*
* Authors: Ben Skeggs
*/
-#include "gf100.h"
+#define gf100_ram(p) container_of((p), struct gf100_ram, base)
+#include "ram.h"
#include "ramfuc.h"
-#include <core/device.h>
#include <core/option.h>
#include <subdev/bios.h>
#include <subdev/bios/pll.h>
gf100_ram_train(struct gf100_ramfuc *fuc, u32 magic)
{
struct gf100_ram *ram = container_of(fuc, typeof(*ram), fuc);
- struct nvkm_fb *pfb = nvkm_fb(ram);
- u32 part = nv_rd32(pfb, 0x022438), i;
- u32 mask = nv_rd32(pfb, 0x022554);
+ struct nvkm_fb *fb = ram->base.fb;
+ struct nvkm_device *device = fb->subdev.device;
+ u32 part = nvkm_rd32(device, 0x022438), i;
+ u32 mask = nvkm_rd32(device, 0x022554);
u32 addr = 0x110974;
ram_wr32(fuc, 0x10f910, magic);
}
static int
-gf100_ram_calc(struct nvkm_fb *pfb, u32 freq)
+gf100_ram_calc(struct nvkm_ram *base, u32 freq)
{
- struct nvkm_clk *clk = nvkm_clk(pfb);
- struct nvkm_bios *bios = nvkm_bios(pfb);
- struct gf100_ram *ram = (void *)pfb->ram;
+ struct gf100_ram *ram = gf100_ram(base);
struct gf100_ramfuc *fuc = &ram->fuc;
+ struct nvkm_subdev *subdev = &ram->base.fb->subdev;
+ struct nvkm_device *device = subdev->device;
+ struct nvkm_clk *clk = device->clk;
+ struct nvkm_bios *bios = device->bios;
struct nvbios_ramcfg cfg;
u8 ver, cnt, len, strap;
struct {
rammap.data = nvbios_rammapEm(bios, freq / 1000, &ver, &rammap.size,
&cnt, &ramcfg.size, &cfg);
if (!rammap.data || ver != 0x10 || rammap.size < 0x0e) {
- nv_error(pfb, "invalid/missing rammap entry\n");
+ nvkm_error(subdev, "invalid/missing rammap entry\n");
return -EINVAL;
}
/* locate specific data set for the attached memory */
- strap = nvbios_ramcfg_index(nv_subdev(pfb));
+ strap = nvbios_ramcfg_index(subdev);
if (strap >= cnt) {
- nv_error(pfb, "invalid ramcfg strap\n");
+ nvkm_error(subdev, "invalid ramcfg strap\n");
return -EINVAL;
}
ramcfg.data = rammap.data + rammap.size + (strap * ramcfg.size);
if (!ramcfg.data || ver != 0x10 || ramcfg.size < 0x0e) {
- nv_error(pfb, "invalid/missing ramcfg entry\n");
+ nvkm_error(subdev, "invalid/missing ramcfg entry\n");
return -EINVAL;
}
/* lookup memory timings, if bios says they're present */
- strap = nv_ro08(bios, ramcfg.data + 0x01);
+ strap = nvbios_rd08(bios, ramcfg.data + 0x01);
if (strap != 0xff) {
timing.data = nvbios_timingEe(bios, strap, &ver, &timing.size,
&cnt, &len);
if (!timing.data || ver != 0x10 || timing.size < 0x19) {
- nv_error(pfb, "invalid/missing timing entry\n");
+ nvkm_error(subdev, "invalid/missing timing entry\n");
return -EINVAL;
}
} else {
timing.data = 0;
}
- ret = ram_init(fuc, pfb);
+ ret = ram_init(fuc, ram->base.fb);
if (ret)
return ret;
/* determine target mclk configuration */
if (!(ram_rd32(fuc, 0x137300) & 0x00000100))
- ref = clk->read(clk, nv_clk_src_sppll0);
+ ref = nvkm_clk_read(clk, nv_clk_src_sppll0);
else
- ref = clk->read(clk, nv_clk_src_sppll1);
+ ref = nvkm_clk_read(clk, nv_clk_src_sppll1);
div = max(min((ref * 2) / freq, (u32)65), (u32)2) - 2;
out = (ref * 2) / (div + 2);
mode = freq != out;
if (mode == 1 && from == 0) {
/* calculate refpll */
- ret = gt215_pll_calc(nv_subdev(pfb), &ram->refpll,
- ram->mempll.refclk, &N1, NULL, &M1, &P);
+ ret = gt215_pll_calc(subdev, &ram->refpll, ram->mempll.refclk,
+ &N1, NULL, &M1, &P);
if (ret <= 0) {
- nv_error(pfb, "unable to calc refpll\n");
+ nvkm_error(subdev, "unable to calc refpll\n");
return ret ? ret : -ERANGE;
}
ram_wait(fuc, 0x137390, 0x00020000, 0x00020000, 64000);
/* calculate mempll */
- ret = gt215_pll_calc(nv_subdev(pfb), &ram->mempll, freq,
+ ret = gt215_pll_calc(subdev, &ram->mempll, freq,
&N1, NULL, &M1, &P);
if (ret <= 0) {
- nv_error(pfb, "unable to calc refpll\n");
+ nvkm_error(subdev, "unable to calc refpll\n");
return ret ? ret : -ERANGE;
}
}
static int
-gf100_ram_prog(struct nvkm_fb *pfb)
+gf100_ram_prog(struct nvkm_ram *base)
{
- struct nvkm_device *device = nv_device(pfb);
- struct gf100_ram *ram = (void *)pfb->ram;
- struct gf100_ramfuc *fuc = &ram->fuc;
- ram_exec(fuc, nvkm_boolopt(device->cfgopt, "NvMemExec", true));
+ struct gf100_ram *ram = gf100_ram(base);
+ struct nvkm_device *device = ram->base.fb->subdev.device;
+ ram_exec(&ram->fuc, nvkm_boolopt(device->cfgopt, "NvMemExec", true));
return 0;
}
static void
-gf100_ram_tidy(struct nvkm_fb *pfb)
+gf100_ram_tidy(struct nvkm_ram *base)
{
- struct gf100_ram *ram = (void *)pfb->ram;
- struct gf100_ramfuc *fuc = &ram->fuc;
- ram_exec(fuc, false);
+ struct gf100_ram *ram = gf100_ram(base);
+ ram_exec(&ram->fuc, false);
}
extern const u8 gf100_pte_storage_type_map[256];
void
-gf100_ram_put(struct nvkm_fb *pfb, struct nvkm_mem **pmem)
+gf100_ram_put(struct nvkm_ram *ram, struct nvkm_mem **pmem)
{
- struct nvkm_ltc *ltc = nvkm_ltc(pfb);
+ struct nvkm_ltc *ltc = ram->fb->subdev.device->ltc;
struct nvkm_mem *mem = *pmem;
*pmem = NULL;
if (unlikely(mem == NULL))
return;
- mutex_lock(&pfb->base.mutex);
+ mutex_lock(&ram->fb->subdev.mutex);
if (mem->tag)
- ltc->tags_free(ltc, &mem->tag);
- __nv50_ram_put(pfb, mem);
- mutex_unlock(&pfb->base.mutex);
+ nvkm_ltc_tags_free(ltc, &mem->tag);
+ __nv50_ram_put(ram, mem);
+ mutex_unlock(&ram->fb->subdev.mutex);
kfree(mem);
}
int
-gf100_ram_get(struct nvkm_fb *pfb, u64 size, u32 align, u32 ncmin,
+gf100_ram_get(struct nvkm_ram *ram, u64 size, u32 align, u32 ncmin,
u32 memtype, struct nvkm_mem **pmem)
{
- struct nvkm_mm *mm = &pfb->vram;
+ struct nvkm_ltc *ltc = ram->fb->subdev.device->ltc;
+ struct nvkm_mm *mm = &ram->vram;
struct nvkm_mm_node *r;
struct nvkm_mem *mem;
int type = (memtype & 0x0ff);
const bool comp = gf100_pte_storage_type_map[type] != type;
int ret;
- size >>= 12;
- align >>= 12;
- ncmin >>= 12;
+ size >>= NVKM_RAM_MM_SHIFT;
+ align >>= NVKM_RAM_MM_SHIFT;
+ ncmin >>= NVKM_RAM_MM_SHIFT;
if (!ncmin)
ncmin = size;
INIT_LIST_HEAD(&mem->regions);
mem->size = size;
- mutex_lock(&pfb->base.mutex);
+ mutex_lock(&ram->fb->subdev.mutex);
if (comp) {
- struct nvkm_ltc *ltc = nvkm_ltc(pfb);
-
/* compression only works with lpages */
- if (align == (1 << (17 - 12))) {
+ if (align == (1 << (17 - NVKM_RAM_MM_SHIFT))) {
int n = size >> 5;
- ltc->tags_alloc(ltc, n, &mem->tag);
+ nvkm_ltc_tags_alloc(ltc, n, &mem->tag);
}
if (unlikely(!mem->tag))
else
ret = nvkm_mm_head(mm, 0, 1, size, ncmin, align, &r);
if (ret) {
- mutex_unlock(&pfb->base.mutex);
- pfb->ram->put(pfb, &mem);
+ mutex_unlock(&ram->fb->subdev.mutex);
+ ram->func->put(ram, &mem);
return ret;
}
list_add_tail(&r->rl_entry, &mem->regions);
size -= r->length;
} while (size);
- mutex_unlock(&pfb->base.mutex);
+ mutex_unlock(&ram->fb->subdev.mutex);
r = list_first_entry(&mem->regions, struct nvkm_mm_node, rl_entry);
- mem->offset = (u64)r->offset << 12;
+ mem->offset = (u64)r->offset << NVKM_RAM_MM_SHIFT;
*pmem = mem;
return 0;
}
-int
-gf100_ram_create_(struct nvkm_object *parent, struct nvkm_object *engine,
- struct nvkm_oclass *oclass, u32 maskaddr, int size,
- void **pobject)
+static int
+gf100_ram_init(struct nvkm_ram *base)
{
- struct nvkm_fb *pfb = nvkm_fb(parent);
- struct nvkm_bios *bios = nvkm_bios(pfb);
- struct nvkm_ram *ram;
- const u32 rsvd_head = ( 256 * 1024) >> 12; /* vga memory */
- const u32 rsvd_tail = (1024 * 1024) >> 12; /* vbios etc */
- u32 parts = nv_rd32(pfb, 0x022438);
- u32 pmask = nv_rd32(pfb, maskaddr);
- u32 bsize = nv_rd32(pfb, 0x10f20c);
- u32 offset, length;
- bool uniform = true;
- int ret, part;
+ static const u8 train0[] = {
+ 0x00, 0xff, 0x55, 0xaa, 0x33, 0xcc,
+ 0x00, 0xff, 0xff, 0x00, 0xff, 0x00,
+ };
+ static const u32 train1[] = {
+ 0x00000000, 0xffffffff,
+ 0x55555555, 0xaaaaaaaa,
+ 0x33333333, 0xcccccccc,
+ 0xf0f0f0f0, 0x0f0f0f0f,
+ 0x00ff00ff, 0xff00ff00,
+ 0x0000ffff, 0xffff0000,
+ };
+ struct gf100_ram *ram = gf100_ram(base);
+ struct nvkm_device *device = ram->base.fb->subdev.device;
+ int i;
- ret = nvkm_ram_create_(parent, engine, oclass, size, pobject);
- ram = *pobject;
- if (ret)
- return ret;
+ switch (ram->base.type) {
+ case NVKM_RAM_TYPE_GDDR5:
+ break;
+ default:
+ return 0;
+ }
- nv_debug(pfb, "0x100800: 0x%08x\n", nv_rd32(pfb, 0x100800));
- nv_debug(pfb, "parts 0x%08x mask 0x%08x\n", parts, pmask);
+ /* prepare for ddr link training, and load training patterns */
+ for (i = 0; i < 0x30; i++) {
+ nvkm_wr32(device, 0x10f968, 0x00000000 | (i << 8));
+ nvkm_wr32(device, 0x10f96c, 0x00000000 | (i << 8));
+ nvkm_wr32(device, 0x10f920, 0x00000100 | train0[i % 12]);
+ nvkm_wr32(device, 0x10f924, 0x00000100 | train0[i % 12]);
+ nvkm_wr32(device, 0x10f918, train1[i % 12]);
+ nvkm_wr32(device, 0x10f91c, train1[i % 12]);
+ nvkm_wr32(device, 0x10f920, 0x00000000 | train0[i % 12]);
+ nvkm_wr32(device, 0x10f924, 0x00000000 | train0[i % 12]);
+ nvkm_wr32(device, 0x10f918, train1[i % 12]);
+ nvkm_wr32(device, 0x10f91c, train1[i % 12]);
+ }
- ram->type = nvkm_fb_bios_memtype(bios);
- ram->ranks = (nv_rd32(pfb, 0x10f200) & 0x00000004) ? 2 : 1;
+ return 0;
+}
+
+static const struct nvkm_ram_func
+gf100_ram_func = {
+ .init = gf100_ram_init,
+ .get = gf100_ram_get,
+ .put = gf100_ram_put,
+ .calc = gf100_ram_calc,
+ .prog = gf100_ram_prog,
+ .tidy = gf100_ram_tidy,
+};
+
+int
+gf100_ram_ctor(const struct nvkm_ram_func *func, struct nvkm_fb *fb,
+ u32 maskaddr, struct nvkm_ram *ram)
+{
+ struct nvkm_subdev *subdev = &fb->subdev;
+ struct nvkm_device *device = subdev->device;
+ struct nvkm_bios *bios = device->bios;
+ const u32 rsvd_head = ( 256 * 1024); /* vga memory */
+ const u32 rsvd_tail = (1024 * 1024); /* vbios etc */
+ u32 parts = nvkm_rd32(device, 0x022438);
+ u32 pmask = nvkm_rd32(device, maskaddr);
+ u64 bsize = (u64)nvkm_rd32(device, 0x10f20c) << 20;
+ u64 psize, size = 0;
+ enum nvkm_ram_type type = nvkm_fb_bios_memtype(bios);
+ bool uniform = true;
+ int ret, i;
+
+ nvkm_debug(subdev, "100800: %08x\n", nvkm_rd32(device, 0x100800));
+ nvkm_debug(subdev, "parts %08x mask %08x\n", parts, pmask);
/* read amount of vram attached to each memory controller */
- for (part = 0; part < parts; part++) {
- if (!(pmask & (1 << part))) {
- u32 psize = nv_rd32(pfb, 0x11020c + (part * 0x1000));
- if (psize != bsize) {
- if (psize < bsize)
- bsize = psize;
- uniform = false;
- }
-
- nv_debug(pfb, "%d: mem_amount 0x%08x\n", part, psize);
- ram->size += (u64)psize << 20;
+ for (i = 0; i < parts; i++) {
+ if (pmask & (1 << i))
+ continue;
+
+ psize = (u64)nvkm_rd32(device, 0x11020c + (i * 0x1000)) << 20;
+ if (psize != bsize) {
+ if (psize < bsize)
+ bsize = psize;
+ uniform = false;
}
+
+ nvkm_debug(subdev, "%d: %d MiB\n", i, (u32)(psize >> 20));
+ size += psize;
}
+ ret = nvkm_ram_ctor(func, fb, type, size, 0, ram);
+ if (ret)
+ return ret;
+
+ nvkm_mm_fini(&ram->vram);
+
/* if all controllers have the same amount attached, there's no holes */
if (uniform) {
- offset = rsvd_head;
- length = (ram->size >> 12) - rsvd_head - rsvd_tail;
- ret = nvkm_mm_init(&pfb->vram, offset, length, 1);
+ ret = nvkm_mm_init(&ram->vram, rsvd_head >> NVKM_RAM_MM_SHIFT,
+ (size - rsvd_head - rsvd_tail) >>
+ NVKM_RAM_MM_SHIFT, 1);
+ if (ret)
+ return ret;
} else {
/* otherwise, address lowest common amount from 0GiB */
- ret = nvkm_mm_init(&pfb->vram, rsvd_head,
- (bsize << 8) * parts - rsvd_head, 1);
+ ret = nvkm_mm_init(&ram->vram, rsvd_head >> NVKM_RAM_MM_SHIFT,
+ ((bsize * parts) - rsvd_head) >>
+ NVKM_RAM_MM_SHIFT, 1);
if (ret)
return ret;
/* and the rest starting from (8GiB + common_size) */
- offset = (0x0200000000ULL >> 12) + (bsize << 8);
- length = (ram->size >> 12) - ((bsize * parts) << 8) - rsvd_tail;
-
- ret = nvkm_mm_init(&pfb->vram, offset, length, 1);
+ ret = nvkm_mm_init(&ram->vram, (0x0200000000ULL + bsize) >>
+ NVKM_RAM_MM_SHIFT,
+ (size - (bsize * parts) - rsvd_tail) >>
+ NVKM_RAM_MM_SHIFT, 1);
if (ret)
- nvkm_mm_fini(&pfb->vram);
- }
-
- if (ret)
- return ret;
-
- ram->get = gf100_ram_get;
- ram->put = gf100_ram_put;
- return 0;
-}
-
-static int
-gf100_ram_init(struct nvkm_object *object)
-{
- struct nvkm_fb *pfb = (void *)object->parent;
- struct gf100_ram *ram = (void *)object;
- int ret, i;
-
- ret = nvkm_ram_init(&ram->base);
- if (ret)
- return ret;
-
- /* prepare for ddr link training, and load training patterns */
- switch (ram->base.type) {
- case NV_MEM_TYPE_GDDR5: {
- static const u8 train0[] = {
- 0x00, 0xff, 0x55, 0xaa, 0x33, 0xcc,
- 0x00, 0xff, 0xff, 0x00, 0xff, 0x00,
- };
- static const u32 train1[] = {
- 0x00000000, 0xffffffff,
- 0x55555555, 0xaaaaaaaa,
- 0x33333333, 0xcccccccc,
- 0xf0f0f0f0, 0x0f0f0f0f,
- 0x00ff00ff, 0xff00ff00,
- 0x0000ffff, 0xffff0000,
- };
-
- for (i = 0; i < 0x30; i++) {
- nv_wr32(pfb, 0x10f968, 0x00000000 | (i << 8));
- nv_wr32(pfb, 0x10f96c, 0x00000000 | (i << 8));
- nv_wr32(pfb, 0x10f920, 0x00000100 | train0[i % 12]);
- nv_wr32(pfb, 0x10f924, 0x00000100 | train0[i % 12]);
- nv_wr32(pfb, 0x10f918, train1[i % 12]);
- nv_wr32(pfb, 0x10f91c, train1[i % 12]);
- nv_wr32(pfb, 0x10f920, 0x00000000 | train0[i % 12]);
- nv_wr32(pfb, 0x10f924, 0x00000000 | train0[i % 12]);
- nv_wr32(pfb, 0x10f918, train1[i % 12]);
- nv_wr32(pfb, 0x10f91c, train1[i % 12]);
- }
- } break;
- default:
- break;
+ return ret;
}
+ ram->ranks = (nvkm_rd32(device, 0x10f200) & 0x00000004) ? 2 : 1;
return 0;
}
-static int
-gf100_ram_ctor(struct nvkm_object *parent, struct nvkm_object *engine,
- struct nvkm_oclass *oclass, void *data, u32 size,
- struct nvkm_object **pobject)
+int
+gf100_ram_new(struct nvkm_fb *fb, struct nvkm_ram **pram)
{
- struct nvkm_bios *bios = nvkm_bios(parent);
+ struct nvkm_subdev *subdev = &fb->subdev;
+ struct nvkm_bios *bios = subdev->device->bios;
struct gf100_ram *ram;
int ret;
- ret = gf100_ram_create(parent, engine, oclass, 0x022554, &ram);
- *pobject = nv_object(ram);
+ if (!(ram = kzalloc(sizeof(*ram), GFP_KERNEL)))
+ return -ENOMEM;
+ *pram = &ram->base;
+
+ ret = gf100_ram_ctor(&gf100_ram_func, fb, 0x022554, &ram->base);
if (ret)
return ret;
ret = nvbios_pll_parse(bios, 0x0c, &ram->refpll);
if (ret) {
- nv_error(ram, "mclk refpll data not found\n");
+ nvkm_error(subdev, "mclk refpll data not found\n");
return ret;
}
ret = nvbios_pll_parse(bios, 0x04, &ram->mempll);
if (ret) {
- nv_error(ram, "mclk pll data not found\n");
+ nvkm_error(subdev, "mclk pll data not found\n");
return ret;
}
- switch (ram->base.type) {
- case NV_MEM_TYPE_GDDR5:
- ram->base.calc = gf100_ram_calc;
- ram->base.prog = gf100_ram_prog;
- ram->base.tidy = gf100_ram_tidy;
- break;
- default:
- nv_warn(ram, "reclocking of this ram type unsupported\n");
- return 0;
- }
-
ram->fuc.r_0x10fe20 = ramfuc_reg(0x10fe20);
ram->fuc.r_0x10fe24 = ramfuc_reg(0x10fe24);
ram->fuc.r_0x137320 = ramfuc_reg(0x137320);
ram->fuc.r_0x13d8f4 = ramfuc_reg(0x13d8f4);
return 0;
}
-
-struct nvkm_oclass
-gf100_ram_oclass = {
- .handle = 0,
- .ofuncs = &(struct nvkm_ofuncs) {
- .ctor = gf100_ram_ctor,
- .dtor = _nvkm_ram_dtor,
- .init = gf100_ram_init,
- .fini = _nvkm_ram_fini,
- }
-};
Code Review / kvmfornfv.git / blobdiff