mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-12-28 11:18:45 +07:00
0df20f9657
The SC block in VPE/VIP contains a SRAM within it. This internal memory requires to be loaded with appropriate scaler coefficients from a contiguous block of memory through VPDMA. The horizontal and vertical scaler each require 2 sets of scaler coefficients for luma and chroma scaling. The horizontal polyphase scaler requires coefficients for a 32 phase and 8 tap filter. Similarly, the vertical scaler requires coefficients for a 5 tap filter. The choice of the scaler coefficients depends on the scaling ratio. Add coefficient tables for different scaling ratios in sc_coeffs.h. In the case of horizontal downscaling, we need to consider the change in ratio caused by decimation performed by the horizontal scaler. In order to load the scaler coefficients via VPDMA, a configuration descriptor is used in block mode. The payload for the descriptor is the scaler coefficients copied to memory. Coefficients for each phase have to be placed in memory in a particular order understood by the scaler hardware. The choice of the scaler coefficients, and the loading of the coefficients from our tables to a contiguous buffer is managed by the functions sc_set_hs_coefficients and sc_set_vs_coefficients. The sc_data handle is now added with some parameters to describe the state of the coefficients loaded in the SC block. 'loaded_coeff_h' and 'loaded_coeff_v' hold the address of the last dma buffer which was used by VPDMA to copy coefficients. This information can be used by a vpe mem-to-mem context to decide whether it should load coefficients or not. 'hs_index' and 'vs_index' provide some optimization by preventing loading of coefficients if the scaling ratio didn't change between 2 contexts. 'load_coeff_h' and 'load_coeff_v' tell the vpe/vip driver whether we need to load the coefficients through VPDMA or not. Signed-off-by: Archit Taneja <archit@ti.com> Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
190 lines
4.0 KiB
C
190 lines
4.0 KiB
C
/*
|
|
* Scaler library
|
|
*
|
|
* Copyright (c) 2013 Texas Instruments Inc.
|
|
*
|
|
* David Griego, <dagriego@biglakesoftware.com>
|
|
* Dale Farnsworth, <dale@farnsworth.org>
|
|
* Archit Taneja, <archit@ti.com>
|
|
*
|
|
* This program is free software; you can redistribute it and/or modify it
|
|
* under the terms of the GNU General Public License version 2 as published by
|
|
* the Free Software Foundation.
|
|
*/
|
|
|
|
#include <linux/err.h>
|
|
#include <linux/io.h>
|
|
#include <linux/platform_device.h>
|
|
#include <linux/slab.h>
|
|
|
|
#include "sc.h"
|
|
#include "sc_coeff.h"
|
|
|
|
void sc_set_regs_bypass(struct sc_data *sc, u32 *sc_reg0)
|
|
{
|
|
*sc_reg0 |= CFG_SC_BYPASS;
|
|
}
|
|
|
|
void sc_dump_regs(struct sc_data *sc)
|
|
{
|
|
struct device *dev = &sc->pdev->dev;
|
|
|
|
u32 read_reg(struct sc_data *sc, int offset)
|
|
{
|
|
return ioread32(sc->base + offset);
|
|
}
|
|
|
|
#define DUMPREG(r) dev_dbg(dev, "%-35s %08x\n", #r, read_reg(sc, CFG_##r))
|
|
|
|
DUMPREG(SC0);
|
|
DUMPREG(SC1);
|
|
DUMPREG(SC2);
|
|
DUMPREG(SC3);
|
|
DUMPREG(SC4);
|
|
DUMPREG(SC5);
|
|
DUMPREG(SC6);
|
|
DUMPREG(SC8);
|
|
DUMPREG(SC9);
|
|
DUMPREG(SC10);
|
|
DUMPREG(SC11);
|
|
DUMPREG(SC12);
|
|
DUMPREG(SC13);
|
|
DUMPREG(SC17);
|
|
DUMPREG(SC18);
|
|
DUMPREG(SC19);
|
|
DUMPREG(SC20);
|
|
DUMPREG(SC21);
|
|
DUMPREG(SC22);
|
|
DUMPREG(SC23);
|
|
DUMPREG(SC24);
|
|
DUMPREG(SC25);
|
|
|
|
#undef DUMPREG
|
|
}
|
|
|
|
/*
|
|
* set the horizontal scaler coefficients according to the ratio of output to
|
|
* input widths, after accounting for up to two levels of decimation
|
|
*/
|
|
void sc_set_hs_coeffs(struct sc_data *sc, void *addr, unsigned int src_w,
|
|
unsigned int dst_w)
|
|
{
|
|
int sixteenths;
|
|
int idx;
|
|
int i, j;
|
|
u16 *coeff_h = addr;
|
|
const u16 *cp;
|
|
|
|
if (dst_w > src_w) {
|
|
idx = HS_UP_SCALE;
|
|
} else {
|
|
if ((dst_w << 1) < src_w)
|
|
dst_w <<= 1; /* first level decimation */
|
|
if ((dst_w << 1) < src_w)
|
|
dst_w <<= 1; /* second level decimation */
|
|
|
|
if (dst_w == src_w) {
|
|
idx = HS_LE_16_16_SCALE;
|
|
} else {
|
|
sixteenths = (dst_w << 4) / src_w;
|
|
if (sixteenths < 8)
|
|
sixteenths = 8;
|
|
idx = HS_LT_9_16_SCALE + sixteenths - 8;
|
|
}
|
|
}
|
|
|
|
if (idx == sc->hs_index)
|
|
return;
|
|
|
|
cp = scaler_hs_coeffs[idx];
|
|
|
|
for (i = 0; i < SC_NUM_PHASES * 2; i++) {
|
|
for (j = 0; j < SC_H_NUM_TAPS; j++)
|
|
*coeff_h++ = *cp++;
|
|
/*
|
|
* for each phase, the scaler expects space for 8 coefficients
|
|
* in it's memory. For the horizontal scaler, we copy the first
|
|
* 7 coefficients and skip the last slot to move to the next
|
|
* row to hold coefficients for the next phase
|
|
*/
|
|
coeff_h += SC_NUM_TAPS_MEM_ALIGN - SC_H_NUM_TAPS;
|
|
}
|
|
|
|
sc->hs_index = idx;
|
|
|
|
sc->load_coeff_h = true;
|
|
}
|
|
|
|
/*
|
|
* set the vertical scaler coefficients according to the ratio of output to
|
|
* input heights
|
|
*/
|
|
void sc_set_vs_coeffs(struct sc_data *sc, void *addr, unsigned int src_h,
|
|
unsigned int dst_h)
|
|
{
|
|
int sixteenths;
|
|
int idx;
|
|
int i, j;
|
|
u16 *coeff_v = addr;
|
|
const u16 *cp;
|
|
|
|
if (dst_h > src_h) {
|
|
idx = VS_UP_SCALE;
|
|
} else if (dst_h == src_h) {
|
|
idx = VS_1_TO_1_SCALE;
|
|
} else {
|
|
sixteenths = (dst_h << 4) / src_h;
|
|
if (sixteenths < 8)
|
|
sixteenths = 8;
|
|
idx = VS_LT_9_16_SCALE + sixteenths - 8;
|
|
}
|
|
|
|
if (idx == sc->vs_index)
|
|
return;
|
|
|
|
cp = scaler_vs_coeffs[idx];
|
|
|
|
for (i = 0; i < SC_NUM_PHASES * 2; i++) {
|
|
for (j = 0; j < SC_V_NUM_TAPS; j++)
|
|
*coeff_v++ = *cp++;
|
|
/*
|
|
* for the vertical scaler, we copy the first 5 coefficients and
|
|
* skip the last 3 slots to move to the next row to hold
|
|
* coefficients for the next phase
|
|
*/
|
|
coeff_v += SC_NUM_TAPS_MEM_ALIGN - SC_V_NUM_TAPS;
|
|
}
|
|
|
|
sc->vs_index = idx;
|
|
sc->load_coeff_v = true;
|
|
}
|
|
|
|
struct sc_data *sc_create(struct platform_device *pdev)
|
|
{
|
|
struct sc_data *sc;
|
|
|
|
dev_dbg(&pdev->dev, "sc_create\n");
|
|
|
|
sc = devm_kzalloc(&pdev->dev, sizeof(*sc), GFP_KERNEL);
|
|
if (!sc) {
|
|
dev_err(&pdev->dev, "couldn't alloc sc_data\n");
|
|
return ERR_PTR(-ENOMEM);
|
|
}
|
|
|
|
sc->pdev = pdev;
|
|
|
|
sc->res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "sc");
|
|
if (!sc->res) {
|
|
dev_err(&pdev->dev, "missing platform resources data\n");
|
|
return ERR_PTR(-ENODEV);
|
|
}
|
|
|
|
sc->base = devm_ioremap_resource(&pdev->dev, sc->res);
|
|
if (!sc->base) {
|
|
dev_err(&pdev->dev, "failed to ioremap\n");
|
|
return ERR_PTR(-ENOMEM);
|
|
}
|
|
|
|
return sc;
|
|
}
|