK_DPARB_CLK_GATE  |
		   ILK_DPFD_CLK_GATE);

	for_each_pipe(pipe) {
		I915_WRITE(DSPCNTR(pipe),
			   I915_READ(DSPCNTR(pipe)) |
			   DISPPLANE_TRICKLE_FEED_DISABLE);
		intel_flush_display_plane(dev_priv, pipe);
	}
}

static void gen7_setup_fixed_func_scheduler(struct drm_i915_private *dev_priv)
{
	uint32_t reg = I915_READ(GEN7_FF_THREAD_MODE);

	reg &= ~GEN7_FF_SCHED_MASK;
	reg |= GEN7_FF_TS_SCHED_HW;
	reg |= GEN7_FF_VS_SCHED_HW;
	reg |= GEN7_FF_DS_SCHED_HW;

	I915_WRITE(GEN7_FF_THREAD_MODE, reg);
}

void ivybridge_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int pipe;
	uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE;

	I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate);

	I915_WRITE(WM3_LP_ILK, 0);
	I915_WRITE(WM2_LP_ILK, 0);
	I915_WRITE(WM1_LP_ILK, 0);

	/* According to the spec, bit 13 (RCZUNIT) must be set on IVB.
	 * This implements the WaDisableRCZUnitClockGating workaround.
	 */
	I915_WRITE(GEN6_UCGCTL2, GEN6_RCZUNIT_CLOCK_GATE_DISABLE);

	I915_WRITE(ILK_DSPCLK_GATE, IVB_VRHUNIT_CLK_GATE);

	I915_WRITE(IVB_CHICKEN3,
		   CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE |
		   CHICKEN3_DGMG_DONE_FIX_DISABLE);

	/* Apply the WaDisableRHWOOptimizationForRenderHang workaround. */
	I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1,
		   GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC);

	/* WaApplyL3ControlAndL3ChickenMode requires those two on Ivy Bridge */
	I915_WRITE(GEN7_L3CNTLREG1,
			GEN7_WA_FOR_GEN7_L3_CONTROL);
	I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER,
			GEN7_WA_L3_CHICKEN_MODE);

	/* This is required by WaCatErrorRejectionIssue */
	I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG,
			I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) |
			GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB);

	for_each_pipe(pipe) {
		I915_WRITE(DSPCNTR(pipe),
			   I915_READ(DSPCNTR(pipe)) |
			   DISPPLANE_TRICKLE_FEED_DISABLE);
		intel_flush_display_plane(dev_priv, pipe);
	}

	gen7_setup_fixed_func_scheduler(dev_priv);
}

void valleyview_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int pipe;
	uint32_t dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE;

	I915_WRITE(PCH_DSPCLK_GATE_D, dspclk_gate);

	I915_WRITE(WM3_LP_ILK, 0);
	I915_WRITE(WM2_LP_ILK, 0);
	I915_WRITE(WM1_LP_ILK, 0);

	/* According to the spec, bit 13 (RCZUNIT) must be set on IVB.
	 * This implements the WaDisableRCZUnitClockGating workaround.
	 */
	I915_WRITE(GEN6_UCGCTL2, GEN6_RCZUNIT_CLOCK_GATE_DISABLE);

	I915_WRITE(ILK_DSPCLK_GATE, IVB_VRHUNIT_CLK_GATE);

	I915_WRITE(IVB_CHICKEN3,
		   CHICKEN3_DGMG_REQ_OUT_FIX_DISABLE |
		   CHICKEN3_DGMG_DONE_FIX_DISABLE);

	/* Apply the WaDisableRHWOOptimizationForRenderHang workaround. */
	I915_WRITE(GEN7_COMMON_SLICE_CHICKEN1,
		   GEN7_CSC1_RHWO_OPT_DISABLE_IN_RCC);

	/* WaApplyL3ControlAndL3ChickenMode requires those two on Ivy Bridge */
	I915_WRITE(GEN7_L3CNTLREG1, GEN7_WA_FOR_GEN7_L3_CONTROL);
	I915_WRITE(GEN7_L3_CHICKEN_MODE_REGISTER, GEN7_WA_L3_CHICKEN_MODE);

	/* This is required by WaCatErrorRejectionIssue */
	I915_WRITE(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG,
		   I915_READ(GEN7_SQ_CHICKEN_MBCUNIT_CONFIG) |
		   GEN7_SQ_CHICKEN_MBCUNIT_SQINTMOB);

	for_each_pipe(pipe) {
		I915_WRITE(DSPCNTR(pipe),
			   I915_READ(DSPCNTR(pipe)) |
			   DISPPLANE_TRICKLE_FEED_DISABLE);
		intel_flush_display_plane(dev_priv, pipe);
	}

	I915_WRITE(CACHE_MODE_1, I915_READ(CACHE_MODE_1) |
		   (PIXEL_SUBSPAN_COLLECT_OPT_DISABLE << 16) |
		   PIXEL_SUBSPAN_COLLECT_OPT_DISABLE);
}

void g4x_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	uint32_t dspclk_gate;

	I915_WRITE(RENCLK_GATE_D1, 0);
	I915_WRITE(RENCLK_GATE_D2, VF_UNIT_CLOCK_GATE_DISABLE |
		   GS_UNIT_CLOCK_GATE_DISABLE |
		   CL_UNIT_CLOCK_GATE_DISABLE);
	I915_WRITE(RAMCLK_GATE_D, 0);
	dspclk_gate = VRHUNIT_CLOCK_GATE_DISABLE |
		OVRUNIT_CLOCK_GATE_DISABLE |
		OVCUNIT_CLOCK_GATE_DISABLE;
	if (IS_GM45(dev))
		dspclk_gate |= DSSUNIT_CLOCK_GATE_DISABLE;
	I915_WRITE(DSPCLK_GATE_D, dspclk_gate);
}

void crestline_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	I915_WRITE(RENCLK_GATE_D1, I965_RCC_CLOCK_GATE_DISABLE);
	I915_WRITE(RENCLK_GATE_D2, 0);
	I915_WRITE(DSPCLK_GATE_D, 0);
	I915_WRITE(RAMCLK_GATE_D, 0);
	I915_WRITE16(DEUC, 0);
}

void broadwater_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	I915_WRITE(RENCLK_GATE_D1, I965_RCZ_CLOCK_GATE_DISABLE |
		   I965_RCC_CLOCK_GATE_DISABLE |
		   I965_RCPB_CLOCK_GATE_DISABLE |
		   I965_ISC_CLOCK_GATE_DISABLE |
		   I965_FBC_CLOCK_GATE_DISABLE);
	I915_WRITE(RENCLK_GATE_D2, 0);
}

void gen3_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	u32 dstate = I915_READ(D_STATE);

	dstate |= DSTATE_PLL_D3_OFF | DSTATE_GFX_CLOCK_GATING |
		DSTATE_DOT_CLOCK_GATING;
	I915_WRITE(D_STATE, dstate);
}

void i85x_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	I915_WRITE(RENCLK_GATE_D1, SV_CLOCK_GATE_DISABLE);
}

void i830_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	I915_WRITE(DSPCLK_GATE_D, OVRUNIT_CLOCK_GATE_DISABLE);
}

void ibx_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	/*
	 * On Ibex Peak and Cougar Point, we need to disable clock
	 * gating for the panel power sequencer or it will fail to
	 * start up when no ports are active.
	 */
	I915_WRITE(SOUTH_DSPCLK_GATE_D, PCH_DPLSUNIT_CLOCK_GATE_DISABLE);
}

void cpt_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;
	int pipe;

	/*
	 * On Ibex Peak and Cougar Point, we need to disable clock
	 * gating for the panel power sequencer or it will fail to
	 * start up when no ports are active.
	 */
	I915_WRITE(SOUTH_DSPCLK_GATE_D, PCH_DPLSUNIT_CLOCK_GATE_DISABLE);
	I915_WRITE(SOUTH_CHICKEN2, I915_READ(SOUTH_CHICKEN2) |
		   DPLS_EDP_PPS_FIX_DIS);
	/* Without this, mode sets may fail silently on FDI */
	for_each_pipe(pipe)
		I915_WRITE(TRANS_CHICKEN2(pipe), TRANS_AUTOTRAIN_GEN_STALL_DIS);
}

void intel_init_clock_gating(struct drm_device *dev)
{
	struct drm_i915_private *dev_priv = dev->dev_private;

	dev_priv->display.init_clock_gating(dev);

	if (dev_priv->display.init_pch_clock_gating)
		dev_priv->display.init_pch_clock_gating(dev);
}