Files
linux-nvgpu/drivers/gpu/nvgpu/gk20a/ltc_common.c
Kevin Huang 7d917f43a5 gpu: nvgpu: gm20b: fix compression sharing
For GM20B alone, the LTC count is already accounted for the HW logic
for the CBC base calculation from the postDivide address. So SW
doesn't have to explicity divide it by the LTC count in the postDivide
address calculation.

Bug 1477079

Change-Id: I558bbe66bbcfb7edfa21210d0dc22c6170149260
Signed-off-by: Kevin Huang <kevinh@nvidia.com>
Reviewed-on: http://git-master/r/414264
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
2015-03-18 12:10:03 -07:00

321 lines
8.7 KiB
C

/*
* drivers/video/tegra/host/gk20a/ltc_common.c
*
* GK20A Graphics
*
* Copyright (c) 2011-2014, NVIDIA CORPORATION. All rights reserved.
*
* This program is free software; you can redistribute it and/or modify it
* under the terms and conditions of the GNU General Public License,
* version 2, as published by the Free Software Foundation.
*
* This program is distributed in the hope it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
* more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
#include <linux/dma-mapping.h>
#include <linux/delay.h>
#include "gk20a.h"
#include "gr_gk20a.h"
static int gk20a_determine_L2_size_bytes(struct gk20a *g)
{
const u32 gpuid = GK20A_GPUID(g->gpu_characteristics.arch,
g->gpu_characteristics.impl);
u32 lts_per_ltc;
u32 ways;
u32 sets;
u32 bytes_per_line;
u32 active_ltcs;
u32 cache_size;
u32 tmp;
u32 active_sets_value;
tmp = gk20a_readl(g, ltc_ltc0_lts0_tstg_cfg1_r());
ways = hweight32(ltc_ltc0_lts0_tstg_cfg1_active_ways_v(tmp));
active_sets_value = ltc_ltc0_lts0_tstg_cfg1_active_sets_v(tmp);
if (active_sets_value == ltc_ltc0_lts0_tstg_cfg1_active_sets_all_v()) {
sets = 64;
} else if (active_sets_value ==
ltc_ltc0_lts0_tstg_cfg1_active_sets_half_v()) {
sets = 32;
} else if (active_sets_value ==
ltc_ltc0_lts0_tstg_cfg1_active_sets_quarter_v()) {
sets = 16;
} else {
dev_err(dev_from_gk20a(g),
"Unknown constant %u for active sets",
(unsigned)active_sets_value);
sets = 0;
}
active_ltcs = g->gr.num_fbps;
/* chip-specific values */
switch (gpuid) {
case GK20A_GPUID_GK20A:
lts_per_ltc = 1;
bytes_per_line = 128;
break;
case GK20A_GPUID_GM20B:
lts_per_ltc = 2;
bytes_per_line = 128;
break;
default:
dev_err(dev_from_gk20a(g), "Unknown GPU id 0x%02x\n",
(unsigned)gpuid);
lts_per_ltc = 0;
bytes_per_line = 0;
}
cache_size = active_ltcs * lts_per_ltc * ways * sets * bytes_per_line;
return cache_size;
}
/*
* Set the maximum number of ways that can have the "EVIST_LAST" class.
*/
static void gk20a_ltc_set_max_ways_evict_last(struct gk20a *g, u32 max_ways)
{
u32 mgmt_reg;
mgmt_reg = gk20a_readl(g, ltc_ltcs_ltss_tstg_set_mgmt_r()) &
~ltc_ltcs_ltss_tstg_set_mgmt_max_ways_evict_last_f(~0);
mgmt_reg |= ltc_ltcs_ltss_tstg_set_mgmt_max_ways_evict_last_f(max_ways);
gk20a_writel(g, ltc_ltcs_ltss_tstg_set_mgmt_r(), mgmt_reg);
}
/*
* Sets the ZBC color for the passed index.
*/
static void gk20a_ltc_set_zbc_color_entry(struct gk20a *g,
struct zbc_entry *color_val,
u32 index)
{
u32 i;
u32 real_index = index + GK20A_STARTOF_ZBC_TABLE;
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_index_r(),
ltc_ltcs_ltss_dstg_zbc_index_address_f(real_index));
for (i = 0;
i < ltc_ltcs_ltss_dstg_zbc_color_clear_value__size_1_v(); i++)
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_color_clear_value_r(i),
color_val->color_l2[i]);
}
/*
* Sets the ZBC depth for the passed index.
*/
static void gk20a_ltc_set_zbc_depth_entry(struct gk20a *g,
struct zbc_entry *depth_val,
u32 index)
{
u32 real_index = index + GK20A_STARTOF_ZBC_TABLE;
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_index_r(),
ltc_ltcs_ltss_dstg_zbc_index_address_f(real_index));
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_depth_clear_value_r(),
depth_val->depth);
}
/*
* Clear the L2 ZBC color table for the passed index.
*/
static void gk20a_ltc_clear_zbc_color_entry(struct gk20a *g, u32 index)
{
u32 i;
u32 real_index = index + GK20A_STARTOF_ZBC_TABLE;
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_index_r(),
ltc_ltcs_ltss_dstg_zbc_index_address_f(real_index));
for (i = 0;
i < ltc_ltcs_ltss_dstg_zbc_color_clear_value__size_1_v(); i++)
gk20a_writel(g,
ltc_ltcs_ltss_dstg_zbc_color_clear_value_r(i), 0);
}
/*
* Clear the L2 ZBC depth entry for the passed index.
*/
static void gk20a_ltc_clear_zbc_depth_entry(struct gk20a *g, u32 index)
{
u32 real_index = index + GK20A_STARTOF_ZBC_TABLE;
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_index_r(),
ltc_ltcs_ltss_dstg_zbc_index_address_f(real_index));
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_depth_clear_value_r(), 0);
}
static int gk20a_ltc_init_zbc(struct gk20a *g, struct gr_gk20a *gr)
{
u32 i, j;
/* reset zbc clear */
for (i = 0; i < GK20A_SIZEOF_ZBC_TABLE -
GK20A_STARTOF_ZBC_TABLE; i++) {
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_index_r(),
(gk20a_readl(g, ltc_ltcs_ltss_dstg_zbc_index_r()) &
~ltc_ltcs_ltss_dstg_zbc_index_address_f(~0)) |
ltc_ltcs_ltss_dstg_zbc_index_address_f(
i + GK20A_STARTOF_ZBC_TABLE));
for (j = 0; j < ltc_ltcs_ltss_dstg_zbc_color_clear_value__size_1_v(); j++)
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_color_clear_value_r(j), 0);
gk20a_writel(g, ltc_ltcs_ltss_dstg_zbc_depth_clear_value_r(), 0);
}
gr_gk20a_clear_zbc_table(g, gr);
gr_gk20a_load_zbc_default_table(g, gr);
return 0;
}
static int gk20a_ltc_alloc_phys_cbc(struct gk20a *g,
size_t compbit_backing_size)
{
struct gr_gk20a *gr = &g->gr;
int order = ffs(compbit_backing_size >> PAGE_SHIFT);
struct page *pages;
struct sg_table *sgt;
int err = 0;
/* allocate few pages */
pages = alloc_pages(GFP_KERNEL, order);
if (!pages) {
gk20a_dbg(gpu_dbg_pte, "alloc_pages failed\n");
err = -ENOMEM;
goto err_alloc_pages;
}
/* clean up the pages */
memset(page_address(pages), 0, compbit_backing_size);
/* allocate room for placing the pages pointer.. */
gr->compbit_store.pages =
kzalloc(sizeof(*gr->compbit_store.pages), GFP_KERNEL);
if (!gr->compbit_store.pages) {
gk20a_dbg(gpu_dbg_pte, "failed to allocate pages struct");
err = -ENOMEM;
goto err_alloc_compbit_store;
}
err = gk20a_get_sgtable_from_pages(&g->dev->dev, &sgt, &pages, 0,
compbit_backing_size);
if (err) {
gk20a_dbg(gpu_dbg_pte, "could not get sg table for pages\n");
goto err_alloc_sg_table;
}
/* store the parameters to gr structure */
*gr->compbit_store.pages = pages;
gr->compbit_store.base_iova = sg_phys(sgt->sgl);
gr->compbit_store.size = compbit_backing_size;
gr->compbit_store.sgt = sgt;
return 0;
err_alloc_sg_table:
kfree(gr->compbit_store.pages);
gr->compbit_store.pages = NULL;
err_alloc_compbit_store:
__free_pages(pages, order);
err_alloc_pages:
return err;
}
static int gk20a_ltc_alloc_virt_cbc(struct gk20a *g,
size_t compbit_backing_size)
{
struct device *d = dev_from_gk20a(g);
struct gr_gk20a *gr = &g->gr;
DEFINE_DMA_ATTRS(attrs);
dma_addr_t iova;
int err;
dma_set_attr(DMA_ATTR_NO_KERNEL_MAPPING, &attrs);
gr->compbit_store.pages =
dma_alloc_attrs(d, compbit_backing_size, &iova,
GFP_KERNEL, &attrs);
if (!gr->compbit_store.pages) {
gk20a_err(dev_from_gk20a(g), "failed to allocate backing store for compbit : size %zu",
compbit_backing_size);
return -ENOMEM;
}
gr->compbit_store.base_iova = iova;
gr->compbit_store.size = compbit_backing_size;
err = gk20a_get_sgtable_from_pages(d,
&gr->compbit_store.sgt,
gr->compbit_store.pages, iova,
compbit_backing_size);
if (err) {
gk20a_err(dev_from_gk20a(g), "failed to allocate sgt for backing store");
return err;
}
return 0;
}
static void gk20a_ltc_init_cbc(struct gk20a *g, struct gr_gk20a *gr)
{
u32 max_size = gr->max_comptag_mem;
u32 max_comptag_lines = max_size << 3;
u32 compbit_base_post_divide;
u64 compbit_base_post_multiply64;
u64 compbit_store_base_iova;
u64 compbit_base_post_divide64;
if (IS_ENABLED(CONFIG_GK20A_PHYS_PAGE_TABLES))
compbit_store_base_iova = gr->compbit_store.base_iova;
else
compbit_store_base_iova = NV_MC_SMMU_VADDR_TRANSLATE(
gr->compbit_store.base_iova);
compbit_base_post_divide64 = compbit_store_base_iova >>
ltc_ltcs_ltss_cbc_base_alignment_shift_v();
do_div(compbit_base_post_divide64, g->ltc_count);
compbit_base_post_divide = u64_lo32(compbit_base_post_divide64);
compbit_base_post_multiply64 = ((u64)compbit_base_post_divide *
gr->num_fbps) << ltc_ltcs_ltss_cbc_base_alignment_shift_v();
if (compbit_base_post_multiply64 < compbit_store_base_iova)
compbit_base_post_divide++;
/* Bug 1477079 indicates sw adjustment on the posted divided base. */
if (g->ops.ltc.cbc_fix_config)
compbit_base_post_divide =
g->ops.ltc.cbc_fix_config(g, compbit_base_post_divide);
gk20a_writel(g, ltc_ltcs_ltss_cbc_base_r(),
compbit_base_post_divide);
gk20a_dbg(gpu_dbg_info | gpu_dbg_map | gpu_dbg_pte,
"compbit base.pa: 0x%x,%08x cbc_base:0x%08x\n",
(u32)(compbit_store_base_iova >> 32),
(u32)(compbit_store_base_iova & 0xffffffff),
compbit_base_post_divide);
g->ops.ltc.cbc_ctrl(g, gk20a_cbc_op_invalidate,
0, max_comptag_lines - 1);
}