diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2015-11-10 20:33:06 +0300 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2015-11-10 20:33:06 +0300 |
commit | 3e82806b97398d542a5e03bd94861f79ce10ecee (patch) | |
tree | 467753d23d422fc42a07992ac25cae7889e48c18 /drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c | |
parent | bd4f203e433387d39be404b67ad02acf6f76b7bc (diff) | |
parent | 816d2206f0f9953ca854e4ff1a2749a5cbd62715 (diff) | |
download | linux-3e82806b97398d542a5e03bd94861f79ce10ecee.tar.xz |
Merge branch 'drm-next' of git://people.freedesktop.org/~airlied/linux
Pull drm updates from Dave Airlie:
"I Was Almost Tempted To Capitalise Every Word, but then I decided I
couldn't read it myself!
I've also got one pull request for the sti driver outstanding. It
relied on a commit in Greg's tree and I didn't find out in time, that
commit is in your tree now so I might send that along once this is
merged.
I also had the accidental misfortune to have access to a Skylake on my
desk for a few days, and I've had to encourage Intel to try harder,
which seems to be happening now.
Here is the main drm-next pull request for 4.4.
Highlights:
New driver:
vc4 driver for the Rasberry Pi VPU.
(From Eric Anholt at Broadcom.)
Core:
Atomic fbdev support
Atomic helpers for runtime pm
dp/aux i2c STATUS_UPDATE handling
struct_mutex usage cleanups.
Generic of probing support.
Documentation:
Kerneldoc for VGA switcheroo code.
Rename to gpu instead of drm to reflect scope.
i915:
Skylake GuC firmware fixes
HPD A support
VBT backlight fallbacks
Fastboot by default for some systems
FBC work
BXT/SKL workarounds
Skylake deeper sleep state fixes
amdgpu:
Enable GPU scheduler by default
New atombios opcodes
GPUVM debugging options
Stoney support.
Fencing cleanups.
radeon:
More efficient CS checking
nouveau:
gk20a instance memory handling improvements.
Improved PGOB detection and GK107 support
Kepler GDDR5 PLL statbility improvement
G8x/GT2xx reclock improvements
new userspace API compatiblity fixes.
virtio-gpu:
Add 3D support - qemu 2.5 has it merged for it's gtk backend.
msm:
Initial msm88896 (snapdragon 8200)
exynos:
HDMI cleanups
Enable mixer driver byt default
Add DECON-TV support
vmwgfx:
Move to using memremap + fixes.
rcar-du:
Add support for R8A7793/4 DU
armada:
Remove support for non-component mode
Improved plane handling
Power savings while in DPMS off.
tda998x:
Remove unused slave encoder support
Use more HDMI helpers
Fix EDID read handling
dwhdmi:
Interlace video mode support for ipu-v3/dw_hdmi
Hotplug state fixes
Audio driver integration
imx:
More color formats support.
tegra:
Minor fixes/improvements"
[ Merge fixup: remove unused variable 'dev' that had all uses removed in
commit 4e270f088011: "drm/gem: Drop struct_mutex requirement from
drm_gem_mmap_obj" ]
* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (764 commits)
drm/vmwgfx: Relax irq locking somewhat
drm/vmwgfx: Properly flush cursor updates and page-flips
drm/i915/skl: disable display side power well support for now
drm/i915: Extend DSL readout fix to BDW and SKL.
drm/i915: Do graphics device reset under forcewake
drm/i915: Skip fence installation for objects with rotated views (v4)
vga_switcheroo: Drop client power state VGA_SWITCHEROO_INIT
drm/amdgpu: group together common fence implementation
drm/amdgpu: remove AMDGPU_FENCE_OWNER_MOVE
drm/amdgpu: remove now unused fence functions
drm/amdgpu: fix fence fallback check
drm/amdgpu: fix stoping the scheduler timeout
drm/amdgpu: cleanup on error in amdgpu_cs_ioctl()
drm/i915: Fix locking around GuC firmware load
drm/amdgpu: update Fiji's Golden setting
drm/amdgpu: update Fiji's rev id
drm/amdgpu: extract common code in vi_common_early_init
drm/amd/scheduler: don't oops on failure to load
drm/amdgpu: don't oops on failure to load (v2)
drm/amdgpu: don't VT switch on suspend
...
Diffstat (limited to 'drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c')
-rw-r--r-- | drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c | 141 |
1 files changed, 111 insertions, 30 deletions
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c index 989355622aac..9df45030ff9f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgk104.c @@ -673,6 +673,25 @@ gk104_ram_calc_gddr5(struct gk104_ram *ram, u32 freq) * DDR3 ******************************************************************************/ +static void +nvkm_sddr3_dll_reset(struct gk104_ramfuc *fuc) +{ + ram_nuke(fuc, mr[0]); + ram_mask(fuc, mr[0], 0x100, 0x100); + ram_mask(fuc, mr[0], 0x100, 0x000); +} + +static void +nvkm_sddr3_dll_disable(struct gk104_ramfuc *fuc) +{ + u32 mr1_old = ram_rd32(fuc, mr[1]); + + if (!(mr1_old & 0x1)) { + ram_mask(fuc, mr[1], 0x1, 0x1); + ram_nsec(fuc, 1000); + } +} + static int gk104_ram_calc_sddr3(struct gk104_ram *ram, u32 freq) { @@ -702,6 +721,10 @@ gk104_ram_calc_sddr3(struct gk104_ram *ram, u32 freq) ram_mask(fuc, 0x10f808, 0x04000000, 0x04000000); ram_wr32(fuc, 0x10f314, 0x00000001); /* PRECHARGE */ + + if (next->bios.ramcfg_DLLoff) + nvkm_sddr3_dll_disable(fuc); + ram_wr32(fuc, 0x10f210, 0x00000000); /* REFRESH_AUTO = 0 */ ram_wr32(fuc, 0x10f310, 0x00000001); /* REFRESH */ ram_mask(fuc, 0x10f200, 0x80000000, 0x80000000); @@ -879,17 +902,20 @@ gk104_ram_calc_sddr3(struct gk104_ram *ram, u32 freq) ram_wr32(fuc, 0x10f210, 0x80000000); /* REFRESH_AUTO = 1 */ ram_nsec(fuc, 1000); - ram_nuke(fuc, mr[0]); - ram_mask(fuc, mr[0], 0x100, 0x100); - ram_mask(fuc, mr[0], 0x100, 0x000); + if (!next->bios.ramcfg_DLLoff) { + ram_mask(fuc, mr[1], 0x1, 0x0); + nvkm_sddr3_dll_reset(fuc); + } - ram_mask(fuc, mr[2], 0xfff, ram->base.mr[2]); + ram_mask(fuc, mr[2], 0x00000fff, ram->base.mr[2]); + ram_mask(fuc, mr[1], 0xffffffff, ram->base.mr[1]); ram_wr32(fuc, mr[0], ram->base.mr[0]); ram_nsec(fuc, 1000); - ram_nuke(fuc, mr[0]); - ram_mask(fuc, mr[0], 0x100, 0x100); - ram_mask(fuc, mr[0], 0x100, 0x000); + if (!next->bios.ramcfg_DLLoff) { + nvkm_sddr3_dll_reset(fuc); + ram_nsec(fuc, 1000); + } if (vc == 0 && ram_have(fuc, gpio2E)) { u32 temp = ram_mask(fuc, gpio2E, 0x3000, fuc->r_func2E[0]); @@ -945,6 +971,67 @@ gk104_ram_calc_data(struct gk104_ram *ram, u32 khz, struct nvkm_ram_data *data) } static int +gk104_calc_pll_output(int fN, int M, int N, int P, int clk) +{ + return ((clk * N) + (((u16)(fN + 4096) * clk) >> 13)) / (M * P); +} + +static int +gk104_pll_calc_hiclk(int target_khz, int crystal, + int *N1, int *fN1, int *M1, int *P1, + int *N2, int *M2, int *P2) +{ + int best_clk = 0, best_err = target_khz, p_ref, n_ref; + bool upper = false; + + *M1 = 1; + /* M has to be 1, otherwise it gets unstable */ + *M2 = 1; + /* can be 1 or 2, sticking with 1 for simplicity */ + *P2 = 1; + + for (p_ref = 0x7; p_ref >= 0x5; --p_ref) { + for (n_ref = 0x25; n_ref <= 0x2b; ++n_ref) { + int cur_N, cur_clk, cur_err; + + cur_clk = gk104_calc_pll_output(0, 1, n_ref, p_ref, crystal); + cur_N = target_khz / cur_clk; + cur_err = target_khz + - gk104_calc_pll_output(0xf000, 1, cur_N, 1, cur_clk); + + /* we found a better combination */ + if (cur_err < best_err) { + best_err = cur_err; + best_clk = cur_clk; + *N2 = cur_N; + *N1 = n_ref; + *P1 = p_ref; + upper = false; + } + + cur_N += 1; + cur_err = gk104_calc_pll_output(0xf000, 1, cur_N, 1, cur_clk) + - target_khz; + if (cur_err < best_err) { + best_err = cur_err; + best_clk = cur_clk; + *N2 = cur_N; + *N1 = n_ref; + *P1 = p_ref; + upper = true; + } + } + } + + /* adjust fN to get closer to the target clock */ + *fN1 = (u16)((((best_err / *N2 * *P2) * (*P1 * *M1)) << 13) / crystal); + if (upper) + *fN1 = (u16)(1 - *fN1); + + return gk104_calc_pll_output(*fN1, 1, *N1, *P1, crystal); +} + +static int gk104_ram_calc_xits(struct gk104_ram *ram, struct nvkm_ram_data *next) { struct gk104_ramfuc *fuc = &ram->fuc; @@ -968,31 +1055,24 @@ gk104_ram_calc_xits(struct gk104_ram *ram, struct nvkm_ram_data *next) * kepler boards, no idea how/why they're chosen. */ refclk = next->freq; - if (ram->mode == 2) - refclk = fuc->mempll.refclk; - - /* calculate refpll coefficients */ - ret = gt215_pll_calc(subdev, &fuc->refpll, refclk, &ram->N1, - &ram->fN1, &ram->M1, &ram->P1); - fuc->mempll.refclk = ret; - if (ret <= 0) { - nvkm_error(subdev, "unable to calc refpll\n"); - return -EINVAL; - } - - /* calculate mempll coefficients, if we're using it */ if (ram->mode == 2) { - /* post-divider doesn't work... the reg takes the values but - * appears to completely ignore it. there *is* a bit at - * bit 28 that appears to divide the clock by 2 if set. - */ - fuc->mempll.min_p = 1; - fuc->mempll.max_p = 2; - - ret = gt215_pll_calc(subdev, &fuc->mempll, next->freq, - &ram->N2, NULL, &ram->M2, &ram->P2); + ret = gk104_pll_calc_hiclk(next->freq, subdev->device->crystal, + &ram->N1, &ram->fN1, &ram->M1, &ram->P1, + &ram->N2, &ram->M2, &ram->P2); + fuc->mempll.refclk = ret; + if (ret <= 0) { + nvkm_error(subdev, "unable to calc plls\n"); + return -EINVAL; + } + nvkm_debug(subdev, "sucessfully calced PLLs for clock %i kHz" + " (refclock: %i kHz)\n", next->freq, ret); + } else { + /* calculate refpll coefficients */ + ret = gt215_pll_calc(subdev, &fuc->refpll, refclk, &ram->N1, + &ram->fN1, &ram->M1, &ram->P1); + fuc->mempll.refclk = ret; if (ret <= 0) { - nvkm_error(subdev, "unable to calc mempll\n"); + nvkm_error(subdev, "unable to calc refpll\n"); return -EINVAL; } } @@ -1600,6 +1680,7 @@ gk104_ram_new(struct nvkm_fb *fb, struct nvkm_ram **pram) break; case NVKM_RAM_TYPE_DDR3: ram->fuc.r_mr[0] = ramfuc_reg(0x10f300); + ram->fuc.r_mr[1] = ramfuc_reg(0x10f304); ram->fuc.r_mr[2] = ramfuc_reg(0x10f320); break; default: |