/* Check pitch constriants for all chips & tiling formats */
static bool
i915_tiling_ok(struct drm_device *dev, int stride, int size, int tiling_mode)
{
int tile_width;
/* Linear is always fine */
if (tiling_mode == I915_TILING_NONE)
return true;
if (IS_GEN2(dev) ||
(tiling_mode == I915_TILING_Y && HAS_128_BYTE_Y_TILING(dev)))
tile_width = 128;
else
tile_width = 512;
/* check maximum stride & object size */
if (INTEL_INFO(dev)->gen >= 4) {
/* i965 stores the end address of the gtt mapping in the fence
* reg, so dont bother to check the size */
if (stride / 128 > I965_FENCE_MAX_PITCH_VAL)
return false;
} else {
if (stride > 8192)
return false;
if (IS_GEN3(dev)) {
if (size > I830_FENCE_MAX_SIZE_VAL << 20)
return false;
} else {
if (size > I830_FENCE_MAX_SIZE_VAL << 19)
return false;
}
}
/* 965+ just needs multiples of tile width */
if (INTEL_INFO(dev)->gen >= 4) {
if (stride & (tile_width - 1))
return false;
return true;
}
/* Pre-965 needs power of two tile widths */
if (stride < tile_width)
return false;
if (stride & (stride - 1))
return false;
return true;
}
static void i8xx_fbc_enable(struct intel_crtc *crtc)
{
struct drm_i915_private *dev_priv = crtc->base.dev->dev_private;
struct drm_framebuffer *fb = crtc->base.primary->fb;
struct drm_i915_gem_object *obj = intel_fb_obj(fb);
int cfb_pitch;
int i;
u32 fbc_ctl;
dev_priv->fbc.enabled = true;
/* Note: fbc.threshold == 1 for i8xx */
cfb_pitch = dev_priv->fbc.uncompressed_size / FBC_LL_SIZE;
if (fb->pitches[0] < cfb_pitch)
cfb_pitch = fb->pitches[0];
/* FBC_CTL wants 32B or 64B units */
if (IS_GEN2(dev_priv))
cfb_pitch = (cfb_pitch / 32) - 1;
else
cfb_pitch = (cfb_pitch / 64) - 1;
/* Clear old tags */
for (i = 0; i < (FBC_LL_SIZE / 32) + 1; i++)
I915_WRITE(FBC_TAG(i), 0);
if (IS_GEN4(dev_priv)) {
u32 fbc_ctl2;
/* Set it up... */
fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
fbc_ctl2 |= FBC_CTL_PLANE(crtc->plane);
I915_WRITE(FBC_CONTROL2, fbc_ctl2);
I915_WRITE(FBC_FENCE_OFF, get_crtc_fence_y_offset(crtc));
}
/* enable it... */
fbc_ctl = I915_READ(FBC_CONTROL);
fbc_ctl &= 0x3fff << FBC_CTL_INTERVAL_SHIFT;
fbc_ctl |= FBC_CTL_EN | FBC_CTL_PERIODIC;
if (IS_I945GM(dev_priv))
fbc_ctl |= FBC_CTL_C3_IDLE; /* 945 needs special SR handling */
fbc_ctl |= (cfb_pitch & 0xff) << FBC_CTL_STRIDE_SHIFT;
fbc_ctl |= obj->fence_reg;
I915_WRITE(FBC_CONTROL, fbc_ctl);
DRM_DEBUG_KMS("enabled FBC, pitch %d, yoff %d, plane %c\n",
cfb_pitch, crtc->base.y, plane_name(crtc->plane));
}
static int get_new_crc_ctl_reg(struct drm_i915_private *dev_priv,
enum pipe pipe,
enum intel_pipe_crc_source *source, u32 *val)
{
if (IS_GEN2(dev_priv))
return i8xx_pipe_crc_ctl_reg(source, val);
else if (INTEL_GEN(dev_priv) < 5)
return i9xx_pipe_crc_ctl_reg(dev_priv, pipe, source, val);
else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
return vlv_pipe_crc_ctl_reg(dev_priv, pipe, source, val);
else if (IS_GEN5(dev_priv) || IS_GEN6(dev_priv))
return ilk_pipe_crc_ctl_reg(source, val);
else
return ivb_pipe_crc_ctl_reg(dev_priv, pipe, source, val);
}
static void i8xx_fbc_activate(struct drm_i915_private *dev_priv)
{
struct intel_fbc_reg_params *params = &dev_priv->fbc.params;
int cfb_pitch;
int i;
u32 fbc_ctl;
dev_priv->fbc.active = true;
/* Note: fbc.threshold == 1 for i8xx */
cfb_pitch = params->cfb_size / FBC_LL_SIZE;
if (params->fb.stride < cfb_pitch)
cfb_pitch = params->fb.stride;
/* FBC_CTL wants 32B or 64B units */
if (IS_GEN2(dev_priv))
cfb_pitch = (cfb_pitch / 32) - 1;
else
cfb_pitch = (cfb_pitch / 64) - 1;
/* Clear old tags */
for (i = 0; i < (FBC_LL_SIZE / 32) + 1; i++)
I915_WRITE(FBC_TAG(i), 0);
if (IS_GEN4(dev_priv)) {
u32 fbc_ctl2;
/* Set it up... */
fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
fbc_ctl2 |= FBC_CTL_PLANE(params->crtc.plane);
I915_WRITE(FBC_CONTROL2, fbc_ctl2);
I915_WRITE(FBC_FENCE_OFF, params->crtc.fence_y_offset);
}
/* enable it... */
fbc_ctl = I915_READ(FBC_CONTROL);
fbc_ctl &= 0x3fff << FBC_CTL_INTERVAL_SHIFT;
fbc_ctl |= FBC_CTL_EN | FBC_CTL_PERIODIC;
if (IS_I945GM(dev_priv))
fbc_ctl |= FBC_CTL_C3_IDLE; /* 945 needs special SR handling */
fbc_ctl |= (cfb_pitch & 0xff) << FBC_CTL_STRIDE_SHIFT;
fbc_ctl |= params->fb.fence_reg;
I915_WRITE(FBC_CONTROL, fbc_ctl);
}
int main(int argc, char** argv)
{
struct pci_device *pci_dev;
uint32_t devid;
int mmio_bar;
pci_dev = intel_get_pci_device();
devid = pci_dev->device_id;
intel_get_mmio(pci_dev);
if (IS_GEN2(devid))
mmio_bar = 1;
else
mmio_bar = 0;
write(1, mmio, pci_dev->regions[mmio_bar].size);
return 0;
}
static bool pixel_format_is_valid(struct drm_i915_private *dev_priv,
uint32_t pixel_format)
{
switch (pixel_format) {
case DRM_FORMAT_XRGB8888:
case DRM_FORMAT_XBGR8888:
return true;
case DRM_FORMAT_XRGB1555:
case DRM_FORMAT_RGB565:
/* 16bpp not supported on gen2 */
if (IS_GEN2(dev_priv))
return false;
/* WaFbcOnly1to1Ratio:ctg */
if (IS_G4X(dev_priv))
return false;
return true;
default:
return false;
}
}
static bool stride_is_valid(struct drm_i915_private *dev_priv,
unsigned int stride)
{
/* These should have been caught earlier. */
WARN_ON(stride < 512);
WARN_ON((stride & (64 - 1)) != 0);
/* Below are the additional FBC restrictions. */
if (IS_GEN2(dev_priv) || IS_GEN3(dev_priv))
return stride == 4096 || stride == 8192;
if (IS_GEN4(dev_priv) && !IS_G4X(dev_priv) && stride < 2048)
return false;
if (stride > 16384)
return false;
return true;
}
Bool intel_check_display_stride(ScrnInfoPtr scrn, int stride, Bool tiling)
{
intel_screen_private *intel = intel_get_screen_private(scrn);
int limit;
/* 8xx spec has always 8K limit, but tests show larger limit in
non-tiling mode, which makes large monitor work. */
if (tiling) {
if (IS_GEN2(intel))
limit = KB(8);
else if (IS_GEN3(intel))
limit = KB(8);
else if (IS_GEN4(intel))
limit = KB(16);
else
limit = KB(32);
} else
limit = KB(32);
return stride <= limit;
}
static void render_copyfunc(struct scratch_buf *src, unsigned src_x, unsigned src_y,
struct scratch_buf *dst, unsigned dst_x, unsigned dst_y,
unsigned logical_tile_no)
{
if (IS_GEN2(devid))
gen2_render_copyfunc(src, src_x, src_y,
dst, dst_x, dst_y,
logical_tile_no);
else if (IS_GEN3(devid))
gen3_render_copyfunc(src, src_x, src_y,
dst, dst_x, dst_y,
logical_tile_no);
else if (IS_GEN6(devid))
gen6_render_copyfunc(src, src_x, src_y,
dst, dst_x, dst_y,
logical_tile_no);
else
blitter_copyfunc(src, src_x, src_y,
dst, dst_x, dst_y,
logical_tile_no);
}
int
main(int argc, char **argv)
{
int fd;
int i, iter = 100;
uint32_t tiling, swizzle;
uint32_t handle;
uint32_t devid;
fd = drm_open_any();
handle = create_bo(fd);
gem_get_tiling(fd, handle, &tiling, &swizzle);
devid = intel_get_drm_devid(fd);
if (IS_GEN2(devid)) {
tile_height = 16;
tile_width = 128;
tile_size = 2048;
} else {
tile_height = 8;
tile_width = 512;
tile_size = PAGE_SIZE;
}
/* Read a bunch of random subsets of the data and check that they come
* out right.
*/
for (i = 0; i < iter; i++) {
int size = WIDTH * HEIGHT * 4;
int offset = (random() % size) & ~3;
int len = (random() % size) & ~3;
int j;
if (len == 0)
len = 4;
if (offset + len > size)
len = size - offset;
if (i == 0) {
offset = 0;
len = size;
}
gem_read(fd, handle, offset, linear, len);
/* Translate from offsets in the read buffer to the swizzled
* address that it corresponds to. This is the opposite of
* what Mesa does (calculate offset to be read given the linear
* offset it's looking for).
*/
for (j = offset; j < offset + len; j += 4) {
uint32_t expected_val, found_val;
int swizzled_offset;
const char *swizzle_str;
switch (swizzle) {
case I915_BIT_6_SWIZZLE_NONE:
swizzled_offset = j;
swizzle_str = "none";
break;
case I915_BIT_6_SWIZZLE_9:
swizzled_offset = j ^
swizzle_bit(9, j);
swizzle_str = "bit9";
break;
case I915_BIT_6_SWIZZLE_9_10:
swizzled_offset = j ^
swizzle_bit(9, j) ^
swizzle_bit(10, j);
swizzle_str = "bit9^10";
break;
case I915_BIT_6_SWIZZLE_9_11:
swizzled_offset = j ^
swizzle_bit(9, j) ^
swizzle_bit(11, j);
swizzle_str = "bit9^11";
break;
case I915_BIT_6_SWIZZLE_9_10_11:
swizzled_offset = j ^
swizzle_bit(9, j) ^
swizzle_bit(10, j) ^
swizzle_bit(11, j);
swizzle_str = "bit9^10^11";
break;
default:
fprintf(stderr, "Bad swizzle bits; %d\n",
swizzle);
abort();
}
expected_val = calculate_expected(swizzled_offset);
found_val = linear[(j - offset) / 4];
if (expected_val != found_val) {
fprintf(stderr,
"Bad read [%d]: %d instead of %d at 0x%08x "
"for read from 0x%08x to 0x%08x, swizzle=%s\n",
i, found_val, expected_val, j,
offset, offset + len,
swizzle_str);
abort();
}
}
}
close(fd);
return 0;
}
/**
* Detects bit 6 swizzling of address lookup between IGD access and CPU
* access through main memory.
*/
void
i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
{
drm_i915_private_t *dev_priv = dev->dev_private;
uint32_t swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
uint32_t swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
if (IS_GEN5(dev) || IS_GEN6(dev)) {
/* On Ironlake whatever DRAM config, GPU always do
* same swizzling setup.
*/
swizzle_x = I915_BIT_6_SWIZZLE_9_10;
swizzle_y = I915_BIT_6_SWIZZLE_9;
} else if (IS_GEN2(dev)) {
/* As far as we know, the 865 doesn't have these bit 6
* swizzling issues.
*/
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
} else if (IS_MOBILE(dev)) {
uint32_t dcc;
/* On mobile 9xx chipsets, channel interleave by the CPU is
* determined by DCC. For single-channel, neither the CPU
* nor the GPU do swizzling. For dual channel interleaved,
* the GPU's interleave is bit 9 and 10 for X tiled, and bit
* 9 for Y tiled. The CPU's interleave is independent, and
* can be based on either bit 11 (haven't seen this yet) or
* bit 17 (common).
*/
dcc = I915_READ(DCC);
switch (dcc & DCC_ADDRESSING_MODE_MASK) {
case DCC_ADDRESSING_MODE_SINGLE_CHANNEL:
case DCC_ADDRESSING_MODE_DUAL_CHANNEL_ASYMMETRIC:
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
break;
case DCC_ADDRESSING_MODE_DUAL_CHANNEL_INTERLEAVED:
if (dcc & DCC_CHANNEL_XOR_DISABLE) {
/* This is the base swizzling by the GPU for
* tiled buffers.
*/
swizzle_x = I915_BIT_6_SWIZZLE_9_10;
swizzle_y = I915_BIT_6_SWIZZLE_9;
} else if ((dcc & DCC_CHANNEL_XOR_BIT_17) == 0) {
/* Bit 11 swizzling by the CPU in addition. */
swizzle_x = I915_BIT_6_SWIZZLE_9_10_11;
swizzle_y = I915_BIT_6_SWIZZLE_9_11;
} else {
/* Bit 17 swizzling by the CPU in addition. */
swizzle_x = I915_BIT_6_SWIZZLE_9_10_17;
swizzle_y = I915_BIT_6_SWIZZLE_9_17;
}
break;
}
if (dcc == 0xffffffff) {
DRM_ERROR("Couldn't read from MCHBAR. "
"Disabling tiling.\n");
swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
}
} else {
/* The 965, G33, and newer, have a very flexible memory
* configuration. It will enable dual-channel mode
* (interleaving) on as much memory as it can, and the GPU
* will additionally sometimes enable different bit 6
* swizzling for tiled objects from the CPU.
*
* Here's what I found on the G965:
* slot fill memory size swizzling
* 0A 0B 1A 1B 1-ch 2-ch
* 512 0 0 0 512 0 O
* 512 0 512 0 16 1008 X
* 512 0 0 512 16 1008 X
* 0 512 0 512 16 1008 X
* 1024 1024 1024 0 2048 1024 O
*
* We could probably detect this based on either the DRB
* matching, which was the case for the swizzling required in
* the table above, or from the 1-ch value being less than
* the minimum size of a rank.
*/
if (I915_READ16(C0DRB3) != I915_READ16(C1DRB3)) {
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
} else {
swizzle_x = I915_BIT_6_SWIZZLE_9_10;
swizzle_y = I915_BIT_6_SWIZZLE_9;
}
}
dev_priv->mm.bit_6_swizzle_x = swizzle_x;
dev_priv->mm.bit_6_swizzle_y = swizzle_y;
}
/* Check pitch constriants for all chips & tiling formats */
static bool
i915_tiling_ok(struct drm_device *dev, int stride, int size, int tiling_mode)
{
int tile_width, tile_height;
/* Linear is always fine */
if (tiling_mode == I915_TILING_NONE)
return true;
if (IS_GEN2(dev) ||
(tiling_mode == I915_TILING_Y && HAS_128_BYTE_Y_TILING(dev)))
tile_width = 128;
else
tile_width = 512;
/* check maximum stride & object size */
if (INTEL_INFO(dev)->gen >= 4) {
/* i965 stores the end address of the gtt mapping in the fence
* reg, so dont bother to check the size */
if (stride / 128 > I965_FENCE_MAX_PITCH_VAL)
return false;
} else {
if (stride > 8192)
return false;
if (IS_GEN3(dev)) {
if (size > I830_FENCE_MAX_SIZE_VAL << 20)
return false;
} else {
if (size > I830_FENCE_MAX_SIZE_VAL << 19)
return false;
}
}
if (IS_GEN2(dev) ||
(tiling_mode == I915_TILING_Y && HAS_128_BYTE_Y_TILING(dev)))
tile_height = 32;
else
tile_height = 8;
/* i8xx is strange: It has 2 interleaved rows of tiles, so needs an even
* number of tile rows. */
if (IS_GEN2(dev))
tile_height *= 2;
/* Size needs to be aligned to a full tile row */
if (size & (tile_height * stride - 1))
return false;
/* 965+ just needs multiples of tile width */
if (INTEL_INFO(dev)->gen >= 4) {
if (stride & (tile_width - 1))
return false;
return true;
}
/* Pre-965 needs power of two tile widths */
if (stride < tile_width)
return false;
if (stride & (stride - 1))
return false;
return true;
}
static void i915_restore_modeset_reg(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
int dpll_a_reg, fpa0_reg, fpa1_reg;
int dpll_b_reg, fpb0_reg, fpb1_reg;
int i;
if (drm_core_check_feature(dev, DRIVER_MODESET))
return;
/* Fences */
switch (INTEL_INFO(dev)->gen) {
case 7:
case 6:
for (i = 0; i < 16; i++)
I915_WRITE64(FENCE_REG_SANDYBRIDGE_0 + (i * 8), dev_priv->saveFENCE[i]);
break;
case 5:
case 4:
for (i = 0; i < 16; i++)
I915_WRITE64(FENCE_REG_965_0 + (i * 8), dev_priv->saveFENCE[i]);
break;
case 3:
case 2:
if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev))
for (i = 0; i < 8; i++)
I915_WRITE(FENCE_REG_945_8 + (i * 4), dev_priv->saveFENCE[i+8]);
for (i = 0; i < 8; i++)
I915_WRITE(FENCE_REG_830_0 + (i * 4), dev_priv->saveFENCE[i]);
break;
}
if (HAS_PCH_SPLIT(dev)) {
dpll_a_reg = _PCH_DPLL_A;
dpll_b_reg = _PCH_DPLL_B;
fpa0_reg = _PCH_FPA0;
fpb0_reg = _PCH_FPB0;
fpa1_reg = _PCH_FPA1;
fpb1_reg = _PCH_FPB1;
} else {
dpll_a_reg = _DPLL_A;
dpll_b_reg = _DPLL_B;
fpa0_reg = _FPA0;
fpb0_reg = _FPB0;
fpa1_reg = _FPA1;
fpb1_reg = _FPB1;
}
if (HAS_PCH_SPLIT(dev)) {
I915_WRITE(PCH_DREF_CONTROL, dev_priv->savePCH_DREF_CONTROL);
I915_WRITE(DISP_ARB_CTL, dev_priv->saveDISP_ARB_CTL);
}
/* Pipe & plane A info */
/* Prime the clock */
if (dev_priv->saveDPLL_A & DPLL_VCO_ENABLE) {
I915_WRITE(dpll_a_reg, dev_priv->saveDPLL_A &
~DPLL_VCO_ENABLE);
POSTING_READ(dpll_a_reg);
DRM_UDELAY(150);
}
I915_WRITE(fpa0_reg, dev_priv->saveFPA0);
I915_WRITE(fpa1_reg, dev_priv->saveFPA1);
/* Actually enable it */
I915_WRITE(dpll_a_reg, dev_priv->saveDPLL_A);
POSTING_READ(dpll_a_reg);
DRM_UDELAY(150);
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev)) {
I915_WRITE(_DPLL_A_MD, dev_priv->saveDPLL_A_MD);
POSTING_READ(_DPLL_A_MD);
}
DRM_UDELAY(150);
/* Restore mode */
I915_WRITE(_HTOTAL_A, dev_priv->saveHTOTAL_A);
I915_WRITE(_HBLANK_A, dev_priv->saveHBLANK_A);
I915_WRITE(_HSYNC_A, dev_priv->saveHSYNC_A);
I915_WRITE(_VTOTAL_A, dev_priv->saveVTOTAL_A);
I915_WRITE(_VBLANK_A, dev_priv->saveVBLANK_A);
I915_WRITE(_VSYNC_A, dev_priv->saveVSYNC_A);
if (!HAS_PCH_SPLIT(dev))
I915_WRITE(_BCLRPAT_A, dev_priv->saveBCLRPAT_A);
if (HAS_PCH_SPLIT(dev)) {
I915_WRITE(_PIPEA_DATA_M1, dev_priv->savePIPEA_DATA_M1);
I915_WRITE(_PIPEA_DATA_N1, dev_priv->savePIPEA_DATA_N1);
I915_WRITE(_PIPEA_LINK_M1, dev_priv->savePIPEA_LINK_M1);
I915_WRITE(_PIPEA_LINK_N1, dev_priv->savePIPEA_LINK_N1);
I915_WRITE(_FDI_RXA_CTL, dev_priv->saveFDI_RXA_CTL);
I915_WRITE(_FDI_TXA_CTL, dev_priv->saveFDI_TXA_CTL);
I915_WRITE(_PFA_CTL_1, dev_priv->savePFA_CTL_1);
I915_WRITE(_PFA_WIN_SZ, dev_priv->savePFA_WIN_SZ);
I915_WRITE(_PFA_WIN_POS, dev_priv->savePFA_WIN_POS);
I915_WRITE(_TRANSACONF, dev_priv->saveTRANSACONF);
I915_WRITE(_TRANS_HTOTAL_A, dev_priv->saveTRANS_HTOTAL_A);
I915_WRITE(_TRANS_HBLANK_A, dev_priv->saveTRANS_HBLANK_A);
I915_WRITE(_TRANS_HSYNC_A, dev_priv->saveTRANS_HSYNC_A);
I915_WRITE(_TRANS_VTOTAL_A, dev_priv->saveTRANS_VTOTAL_A);
I915_WRITE(_TRANS_VBLANK_A, dev_priv->saveTRANS_VBLANK_A);
I915_WRITE(_TRANS_VSYNC_A, dev_priv->saveTRANS_VSYNC_A);
}
/* Restore plane info */
I915_WRITE(_DSPASIZE, dev_priv->saveDSPASIZE);
I915_WRITE(_DSPAPOS, dev_priv->saveDSPAPOS);
I915_WRITE(_PIPEASRC, dev_priv->savePIPEASRC);
I915_WRITE(_DSPAADDR, dev_priv->saveDSPAADDR);
I915_WRITE(_DSPASTRIDE, dev_priv->saveDSPASTRIDE);
if (INTEL_INFO(dev)->gen >= 4) {
I915_WRITE(_DSPASURF, dev_priv->saveDSPASURF);
I915_WRITE(_DSPATILEOFF, dev_priv->saveDSPATILEOFF);
}
I915_WRITE(_PIPEACONF, dev_priv->savePIPEACONF);
i915_restore_palette(dev, PIPE_A);
/* Enable the plane */
I915_WRITE(_DSPACNTR, dev_priv->saveDSPACNTR);
I915_WRITE(_DSPAADDR, I915_READ(_DSPAADDR));
/* Pipe & plane B info */
if (dev_priv->saveDPLL_B & DPLL_VCO_ENABLE) {
I915_WRITE(dpll_b_reg, dev_priv->saveDPLL_B &
~DPLL_VCO_ENABLE);
POSTING_READ(dpll_b_reg);
DRM_UDELAY(150);
}
I915_WRITE(fpb0_reg, dev_priv->saveFPB0);
I915_WRITE(fpb1_reg, dev_priv->saveFPB1);
/* Actually enable it */
I915_WRITE(dpll_b_reg, dev_priv->saveDPLL_B);
POSTING_READ(dpll_b_reg);
DRM_UDELAY(150);
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev)) {
I915_WRITE(_DPLL_B_MD, dev_priv->saveDPLL_B_MD);
POSTING_READ(_DPLL_B_MD);
}
DRM_UDELAY(150);
/* Restore mode */
I915_WRITE(_HTOTAL_B, dev_priv->saveHTOTAL_B);
I915_WRITE(_HBLANK_B, dev_priv->saveHBLANK_B);
I915_WRITE(_HSYNC_B, dev_priv->saveHSYNC_B);
I915_WRITE(_VTOTAL_B, dev_priv->saveVTOTAL_B);
I915_WRITE(_VBLANK_B, dev_priv->saveVBLANK_B);
I915_WRITE(_VSYNC_B, dev_priv->saveVSYNC_B);
if (!HAS_PCH_SPLIT(dev))
I915_WRITE(_BCLRPAT_B, dev_priv->saveBCLRPAT_B);
if (HAS_PCH_SPLIT(dev)) {
I915_WRITE(_PIPEB_DATA_M1, dev_priv->savePIPEB_DATA_M1);
I915_WRITE(_PIPEB_DATA_N1, dev_priv->savePIPEB_DATA_N1);
I915_WRITE(_PIPEB_LINK_M1, dev_priv->savePIPEB_LINK_M1);
I915_WRITE(_PIPEB_LINK_N1, dev_priv->savePIPEB_LINK_N1);
I915_WRITE(_FDI_RXB_CTL, dev_priv->saveFDI_RXB_CTL);
I915_WRITE(_FDI_TXB_CTL, dev_priv->saveFDI_TXB_CTL);
I915_WRITE(_PFB_CTL_1, dev_priv->savePFB_CTL_1);
I915_WRITE(_PFB_WIN_SZ, dev_priv->savePFB_WIN_SZ);
I915_WRITE(_PFB_WIN_POS, dev_priv->savePFB_WIN_POS);
I915_WRITE(_TRANSBCONF, dev_priv->saveTRANSBCONF);
I915_WRITE(_TRANS_HTOTAL_B, dev_priv->saveTRANS_HTOTAL_B);
I915_WRITE(_TRANS_HBLANK_B, dev_priv->saveTRANS_HBLANK_B);
I915_WRITE(_TRANS_HSYNC_B, dev_priv->saveTRANS_HSYNC_B);
I915_WRITE(_TRANS_VTOTAL_B, dev_priv->saveTRANS_VTOTAL_B);
I915_WRITE(_TRANS_VBLANK_B, dev_priv->saveTRANS_VBLANK_B);
I915_WRITE(_TRANS_VSYNC_B, dev_priv->saveTRANS_VSYNC_B);
}
/* Restore plane info */
I915_WRITE(_DSPBSIZE, dev_priv->saveDSPBSIZE);
I915_WRITE(_DSPBPOS, dev_priv->saveDSPBPOS);
I915_WRITE(_PIPEBSRC, dev_priv->savePIPEBSRC);
I915_WRITE(_DSPBADDR, dev_priv->saveDSPBADDR);
I915_WRITE(_DSPBSTRIDE, dev_priv->saveDSPBSTRIDE);
if (INTEL_INFO(dev)->gen >= 4) {
I915_WRITE(_DSPBSURF, dev_priv->saveDSPBSURF);
I915_WRITE(_DSPBTILEOFF, dev_priv->saveDSPBTILEOFF);
}
I915_WRITE(_PIPEBCONF, dev_priv->savePIPEBCONF);
i915_restore_palette(dev, PIPE_B);
/* Enable the plane */
I915_WRITE(_DSPBCNTR, dev_priv->saveDSPBCNTR);
I915_WRITE(_DSPBADDR, I915_READ(_DSPBADDR));
/* Cursor state */
I915_WRITE(_CURAPOS, dev_priv->saveCURAPOS);
I915_WRITE(_CURACNTR, dev_priv->saveCURACNTR);
I915_WRITE(_CURABASE, dev_priv->saveCURABASE);
I915_WRITE(_CURBPOS, dev_priv->saveCURBPOS);
I915_WRITE(_CURBCNTR, dev_priv->saveCURBCNTR);
I915_WRITE(_CURBBASE, dev_priv->saveCURBBASE);
if (IS_GEN2(dev))
I915_WRITE(CURSIZE, dev_priv->saveCURSIZE);
return;
}
static void i915_save_modeset_reg(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
int i;
if (drm_core_check_feature(dev, DRIVER_MODESET))
return;
/* Cursor state */
dev_priv->saveCURACNTR = I915_READ(_CURACNTR);
dev_priv->saveCURAPOS = I915_READ(_CURAPOS);
dev_priv->saveCURABASE = I915_READ(_CURABASE);
dev_priv->saveCURBCNTR = I915_READ(_CURBCNTR);
dev_priv->saveCURBPOS = I915_READ(_CURBPOS);
dev_priv->saveCURBBASE = I915_READ(_CURBBASE);
if (IS_GEN2(dev))
dev_priv->saveCURSIZE = I915_READ(CURSIZE);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->savePCH_DREF_CONTROL = I915_READ(PCH_DREF_CONTROL);
dev_priv->saveDISP_ARB_CTL = I915_READ(DISP_ARB_CTL);
}
/* Pipe & plane A info */
dev_priv->savePIPEACONF = I915_READ(_PIPEACONF);
dev_priv->savePIPEASRC = I915_READ(_PIPEASRC);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->saveFPA0 = I915_READ(_PCH_FPA0);
dev_priv->saveFPA1 = I915_READ(_PCH_FPA1);
dev_priv->saveDPLL_A = I915_READ(_PCH_DPLL_A);
} else {
dev_priv->saveFPA0 = I915_READ(_FPA0);
dev_priv->saveFPA1 = I915_READ(_FPA1);
dev_priv->saveDPLL_A = I915_READ(_DPLL_A);
}
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev))
dev_priv->saveDPLL_A_MD = I915_READ(_DPLL_A_MD);
dev_priv->saveHTOTAL_A = I915_READ(_HTOTAL_A);
dev_priv->saveHBLANK_A = I915_READ(_HBLANK_A);
dev_priv->saveHSYNC_A = I915_READ(_HSYNC_A);
dev_priv->saveVTOTAL_A = I915_READ(_VTOTAL_A);
dev_priv->saveVBLANK_A = I915_READ(_VBLANK_A);
dev_priv->saveVSYNC_A = I915_READ(_VSYNC_A);
if (!HAS_PCH_SPLIT(dev))
dev_priv->saveBCLRPAT_A = I915_READ(_BCLRPAT_A);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->savePIPEA_DATA_M1 = I915_READ(_PIPEA_DATA_M1);
dev_priv->savePIPEA_DATA_N1 = I915_READ(_PIPEA_DATA_N1);
dev_priv->savePIPEA_LINK_M1 = I915_READ(_PIPEA_LINK_M1);
dev_priv->savePIPEA_LINK_N1 = I915_READ(_PIPEA_LINK_N1);
dev_priv->saveFDI_TXA_CTL = I915_READ(_FDI_TXA_CTL);
dev_priv->saveFDI_RXA_CTL = I915_READ(_FDI_RXA_CTL);
dev_priv->savePFA_CTL_1 = I915_READ(_PFA_CTL_1);
dev_priv->savePFA_WIN_SZ = I915_READ(_PFA_WIN_SZ);
dev_priv->savePFA_WIN_POS = I915_READ(_PFA_WIN_POS);
dev_priv->saveTRANSACONF = I915_READ(_TRANSACONF);
dev_priv->saveTRANS_HTOTAL_A = I915_READ(_TRANS_HTOTAL_A);
dev_priv->saveTRANS_HBLANK_A = I915_READ(_TRANS_HBLANK_A);
dev_priv->saveTRANS_HSYNC_A = I915_READ(_TRANS_HSYNC_A);
dev_priv->saveTRANS_VTOTAL_A = I915_READ(_TRANS_VTOTAL_A);
dev_priv->saveTRANS_VBLANK_A = I915_READ(_TRANS_VBLANK_A);
dev_priv->saveTRANS_VSYNC_A = I915_READ(_TRANS_VSYNC_A);
}
dev_priv->saveDSPACNTR = I915_READ(_DSPACNTR);
dev_priv->saveDSPASTRIDE = I915_READ(_DSPASTRIDE);
dev_priv->saveDSPASIZE = I915_READ(_DSPASIZE);
dev_priv->saveDSPAPOS = I915_READ(_DSPAPOS);
dev_priv->saveDSPAADDR = I915_READ(_DSPAADDR);
if (INTEL_INFO(dev)->gen >= 4) {
dev_priv->saveDSPASURF = I915_READ(_DSPASURF);
dev_priv->saveDSPATILEOFF = I915_READ(_DSPATILEOFF);
}
i915_save_palette(dev, PIPE_A);
dev_priv->savePIPEASTAT = I915_READ(_PIPEASTAT);
/* Pipe & plane B info */
dev_priv->savePIPEBCONF = I915_READ(_PIPEBCONF);
dev_priv->savePIPEBSRC = I915_READ(_PIPEBSRC);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->saveFPB0 = I915_READ(_PCH_FPB0);
dev_priv->saveFPB1 = I915_READ(_PCH_FPB1);
dev_priv->saveDPLL_B = I915_READ(_PCH_DPLL_B);
} else {
dev_priv->saveFPB0 = I915_READ(_FPB0);
dev_priv->saveFPB1 = I915_READ(_FPB1);
dev_priv->saveDPLL_B = I915_READ(_DPLL_B);
}
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev))
dev_priv->saveDPLL_B_MD = I915_READ(_DPLL_B_MD);
dev_priv->saveHTOTAL_B = I915_READ(_HTOTAL_B);
dev_priv->saveHBLANK_B = I915_READ(_HBLANK_B);
dev_priv->saveHSYNC_B = I915_READ(_HSYNC_B);
dev_priv->saveVTOTAL_B = I915_READ(_VTOTAL_B);
dev_priv->saveVBLANK_B = I915_READ(_VBLANK_B);
dev_priv->saveVSYNC_B = I915_READ(_VSYNC_B);
if (!HAS_PCH_SPLIT(dev))
dev_priv->saveBCLRPAT_B = I915_READ(_BCLRPAT_B);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->savePIPEB_DATA_M1 = I915_READ(_PIPEB_DATA_M1);
dev_priv->savePIPEB_DATA_N1 = I915_READ(_PIPEB_DATA_N1);
dev_priv->savePIPEB_LINK_M1 = I915_READ(_PIPEB_LINK_M1);
dev_priv->savePIPEB_LINK_N1 = I915_READ(_PIPEB_LINK_N1);
dev_priv->saveFDI_TXB_CTL = I915_READ(_FDI_TXB_CTL);
dev_priv->saveFDI_RXB_CTL = I915_READ(_FDI_RXB_CTL);
dev_priv->savePFB_CTL_1 = I915_READ(_PFB_CTL_1);
dev_priv->savePFB_WIN_SZ = I915_READ(_PFB_WIN_SZ);
dev_priv->savePFB_WIN_POS = I915_READ(_PFB_WIN_POS);
dev_priv->saveTRANSBCONF = I915_READ(_TRANSBCONF);
dev_priv->saveTRANS_HTOTAL_B = I915_READ(_TRANS_HTOTAL_B);
dev_priv->saveTRANS_HBLANK_B = I915_READ(_TRANS_HBLANK_B);
dev_priv->saveTRANS_HSYNC_B = I915_READ(_TRANS_HSYNC_B);
dev_priv->saveTRANS_VTOTAL_B = I915_READ(_TRANS_VTOTAL_B);
dev_priv->saveTRANS_VBLANK_B = I915_READ(_TRANS_VBLANK_B);
dev_priv->saveTRANS_VSYNC_B = I915_READ(_TRANS_VSYNC_B);
}
dev_priv->saveDSPBCNTR = I915_READ(_DSPBCNTR);
dev_priv->saveDSPBSTRIDE = I915_READ(_DSPBSTRIDE);
dev_priv->saveDSPBSIZE = I915_READ(_DSPBSIZE);
dev_priv->saveDSPBPOS = I915_READ(_DSPBPOS);
dev_priv->saveDSPBADDR = I915_READ(_DSPBADDR);
if (INTEL_INFO(dev)->gen >= 4) {
dev_priv->saveDSPBSURF = I915_READ(_DSPBSURF);
dev_priv->saveDSPBTILEOFF = I915_READ(_DSPBTILEOFF);
}
i915_save_palette(dev, PIPE_B);
dev_priv->savePIPEBSTAT = I915_READ(_PIPEBSTAT);
/* Fences */
switch (INTEL_INFO(dev)->gen) {
case 7:
case 6:
for (i = 0; i < 16; i++)
dev_priv->saveFENCE[i] = I915_READ64(FENCE_REG_SANDYBRIDGE_0 + (i * 8));
break;
case 5:
case 4:
for (i = 0; i < 16; i++)
dev_priv->saveFENCE[i] = I915_READ64(FENCE_REG_965_0 + (i * 8));
break;
case 3:
if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev))
for (i = 0; i < 8; i++)
dev_priv->saveFENCE[i+8] = I915_READ(FENCE_REG_945_8 + (i * 4));
case 2:
for (i = 0; i < 8; i++)
dev_priv->saveFENCE[i] = I915_READ(FENCE_REG_830_0 + (i * 4));
break;
}
return;
}
static unsigned int
intel_uxa_pixmap_compute_size(PixmapPtr pixmap,
int w, int h,
uint32_t *tiling,
int *stride,
unsigned usage)
{
ScrnInfoPtr scrn = xf86Screens[pixmap->drawable.pScreen->myNum];
intel_screen_private *intel = intel_get_screen_private(scrn);
int pitch, size;
if (*tiling != I915_TILING_NONE) {
/* First check whether tiling is necessary. */
pitch = (w * pixmap->drawable.bitsPerPixel + 7) / 8;
pitch = ALIGN(pitch, 64);
size = pitch * ALIGN (h, 2);
if (INTEL_INFO(intel)->gen < 40) {
/* Gen 2/3 has a maximum stride for tiling of
* 8192 bytes.
*/
if (pitch > KB(8))
*tiling = I915_TILING_NONE;
/* Narrower than half a tile? */
if (pitch < 256)
*tiling = I915_TILING_NONE;
/* Older hardware requires fences to be pot size
* aligned with a minimum of 1 MiB, so causes
* massive overallocation for small textures.
*/
if (size < 1024*1024/2 && !intel->has_relaxed_fencing)
*tiling = I915_TILING_NONE;
} else if (!(usage & INTEL_CREATE_PIXMAP_DRI2) && size <= 4096) {
/* Disable tiling beneath a page size, we will not see
* any benefit from reducing TLB misses and instead
* just incur extra cost when we require a fence.
*/
*tiling = I915_TILING_NONE;
}
}
pitch = (w * pixmap->drawable.bitsPerPixel + 7) / 8;
if (!(usage & INTEL_CREATE_PIXMAP_DRI2) && pitch <= 256)
*tiling = I915_TILING_NONE;
if (*tiling != I915_TILING_NONE) {
int aligned_h, tile_height;
if (IS_GEN2(intel))
tile_height = 16;
else if (*tiling == I915_TILING_X)
tile_height = 8;
else
tile_height = 32;
aligned_h = ALIGN(h, tile_height);
*stride = intel_get_fence_pitch(intel,
ALIGN(pitch, 512),
*tiling);
/* Round the object up to the size of the fence it will live in
* if necessary. We could potentially make the kernel allocate
* a larger aperture space and just bind the subset of pages in,
* but this is easier and also keeps us out of trouble (as much)
* with drm_intel_bufmgr_check_aperture().
*/
size = intel_get_fence_size(intel, *stride * aligned_h);
if (size > intel->max_tiling_size)
*tiling = I915_TILING_NONE;
}
if (*tiling == I915_TILING_NONE) {
/* Round the height up so that the GPU's access to a 2x2 aligned
* subspan doesn't address an invalid page offset beyond the
* end of the GTT.
*/
*stride = ALIGN(pitch, 64);
size = *stride * ALIGN(h, 2);
}
return size;
}
static int intelfb_create(struct intel_fbdev *ifbdev,
struct drm_fb_helper_surface_size *sizes)
{
struct drm_device *dev = ifbdev->helper.dev;
struct fb_info *info;
struct drm_framebuffer *fb;
struct drm_mode_fb_cmd mode_cmd;
struct drm_i915_gem_object *obj;
struct device *device = &dev->pdev->dev;
int size, ret;
/* we don't do packed 24bpp */
if (sizes->surface_bpp == 24)
sizes->surface_bpp = 32;
mode_cmd.width = sizes->surface_width;
mode_cmd.height = sizes->surface_height;
mode_cmd.bpp = sizes->surface_bpp;
mode_cmd.pitch = ALIGN(mode_cmd.width * ((mode_cmd.bpp + 7) / 8), 64);
mode_cmd.depth = sizes->surface_depth;
size = mode_cmd.pitch * mode_cmd.height;
size = ALIGN(size, PAGE_SIZE);
obj = i915_gem_alloc_object(dev, size);
if (!obj) {
DRM_ERROR("failed to allocate framebuffer\n");
ret = -ENOMEM;
goto out;
}
mutex_lock(&dev->struct_mutex);
/* Flush everything out, we'll be doing GTT only from now on */
ret = intel_pin_and_fence_fb_obj(dev, obj, false);
if (ret) {
DRM_ERROR("failed to pin fb: %d\n", ret);
goto out_unref;
}
info = framebuffer_alloc(0, device);
if (!info) {
ret = -ENOMEM;
goto out_unpin;
}
info->par = ifbdev;
ret = intel_framebuffer_init(dev, &ifbdev->ifb, &mode_cmd, obj);
if (ret)
goto out_unpin;
fb = &ifbdev->ifb.base;
ifbdev->helper.fb = fb;
ifbdev->helper.fbdev = info;
strcpy(info->fix.id, "inteldrmfb");
info->flags = FBINFO_DEFAULT | FBINFO_CAN_FORCE_OUTPUT;
info->fbops = &intelfb_ops;
ret = fb_alloc_cmap(&info->cmap, 256, 0);
if (ret) {
ret = -ENOMEM;
goto out_unpin;
}
/* setup aperture base/size for vesafb takeover */
info->aperture_base = dev->mode_config.fb_base;
if (!IS_GEN2(dev))
info->aperture_size = pci_resource_len(dev->pdev, 2);
else
info->aperture_size = pci_resource_len(dev->pdev, 0);
info->fix.smem_start = dev->mode_config.fb_base + obj->gtt_offset;
info->fix.smem_len = size;
info->screen_base = ioremap_wc(dev->agp->base + obj->gtt_offset, size);
if (!info->screen_base) {
ret = -ENOSPC;
goto out_unpin;
}
info->screen_size = size;
// memset(info->screen_base, 0, size);
drm_fb_helper_fill_fix(info, fb->pitch, fb->depth);
drm_fb_helper_fill_var(info, &ifbdev->helper, sizes->fb_width, sizes->fb_height);
info->pixmap.size = 64*1024;
info->pixmap.buf_align = 8;
info->pixmap.access_align = 32;
info->pixmap.flags = FB_PIXMAP_SYSTEM;
info->pixmap.scan_align = 1;
DRM_DEBUG_KMS("allocated %dx%d fb: 0x%08x, bo %p\n",
fb->width, fb->height,
obj->gtt_offset, obj);
mutex_unlock(&dev->struct_mutex);
vga_switcheroo_client_fb_set(dev->pdev, info);
return 0;
out_unpin:
i915_gem_object_unpin(obj);
out_unref:
drm_gem_object_unreference(&obj->base);
mutex_unlock(&dev->struct_mutex);
out:
return ret;
}
/**
* Detects bit 6 swizzling of address lookup between IGD access and CPU
* access through main memory.
*/
void
i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
uint32_t swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
uint32_t swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
if (INTEL_INFO(dev)->gen >= 8 || IS_VALLEYVIEW(dev)) {
/*
* On BDW+, swizzling is not used. We leave the CPU memory
* controller in charge of optimizing memory accesses without
* the extra address manipulation GPU side.
*
* VLV and CHV don't have GPU swizzling.
*/
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
} else if (INTEL_INFO(dev)->gen >= 6) {
uint32_t dimm_c0, dimm_c1;
dimm_c0 = I915_READ(MAD_DIMM_C0);
dimm_c1 = I915_READ(MAD_DIMM_C1);
dimm_c0 &= MAD_DIMM_A_SIZE_MASK | MAD_DIMM_B_SIZE_MASK;
dimm_c1 &= MAD_DIMM_A_SIZE_MASK | MAD_DIMM_B_SIZE_MASK;
/* Enable swizzling when the channels are populated with
* identically sized dimms. We don't need to check the 3rd
* channel because no cpu with gpu attached ships in that
* configuration. Also, swizzling only makes sense for 2
* channels anyway. */
if (dimm_c0 == dimm_c1) {
swizzle_x = I915_BIT_6_SWIZZLE_9_10;
swizzle_y = I915_BIT_6_SWIZZLE_9;
} else {
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
}
} else if (IS_GEN5(dev)) {
/* On Ironlake whatever DRAM config, GPU always do
* same swizzling setup.
*/
swizzle_x = I915_BIT_6_SWIZZLE_9_10;
swizzle_y = I915_BIT_6_SWIZZLE_9;
} else if (IS_GEN2(dev)) {
/* As far as we know, the 865 doesn't have these bit 6
* swizzling issues.
*/
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
} else if (IS_MOBILE(dev) || (IS_GEN3(dev) && !IS_G33(dev))) {
uint32_t dcc;
/* On 9xx chipsets, channel interleave by the CPU is
* determined by DCC. For single-channel, neither the CPU
* nor the GPU do swizzling. For dual channel interleaved,
* the GPU's interleave is bit 9 and 10 for X tiled, and bit
* 9 for Y tiled. The CPU's interleave is independent, and
* can be based on either bit 11 (haven't seen this yet) or
* bit 17 (common).
*/
dcc = I915_READ(DCC);
switch (dcc & DCC_ADDRESSING_MODE_MASK) {
case DCC_ADDRESSING_MODE_SINGLE_CHANNEL:
case DCC_ADDRESSING_MODE_DUAL_CHANNEL_ASYMMETRIC:
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
break;
case DCC_ADDRESSING_MODE_DUAL_CHANNEL_INTERLEAVED:
if (dcc & DCC_CHANNEL_XOR_DISABLE) {
/* This is the base swizzling by the GPU for
* tiled buffers.
*/
swizzle_x = I915_BIT_6_SWIZZLE_9_10;
swizzle_y = I915_BIT_6_SWIZZLE_9;
} else if ((dcc & DCC_CHANNEL_XOR_BIT_17) == 0) {
/* Bit 11 swizzling by the CPU in addition. */
swizzle_x = I915_BIT_6_SWIZZLE_9_10_11;
swizzle_y = I915_BIT_6_SWIZZLE_9_11;
} else {
/* Bit 17 swizzling by the CPU in addition. */
swizzle_x = I915_BIT_6_SWIZZLE_9_10_17;
swizzle_y = I915_BIT_6_SWIZZLE_9_17;
}
break;
}
if (dcc == 0xffffffff) {
DRM_ERROR("Couldn't read from MCHBAR. "
"Disabling tiling.\n");
swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN;
swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN;
}
} else {
/* The 965, G33, and newer, have a very flexible memory
* configuration. It will enable dual-channel mode
* (interleaving) on as much memory as it can, and the GPU
* will additionally sometimes enable different bit 6
* swizzling for tiled objects from the CPU.
*
* Here's what I found on the G965:
* slot fill memory size swizzling
* 0A 0B 1A 1B 1-ch 2-ch
* 512 0 0 0 512 0 O
* 512 0 512 0 16 1008 X
* 512 0 0 512 16 1008 X
* 0 512 0 512 16 1008 X
* 1024 1024 1024 0 2048 1024 O
*
* We could probably detect this based on either the DRB
* matching, which was the case for the swizzling required in
* the table above, or from the 1-ch value being less than
* the minimum size of a rank.
*/
if (I915_READ16(C0DRB3) != I915_READ16(C1DRB3)) {
swizzle_x = I915_BIT_6_SWIZZLE_NONE;
swizzle_y = I915_BIT_6_SWIZZLE_NONE;
} else {
swizzle_x = I915_BIT_6_SWIZZLE_9_10;
swizzle_y = I915_BIT_6_SWIZZLE_9;
}
}
dev_priv->mm.bit_6_swizzle_x = swizzle_x;
dev_priv->mm.bit_6_swizzle_y = swizzle_y;
}
void
init_instdone_definitions(uint32_t devid)
{
if (IS_GEN7(devid)) {
init_gen7_instdone();
} else if (IS_GEN6(devid)) {
/* Now called INSTDONE_1 in the docs. */
gen6_instdone1_bit(GEN6_MA_3_DONE, "Message Arbiter 3");
gen6_instdone1_bit(GEN6_EU_32_DONE, "EU 32");
gen6_instdone1_bit(GEN6_EU_31_DONE, "EU 31");
gen6_instdone1_bit(GEN6_EU_30_DONE, "EU 30");
gen6_instdone1_bit(GEN6_MA_3_DONE, "Message Arbiter 2");
gen6_instdone1_bit(GEN6_EU_22_DONE, "EU 22");
gen6_instdone1_bit(GEN6_EU_21_DONE, "EU 21");
gen6_instdone1_bit(GEN6_EU_20_DONE, "EU 20");
gen6_instdone1_bit(GEN6_MA_3_DONE, "Message Arbiter 1");
gen6_instdone1_bit(GEN6_EU_12_DONE, "EU 12");
gen6_instdone1_bit(GEN6_EU_11_DONE, "EU 11");
gen6_instdone1_bit(GEN6_EU_10_DONE, "EU 10");
gen6_instdone1_bit(GEN6_MA_3_DONE, "Message Arbiter 0");
gen6_instdone1_bit(GEN6_EU_02_DONE, "EU 02");
gen6_instdone1_bit(GEN6_EU_01_DONE, "EU 01");
gen6_instdone1_bit(GEN6_EU_00_DONE, "EU 00");
gen6_instdone1_bit(GEN6_IC_3_DONE, "IC 3");
gen6_instdone1_bit(GEN6_IC_2_DONE, "IC 2");
gen6_instdone1_bit(GEN6_IC_1_DONE, "IC 1");
gen6_instdone1_bit(GEN6_IC_0_DONE, "IC 0");
gen6_instdone1_bit(GEN6_ISC_10_DONE, "ISC 1/0");
gen6_instdone1_bit(GEN6_ISC_32_DONE, "ISC 3/2");
gen6_instdone1_bit(GEN6_VSC_DONE, "VSC");
gen6_instdone1_bit(GEN6_IEF_DONE, "IEF");
gen6_instdone1_bit(GEN6_VFE_DONE, "VFE");
gen6_instdone1_bit(GEN6_TD_DONE, "TD");
gen6_instdone1_bit(GEN6_TS_DONE, "TS");
gen6_instdone1_bit(GEN6_GW_DONE, "GW");
gen6_instdone1_bit(GEN6_HIZ_DONE, "HIZ");
gen6_instdone1_bit(GEN6_AVS_DONE, "AVS");
/* Now called INSTDONE_2 in the docs. */
gen6_instdone2_bit(GEN6_GAM_DONE, "GAM");
gen6_instdone2_bit(GEN6_CS_DONE, "CS");
gen6_instdone2_bit(GEN6_WMBE_DONE, "WMBE");
gen6_instdone2_bit(GEN6_SVRW_DONE, "SVRW");
gen6_instdone2_bit(GEN6_RCC_DONE, "RCC");
gen6_instdone2_bit(GEN6_SVG_DONE, "SVG");
gen6_instdone2_bit(GEN6_ISC_DONE, "ISC");
gen6_instdone2_bit(GEN6_MT_DONE, "MT");
gen6_instdone2_bit(GEN6_RCPFE_DONE, "RCPFE");
gen6_instdone2_bit(GEN6_RCPBE_DONE, "RCPBE");
gen6_instdone2_bit(GEN6_VDI_DONE, "VDI");
gen6_instdone2_bit(GEN6_RCZ_DONE, "RCZ");
gen6_instdone2_bit(GEN6_DAP_DONE, "DAP");
gen6_instdone2_bit(GEN6_PSD_DONE, "PSD");
gen6_instdone2_bit(GEN6_IZ_DONE, "IZ");
gen6_instdone2_bit(GEN6_WMFE_DONE, "WMFE");
gen6_instdone2_bit(GEN6_SVSM_DONE, "SVSM");
gen6_instdone2_bit(GEN6_QC_DONE, "QC");
gen6_instdone2_bit(GEN6_FL_DONE, "FL");
gen6_instdone2_bit(GEN6_SC_DONE, "SC");
gen6_instdone2_bit(GEN6_DM_DONE, "DM");
gen6_instdone2_bit(GEN6_FT_DONE, "FT");
gen6_instdone2_bit(GEN6_DG_DONE, "DG");
gen6_instdone2_bit(GEN6_SI_DONE, "SI");
gen6_instdone2_bit(GEN6_SO_DONE, "SO");
gen6_instdone2_bit(GEN6_PL_DONE, "PL");
gen6_instdone2_bit(GEN6_VME_DONE, "VME");
gen6_instdone2_bit(GEN6_SF_DONE, "SF");
gen6_instdone2_bit(GEN6_CL_DONE, "CL");
gen6_instdone2_bit(GEN6_GS_DONE, "GS");
gen6_instdone2_bit(GEN6_VS0_DONE, "VS0");
gen6_instdone2_bit(GEN6_VF_DONE, "VF");
} else if (IS_GEN5(devid)) {
gen4_instdone_bit(ILK_ROW_0_EU_0_DONE, "Row 0, EU 0");
gen4_instdone_bit(ILK_ROW_0_EU_1_DONE, "Row 0, EU 1");
gen4_instdone_bit(ILK_ROW_0_EU_2_DONE, "Row 0, EU 2");
gen4_instdone_bit(ILK_ROW_0_EU_3_DONE, "Row 0, EU 3");
gen4_instdone_bit(ILK_ROW_1_EU_0_DONE, "Row 1, EU 0");
gen4_instdone_bit(ILK_ROW_1_EU_1_DONE, "Row 1, EU 1");
gen4_instdone_bit(ILK_ROW_1_EU_2_DONE, "Row 1, EU 2");
gen4_instdone_bit(ILK_ROW_1_EU_3_DONE, "Row 1, EU 3");
gen4_instdone_bit(ILK_ROW_2_EU_0_DONE, "Row 2, EU 0");
gen4_instdone_bit(ILK_ROW_2_EU_1_DONE, "Row 2, EU 1");
gen4_instdone_bit(ILK_ROW_2_EU_2_DONE, "Row 2, EU 2");
gen4_instdone_bit(ILK_ROW_2_EU_3_DONE, "Row 2, EU 3");
gen4_instdone_bit(ILK_VCP_DONE, "VCP");
gen4_instdone_bit(ILK_ROW_0_MATH_DONE, "Row 0 math");
gen4_instdone_bit(ILK_ROW_1_MATH_DONE, "Row 1 math");
gen4_instdone_bit(ILK_ROW_2_MATH_DONE, "Row 2 math");
gen4_instdone_bit(ILK_VC1_DONE, "VC1");
gen4_instdone_bit(ILK_ROW_0_MA_DONE, "Row 0 MA");
gen4_instdone_bit(ILK_ROW_1_MA_DONE, "Row 1 MA");
gen4_instdone_bit(ILK_ROW_2_MA_DONE, "Row 2 MA");
gen4_instdone_bit(ILK_ROW_0_ISC_DONE, "Row 0 ISC");
gen4_instdone_bit(ILK_ROW_1_ISC_DONE, "Row 1 ISC");
gen4_instdone_bit(ILK_ROW_2_ISC_DONE, "Row 2 ISC");
gen4_instdone_bit(ILK_VFE_DONE, "VFE");
gen4_instdone_bit(ILK_TD_DONE, "TD");
gen4_instdone_bit(ILK_SVTS_DONE, "SVTS");
gen4_instdone_bit(ILK_TS_DONE, "TS");
gen4_instdone_bit(ILK_GW_DONE, "GW");
gen4_instdone_bit(ILK_AI_DONE, "AI");
gen4_instdone_bit(ILK_AC_DONE, "AC");
gen4_instdone_bit(ILK_AM_DONE, "AM");
init_g4x_instdone1();
} else if (IS_GEN4(devid)) {
gen4_instdone_bit(I965_ROW_0_EU_0_DONE, "Row 0, EU 0");
gen4_instdone_bit(I965_ROW_0_EU_1_DONE, "Row 0, EU 1");
gen4_instdone_bit(I965_ROW_0_EU_2_DONE, "Row 0, EU 2");
gen4_instdone_bit(I965_ROW_0_EU_3_DONE, "Row 0, EU 3");
gen4_instdone_bit(I965_ROW_1_EU_0_DONE, "Row 1, EU 0");
gen4_instdone_bit(I965_ROW_1_EU_1_DONE, "Row 1, EU 1");
gen4_instdone_bit(I965_ROW_1_EU_2_DONE, "Row 1, EU 2");
gen4_instdone_bit(I965_ROW_1_EU_3_DONE, "Row 1, EU 3");
gen4_instdone_bit(I965_SF_DONE, "Strips and Fans");
gen4_instdone_bit(I965_SE_DONE, "Setup Engine");
gen4_instdone_bit(I965_WM_DONE, "Windowizer");
gen4_instdone_bit(I965_DISPATCHER_DONE, "Dispatcher");
gen4_instdone_bit(I965_PROJECTION_DONE, "Projection and LOD");
gen4_instdone_bit(I965_DG_DONE, "Dependent address generator");
gen4_instdone_bit(I965_QUAD_CACHE_DONE, "Texture fetch");
gen4_instdone_bit(I965_TEXTURE_FETCH_DONE, "Texture fetch");
gen4_instdone_bit(I965_TEXTURE_DECOMPRESS_DONE, "Texture decompress");
gen4_instdone_bit(I965_SAMPLER_CACHE_DONE, "Sampler cache");
gen4_instdone_bit(I965_FILTER_DONE, "Filtering");
gen4_instdone_bit(I965_BYPASS_DONE, "Bypass FIFO");
gen4_instdone_bit(I965_PS_DONE, "Pixel shader");
gen4_instdone_bit(I965_CC_DONE, "Color calculator");
gen4_instdone_bit(I965_MAP_FILTER_DONE, "Map filter");
gen4_instdone_bit(I965_MAP_L2_IDLE, "Map L2");
gen4_instdone_bit(I965_MA_ROW_0_DONE, "Message Arbiter row 0");
gen4_instdone_bit(I965_MA_ROW_1_DONE, "Message Arbiter row 1");
gen4_instdone_bit(I965_IC_ROW_0_DONE, "Instruction cache row 0");
gen4_instdone_bit(I965_IC_ROW_1_DONE, "Instruction cache row 1");
gen4_instdone_bit(I965_CP_DONE, "Command Processor");
if (IS_G4X(devid)) {
init_g4x_instdone1();
} else {
init_g965_instdone1();
}
} else if (IS_GEN3(devid)) {
gen3_instdone_bit(IDCT_DONE, "IDCT");
gen3_instdone_bit(IQ_DONE, "IQ");
gen3_instdone_bit(PR_DONE, "PR");
gen3_instdone_bit(VLD_DONE, "VLD");
gen3_instdone_bit(IP_DONE, "Instruction parser");
gen3_instdone_bit(FBC_DONE, "Framebuffer Compression");
gen3_instdone_bit(BINNER_DONE, "Binner");
gen3_instdone_bit(SF_DONE, "Strips and fans");
gen3_instdone_bit(SE_DONE, "Setup engine");
gen3_instdone_bit(WM_DONE, "Windowizer");
gen3_instdone_bit(IZ_DONE, "Intermediate Z");
gen3_instdone_bit(PERSPECTIVE_INTERP_DONE, "Perspective interpolation");
gen3_instdone_bit(DISPATCHER_DONE, "Dispatcher");
gen3_instdone_bit(PROJECTION_DONE, "Projection and LOD");
gen3_instdone_bit(DEPENDENT_ADDRESS_DONE, "Dependent address calculation");
gen3_instdone_bit(TEXTURE_FETCH_DONE, "Texture fetch");
gen3_instdone_bit(TEXTURE_DECOMPRESS_DONE, "Texture decompression");
gen3_instdone_bit(SAMPLER_CACHE_DONE, "Sampler Cache");
gen3_instdone_bit(FILTER_DONE, "Filtering");
gen3_instdone_bit(BYPASS_FIFO_DONE, "Bypass FIFO");
gen3_instdone_bit(PS_DONE, "Pixel shader");
gen3_instdone_bit(CC_DONE, "Color calculator");
gen3_instdone_bit(MAP_FILTER_DONE, "Map filter");
gen3_instdone_bit(MAP_L2_IDLE, "Map L2");
} else {
assert(IS_GEN2(devid));
gen3_instdone_bit(I830_GMBUS_DONE, "GMBUS");
gen3_instdone_bit(I830_FBC_DONE, "FBC");
gen3_instdone_bit(I830_BINNER_DONE, "BINNER");
gen3_instdone_bit(I830_MPEG_DONE, "MPEG");
gen3_instdone_bit(I830_MECO_DONE, "MECO");
gen3_instdone_bit(I830_MCD_DONE, "MCD");
gen3_instdone_bit(I830_MCSTP_DONE, "MCSTP");
gen3_instdone_bit(I830_CC_DONE, "CC");
gen3_instdone_bit(I830_DG_DONE, "DG");
gen3_instdone_bit(I830_DCMP_DONE, "DCMP");
gen3_instdone_bit(I830_FTCH_DONE, "FTCH");
gen3_instdone_bit(I830_IT_DONE, "IT");
gen3_instdone_bit(I830_MG_DONE, "MG");
gen3_instdone_bit(I830_MEC_DONE, "MEC");
gen3_instdone_bit(I830_PC_DONE, "PC");
gen3_instdone_bit(I830_QCC_DONE, "QCC");
gen3_instdone_bit(I830_TB_DONE, "TB");
gen3_instdone_bit(I830_WM_DONE, "WM");
gen3_instdone_bit(I830_EF_DONE, "EF");
gen3_instdone_bit(I830_BLITTER_DONE, "Blitter");
gen3_instdone_bit(I830_MAP_L2_DONE, "Map L2 cache");
gen3_instdone_bit(I830_SECONDARY_RING_3_DONE, "Secondary ring 3");
gen3_instdone_bit(I830_SECONDARY_RING_2_DONE, "Secondary ring 2");
gen3_instdone_bit(I830_SECONDARY_RING_1_DONE, "Secondary ring 1");
gen3_instdone_bit(I830_SECONDARY_RING_0_DONE, "Secondary ring 0");
gen3_instdone_bit(I830_PRIMARY_RING_1_DONE, "Primary ring 1");
gen3_instdone_bit(I830_PRIMARY_RING_0_DONE, "Primary ring 0");
}
}
void i915_save_display_reg(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
int i;
/* Cursor state */
dev_priv->regfile.saveCURACNTR = I915_READ(_CURACNTR);
dev_priv->regfile.saveCURAPOS = I915_READ(_CURAPOS);
dev_priv->regfile.saveCURABASE = I915_READ(_CURABASE);
dev_priv->regfile.saveCURBCNTR = I915_READ(_CURBCNTR);
dev_priv->regfile.saveCURBPOS = I915_READ(_CURBPOS);
dev_priv->regfile.saveCURBBASE = I915_READ(_CURBBASE);
if (IS_GEN2(dev))
dev_priv->regfile.saveCURSIZE = I915_READ(CURSIZE);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->regfile.savePCH_DREF_CONTROL = I915_READ(PCH_DREF_CONTROL);
dev_priv->regfile.saveDISP_ARB_CTL = I915_READ(DISP_ARB_CTL);
}
/* Pipe & plane A info */
dev_priv->regfile.savePIPEACONF = I915_READ(_PIPEACONF);
dev_priv->regfile.savePIPEASRC = I915_READ(_PIPEASRC);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->regfile.saveFPA0 = I915_READ(_PCH_FPA0);
dev_priv->regfile.saveFPA1 = I915_READ(_PCH_FPA1);
dev_priv->regfile.saveDPLL_A = I915_READ(_PCH_DPLL_A);
} else {
dev_priv->regfile.saveFPA0 = I915_READ(_FPA0);
dev_priv->regfile.saveFPA1 = I915_READ(_FPA1);
dev_priv->regfile.saveDPLL_A = I915_READ(_DPLL_A);
}
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev))
dev_priv->regfile.saveDPLL_A_MD = I915_READ(_DPLL_A_MD);
dev_priv->regfile.saveHTOTAL_A = I915_READ(_HTOTAL_A);
dev_priv->regfile.saveHBLANK_A = I915_READ(_HBLANK_A);
dev_priv->regfile.saveHSYNC_A = I915_READ(_HSYNC_A);
dev_priv->regfile.saveVTOTAL_A = I915_READ(_VTOTAL_A);
dev_priv->regfile.saveVBLANK_A = I915_READ(_VBLANK_A);
dev_priv->regfile.saveVSYNC_A = I915_READ(_VSYNC_A);
if (!HAS_PCH_SPLIT(dev))
dev_priv->regfile.saveBCLRPAT_A = I915_READ(_BCLRPAT_A);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->regfile.savePIPEA_DATA_M1 = I915_READ(_PIPEA_DATA_M1);
dev_priv->regfile.savePIPEA_DATA_N1 = I915_READ(_PIPEA_DATA_N1);
dev_priv->regfile.savePIPEA_LINK_M1 = I915_READ(_PIPEA_LINK_M1);
dev_priv->regfile.savePIPEA_LINK_N1 = I915_READ(_PIPEA_LINK_N1);
dev_priv->regfile.saveFDI_TXA_CTL = I915_READ(_FDI_TXA_CTL);
dev_priv->regfile.saveFDI_RXA_CTL = I915_READ(_FDI_RXA_CTL);
dev_priv->regfile.savePFA_CTL_1 = I915_READ(_PFA_CTL_1);
dev_priv->regfile.savePFA_WIN_SZ = I915_READ(_PFA_WIN_SZ);
dev_priv->regfile.savePFA_WIN_POS = I915_READ(_PFA_WIN_POS);
dev_priv->regfile.saveTRANSACONF = I915_READ(_PCH_TRANSACONF);
dev_priv->regfile.saveTRANS_HTOTAL_A = I915_READ(_PCH_TRANS_HTOTAL_A);
dev_priv->regfile.saveTRANS_HBLANK_A = I915_READ(_PCH_TRANS_HBLANK_A);
dev_priv->regfile.saveTRANS_HSYNC_A = I915_READ(_PCH_TRANS_HSYNC_A);
dev_priv->regfile.saveTRANS_VTOTAL_A = I915_READ(_PCH_TRANS_VTOTAL_A);
dev_priv->regfile.saveTRANS_VBLANK_A = I915_READ(_PCH_TRANS_VBLANK_A);
dev_priv->regfile.saveTRANS_VSYNC_A = I915_READ(_PCH_TRANS_VSYNC_A);
}
dev_priv->regfile.saveDSPACNTR = I915_READ(_DSPACNTR);
dev_priv->regfile.saveDSPASTRIDE = I915_READ(_DSPASTRIDE);
dev_priv->regfile.saveDSPASIZE = I915_READ(_DSPASIZE);
dev_priv->regfile.saveDSPAPOS = I915_READ(_DSPAPOS);
dev_priv->regfile.saveDSPAADDR = I915_READ(_DSPAADDR);
if (INTEL_INFO(dev)->gen >= 4) {
dev_priv->regfile.saveDSPASURF = I915_READ(_DSPASURF);
dev_priv->regfile.saveDSPATILEOFF = I915_READ(_DSPATILEOFF);
}
i915_save_palette(dev, PIPE_A);
dev_priv->regfile.savePIPEASTAT = I915_READ(_PIPEASTAT);
/* Pipe & plane B info */
dev_priv->regfile.savePIPEBCONF = I915_READ(_PIPEBCONF);
dev_priv->regfile.savePIPEBSRC = I915_READ(_PIPEBSRC);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->regfile.saveFPB0 = I915_READ(_PCH_FPB0);
dev_priv->regfile.saveFPB1 = I915_READ(_PCH_FPB1);
dev_priv->regfile.saveDPLL_B = I915_READ(_PCH_DPLL_B);
} else {
dev_priv->regfile.saveFPB0 = I915_READ(_FPB0);
dev_priv->regfile.saveFPB1 = I915_READ(_FPB1);
dev_priv->regfile.saveDPLL_B = I915_READ(_DPLL_B);
}
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev))
dev_priv->regfile.saveDPLL_B_MD = I915_READ(_DPLL_B_MD);
dev_priv->regfile.saveHTOTAL_B = I915_READ(_HTOTAL_B);
dev_priv->regfile.saveHBLANK_B = I915_READ(_HBLANK_B);
dev_priv->regfile.saveHSYNC_B = I915_READ(_HSYNC_B);
dev_priv->regfile.saveVTOTAL_B = I915_READ(_VTOTAL_B);
dev_priv->regfile.saveVBLANK_B = I915_READ(_VBLANK_B);
dev_priv->regfile.saveVSYNC_B = I915_READ(_VSYNC_B);
if (!HAS_PCH_SPLIT(dev))
dev_priv->regfile.saveBCLRPAT_B = I915_READ(_BCLRPAT_B);
if (HAS_PCH_SPLIT(dev)) {
dev_priv->regfile.savePIPEB_DATA_M1 = I915_READ(_PIPEB_DATA_M1);
dev_priv->regfile.savePIPEB_DATA_N1 = I915_READ(_PIPEB_DATA_N1);
dev_priv->regfile.savePIPEB_LINK_M1 = I915_READ(_PIPEB_LINK_M1);
dev_priv->regfile.savePIPEB_LINK_N1 = I915_READ(_PIPEB_LINK_N1);
dev_priv->regfile.saveFDI_TXB_CTL = I915_READ(_FDI_TXB_CTL);
dev_priv->regfile.saveFDI_RXB_CTL = I915_READ(_FDI_RXB_CTL);
dev_priv->regfile.savePFB_CTL_1 = I915_READ(_PFB_CTL_1);
dev_priv->regfile.savePFB_WIN_SZ = I915_READ(_PFB_WIN_SZ);
dev_priv->regfile.savePFB_WIN_POS = I915_READ(_PFB_WIN_POS);
dev_priv->regfile.saveTRANSBCONF = I915_READ(_PCH_TRANSBCONF);
dev_priv->regfile.saveTRANS_HTOTAL_B = I915_READ(_PCH_TRANS_HTOTAL_B);
dev_priv->regfile.saveTRANS_HBLANK_B = I915_READ(_PCH_TRANS_HBLANK_B);
dev_priv->regfile.saveTRANS_HSYNC_B = I915_READ(_PCH_TRANS_HSYNC_B);
dev_priv->regfile.saveTRANS_VTOTAL_B = I915_READ(_PCH_TRANS_VTOTAL_B);
dev_priv->regfile.saveTRANS_VBLANK_B = I915_READ(_PCH_TRANS_VBLANK_B);
dev_priv->regfile.saveTRANS_VSYNC_B = I915_READ(_PCH_TRANS_VSYNC_B);
}
dev_priv->regfile.saveDSPBCNTR = I915_READ(_DSPBCNTR);
dev_priv->regfile.saveDSPBSTRIDE = I915_READ(_DSPBSTRIDE);
dev_priv->regfile.saveDSPBSIZE = I915_READ(_DSPBSIZE);
dev_priv->regfile.saveDSPBPOS = I915_READ(_DSPBPOS);
dev_priv->regfile.saveDSPBADDR = I915_READ(_DSPBADDR);
if (INTEL_INFO(dev)->gen >= 4) {
dev_priv->regfile.saveDSPBSURF = I915_READ(_DSPBSURF);
dev_priv->regfile.saveDSPBTILEOFF = I915_READ(_DSPBTILEOFF);
}
i915_save_palette(dev, PIPE_B);
dev_priv->regfile.savePIPEBSTAT = I915_READ(_PIPEBSTAT);
/* Fences */
switch (INTEL_INFO(dev)->gen) {
case 7:
case 6:
for (i = 0; i < 16; i++)
dev_priv->regfile.saveFENCE[i] = I915_READ64(FENCE_REG_SANDYBRIDGE_0 + (i * 8));
break;
case 5:
case 4:
for (i = 0; i < 16; i++)
dev_priv->regfile.saveFENCE[i] = I915_READ64(FENCE_REG_965_0 + (i * 8));
break;
case 3:
if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev))
for (i = 0; i < 8; i++)
dev_priv->regfile.saveFENCE[i+8] = I915_READ(FENCE_REG_945_8 + (i * 4));
case 2:
for (i = 0; i < 8; i++)
dev_priv->regfile.saveFENCE[i] = I915_READ(FENCE_REG_830_0 + (i * 4));
break;
}
/* CRT state */
if (HAS_PCH_SPLIT(dev))
dev_priv->regfile.saveADPA = I915_READ(PCH_ADPA);
else
dev_priv->regfile.saveADPA = I915_READ(ADPA);
/* Display Port state */
if (SUPPORTS_INTEGRATED_DP(dev)) {
dev_priv->regfile.saveDP_B = I915_READ(DP_B);
dev_priv->regfile.saveDP_C = I915_READ(DP_C);
dev_priv->regfile.saveDP_D = I915_READ(DP_D);
dev_priv->regfile.savePIPEA_GMCH_DATA_M = I915_READ(_PIPEA_DATA_M_G4X);
dev_priv->regfile.savePIPEB_GMCH_DATA_M = I915_READ(_PIPEB_DATA_M_G4X);
dev_priv->regfile.savePIPEA_GMCH_DATA_N = I915_READ(_PIPEA_DATA_N_G4X);
dev_priv->regfile.savePIPEB_GMCH_DATA_N = I915_READ(_PIPEB_DATA_N_G4X);
dev_priv->regfile.savePIPEA_DP_LINK_M = I915_READ(_PIPEA_LINK_M_G4X);
dev_priv->regfile.savePIPEB_DP_LINK_M = I915_READ(_PIPEB_LINK_M_G4X);
dev_priv->regfile.savePIPEA_DP_LINK_N = I915_READ(_PIPEA_LINK_N_G4X);
dev_priv->regfile.savePIPEB_DP_LINK_N = I915_READ(_PIPEB_LINK_N_G4X);
}
/* FIXME: regfile.save TV & SDVO state */
/* Backlight */
if (HAS_PCH_SPLIT(dev)) {
dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_PCH_CTL1);
dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_PCH_CTL2);
dev_priv->regfile.saveBLC_CPU_PWM_CTL = I915_READ(BLC_PWM_CPU_CTL);
dev_priv->regfile.saveBLC_CPU_PWM_CTL2 = I915_READ(BLC_PWM_CPU_CTL2);
} else {
dev_priv->regfile.saveBLC_PWM_CTL = I915_READ(BLC_PWM_CTL);
if (INTEL_INFO(dev)->gen >= 4)
dev_priv->regfile.saveBLC_PWM_CTL2 = I915_READ(BLC_PWM_CTL2);
}
return;
}
void i915_restore_display_reg(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = dev->dev_private;
int dpll_a_reg, fpa0_reg, fpa1_reg;
int dpll_b_reg, fpb0_reg, fpb1_reg;
int i;
/* Backlight */
if (HAS_PCH_SPLIT(dev)) {
I915_WRITE(BLC_PWM_PCH_CTL1, dev_priv->regfile.saveBLC_PWM_CTL);
I915_WRITE(BLC_PWM_PCH_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
/* NOTE: BLC_PWM_CPU_CTL must be written after BLC_PWM_CPU_CTL2;
* otherwise we get blank eDP screen after S3 on some machines
*/
I915_WRITE(BLC_PWM_CPU_CTL2, dev_priv->regfile.saveBLC_CPU_PWM_CTL2);
I915_WRITE(BLC_PWM_CPU_CTL, dev_priv->regfile.saveBLC_CPU_PWM_CTL);
} else {
if (INTEL_INFO(dev)->gen >= 4)
I915_WRITE(BLC_PWM_CTL2, dev_priv->regfile.saveBLC_PWM_CTL2);
I915_WRITE(BLC_PWM_CTL, dev_priv->regfile.saveBLC_PWM_CTL);
}
/* Display port ratios (must be done before clock is set) */
if (SUPPORTS_INTEGRATED_DP(dev)) {
I915_WRITE(_PIPEA_DATA_M_G4X, dev_priv->regfile.savePIPEA_GMCH_DATA_M);
I915_WRITE(_PIPEB_DATA_M_G4X, dev_priv->regfile.savePIPEB_GMCH_DATA_M);
I915_WRITE(_PIPEA_DATA_N_G4X, dev_priv->regfile.savePIPEA_GMCH_DATA_N);
I915_WRITE(_PIPEB_DATA_N_G4X, dev_priv->regfile.savePIPEB_GMCH_DATA_N);
I915_WRITE(_PIPEA_LINK_M_G4X, dev_priv->regfile.savePIPEA_DP_LINK_M);
I915_WRITE(_PIPEB_LINK_M_G4X, dev_priv->regfile.savePIPEB_DP_LINK_M);
I915_WRITE(_PIPEA_LINK_N_G4X, dev_priv->regfile.savePIPEA_DP_LINK_N);
I915_WRITE(_PIPEB_LINK_N_G4X, dev_priv->regfile.savePIPEB_DP_LINK_N);
}
/* Fences */
switch (INTEL_INFO(dev)->gen) {
case 7:
case 6:
for (i = 0; i < 16; i++)
I915_WRITE64(FENCE_REG_SANDYBRIDGE_0 + (i * 8), dev_priv->regfile.saveFENCE[i]);
break;
case 5:
case 4:
for (i = 0; i < 16; i++)
I915_WRITE64(FENCE_REG_965_0 + (i * 8), dev_priv->regfile.saveFENCE[i]);
break;
case 3:
case 2:
if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev))
for (i = 0; i < 8; i++)
I915_WRITE(FENCE_REG_945_8 + (i * 4), dev_priv->regfile.saveFENCE[i+8]);
for (i = 0; i < 8; i++)
I915_WRITE(FENCE_REG_830_0 + (i * 4), dev_priv->regfile.saveFENCE[i]);
break;
}
if (HAS_PCH_SPLIT(dev)) {
dpll_a_reg = _PCH_DPLL_A;
dpll_b_reg = _PCH_DPLL_B;
fpa0_reg = _PCH_FPA0;
fpb0_reg = _PCH_FPB0;
fpa1_reg = _PCH_FPA1;
fpb1_reg = _PCH_FPB1;
} else {
dpll_a_reg = _DPLL_A;
dpll_b_reg = _DPLL_B;
fpa0_reg = _FPA0;
fpb0_reg = _FPB0;
fpa1_reg = _FPA1;
fpb1_reg = _FPB1;
}
if (HAS_PCH_SPLIT(dev)) {
I915_WRITE(PCH_DREF_CONTROL, dev_priv->regfile.savePCH_DREF_CONTROL);
I915_WRITE(DISP_ARB_CTL, dev_priv->regfile.saveDISP_ARB_CTL);
}
/* Pipe & plane A info */
/* Prime the clock */
if (dev_priv->regfile.saveDPLL_A & DPLL_VCO_ENABLE) {
I915_WRITE(dpll_a_reg, dev_priv->regfile.saveDPLL_A &
~DPLL_VCO_ENABLE);
POSTING_READ(dpll_a_reg);
udelay(150);
}
I915_WRITE(fpa0_reg, dev_priv->regfile.saveFPA0);
I915_WRITE(fpa1_reg, dev_priv->regfile.saveFPA1);
/* Actually enable it */
I915_WRITE(dpll_a_reg, dev_priv->regfile.saveDPLL_A);
POSTING_READ(dpll_a_reg);
udelay(150);
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev)) {
I915_WRITE(_DPLL_A_MD, dev_priv->regfile.saveDPLL_A_MD);
POSTING_READ(_DPLL_A_MD);
}
udelay(150);
/* Restore mode */
I915_WRITE(_HTOTAL_A, dev_priv->regfile.saveHTOTAL_A);
I915_WRITE(_HBLANK_A, dev_priv->regfile.saveHBLANK_A);
I915_WRITE(_HSYNC_A, dev_priv->regfile.saveHSYNC_A);
I915_WRITE(_VTOTAL_A, dev_priv->regfile.saveVTOTAL_A);
I915_WRITE(_VBLANK_A, dev_priv->regfile.saveVBLANK_A);
I915_WRITE(_VSYNC_A, dev_priv->regfile.saveVSYNC_A);
if (!HAS_PCH_SPLIT(dev))
I915_WRITE(_BCLRPAT_A, dev_priv->regfile.saveBCLRPAT_A);
if (HAS_PCH_SPLIT(dev)) {
I915_WRITE(_PIPEA_DATA_M1, dev_priv->regfile.savePIPEA_DATA_M1);
I915_WRITE(_PIPEA_DATA_N1, dev_priv->regfile.savePIPEA_DATA_N1);
I915_WRITE(_PIPEA_LINK_M1, dev_priv->regfile.savePIPEA_LINK_M1);
I915_WRITE(_PIPEA_LINK_N1, dev_priv->regfile.savePIPEA_LINK_N1);
I915_WRITE(_FDI_RXA_CTL, dev_priv->regfile.saveFDI_RXA_CTL);
I915_WRITE(_FDI_TXA_CTL, dev_priv->regfile.saveFDI_TXA_CTL);
I915_WRITE(_PFA_CTL_1, dev_priv->regfile.savePFA_CTL_1);
I915_WRITE(_PFA_WIN_SZ, dev_priv->regfile.savePFA_WIN_SZ);
I915_WRITE(_PFA_WIN_POS, dev_priv->regfile.savePFA_WIN_POS);
I915_WRITE(_PCH_TRANSACONF, dev_priv->regfile.saveTRANSACONF);
I915_WRITE(_PCH_TRANS_HTOTAL_A, dev_priv->regfile.saveTRANS_HTOTAL_A);
I915_WRITE(_PCH_TRANS_HBLANK_A, dev_priv->regfile.saveTRANS_HBLANK_A);
I915_WRITE(_PCH_TRANS_HSYNC_A, dev_priv->regfile.saveTRANS_HSYNC_A);
I915_WRITE(_PCH_TRANS_VTOTAL_A, dev_priv->regfile.saveTRANS_VTOTAL_A);
I915_WRITE(_PCH_TRANS_VBLANK_A, dev_priv->regfile.saveTRANS_VBLANK_A);
I915_WRITE(_PCH_TRANS_VSYNC_A, dev_priv->regfile.saveTRANS_VSYNC_A);
}
/* Restore plane info */
I915_WRITE(_DSPASIZE, dev_priv->regfile.saveDSPASIZE);
I915_WRITE(_DSPAPOS, dev_priv->regfile.saveDSPAPOS);
I915_WRITE(_PIPEASRC, dev_priv->regfile.savePIPEASRC);
I915_WRITE(_DSPAADDR, dev_priv->regfile.saveDSPAADDR);
I915_WRITE(_DSPASTRIDE, dev_priv->regfile.saveDSPASTRIDE);
if (INTEL_INFO(dev)->gen >= 4) {
I915_WRITE(_DSPASURF, dev_priv->regfile.saveDSPASURF);
I915_WRITE(_DSPATILEOFF, dev_priv->regfile.saveDSPATILEOFF);
}
I915_WRITE(_PIPEACONF, dev_priv->regfile.savePIPEACONF);
i915_restore_palette(dev, PIPE_A);
/* Enable the plane */
I915_WRITE(_DSPACNTR, dev_priv->regfile.saveDSPACNTR);
I915_WRITE(_DSPAADDR, I915_READ(_DSPAADDR));
/* Pipe & plane B info */
if (dev_priv->regfile.saveDPLL_B & DPLL_VCO_ENABLE) {
I915_WRITE(dpll_b_reg, dev_priv->regfile.saveDPLL_B &
~DPLL_VCO_ENABLE);
POSTING_READ(dpll_b_reg);
udelay(150);
}
I915_WRITE(fpb0_reg, dev_priv->regfile.saveFPB0);
I915_WRITE(fpb1_reg, dev_priv->regfile.saveFPB1);
/* Actually enable it */
I915_WRITE(dpll_b_reg, dev_priv->regfile.saveDPLL_B);
POSTING_READ(dpll_b_reg);
udelay(150);
if (INTEL_INFO(dev)->gen >= 4 && !HAS_PCH_SPLIT(dev)) {
I915_WRITE(_DPLL_B_MD, dev_priv->regfile.saveDPLL_B_MD);
POSTING_READ(_DPLL_B_MD);
}
udelay(150);
/* Restore mode */
I915_WRITE(_HTOTAL_B, dev_priv->regfile.saveHTOTAL_B);
I915_WRITE(_HBLANK_B, dev_priv->regfile.saveHBLANK_B);
I915_WRITE(_HSYNC_B, dev_priv->regfile.saveHSYNC_B);
I915_WRITE(_VTOTAL_B, dev_priv->regfile.saveVTOTAL_B);
I915_WRITE(_VBLANK_B, dev_priv->regfile.saveVBLANK_B);
I915_WRITE(_VSYNC_B, dev_priv->regfile.saveVSYNC_B);
if (!HAS_PCH_SPLIT(dev))
I915_WRITE(_BCLRPAT_B, dev_priv->regfile.saveBCLRPAT_B);
if (HAS_PCH_SPLIT(dev)) {
I915_WRITE(_PIPEB_DATA_M1, dev_priv->regfile.savePIPEB_DATA_M1);
I915_WRITE(_PIPEB_DATA_N1, dev_priv->regfile.savePIPEB_DATA_N1);
I915_WRITE(_PIPEB_LINK_M1, dev_priv->regfile.savePIPEB_LINK_M1);
I915_WRITE(_PIPEB_LINK_N1, dev_priv->regfile.savePIPEB_LINK_N1);
I915_WRITE(_FDI_RXB_CTL, dev_priv->regfile.saveFDI_RXB_CTL);
I915_WRITE(_FDI_TXB_CTL, dev_priv->regfile.saveFDI_TXB_CTL);
I915_WRITE(_PFB_CTL_1, dev_priv->regfile.savePFB_CTL_1);
I915_WRITE(_PFB_WIN_SZ, dev_priv->regfile.savePFB_WIN_SZ);
I915_WRITE(_PFB_WIN_POS, dev_priv->regfile.savePFB_WIN_POS);
I915_WRITE(_PCH_TRANSBCONF, dev_priv->regfile.saveTRANSBCONF);
I915_WRITE(_PCH_TRANS_HTOTAL_B, dev_priv->regfile.saveTRANS_HTOTAL_B);
I915_WRITE(_PCH_TRANS_HBLANK_B, dev_priv->regfile.saveTRANS_HBLANK_B);
I915_WRITE(_PCH_TRANS_HSYNC_B, dev_priv->regfile.saveTRANS_HSYNC_B);
I915_WRITE(_PCH_TRANS_VTOTAL_B, dev_priv->regfile.saveTRANS_VTOTAL_B);
I915_WRITE(_PCH_TRANS_VBLANK_B, dev_priv->regfile.saveTRANS_VBLANK_B);
I915_WRITE(_PCH_TRANS_VSYNC_B, dev_priv->regfile.saveTRANS_VSYNC_B);
}
/* Restore plane info */
I915_WRITE(_DSPBSIZE, dev_priv->regfile.saveDSPBSIZE);
I915_WRITE(_DSPBPOS, dev_priv->regfile.saveDSPBPOS);
I915_WRITE(_PIPEBSRC, dev_priv->regfile.savePIPEBSRC);
I915_WRITE(_DSPBADDR, dev_priv->regfile.saveDSPBADDR);
I915_WRITE(_DSPBSTRIDE, dev_priv->regfile.saveDSPBSTRIDE);
if (INTEL_INFO(dev)->gen >= 4) {
I915_WRITE(_DSPBSURF, dev_priv->regfile.saveDSPBSURF);
I915_WRITE(_DSPBTILEOFF, dev_priv->regfile.saveDSPBTILEOFF);
}
I915_WRITE(_PIPEBCONF, dev_priv->regfile.savePIPEBCONF);
i915_restore_palette(dev, PIPE_B);
/* Enable the plane */
I915_WRITE(_DSPBCNTR, dev_priv->regfile.saveDSPBCNTR);
I915_WRITE(_DSPBADDR, I915_READ(_DSPBADDR));
/* Cursor state */
I915_WRITE(_CURAPOS, dev_priv->regfile.saveCURAPOS);
I915_WRITE(_CURACNTR, dev_priv->regfile.saveCURACNTR);
I915_WRITE(_CURABASE, dev_priv->regfile.saveCURABASE);
I915_WRITE(_CURBPOS, dev_priv->regfile.saveCURBPOS);
I915_WRITE(_CURBCNTR, dev_priv->regfile.saveCURBCNTR);
I915_WRITE(_CURBBASE, dev_priv->regfile.saveCURBBASE);
if (IS_GEN2(dev))
I915_WRITE(CURSIZE, dev_priv->regfile.saveCURSIZE);
/* CRT state */
if (HAS_PCH_SPLIT(dev))
I915_WRITE(PCH_ADPA, dev_priv->regfile.saveADPA);
else
I915_WRITE(ADPA, dev_priv->regfile.saveADPA);
/* Display Port state */
if (SUPPORTS_INTEGRATED_DP(dev)) {
I915_WRITE(DP_B, dev_priv->regfile.saveDP_B);
I915_WRITE(DP_C, dev_priv->regfile.saveDP_C);
I915_WRITE(DP_D, dev_priv->regfile.saveDP_D);
}
/* FIXME: restore TV & SDVO state */
return;
}
int i915_driver_load(struct drm_device *dev, unsigned long flags)
{
struct drm_i915_private *dev_priv;
struct intel_device_info *info, *device_info;
int ret = 0, mmio_bar, mmio_size;
uint32_t aperture_size;
info = (struct intel_device_info *) flags;
/* Refuse to load on gen6+ without kms enabled. */
if (info->gen >= 6 && !drm_core_check_feature(dev, DRIVER_MODESET)) {
DRM_INFO("Your hardware requires kernel modesetting (KMS)\n");
DRM_INFO("See CONFIG_DRM_I915_KMS, nomodeset, and i915.modeset parameters\n");
return -ENODEV;
}
/* UMS needs agp support. */
if (!drm_core_check_feature(dev, DRIVER_MODESET) && !dev->agp)
return -EINVAL;
dev_priv = kzalloc(sizeof(*dev_priv), GFP_KERNEL);
if (dev_priv == NULL)
return -ENOMEM;
dev->dev_private = (void *)dev_priv;
gpu_perf_dev_priv = (void *)dev_priv;
dev_priv->dev = dev;
/* Setup the write-once "constant" device info */
device_info = (struct intel_device_info *)&dev_priv->info;
memcpy(device_info, info, sizeof(dev_priv->info));
device_info->device_id = dev->pdev->device;
spin_lock_init(&dev_priv->irq_lock);
spin_lock_init(&dev_priv->gpu_error.lock);
mutex_init(&dev_priv->backlight_lock);
spin_lock_init(&dev_priv->uncore.lock);
spin_lock_init(&dev_priv->mm.object_stat_lock);
spin_lock_init(&dev_priv->mmio_flip_lock);
mutex_init(&dev_priv->dpio_lock);
mutex_init(&dev_priv->modeset_restore_lock);
intel_pm_setup(dev);
intel_display_crc_init(dev);
i915_dump_device_info(dev_priv);
/* Not all pre-production machines fall into this category, only the
* very first ones. Almost everything should work, except for maybe
* suspend/resume. And we don't implement workarounds that affect only
* pre-production machines. */
if (IS_HSW_EARLY_SDV(dev))
DRM_INFO("This is an early pre-production Haswell machine. "
"It may not be fully functional.\n");
if (i915_get_bridge_dev(dev)) {
ret = -EIO;
goto free_priv;
}
mmio_bar = IS_GEN2(dev) ? 1 : 0;
/* Before gen4, the registers and the GTT are behind different BARs.
* However, from gen4 onwards, the registers and the GTT are shared
* in the same BAR, so we want to restrict this ioremap from
* clobbering the GTT which we want ioremap_wc instead. Fortunately,
* the register BAR remains the same size for all the earlier
* generations up to Ironlake.
*/
if (info->gen < 5)
mmio_size = 512*1024;
else
mmio_size = 2*1024*1024;
dev_priv->regs = pci_iomap(dev->pdev, mmio_bar, mmio_size);
if (!dev_priv->regs) {
DRM_ERROR("failed to map registers\n");
ret = -EIO;
goto put_bridge;
}
/* This must be called before any calls to HAS_PCH_* */
intel_detect_pch(dev);
intel_uncore_init(dev);
if (i915_start_vgt(dev->pdev))
i915_host_mediate = true;
printk("i915_start_vgt: %s\n", i915_host_mediate ? "success" : "fail");
i915_check_vgt(dev_priv);
if (USES_VGT(dev))
i915.enable_fbc = 0;
ret = i915_gem_gtt_init(dev);
if (ret)
goto out_regs;
if (drm_core_check_feature(dev, DRIVER_MODESET)) {
/* WARNING: Apparently we must kick fbdev drivers before vgacon,
* otherwise the vga fbdev driver falls over. */
ret = i915_kick_out_firmware_fb(dev_priv);
if (ret) {
DRM_ERROR("failed to remove conflicting framebuffer drivers\n");
goto out_gtt;
}
ret = i915_kick_out_vgacon(dev_priv);
if (ret) {
DRM_ERROR("failed to remove conflicting VGA console\n");
goto out_gtt;
}
}
pci_set_master(dev->pdev);
/* overlay on gen2 is broken and can't address above 1G */
if (IS_GEN2(dev))
dma_set_coherent_mask(&dev->pdev->dev, DMA_BIT_MASK(30));
/* 965GM sometimes incorrectly writes to hardware status page (HWS)
* using 32bit addressing, overwriting memory if HWS is located
* above 4GB.
*
* The documentation also mentions an issue with undefined
* behaviour if any general state is accessed within a page above 4GB,
* which also needs to be handled carefully.
*/
if (IS_BROADWATER(dev) || IS_CRESTLINE(dev))
dma_set_coherent_mask(&dev->pdev->dev, DMA_BIT_MASK(32));
aperture_size = dev_priv->gtt.mappable_end;
dev_priv->gtt.mappable =
io_mapping_create_wc(dev_priv->gtt.mappable_base,
aperture_size);
if (dev_priv->gtt.mappable == NULL) {
ret = -EIO;
goto out_gtt;
}
dev_priv->gtt.mtrr = arch_phys_wc_add(dev_priv->gtt.mappable_base,
aperture_size);
/* The i915 workqueue is primarily used for batched retirement of
* requests (and thus managing bo) once the task has been completed
* by the GPU. i915_gem_retire_requests() is called directly when we
* need high-priority retirement, such as waiting for an explicit
* bo.
*
* It is also used for periodic low-priority events, such as
* idle-timers and recording error state.
*
* All tasks on the workqueue are expected to acquire the dev mutex
* so there is no point in running more than one instance of the
* workqueue at any time. Use an ordered one.
*/
dev_priv->wq = alloc_ordered_workqueue("i915", 0);
if (dev_priv->wq == NULL) {
DRM_ERROR("Failed to create our workqueue.\n");
ret = -ENOMEM;
goto out_mtrrfree;
}
dev_priv->dp_wq = alloc_ordered_workqueue("i915-dp", 0);
if (dev_priv->dp_wq == NULL) {
DRM_ERROR("Failed to create our dp workqueue.\n");
ret = -ENOMEM;
goto out_freewq;
}
intel_irq_init(dev_priv);
intel_uncore_sanitize(dev);
/* Try to make sure MCHBAR is enabled before poking at it */
intel_setup_mchbar(dev);
intel_setup_gmbus(dev);
intel_opregion_setup(dev);
intel_setup_bios(dev);
i915_gem_load(dev);
/* On the 945G/GM, the chipset reports the MSI capability on the
* integrated graphics even though the support isn't actually there
* according to the published specs. It doesn't appear to function
* correctly in testing on 945G.
* This may be a side effect of MSI having been made available for PEG
* and the registers being closely associated.
*
* According to chipset errata, on the 965GM, MSI interrupts may
* be lost or delayed, but we use them anyways to avoid
* stuck interrupts on some machines.
*/
if (!IS_I945G(dev) && !IS_I945GM(dev))
pci_enable_msi(dev->pdev);
intel_device_info_runtime_init(dev);
if (INTEL_INFO(dev)->num_pipes) {
ret = drm_vblank_init(dev, INTEL_INFO(dev)->num_pipes);
if (ret)
goto out_gem_unload;
}
intel_power_domains_init(dev_priv);
if (drm_core_check_feature(dev, DRIVER_MODESET)) {
ret = i915_load_modeset_init(dev);
if (ret < 0) {
DRM_ERROR("failed to init modeset\n");
goto out_power_well;
}
#ifdef DRM_I915_VGT_SUPPORT
if (USES_VGT(dev)) {
/*
* Tell VGT that we have a valid surface to show
* after modesetting. We doesn't distinguish DOM0 and
* Linux guest here, The PVINFO write handler will
* handle this.
*/
I915_WRITE(vgt_info_off(display_ready), 1);
}
#endif
}
i915_setup_sysfs(dev);
if (INTEL_INFO(dev)->num_pipes) {
/* Must be done after probing outputs */
intel_opregion_init(dev);
acpi_video_register();
}
if (IS_GEN5(dev))
intel_gpu_ips_init(dev_priv);
intel_runtime_pm_enable(dev_priv);
return 0;
out_power_well:
intel_power_domains_fini(dev_priv);
drm_vblank_cleanup(dev);
out_gem_unload:
WARN_ON(unregister_oom_notifier(&dev_priv->mm.oom_notifier));
unregister_shrinker(&dev_priv->mm.shrinker);
if (dev->pdev->msi_enabled)
pci_disable_msi(dev->pdev);
intel_teardown_gmbus(dev);
intel_teardown_mchbar(dev);
pm_qos_remove_request(&dev_priv->pm_qos);
destroy_workqueue(dev_priv->dp_wq);
out_freewq:
destroy_workqueue(dev_priv->wq);
out_mtrrfree:
arch_phys_wc_del(dev_priv->gtt.mtrr);
io_mapping_free(dev_priv->gtt.mappable);
out_gtt:
i915_global_gtt_cleanup(dev);
out_regs:
intel_uncore_fini(dev);
pci_iounmap(dev->pdev, dev_priv->regs);
put_bridge:
pci_dev_put(dev_priv->bridge_dev);
free_priv:
if (dev_priv->slab)
kmem_cache_destroy(dev_priv->slab);
kfree(dev_priv);
return ret;
}
static void parse_options(int argc, char **argv)
{
int c, tmp;
int option_index = 0;
static struct option long_options[] = {
{"no-hw", 0, 0, 'd'},
{"buf-size", 1, 0, 's'},
{"gpu-busy-load", 1, 0, 'g'},
{"no-signals", 0, 0, 'S'},
{"buffer-count", 1, 0, 'c'},
{"trace-tile", 1, 0, 't'},
{"disable-blt", 0, 0, 'b'},
{"disable-render", 0, 0, 'r'},
{"untiled", 0, 0, 'u'},
{"x-tiled", 0, 0, 'x'},
{"use-cpu-maps", 0, 0, 'm'},
{"rounds", 1, 0, 'o'},
{"no-fail", 0, 0, 'f'},
{"tiles-per-buf", 0, 0, 'p'},
#define DUCTAPE 0xdead0001
{"remove-duct-tape", 0, 0, DUCTAPE},
#define TILESZ 0xdead0002
{"tile-size", 1, 0, TILESZ},
#define CHCK_RENDER 0xdead0003
{"check-render-cpyfn", 0, 0, CHCK_RENDER},
};
options.scratch_buf_size = 256*4096;
options.no_hw = 0;
options.use_signal_helper = 1;
options.gpu_busy_load = 0;
options.num_buffers = 0;
options.trace_tile = -1;
options.use_render = 1;
options.use_blt = 1;
options.forced_tiling = -1;
options.use_cpu_maps = 0;
options.total_rounds = 512;
options.fail = 1;
options.ducttape = 1;
options.tile_size = 16;
options.tiles_per_buf = options.scratch_buf_size / TILE_BYTES(options.tile_size);
options.check_render_cpyfn = 0;
while((c = getopt_long(argc, argv, "ds:g:c:t:rbuxmo:fp:",
long_options, &option_index)) != -1) {
switch(c) {
case 'd':
options.no_hw = 1;
printf("no-hw debug mode\n");
break;
case 'S':
options.use_signal_helper = 0;
printf("disabling that pesky nuisance who keeps interrupting us\n");
break;
case 's':
tmp = atoi(optarg);
if (tmp < options.tile_size*8192)
printf("scratch buffer size needs to be at least %i\n",
options.tile_size*8192);
else if (tmp & (tmp - 1)) {
printf("scratch buffer size needs to be a power-of-two\n");
} else {
printf("fixed scratch buffer size to %u\n", tmp);
options.scratch_buf_size = tmp;
sanitize_tiles_per_buf();
}
break;
case 'g':
tmp = atoi(optarg);
if (tmp < 0 || tmp > 10)
printf("gpu busy load needs to be bigger than 0 and smaller than 10\n");
else {
printf("gpu busy load factor set to %i\n", tmp);
gpu_busy_load = options.gpu_busy_load = tmp;
}
break;
case 'c':
options.num_buffers = atoi(optarg);
printf("buffer count set to %i\n", options.num_buffers);
break;
case 't':
options.trace_tile = atoi(optarg);
printf("tracing tile %i\n", options.trace_tile);
break;
case 'r':
options.use_render = 0;
printf("disabling render copy\n");
break;
case 'b':
options.use_blt = 0;
printf("disabling blt copy\n");
break;
case 'u':
options.forced_tiling = I915_TILING_NONE;
printf("disabling tiling\n");
break;
case 'x':
if (options.use_cpu_maps) {
printf("tiling not possible with cpu maps\n");
} else {
options.forced_tiling = I915_TILING_X;
printf("using only X-tiling\n");
}
break;
case 'm':
options.use_cpu_maps = 1;
options.forced_tiling = I915_TILING_NONE;
printf("disabling tiling\n");
break;
case 'o':
options.total_rounds = atoi(optarg);
printf("total rounds %i\n", options.total_rounds);
break;
case 'f':
options.fail = 0;
printf("not failing when detecting errors\n");
break;
case 'p':
options.tiles_per_buf = atoi(optarg);
printf("tiles per buffer %i\n", options.tiles_per_buf);
break;
case DUCTAPE:
options.ducttape = 0;
printf("applying duct-tape\n");
break;
case TILESZ:
options.tile_size = atoi(optarg);
sanitize_tiles_per_buf();
printf("til size %i\n", options.tile_size);
break;
case CHCK_RENDER:
options.check_render_cpyfn = 1;
printf("checking render copy function\n");
break;
default:
printf("unkown command options\n");
break;
}
}
if (optind < argc)
printf("unkown command options\n");
/* actually 32767, according to docs, but that kills our nice pot calculations. */
options.max_dimension = 16*1024;
if (options.use_render) {
if (IS_GEN2(devid) || IS_GEN3(devid))
options.max_dimension = 2048;
else
options.max_dimension = 8192;
}
printf("Limiting buffer to %dx%d\n",
options.max_dimension, options.max_dimension);
}
