Re: [PATCH RFC 4/8] drm/sprd: add Unisoc's drm display controller driver

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 





On Tue, Dec 10, 2019 at 6:45 PM Thomas Zimmermann <tzimmermann@xxxxxxx> wrote:
Hi

Am 10.12.19 um 09:36 schrieb Kevin Tang:
> From: Kevin Tang <kevin.tang@xxxxxxxxxx>
>
> Adds DPU(Display Processor Unit) support for the Unisoc's display subsystem.
> It's support multi planes, scaler, rotation, PQ(Picture Quality) and more.
>
> Cc: Orson Zhai <orsonzhai@xxxxxxxxx>
> Cc: Baolin Wang <baolin.wang@xxxxxxxxxx>
> Cc: Chunyan Zhang <zhang.lyra@xxxxxxxxx>
> Signed-off-by: Kevin Tang <kevin.tang@xxxxxxxxxx>
> ---
>  drivers/gpu/drm/sprd/Makefile       |    6 +-
>  drivers/gpu/drm/sprd/disp_lib.c     |  290 +++++++
>  drivers/gpu/drm/sprd/disp_lib.h     |   40 +
>  drivers/gpu/drm/sprd/dpu/Makefile   |    8 +
>  drivers/gpu/drm/sprd/dpu/dpu_r2p0.c | 1464 +++++++++++++++++++++++++++++++++++
>  drivers/gpu/drm/sprd/sprd_dpu.c     | 1152 +++++++++++++++++++++++++++
>  drivers/gpu/drm/sprd/sprd_dpu.h     |  217 ++++++
>  7 files changed, 3176 insertions(+), 1 deletion(-)
>  create mode 100644 drivers/gpu/drm/sprd/disp_lib.c
>  create mode 100644 drivers/gpu/drm/sprd/disp_lib.h
>  create mode 100644 drivers/gpu/drm/sprd/dpu/Makefile
>  create mode 100644 drivers/gpu/drm/sprd/dpu/dpu_r2p0.c
>  create mode 100644 drivers/gpu/drm/sprd/sprd_dpu.c
>  create mode 100644 drivers/gpu/drm/sprd/sprd_dpu.h
>
> diff --git a/drivers/gpu/drm/sprd/Makefile b/drivers/gpu/drm/sprd/Makefile
> index df0b316..3f188ab 100644
> --- a/drivers/gpu/drm/sprd/Makefile
> +++ b/drivers/gpu/drm/sprd/Makefile
> @@ -5,4 +5,8 @@ ccflags-y += -Iinclude/drm
>  subdir-ccflags-y += -I$(src)

>  obj-y := sprd_drm.o \
> -     sprd_gem.o
> \ No newline at end of file
> +     sprd_gem.o \
> +     sprd_dpu.o
> +
> +obj-y += disp_lib.o
> +obj-y += dpu/
> \ No newline at end of file
> diff --git a/drivers/gpu/drm/sprd/disp_lib.c b/drivers/gpu/drm/sprd/disp_lib.c
> new file mode 100644
> index 0000000..cadd1ad
> --- /dev/null
> +++ b/drivers/gpu/drm/sprd/disp_lib.c
> @@ -0,0 +1,290 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (C) 2019 Unisoc Inc.
> + */
> +
> +#define pr_fmt(__fmt) "[drm][%20s] "__fmt, __func__
> +
> +#include <linux/device.h>
> +#include <linux/libfdt.h>
> +#include <linux/module.h>
> +#include <linux/of_graph.h>
> +#include <linux/of_platform.h>
> +#include <linux/slab.h>
> +#include <linux/uaccess.h>
> +#include <linux/fs.h>
> +
> +#include "disp_lib.h"
> +
> +struct bmp_header {
> +     u16 magic;
> +     u32 size;
> +     u32 unused;
> +     u32 start;
> +} __attribute__((__packed__));
> +
> +struct dib_header {
> +     u32 size;
> +     u32 width;
> +     u32 height;
> +     u16 planes;
> +     u16 bpp;
> +     u32 compression;
> +     u32 data_size;
> +     u32 h_res;
> +     u32 v_res;
> +     u32 colours;
> +     u32 important_colours;
> +     u32 red_mask;
> +     u32 green_mask;
> +     u32 blue_mask;
> +     u32 alpha_mask;
> +     u32 colour_space;
> +     u32 unused[12];
> +} __attribute__((__packed__));
> +
> +int str_to_u32_array(const char *p, u32 base, u32 array[])
> +{
> +     const char *start = p;
> +     char str[12];
> +     int length = 0;
> +     int i, ret;
> +
> +     pr_info("input: %s", p);
> +
> +     for (i = 0 ; i < 255; i++) {
> +             while (*p == ' ')
> +                     p++;
> +             if (*p == '\0')
> +                     break;
> +             start = p;
> +
> +             while ((*p != ' ') && (*p != '\0'))
> +                     p++;
> +
> +             if ((p - start) >= sizeof(str))
> +                     break;
> +
> +             memset(str, 0, sizeof(str));
> +             memcpy(str, start, p - start);
> +
> +             ret = kstrtou32(str, base, &array[i]);
> +             if (ret) {
> +                     DRM_ERROR("input format error\n");
> +                     break;
> +             }
> +
> +             length++;
> +     }
> +
> +     return length;
> +}
> +EXPORT_SYMBOL_GPL(str_to_u32_array);
> +
> +int load_dtb_to_mem(const char *name, void **blob)
> +{
> +     ssize_t ret;
> +     u32 count;
> +     struct fdt_header dtbhead;
> +     loff_t pos = 0;
> +     struct file *fdtb;
> +
> +
> +     fdtb = filp_open(name, O_RDONLY, 0644);
> +     if (IS_ERR(fdtb)) {
> +             DRM_ERROR("%s open file error\n", __func__);
> +             return PTR_ERR(fdtb);
> +     }
> +
> +     ret = kernel_read(fdtb, &dtbhead, sizeof(dtbhead), &pos);
> +     pos = 0;
> +     count = ntohl(dtbhead.totalsize);
> +     *blob = kzalloc(count, GFP_KERNEL);
> +     if (*blob == NULL) {
> +             filp_close(fdtb, NULL);
> +             return -ENOMEM;
> +     }
> +     ret = kernel_read(fdtb, *blob, count, &pos);
> +
> +     if (ret != count) {
> +             DRM_ERROR("Read to mem fail: ret %zd size%x\n", ret, count);
> +             kfree(*blob);
> +             *blob = NULL;
> +             filp_close(fdtb, NULL);
> +             return ret < 0 ? ret : -ENODEV;
> +     }
> +
> +     filp_close(fdtb, NULL);
> +
> +     return 0;
> +}
> +EXPORT_SYMBOL_GPL(load_dtb_to_mem);
> +
> +int str_to_u8_array(const char *p, u32 base, u8 array[])
> +{
> +     const char *start = p;
> +     char str[12];
> +     int length = 0;
> +     int i, ret;
> +
> +     pr_info("input: %s", p);
> +
> +     for (i = 0 ; i < 255; i++) {
> +             while (*p == ' ')
> +                     p++;
> +             if (*p == '\0')
> +                     break;
> +             start = p;
> +
> +             while ((*p != ' ') && (*p != '\0'))
> +                     p++;
> +
> +             if ((p - start) >= sizeof(str))
> +                     break;
> +
> +             memset(str, 0, sizeof(str));
> +             memcpy(str, start, p - start);
> +
> +             ret = kstrtou8(str, base, &array[i]);
> +             if (ret) {
> +                     DRM_ERROR("input format error\n");
> +                     break;
> +             }
> +
> +             length++;
> +     }
> +
> +     return length;
> +}
> +EXPORT_SYMBOL_GPL(str_to_u8_array);
> +
> +int dump_bmp32(const char *p, u32 width, u32 height,
> +             bool noflip, const char *filename)
> +{
> +     struct file *fp;
> +     mm_segment_t fs;
> +     loff_t pos;
> +     struct dib_header dib_header = {
> +             .size = sizeof(dib_header),
> +             .width = width,
> +             .height = noflip ? -height : height,
> +             .planes = 1,
> +             .bpp = 32,
> +             .compression = 3,
> +             .data_size = 4 * width * height,
> +             .h_res = 0xB13,
> +             .v_res = 0xB13,
> +             .colours = 0,
> +             .important_colours = 0,
> +             .red_mask = 0x000000FF,
> +             .green_mask = 0x0000FF00,
> +             .blue_mask = 0x00FF0000,
> +             .alpha_mask = 0xFF000000,
> +             .colour_space = 0x57696E20,
> +     };
> +     struct bmp_header bmp_header = {
> +             .magic = 0x4d42,
> +             .size = (width * height * 4) +
> +             sizeof(bmp_header) + sizeof(dib_header),
> +             .start = sizeof(bmp_header) + sizeof(dib_header),
> +     };
> +
> +     fp = filp_open(filename, O_RDWR | O_CREAT, 0644);
> +     if (IS_ERR(fp)) {
> +             DRM_ERROR("failed to open %s: %ld\n", filename, PTR_ERR(fp));
> +             return PTR_ERR(fp);
> +     }
> +
> +     fs = get_fs();
> +     set_fs(KERNEL_DS);
> +     pos = 0;
> +
> +     vfs_write(fp, (const char *)&bmp_header, sizeof(bmp_header), &pos);
> +     vfs_write(fp, (const char *)&dib_header, sizeof(dib_header), &pos);
> +     vfs_write(fp, p, width * height * 4, &pos);
> +
> +     filp_close(fp, NULL);
> +     set_fs(fs);
> +
> +     return 0;
> +}
> +EXPORT_SYMBOL_GPL(dump_bmp32);
> +
> +void *disp_ops_attach(const char *str, struct list_head *head)
> +{
> +     struct ops_list *list;
> +     const char *ver;
> +
> +     list_for_each_entry(list, head, head) {
> +             ver = list->entry->ver;
> +             if (!strcmp(str, ver))
> +                     return list->entry->ops;
> +     }
> +
> +     DRM_ERROR("attach disp ops %s failed\n", str);
> +     return NULL;
> +}
> +EXPORT_SYMBOL_GPL(disp_ops_attach);
> +
> +int disp_ops_register(struct ops_entry *entry, struct list_head *head)
> +{
> +     struct ops_list *list;
> +
> +     list = kzalloc(sizeof(struct ops_list), GFP_KERNEL);
> +     if (!list)
> +             return -ENOMEM;
> +
> +     list->entry = entry;
> +     list_add(&list->head, head);
> +
> +     return 0;
> +}
> +EXPORT_SYMBOL_GPL(disp_ops_register);
> +
> +struct device *sprd_disp_pipe_get_by_port(struct device *dev, int port)
> +{
> +     struct device_node *np = dev->of_node;
> +     struct device_node *endpoint;
> +     struct device_node *remote_node;
> +     struct platform_device *remote_pdev;
> +
> +     endpoint = of_graph_get_endpoint_by_regs(np, port, 0);
> +     if (!endpoint) {
> +             DRM_ERROR("%s/port%d/endpoint0 was not found\n",
> +                       np->full_name, port);
> +             return NULL;
> +     }
> +
> +     remote_node = of_graph_get_remote_port_parent(endpoint);
> +     if (!remote_node) {
> +             DRM_ERROR("device node was not found by endpoint0\n");
> +             return NULL;
> +     }
> +
> +     remote_pdev = of_find_device_by_node(remote_node);
> +     if (remote_pdev == NULL) {
> +             DRM_ERROR("find %s platform device failed\n",
> +                       remote_node->full_name);
> +             return NULL;
> +     }
> +
> +     return &remote_pdev->dev;
> +}
> +EXPORT_SYMBOL_GPL(sprd_disp_pipe_get_by_port);
> +
> +struct device *sprd_disp_pipe_get_input(struct device *dev)
> +{
> +     return sprd_disp_pipe_get_by_port(dev, 1);
> +}
> +EXPORT_SYMBOL_GPL(sprd_disp_pipe_get_input);
> +
> +struct device *sprd_disp_pipe_get_output(struct device *dev)
> +{
> +     return sprd_disp_pipe_get_by_port(dev, 0);
> +}
> +EXPORT_SYMBOL_GPL(sprd_disp_pipe_get_output);
> +
> +MODULE_AUTHOR("Leon He <leon.he@xxxxxxxxxx>");
> +MODULE_AUTHOR("Kevin Tang <kevin.tang@xxxxxxxxxx>");
> +MODULE_DESCRIPTION("Unisoc display common API library");
> +MODULE_LICENSE("GPL");
> diff --git a/drivers/gpu/drm/sprd/disp_lib.h b/drivers/gpu/drm/sprd/disp_lib.h
> new file mode 100644
> index 0000000..7900b89
> --- /dev/null
> +++ b/drivers/gpu/drm/sprd/disp_lib.h
> @@ -0,0 +1,40 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +/*
> + * Copyright (C) 2019 Unisoc Inc.
> + */
> +
> +#ifndef _DISP_LIB_H_
> +#define _DISP_LIB_H_
> +
> +#include <linux/list.h>
> +#include <drm/drm_print.h>
> +
> +#ifdef pr_fmt
> +#undef pr_fmt
> +#define pr_fmt(__fmt) "[drm][%20s] "__fmt, __func__
> +#endif
> +
> +struct ops_entry {
> +     const char *ver;
> +     void *ops;
> +};
> +
> +struct ops_list {
> +     struct list_head head;
> +     struct ops_entry *entry;
> +};
> +
> +int str_to_u32_array(const char *p, u32 base, u32 array[]);
> +int str_to_u8_array(const char *p, u32 base, u8 array[]);
> +int dump_bmp32(const char *p, u32 width, u32 height,
> +             bool bgra, const char *filename);
> +int load_dtb_to_mem(const char *name, void **blob);

The bitmap-loading code seems out of place in a driver. It should be a
helper. And efifb has similar code already. Is there an opportunity to
share the implementation?

I cannot find any caller of load_dtb_to_mem(). Where is it being used?

Best regards
Thomas

> +
> +void *disp_ops_attach(const char *str, struct list_head *head);
> +int disp_ops_register(struct ops_entry *entry, struct list_head *head);
> +
> +struct device *sprd_disp_pipe_get_by_port(struct device *dev, int port);
> +struct device *sprd_disp_pipe_get_input(struct device *dev);
> +struct device *sprd_disp_pipe_get_output(struct device *dev);
> +
> +#endif
> diff --git a/drivers/gpu/drm/sprd/dpu/Makefile b/drivers/gpu/drm/sprd/dpu/Makefile
> new file mode 100644
> index 0000000..d960107
> --- /dev/null
> +++ b/drivers/gpu/drm/sprd/dpu/Makefile
> @@ -0,0 +1,8 @@
> +# SPDX-License-Identifier: GPL-2.0
> +
> +ifdef CONFIG_ARM64
> +KBUILD_CFLAGS += -mstrict-align
> +endif
> +
> +obj-y += dpu_r2p0.o
> +
> diff --git a/drivers/gpu/drm/sprd/dpu/dpu_r2p0.c b/drivers/gpu/drm/sprd/dpu/dpu_r2p0.c
> new file mode 100644
> index 0000000..4c0a539
> --- /dev/null
> +++ b/drivers/gpu/drm/sprd/dpu/dpu_r2p0.c
> @@ -0,0 +1,1464 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (C) 2019 Unisoc Inc.
> + */
> +
> +#include <linux/delay.h>
> +#include <linux/wait.h>
> +#include <linux/workqueue.h>
> +#include "sprd_dpu.h"
> +
> +#define DISPC_INT_FBC_PLD_ERR_MASK   BIT(8)
> +#define DISPC_INT_FBC_HDR_ERR_MASK   BIT(9)
> +
> +#define DISPC_INT_MMU_INV_WR_MASK    BIT(19)
> +#define DISPC_INT_MMU_INV_RD_MASK    BIT(18)
> +#define DISPC_INT_MMU_VAOR_WR_MASK   BIT(17)
> +#define DISPC_INT_MMU_VAOR_RD_MASK   BIT(16)
> +
> +#define XFBC8888_HEADER_SIZE(w, h) (ALIGN((w) * (h) / (8 * 8) / 2, 128))
> +#define XFBC8888_PAYLOAD_SIZE(w, h) (w * h * 4)
> +#define XFBC8888_BUFFER_SIZE(w, h) (XFBC8888_HEADER_SIZE(w, h) \
> +                             + XFBC8888_PAYLOAD_SIZE(w, h))
> +
> +#define XFBC565_HEADER_SIZE(w, h) (ALIGN((w) * (h) / (16 * 8) / 2, 128))
> +#define XFBC565_PAYLOAD_SIZE(w, h) (w * h * 2)
> +#define XFBC565_BUFFER_SIZE(w, h) (XFBC565_HEADER_SIZE(w, h) \
> +                             + XFBC565_PAYLOAD_SIZE(w, h))
> +
> +#define SLP_BRIGHTNESS_THRESHOLD 0x20
> +
> +struct layer_reg {
> +     u32 addr[4];
> +     u32 ctrl;
> +     u32 size;
> +     u32 pitch;
> +     u32 pos;
> +     u32 alpha;
> +     u32 ck;
> +     u32 pallete;
> +     u32 crop_start;
> +};
> +
> +struct wb_region_reg {
> +     u32 pos;
> +     u32 size;
> +};
> +
> +struct dpu_reg {
> +     u32 dpu_version;
> +     u32 dpu_ctrl;
> +     u32 dpu_cfg0;
> +     u32 dpu_cfg1;
> +     u32 dpu_cfg2;
> +     u32 dpu_secure;
> +     u32 reserved_0x0018_0x001C[2];
> +     u32 panel_size;
> +     u32 blend_size;
> +     u32 reserved_0x0028;
> +     u32 bg_color;
> +     struct layer_reg layers[8];
> +     u32 wb_base_addr;
> +     u32 wb_ctrl;
> +     u32 wb_cfg;
> +     u32 wb_pitch;
> +     struct wb_region_reg region[3];
> +     u32 reserved_0x01D8_0x01DC[2];
> +     u32 dpu_int_en;
> +     u32 dpu_int_clr;
> +     u32 dpu_int_sts;
> +     u32 dpu_int_raw;
> +     u32 dpi_ctrl;
> +     u32 dpi_h_timing;
> +     u32 dpi_v_timing;
> +     u32 reserved_0x01FC;
> +     u32 dpu_enhance_cfg;
> +     u32 reserved_0x0204_0x020C[3];
> +     u32 epf_epsilon;
> +     u32 epf_gain0_3;
> +     u32 epf_gain4_7;
> +     u32 epf_diff;
> +     u32 reserved_0x0220_0x023C[8];
> +     u32 hsv_lut_addr;
> +     u32 hsv_lut_wdata;
> +     u32 hsv_lut_rdata;
> +     u32 reserved_0x024C_0x027C[13];
> +     u32 cm_coef01_00;
> +     u32 cm_coef03_02;
> +     u32 cm_coef11_10;
> +     u32 cm_coef13_12;
> +     u32 cm_coef21_20;
> +     u32 cm_coef23_22;
> +     u32 reserved_0x0298_0x02BC[10];
> +     u32 slp_cfg0;
> +     u32 slp_cfg1;
> +     u32 reserved_0x02C8_0x02FC[14];
> +     u32 gamma_lut_addr;
> +     u32 gamma_lut_wdata;
> +     u32 gamma_lut_rdata;
> +     u32 reserved_0x030C_0x033C[13];
> +     u32 checksum_en;
> +     u32 checksum0_start_pos;
> +     u32 checksum0_end_pos;
> +     u32 checksum1_start_pos;
> +     u32 checksum1_end_pos;
> +     u32 checksum0_result;
> +     u32 checksum1_result;
> +     u32 reserved_0x035C;
> +     u32 dpu_sts[18];
> +     u32 reserved_0x03A8_0x03AC[2];
> +     u32 dpu_fbc_cfg0;
> +     u32 dpu_fbc_cfg1;
> +     u32 reserved_0x03B8_0x03EC[14];
> +     u32 rf_ram_addr;
> +     u32 rf_ram_rdata_low;
> +     u32 rf_ram_rdata_high;
> +     u32 reserved_0x03FC_0x07FC[257];
> +     u32 mmu_en;
> +     u32 mmu_update;
> +     u32 mmu_min_vpn;
> +     u32 mmu_vpn_range;
> +     u32 mmu_pt_addr;
> +     u32 mmu_default_page;
> +     u32 mmu_vaor_addr_rd;
> +     u32 mmu_vaor_addr_wr;
> +     u32 mmu_inv_addr_rd;
> +     u32 mmu_inv_addr_wr;
> +     u32 mmu_uns_addr_rd;
> +     u32 mmu_uns_addr_wr;
> +     u32 mmu_miss_cnt;
> +     u32 mmu_pt_update_qos;
> +     u32 mmu_version;
> +     u32 mmu_min_ppn1;
> +     u32 mmu_ppn_range1;
> +     u32 mmu_min_ppn2;
> +     u32 mmu_ppn_range2;
> +     u32 mmu_vpn_paor_rd;
> +     u32 mmu_vpn_paor_wr;
> +     u32 mmu_ppn_paor_rd;
> +     u32 mmu_ppn_paor_wr;
> +     u32 mmu_reg_au_manage;
> +     u32 mmu_page_rd_ch;
> +     u32 mmu_page_wr_ch;
> +     u32 mmu_read_page_cmd_cnt;
> +     u32 mmu_read_page_latency_cnt;
> +     u32 mmu_page_max_latency;
> +};
> +
> +struct wb_region {
> +     u32 index;
> +     u16 pos_x;
> +     u16 pos_y;
> +     u16 size_w;
> +     u16 size_h;
> +};
> +
> +struct enhance_module {
> +     u32 scl_en: 1;
> +     u32 epf_en: 1;
> +     u32 hsv_en: 1;
> +     u32 cm_en: 1;
> +     u32 slp_en: 1;
> +     u32 gamma_en: 1;
> +     u32 blp_en: 1;
> +};
> +
> +struct scale_cfg {
> +     u32 in_w;
> +     u32 in_h;
> +};
> +
> +struct epf_cfg {
> +     u16 epsilon0;
> +     u16 epsilon1;
> +     u8 gain0;
> +     u8 gain1;
> +     u8 gain2;
> +     u8 gain3;
> +     u8 gain4;
> +     u8 gain5;
> +     u8 gain6;
> +     u8 gain7;
> +     u8 max_diff;
> +     u8 min_diff;
> +};
> +
> +struct hsv_entry {
> +     u16 hue;
> +     u16 sat;
> +};
> +
> +struct hsv_lut {
> +     struct hsv_entry table[360];
> +};
> +
> +struct gamma_entry {
> +     u16 r;
> +     u16 g;
> +     u16 b;
> +};
> +
> +struct gamma_lut {
> +     u16 r[256];
> +     u16 g[256];
> +     u16 b[256];
> +};
> +
> +struct cm_cfg {
> +     short coef00;
> +     short coef01;
> +     short coef02;
> +     short coef03;
> +     short coef10;
> +     short coef11;
> +     short coef12;
> +     short coef13;
> +     short coef20;
> +     short coef21;
> +     short coef22;
> +     short coef23;
> +};
> +
> +struct slp_cfg {
> +     u8 brightness;
> +     u8 conversion_matrix;
> +     u8 brightness_step;
> +     u8 second_bright_factor;
> +     u8 first_percent_th;
> +     u8 first_max_bright_th;
> +};
> +
> +static struct scale_cfg scale_copy;
> +static struct cm_cfg cm_copy;
> +static struct slp_cfg slp_copy;
> +static struct gamma_lut gamma_copy;
> +static struct hsv_lut hsv_copy;
> +static struct epf_cfg epf_copy;
> +static u32 enhance_en;
> +
> +static DECLARE_WAIT_QUEUE_HEAD(wait_queue);
> +static bool panel_ready = true;
> +static bool need_scale;
> +static bool mode_changed;
> +static bool evt_update;
> +static bool evt_stop;
> +static u32 prev_y2r_coef;
> +
> +static void dpu_sr_config(struct dpu_context *ctx);
> +static void dpu_enhance_reload(struct dpu_context *ctx);
> +static void dpu_clean_all(struct dpu_context *ctx);
> +static void dpu_layer(struct dpu_context *ctx,
> +                 struct sprd_dpu_layer *hwlayer);
> +
> +static u32 dpu_get_version(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     return reg->dpu_version;
> +}
> +
> +static bool dpu_check_raw_int(struct dpu_context *ctx, u32 mask)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     u32 val;
> +
> +     down(&ctx->refresh_lock);
> +     if (!ctx->is_inited) {
> +             up(&ctx->refresh_lock);
> +             pr_err("dpu is not initialized\n");
> +             return false;
> +     }
> +
> +     val = reg->dpu_int_raw;
> +     up(&ctx->refresh_lock);
> +
> +     if (val & mask)
> +             return true;
> +
> +     pr_err("dpu_int_raw:0x%x\n", val);
> +     return false;
> +}
> +
> +static int dpu_parse_dt(struct dpu_context *ctx,
> +                             struct device_node *np)
> +{
> +     return 0;
> +}
> +
> +static void dpu_dump(struct dpu_context *ctx)
> +{
> +     u32 *reg = (u32 *)ctx->base;
> +     int i;
> +
> +     pr_info("      0          4          8          C\n");
> +     for (i = 0; i < 256; i += 4) {
> +             pr_info("%04x: 0x%08x 0x%08x 0x%08x 0x%08x\n",
> +                     i * 4, reg[i], reg[i + 1], reg[i + 2], reg[i + 3]);
> +     }
> +}
> +
> +static u32 check_mmu_isr(struct dpu_context *ctx, u32 reg_val)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     u32 mmu_mask = DISPC_INT_MMU_VAOR_RD_MASK |
> +                     DISPC_INT_MMU_VAOR_WR_MASK |
> +                     DISPC_INT_MMU_INV_RD_MASK |
> +                     DISPC_INT_MMU_INV_WR_MASK;
> +     u32 val = reg_val & mmu_mask;
> +
> +     if (val) {
> +             pr_err("--- iommu interrupt err: 0x%04x ---\n", val);
> +
> +             pr_err("iommu invalid read error, addr: 0x%08x\n",
> +                     reg->mmu_inv_addr_rd);
> +             pr_err("iommu invalid write error, addr: 0x%08x\n",
> +                     reg->mmu_inv_addr_wr);
> +             pr_err("iommu va out of range read error, addr: 0x%08x\n",
> +                     reg->mmu_vaor_addr_rd);
> +             pr_err("iommu va out of range write error, addr: 0x%08x\n",
> +                     reg->mmu_vaor_addr_wr);
> +             pr_err("BUG: iommu failure at %s:%d/%s()!\n",
> +                     __FILE__, __LINE__, __func__);
> +
> +             dpu_dump(ctx);
> +     }
> +
> +     return val;
> +}
> +
> +static u32 dpu_isr(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     u32 reg_val, int_mask = 0;
> +
> +     reg_val = reg->dpu_int_sts;
> +
> +     /* disable err interrupt */
> +     if (reg_val & DISPC_INT_ERR_MASK)
> +             int_mask |= DISPC_INT_ERR_MASK;
> +
> +     /* dpu update done isr */
> +     if (reg_val & DISPC_INT_UPDATE_DONE_MASK) {
> +             evt_update = true;
> +             wake_up_interruptible_all(&wait_queue);
> +     }
> +
> +     /* dpu stop done isr */
> +     if (reg_val & DISPC_INT_DONE_MASK) {
> +             evt_stop = true;
> +             wake_up_interruptible_all(&wait_queue);
> +     }
> +
> +     /* dpu ifbc payload error isr */
> +     if (reg_val & DISPC_INT_FBC_PLD_ERR_MASK) {
> +             int_mask |= DISPC_INT_FBC_PLD_ERR_MASK;
> +             pr_err("dpu ifbc payload error\n");
> +     }
> +
> +     /* dpu ifbc header error isr */
> +     if (reg_val & DISPC_INT_FBC_HDR_ERR_MASK) {
> +             int_mask |= DISPC_INT_FBC_HDR_ERR_MASK;
> +             pr_err("dpu ifbc header error\n");
> +     }
> +
> +     int_mask |= check_mmu_isr(ctx, reg_val);
> +
> +     reg->dpu_int_clr = reg_val;
> +     reg->dpu_int_en &= ~int_mask;
> +
> +     return reg_val;
> +}
> +
> +static int dpu_wait_stop_done(struct dpu_context *ctx)
> +{
> +     int rc;
> +
> +     if (ctx->is_stopped)
> +             return 0;
> +
> +     rc = wait_event_interruptible_timeout(wait_queue, evt_stop,
> +                                            msecs_to_jiffies(500));
> +     evt_stop = false;
> +
> +     ctx->is_stopped = true;
> +
> +     if (!rc) {
> +             pr_err("dpu wait for stop done time out!\n");
> +             return -ETIMEDOUT;
> +     }
> +
> +     return 0;
> +}
> +
> +static int dpu_wait_update_done(struct dpu_context *ctx)
> +{
> +     int rc;
> +
> +     evt_update = false;
> +
> +     rc = wait_event_interruptible_timeout(wait_queue, evt_update,
> +                                            msecs_to_jiffies(500));
> +
> +     if (!rc) {
> +             pr_err("dpu wait for reg update done time out!\n");
> +             return -ETIMEDOUT;
> +     }
> +
> +     return 0;
> +}
> +
> +static void dpu_stop(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     if (ctx->if_type == SPRD_DISPC_IF_DPI)
> +             reg->dpu_ctrl |= BIT(1);
> +
> +     dpu_wait_stop_done(ctx);
> +     pr_info("dpu stop\n");
> +}
> +
> +static void dpu_run(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     reg->dpu_ctrl |= BIT(0);
> +
> +     ctx->is_stopped = false;
> +
> +     pr_info("dpu run\n");
> +
> +     if (ctx->if_type == SPRD_DISPC_IF_EDPI) {
> +             /*
> +              * If the panel read GRAM speed faster than
> +              * DSI write GRAM speed, it will display some
> +              * mass on screen when backlight on. So wait
> +              * a TE period after flush the GRAM.
> +              */
> +             if (!panel_ready) {
> +                     dpu_wait_stop_done(ctx);
> +                     /* wait for TE again */
> +                     mdelay(20);
> +                     panel_ready = true;
> +             }
> +     }
> +}
> +
> +static int dpu_init(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     u32 size;
> +
> +     reg->bg_color = 0;
> +
> +     size = (ctx->vm.vactive << 16) | ctx->vm.hactive;
> +     reg->panel_size = size;
> +     reg->blend_size = size;
> +
> +     reg->dpu_cfg0 = BIT(4) | BIT(5);
> +     prev_y2r_coef = 3;
> +
> +     reg->dpu_cfg1 = 0x004466da;
> +     reg->dpu_cfg2 = 0;
> +
> +     if (ctx->is_stopped)
> +             dpu_clean_all(ctx);
> +
> +     reg->mmu_en = 0;
> +     reg->mmu_min_ppn1 = 0;
> +     reg->mmu_ppn_range1 = 0xffff;
> +     reg->mmu_min_ppn2 = 0;
> +     reg->mmu_ppn_range2 = 0xffff;
> +     reg->mmu_vpn_range = 0x1ffff;
> +
> +     reg->dpu_int_clr = 0xffff;
> +
> +     dpu_enhance_reload(ctx);
> +
> +     return 0;
> +}
> +
> +static void dpu_uninit(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     reg->dpu_int_en = 0;
> +     reg->dpu_int_clr = 0xff;
> +
> +     panel_ready = false;
> +}
> +
> +enum {
> +     DPU_LAYER_FORMAT_YUV422_2PLANE,
> +     DPU_LAYER_FORMAT_YUV420_2PLANE,
> +     DPU_LAYER_FORMAT_YUV420_3PLANE,
> +     DPU_LAYER_FORMAT_ARGB8888,
> +     DPU_LAYER_FORMAT_RGB565,
> +     DPU_LAYER_FORMAT_XFBC_ARGB8888 = 8,
> +     DPU_LAYER_FORMAT_XFBC_RGB565,
> +     DPU_LAYER_FORMAT_MAX_TYPES,
> +};
> +
> +enum {
> +     DPU_LAYER_ROTATION_0,
> +     DPU_LAYER_ROTATION_90,
> +     DPU_LAYER_ROTATION_180,
> +     DPU_LAYER_ROTATION_270,
> +     DPU_LAYER_ROTATION_0_M,
> +     DPU_LAYER_ROTATION_90_M,
> +     DPU_LAYER_ROTATION_180_M,
> +     DPU_LAYER_ROTATION_270_M,
> +};
> +
> +static u32 to_dpu_rotation(u32 angle)
> +{
> +     u32 rot = DPU_LAYER_ROTATION_0;
> +
> +     switch (angle) {
> +     case 0:
> +     case DRM_MODE_ROTATE_0:
> +             rot = DPU_LAYER_ROTATION_0;
> +             break;
> +     case DRM_MODE_ROTATE_90:
> +             rot = DPU_LAYER_ROTATION_90;
> +             break;
> +     case DRM_MODE_ROTATE_180:
> +             rot = DPU_LAYER_ROTATION_180;
> +             break;
> +     case DRM_MODE_ROTATE_270:
> +             rot = DPU_LAYER_ROTATION_270;
> +             break;
> +     case DRM_MODE_REFLECT_Y:
> +             rot = DPU_LAYER_ROTATION_180_M;
> +             break;
> +     case (DRM_MODE_REFLECT_Y | DRM_MODE_ROTATE_90):
> +             rot = DPU_LAYER_ROTATION_90_M;
> +             break;
> +     case DRM_MODE_REFLECT_X:
> +             rot = DPU_LAYER_ROTATION_0_M;
> +             break;
> +     case (DRM_MODE_REFLECT_X | DRM_MODE_ROTATE_90):
> +             rot = DPU_LAYER_ROTATION_270_M;
> +             break;
> +     default:
> +             pr_err("rotation convert unsupport angle (drm)= 0x%x\n", angle);
> +             break;
> +     }
> +
> +     return rot;
> +}
> +
> +static u32 dpu_img_ctrl(u32 format, u32 blending, u32 compression, u32 rotation)
> +{
> +     int reg_val = 0;
> +
> +     /* layer enable */
> +     reg_val |= BIT(0);
> +
> +     switch (format) {
> +     case DRM_FORMAT_BGRA8888:
> +             /* BGRA8888 -> ARGB8888 */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B3B2B1B0 << 8;
> +             if (compression)
> +                     /* XFBC-ARGB8888 */
> +                     reg_val |= (DPU_LAYER_FORMAT_XFBC_ARGB8888 << 4);
> +             else
> +                     reg_val |= (DPU_LAYER_FORMAT_ARGB8888 << 4);
> +             break;
> +     case DRM_FORMAT_RGBX8888:
> +     case DRM_FORMAT_RGBA8888:
> +             /* RGBA8888 -> ABGR8888 */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B3B2B1B0 << 8;
> +     case DRM_FORMAT_ABGR8888:
> +             /* rb switch */
> +             reg_val |= BIT(10);
> +     case DRM_FORMAT_ARGB8888:
> +             if (compression)
> +                     /* XFBC-ARGB8888 */
> +                     reg_val |= (DPU_LAYER_FORMAT_XFBC_ARGB8888 << 4);
> +             else
> +                     reg_val |= (DPU_LAYER_FORMAT_ARGB8888 << 4);
> +             break;
> +     case DRM_FORMAT_XBGR8888:
> +             /* rb switch */
> +             reg_val |= BIT(10);
> +     case DRM_FORMAT_XRGB8888:
> +             if (compression)
> +                     /* XFBC-ARGB8888 */
> +                     reg_val |= (DPU_LAYER_FORMAT_XFBC_ARGB8888 << 4);
> +             else
> +                     reg_val |= (DPU_LAYER_FORMAT_ARGB8888 << 4);
> +             break;
> +     case DRM_FORMAT_BGR565:
> +             /* rb switch */
> +             reg_val |= BIT(10);
> +     case DRM_FORMAT_RGB565:
> +             if (compression)
> +                     /* XFBC-RGB565 */
> +                     reg_val |= (DPU_LAYER_FORMAT_XFBC_RGB565 << 4);
> +             else
> +                     reg_val |= (DPU_LAYER_FORMAT_RGB565 << 4);
> +             break;
> +     case DRM_FORMAT_NV12:
> +             /* 2-Lane: Yuv420 */
> +             reg_val |= DPU_LAYER_FORMAT_YUV420_2PLANE << 4;
> +             /* Y endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 8;
> +             /* UV endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 10;
> +             break;
> +     case DRM_FORMAT_NV21:
> +             /* 2-Lane: Yuv420 */
> +             reg_val |= DPU_LAYER_FORMAT_YUV420_2PLANE << 4;
> +             /* Y endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 8;
> +             /* UV endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B3B2B1B0 << 10;
> +             break;
> +     case DRM_FORMAT_NV16:
> +             /* 2-Lane: Yuv422 */
> +             reg_val |= DPU_LAYER_FORMAT_YUV422_2PLANE << 4;
> +             /* Y endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B3B2B1B0 << 8;
> +             /* UV endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B3B2B1B0 << 10;
> +             break;
> +     case DRM_FORMAT_NV61:
> +             /* 2-Lane: Yuv422 */
> +             reg_val |= DPU_LAYER_FORMAT_YUV422_2PLANE << 4;
> +             /* Y endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 8;
> +             /* UV endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 10;
> +             break;
> +     case DRM_FORMAT_YUV420:
> +             reg_val |= DPU_LAYER_FORMAT_YUV420_3PLANE << 4;
> +             /* Y endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 8;
> +             /* UV endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 10;
> +             break;
> +     case DRM_FORMAT_YVU420:
> +             reg_val |= DPU_LAYER_FORMAT_YUV420_3PLANE << 4;
> +             /* Y endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B0B1B2B3 << 8;
> +             /* UV endian */
> +             reg_val |= SPRD_IMG_DATA_ENDIAN_B3B2B1B0 << 10;
> +             break;
> +     default:
> +             pr_err("error: invalid format %c%c%c%c\n", format,
> +                                             format >> 8,
> +                                             format >> 16,
> +                                             format >> 24);
> +             break;
> +     }
> +
> +     switch (blending) {
> +     case DRM_MODE_BLEND_PIXEL_NONE:
> +             /* don't do blending, maybe RGBX */
> +             /* alpha mode select - layer alpha */
> +             reg_val |= BIT(2);
> +             break;
> +     case DRM_MODE_BLEND_COVERAGE:
> +             /* alpha mode select - combo alpha */
> +             reg_val |= BIT(3);
> +             /*Normal mode*/
> +             reg_val &= (~BIT(16));
> +             break;
> +     case DRM_MODE_BLEND_PREMULTI:
> +             /* alpha mode select - combo alpha */
> +             reg_val |= BIT(3);
> +             /*Pre-mult mode*/
> +             reg_val |= BIT(16);
> +             break;
> +     default:
> +             /* alpha mode select - layer alpha */
> +             reg_val |= BIT(2);
> +             break;
> +     }
> +
> +     rotation = to_dpu_rotation(rotation);
> +     reg_val |= (rotation & 0x7) << 20;
> +
> +     return reg_val;
> +}
> +
> +static int check_layer_y2r_coef(struct sprd_dpu_layer layers[], u8 count)
> +{
> +     int i;
> +
> +     for (i = (count - 1); i >= 0; i--) {
> +             switch (layers[i].format) {
> +             case DRM_FORMAT_NV12:
> +             case DRM_FORMAT_NV21:
> +             case DRM_FORMAT_NV16:
> +             case DRM_FORMAT_NV61:
> +             case DRM_FORMAT_YUV420:
> +             case DRM_FORMAT_YVU420:
> +                     if (layers[i].y2r_coef == prev_y2r_coef)
> +                             return -EINVAL;
> +
> +                     /* need to config dpu y2r coef */
> +                     prev_y2r_coef = layers[i].y2r_coef;
> +                     return prev_y2r_coef;
> +             default:
> +                     break;
> +             }
> +     }
> +
> +     /* not find yuv layer */
> +     return -EINVAL;
> +}
> +
> +static void dpu_clean_all(struct dpu_context *ctx)
> +{
> +     int i;
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     for (i = 0; i < 8; i++)
> +             reg->layers[i].ctrl = 0;
> +}
> +
> +static void dpu_bgcolor(struct dpu_context *ctx, u32 color)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     if (ctx->if_type == SPRD_DISPC_IF_EDPI)
> +             dpu_wait_stop_done(ctx);
> +
> +     reg->bg_color = color;
> +
> +     dpu_clean_all(ctx);
> +
> +     if ((ctx->if_type == SPRD_DISPC_IF_DPI) && !ctx->is_stopped) {
> +             reg->dpu_ctrl |= BIT(2);
> +             dpu_wait_update_done(ctx);
> +     } else if (ctx->if_type == SPRD_DISPC_IF_EDPI) {
> +             reg->dpu_ctrl |= BIT(0);
> +             ctx->is_stopped = false;
> +     }
> +}
> +
> +static void dpu_layer(struct dpu_context *ctx,
> +                 struct sprd_dpu_layer *hwlayer)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     const struct drm_format_info *info;
> +     struct layer_reg *layer;
> +     u32 addr, size, offset;
> +     int i;
> +
> +     layer = &reg->layers[hwlayer->index];
> +     offset = (hwlayer->dst_x & 0xffff) | ((hwlayer->dst_y) << 16);
> +
> +     if (hwlayer->pallete_en) {
> +             size = (hwlayer->dst_w & 0xffff) | ((hwlayer->dst_h) << 16);
> +             layer->pos = offset;
> +             layer->size = size;
> +             layer->alpha = hwlayer->alpha;
> +             layer->pallete = hwlayer->pallete_color;
> +
> +             /* pallete layer enable */
> +             layer->ctrl = 0x1005;
> +
> +             pr_debug("dst_x = %d, dst_y = %d, dst_w = %d, dst_h = %d\n",
> +                     hwlayer->dst_x, hwlayer->dst_y,
> +                     hwlayer->dst_w, hwlayer->dst_h);
> +             return;
> +     }
> +
> +     if (hwlayer->src_w && hwlayer->src_h)
> +             size = (hwlayer->src_w & 0xffff) | ((hwlayer->src_h) << 16);
> +     else
> +             size = (hwlayer->dst_w & 0xffff) | ((hwlayer->dst_h) << 16);
> +
> +     for (i = 0; i < hwlayer->planes; i++) {
> +             addr = hwlayer->addr[i];
> +
> +             /* dpu r2p0 just support xfbc-rgb */
> +             if (hwlayer->xfbc)
> +                     addr += hwlayer->header_size_r;
> +
> +             if (addr % 16)
> +                     pr_err("layer addr[%d] is not 16 bytes align, it's 0x%08x\n",
> +                             i, addr);
> +             layer->addr[i] = addr;
> +     }
> +
> +     layer->pos = offset;
> +     layer->size = size;
> +     layer->crop_start = (hwlayer->src_y << 16) | hwlayer->src_x;
> +     layer->alpha = hwlayer->alpha;
> +
> +     info = drm_format_info(hwlayer->format);
> +     if (info->cpp[0] == 0) {
> +             pr_err("layer[%d] bytes per pixel is invalid\n", hwlayer->index);
> +             return;
> +     }
> +
> +     if (hwlayer->planes == 3)
> +             /* UV pitch is 1/2 of Y pitch*/
> +             layer->pitch = (hwlayer->pitch[0] / info->cpp[0]) |
> +                             (hwlayer->pitch[0] / info->cpp[0] << 15);
> +     else
> +             layer->pitch = hwlayer->pitch[0] / info->cpp[0];
> +
> +     layer->ctrl = dpu_img_ctrl(hwlayer->format, hwlayer->blending,
> +             hwlayer->xfbc, hwlayer->rotation);
> +
> +     pr_debug("dst_x = %d, dst_y = %d, dst_w = %d, dst_h = %d\n",
> +                             hwlayer->dst_x, hwlayer->dst_y,
> +                             hwlayer->dst_w, hwlayer->dst_h);
> +     pr_debug("start_x = %d, start_y = %d, start_w = %d, start_h = %d\n",
> +                             hwlayer->src_x, hwlayer->src_y,
> +                             hwlayer->src_w, hwlayer->src_h);
> +}
> +
> +static void dpu_scaling(struct dpu_context *ctx,
> +                     struct sprd_dpu_layer layers[], u8 count)
> +{
> +     int i;
> +     struct sprd_dpu_layer *top_layer;
> +
> +     if (mode_changed) {
> +             top_layer = &layers[count - 1];
> +             pr_debug("------------------------------------\n");
> +             for (i = 0; i < count; i++) {
> +                     pr_debug("layer[%d] : %dx%d --- (%d)\n", i,
> +                             layers[i].dst_w, layers[i].dst_h,
> +                             scale_copy.in_w);
> +             }
> +
> +             if  (top_layer->dst_w <= scale_copy.in_w) {
> +                     dpu_sr_config(ctx);
> +                     mode_changed = false;
> +
> +                     pr_info("do scaling enhace: 0x%x, top layer(%dx%d)\n",
> +                             enhance_en, top_layer->dst_w,
> +                             top_layer->dst_h);
> +             }
> +     }
> +}
> +
> +static void dpu_flip(struct dpu_context *ctx,
> +                  struct sprd_dpu_layer layers[], u8 count)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     int i;
> +     int y2r_coef;
> +
> +     /*
> +      * Make sure the dpu is in stop status. DPU_R2P0 has no shadow
> +      * registers in EDPI mode. So the config registers can only be
> +      * updated in the rising edge of DPU_RUN bit.
> +      */
> +     if (ctx->if_type == SPRD_DISPC_IF_EDPI)
> +             dpu_wait_stop_done(ctx);
> +
> +     /* set Y2R conversion coef */
> +     y2r_coef = check_layer_y2r_coef(layers, count);
> +     if (y2r_coef >= 0) {
> +             /* write dpu_cfg0 register after dpu is in idle status */
> +             if (ctx->if_type == SPRD_DISPC_IF_DPI)
> +                     dpu_stop(ctx);
> +
> +             reg->dpu_cfg0 &= ~(0x7 << 4);
> +             reg->dpu_cfg0 |= (y2r_coef << 4);
> +     }
> +
> +     /* reset the bgcolor to black */
> +     reg->bg_color = 0;
> +
> +     /* disable all the layers */
> +     dpu_clean_all(ctx);
> +
> +     /* to check if dpu need scaling the frame for SR */
> +     dpu_scaling(ctx, layers, count);
> +
> +     /* start configure dpu layers */
> +     for (i = 0; i < count; i++)
> +             dpu_layer(ctx, &layers[i]);
> +
> +     /* update trigger and wait */
> +     if (ctx->if_type == SPRD_DISPC_IF_DPI) {
> +             if (!ctx->is_stopped) {
> +                     reg->dpu_ctrl |= BIT(2);
> +                     dpu_wait_update_done(ctx);
> +             } else if (y2r_coef >= 0) {
> +                     reg->dpu_ctrl |= BIT(0);
> +                     ctx->is_stopped = false;
> +                     pr_info("dpu start\n");
> +             }
> +
> +             reg->dpu_int_en |= DISPC_INT_ERR_MASK;
> +
> +     } else if (ctx->if_type == SPRD_DISPC_IF_EDPI) {
> +             reg->dpu_ctrl |= BIT(0);
> +
> +             ctx->is_stopped = false;
> +     }
> +
> +     /*
> +      * If the following interrupt was disabled in isr,
> +      * re-enable it.
> +      */
> +     reg->dpu_int_en |= DISPC_INT_FBC_PLD_ERR_MASK |
> +                        DISPC_INT_FBC_HDR_ERR_MASK |
> +                        DISPC_INT_MMU_VAOR_RD_MASK |
> +                        DISPC_INT_MMU_VAOR_WR_MASK |
> +                        DISPC_INT_MMU_INV_RD_MASK |
> +                        DISPC_INT_MMU_INV_WR_MASK;
> +}
> +
> +static void dpu_dpi_init(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     u32 int_mask = 0;
> +
> +     if (ctx->if_type == SPRD_DISPC_IF_DPI) {
> +             /* use dpi as interface */
> +             reg->dpu_cfg0 &= ~BIT(0);
> +
> +             /* disable Halt function for SPRD DSI */
> +             reg->dpi_ctrl &= ~BIT(16);
> +
> +             /* select te from external pad */
> +             reg->dpi_ctrl |= BIT(10);
> +
> +             /* set dpi timing */
> +             reg->dpi_h_timing = (ctx->vm.hsync_len << 0) |
> +                                 (ctx->vm.hback_porch << 8) |
> +                                 (ctx->vm.hfront_porch << 20);
> +             reg->dpi_v_timing = (ctx->vm.vsync_len << 0) |
> +                                 (ctx->vm.vback_porch << 8) |
> +                                 (ctx->vm.vfront_porch << 20);
> +             if (ctx->vm.vsync_len + ctx->vm.vback_porch < 32)
> +                     pr_warn("Warning: (vsync + vbp) < 32, "
> +                             "underflow risk!\n");
> +
> +             /* enable dpu update done INT */
> +             int_mask |= DISPC_INT_UPDATE_DONE_MASK;
> +             /* enable dpu DONE  INT */
> +             int_mask |= DISPC_INT_DONE_MASK;
> +             /* enable dpu dpi vsync */
> +             int_mask |= DISPC_INT_DPI_VSYNC_MASK;
> +             /* enable dpu TE INT */
> +             int_mask |= DISPC_INT_TE_MASK;
> +             /* enable underflow err INT */
> +             int_mask |= DISPC_INT_ERR_MASK;
> +             /* enable write back done INT */
> +             int_mask |= DISPC_INT_WB_DONE_MASK;
> +             /* enable write back fail INT */
> +             int_mask |= DISPC_INT_WB_FAIL_MASK;
> +
> +     } else if (ctx->if_type == SPRD_DISPC_IF_EDPI) {
> +             /* use edpi as interface */
> +             reg->dpu_cfg0 |= BIT(0);
> +
> +             /* use external te */
> +             reg->dpi_ctrl |= BIT(10);
> +
> +             /* enable te */
> +             reg->dpi_ctrl |= BIT(8);
> +
> +             /* enable stop DONE INT */
> +             int_mask |= DISPC_INT_DONE_MASK;
> +             /* enable TE INT */
> +             int_mask |= DISPC_INT_TE_MASK;
> +     }
> +
> +     /* enable ifbc payload error INT */
> +     int_mask |= DISPC_INT_FBC_PLD_ERR_MASK;
> +     /* enable ifbc header error INT */
> +     int_mask |= DISPC_INT_FBC_HDR_ERR_MASK;
> +     /* enable iommu va out of range read error INT */
> +     int_mask |= DISPC_INT_MMU_VAOR_RD_MASK;
> +     /* enable iommu va out of range write error INT */
> +     int_mask |= DISPC_INT_MMU_VAOR_WR_MASK;
> +     /* enable iommu invalid read error INT */
> +     int_mask |= DISPC_INT_MMU_INV_RD_MASK;
> +     /* enable iommu invalid write error INT */
> +     int_mask |= DISPC_INT_MMU_INV_WR_MASK;
> +
> +     reg->dpu_int_en = int_mask;
> +}
> +
> +static void enable_vsync(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     reg->dpu_int_en |= DISPC_INT_DPI_VSYNC_MASK;
> +}
> +
> +static void disable_vsync(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     reg->dpu_int_en &= ~DISPC_INT_DPI_VSYNC_MASK;
> +}
> +
> +static void dpu_enhance_backup(u32 id, void *param)
> +{
> +     u32 *p;
> +
> +     switch (id) {
> +     case ENHANCE_CFG_ID_ENABLE:
> +             p = param;
> +             enhance_en |= *p;
> +             pr_info("enhance enable backup: 0x%x\n", *p);
> +             break;
> +     case ENHANCE_CFG_ID_DISABLE:
> +             p = param;
> +             enhance_en &= ~(*p);
> +             if (*p & BIT(1))
> +                     memset(&epf_copy, 0, sizeof(epf_copy));
> +             pr_info("enhance disable backup: 0x%x\n", *p);
> +             break;
> +     case ENHANCE_CFG_ID_SCL:
> +             memcpy(&scale_copy, param, sizeof(scale_copy));
> +             enhance_en |= BIT(0);
> +             pr_info("enhance scaling backup\n");
> +             break;
> +     case ENHANCE_CFG_ID_HSV:
> +             memcpy(&hsv_copy, param, sizeof(hsv_copy));
> +             enhance_en |= BIT(2);
> +             pr_info("enhance hsv backup\n");
> +             break;
> +     case ENHANCE_CFG_ID_CM:
> +             memcpy(&cm_copy, param, sizeof(cm_copy));
> +             enhance_en |= BIT(3);
> +             pr_info("enhance cm backup\n");
> +             break;
> +     case ENHANCE_CFG_ID_SLP:
> +             memcpy(&slp_copy, param, sizeof(slp_copy));
> +             enhance_en |= BIT(4);
> +             pr_info("enhance slp backup\n");
> +             break;
> +     case ENHANCE_CFG_ID_GAMMA:
> +             memcpy(&gamma_copy, param, sizeof(gamma_copy));
> +             enhance_en |= BIT(5);
> +             pr_info("enhance gamma backup\n");
> +             break;
> +     case ENHANCE_CFG_ID_EPF:
> +             memcpy(&epf_copy, param, sizeof(epf_copy));
> +             if ((slp_copy.brightness > SLP_BRIGHTNESS_THRESHOLD) ||
> +                     (enhance_en & BIT(0))) {
> +                     enhance_en |= BIT(1);
> +                     pr_info("enhance epf backup\n");
> +             }
> +             break;
> +     default:
> +             break;
> +     }
> +}
> +
> +static void dpu_epf_set(struct dpu_reg *reg, struct epf_cfg *epf)
> +{
> +     reg->epf_epsilon = (epf->epsilon1 << 16) | epf->epsilon0;
> +     reg->epf_gain0_3 = (epf->gain3 << 24) | (epf->gain2 << 16) |
> +                        (epf->gain1 << 8) | epf->gain0;
> +     reg->epf_gain4_7 = (epf->gain7 << 24) | (epf->gain6 << 16) |
> +                        (epf->gain5 << 8) | epf->gain4;
> +     reg->epf_diff = (epf->max_diff << 8) | epf->min_diff;
> +}
> +
> +static void dpu_enhance_set(struct dpu_context *ctx, u32 id, void *param)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     struct scale_cfg *scale;
> +     struct cm_cfg *cm;
> +     struct slp_cfg *slp;
> +     struct gamma_lut *gamma;
> +     struct hsv_lut *hsv;
> +     struct epf_cfg *epf;
> +     u32 *p, i;
> +
> +     if (!ctx->is_inited) {
> +             dpu_enhance_backup(id, param);
> +             return;
> +     }
> +
> +     if (ctx->if_type == SPRD_DISPC_IF_EDPI)
> +             dpu_wait_stop_done(ctx);
> +
> +     switch (id) {
> +     case ENHANCE_CFG_ID_ENABLE:
> +             p = param;
> +             reg->dpu_enhance_cfg |= *p;
> +             pr_info("enhance module enable: 0x%x\n", *p);
> +             break;
> +     case ENHANCE_CFG_ID_DISABLE:
> +             p = param;
> +             reg->dpu_enhance_cfg &= ~(*p);
> +             if (*p & BIT(1))
> +                     memset(&epf_copy, 0, sizeof(epf_copy));
> +             pr_info("enhance module disable: 0x%x\n", *p);
> +             break;
> +     case ENHANCE_CFG_ID_SCL:
> +             memcpy(&scale_copy, param, sizeof(scale_copy));
> +             scale = &scale_copy;
> +             reg->blend_size = (scale->in_h << 16) | scale->in_w;
> +             reg->dpu_enhance_cfg |= BIT(0);
> +             pr_info("enhance scaling: %ux%u\n", scale->in_w, scale->in_h);
> +             break;
> +     case ENHANCE_CFG_ID_HSV:
> +             memcpy(&hsv_copy, param, sizeof(hsv_copy));
> +             hsv = &hsv_copy;
> +             for (i = 0; i < 360; i++) {
> +                     reg->hsv_lut_addr = i;
> +                     udelay(1);
> +                     reg->hsv_lut_wdata = (hsv->table[i].sat << 16) |
> +                                             hsv->table[i].hue;
> +             }
> +             reg->dpu_enhance_cfg |= BIT(2);
> +             pr_info("enhance hsv set\n");
> +             break;
> +     case ENHANCE_CFG_ID_CM:
> +             memcpy(&cm_copy, param, sizeof(cm_copy));
> +             cm = &cm_copy;
> +             reg->cm_coef01_00 = (cm->coef01 << 16) | cm->coef00;
> +             reg->cm_coef03_02 = (cm->coef03 << 16) | cm->coef02;
> +             reg->cm_coef11_10 = (cm->coef11 << 16) | cm->coef10;
> +             reg->cm_coef13_12 = (cm->coef13 << 16) | cm->coef12;
> +             reg->cm_coef21_20 = (cm->coef21 << 16) | cm->coef20;
> +             reg->cm_coef23_22 = (cm->coef23 << 16) | cm->coef22;
> +             reg->dpu_enhance_cfg |= BIT(3);
> +             pr_info("enhance cm set\n");
> +             break;
> +     case ENHANCE_CFG_ID_SLP:
> +             memcpy(&slp_copy, param, sizeof(slp_copy));
> +             slp = &slp_copy;
> +             reg->slp_cfg0 = (slp->second_bright_factor << 24) |
> +                             (slp->brightness_step << 16) |
> +                             (slp->conversion_matrix << 8) |
> +                             slp->brightness;
> +             reg->slp_cfg1 = (slp->first_max_bright_th << 8) |
> +                             slp->first_percent_th;
> +             reg->dpu_enhance_cfg |= BIT(4);
> +             pr_info("enhance slp set\n");
> +             break;
> +     case ENHANCE_CFG_ID_GAMMA:
> +             memcpy(&gamma_copy, param, sizeof(gamma_copy));
> +             gamma = &gamma_copy;
> +             for (i = 0; i < 256; i++) {
> +                     reg->gamma_lut_addr = i;
> +                     udelay(1);
> +                     reg->gamma_lut_wdata = (gamma->r[i] << 20) |
> +                                             (gamma->g[i] << 10) |
> +                                             gamma->b[i];
> +                     pr_debug("0x%02x: r=%u, g=%u, b=%u\n", i,
> +                             gamma->r[i], gamma->g[i], gamma->b[i]);
> +             }
> +             reg->dpu_enhance_cfg |= BIT(5);
> +             pr_info("enhance gamma set\n");
> +             break;
> +     case ENHANCE_CFG_ID_EPF:
> +             memcpy(&epf_copy, param, sizeof(epf_copy));
> +             if ((slp_copy.brightness > SLP_BRIGHTNESS_THRESHOLD) ||
> +                     (enhance_en & BIT(0))) {
> +                     epf = &epf_copy;
> +                     dpu_epf_set(reg, epf);
> +                     reg->dpu_enhance_cfg |= BIT(1);
> +                     pr_info("enhance epf set\n");
> +                     break;
> +             }
> +             return;
> +     default:
> +             break;
> +     }
> +
> +     if ((ctx->if_type == SPRD_DISPC_IF_DPI) && !ctx->is_stopped) {
> +             reg->dpu_ctrl |= BIT(2);
> +             dpu_wait_update_done(ctx);
> +     } else if ((ctx->if_type == SPRD_DISPC_IF_EDPI) && panel_ready) {
> +             /*
> +              * In EDPI mode, we need to wait panel initializatin
> +              * completed. Otherwise, the dpu enhance settings may
> +              * start before panel initialization.
> +              */
> +             reg->dpu_ctrl |= BIT(0);
> +             ctx->is_stopped = false;
> +     }
> +
> +     enhance_en = reg->dpu_enhance_cfg;
> +}
> +
> +static void dpu_enhance_get(struct dpu_context *ctx, u32 id, void *param)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     struct scale_cfg *scale;
> +     struct epf_cfg *ep;
> +     struct slp_cfg *slp;
> +     struct gamma_lut *gamma;
> +     u32 *p32;
> +     int i, val;
> +
> +     switch (id) {
> +     case ENHANCE_CFG_ID_ENABLE:
> +             p32 = param;
> +             *p32 = reg->dpu_enhance_cfg;
> +             pr_info("enhance module enable get\n");
> +             break;
> +     case ENHANCE_CFG_ID_SCL:
> +             scale = param;
> +             val = reg->blend_size;
> +             scale->in_w = val & 0xffff;
> +             scale->in_h = val >> 16;
> +             pr_info("enhance scaling get\n");
> +             break;
> +     case ENHANCE_CFG_ID_EPF:
> +             ep = param;
> +
> +             val = reg->epf_epsilon;
> +             ep->epsilon0 = val;
> +             ep->epsilon1 = val >> 16;
> +
> +             val = reg->epf_gain0_3;
> +             ep->gain0 = val;
> +             ep->gain1 = val >> 8;
> +             ep->gain2 = val >> 16;
> +             ep->gain3 = val >> 24;
> +
> +             val = reg->epf_gain4_7;
> +             ep->gain4 = val;
> +             ep->gain5 = val >> 8;
> +             ep->gain6 = val >> 16;
> +             ep->gain7 = val >> 24;
> +
> +             val = reg->epf_diff;
> +             ep->min_diff = val;
> +             ep->max_diff = val >> 8;
> +             pr_info("enhance epf get\n");
> +             break;
> +     case ENHANCE_CFG_ID_HSV:
> +             dpu_stop(ctx);
> +             p32 = param;
> +             for (i = 0; i < 360; i++) {
> +                     reg->hsv_lut_addr = i;
> +                     udelay(1);
> +                     *p32++ = reg->hsv_lut_rdata;
> +             }
> +             dpu_run(ctx);
> +             pr_info("enhance hsv get\n");
> +             break;
> +     case ENHANCE_CFG_ID_CM:
> +             p32 = param;
> +             *p32++ = reg->cm_coef01_00;
> +             *p32++ = reg->cm_coef03_02;
> +             *p32++ = reg->cm_coef11_10;
> +             *p32++ = reg->cm_coef13_12;
> +             *p32++ = reg->cm_coef21_20;
> +             *p32++ = reg->cm_coef23_22;
> +             pr_info("enhance cm get\n");
> +             break;
> +     case ENHANCE_CFG_ID_SLP:
> +             slp = param;
> +
> +             val = reg->slp_cfg0;
> +             slp->brightness = val;
> +             slp->conversion_matrix = val >> 8;
> +             slp->brightness_step = val >> 16;
> +             slp->second_bright_factor = val >> 24;
> +
> +             val = reg->slp_cfg1;
> +             slp->first_percent_th = val;
> +             slp->first_max_bright_th = val >> 8;
> +             pr_info("enhance slp get\n");
> +             break;
> +     case ENHANCE_CFG_ID_GAMMA:
> +             dpu_stop(ctx);
> +             gamma = param;
> +             for (i = 0; i < 256; i++) {
> +                     reg->gamma_lut_addr = i;
> +                     udelay(1);
> +                     val = reg->gamma_lut_rdata;
> +                     gamma->r[i] = (val >> 20) & 0x3FF;
> +                     gamma->g[i] = (val >> 10) & 0x3FF;
> +                     gamma->b[i] = val & 0x3FF;
> +                     pr_debug("0x%02x: r=%u, g=%u, b=%u\n", i,
> +                             gamma->r[i], gamma->g[i], gamma->b[i]);
> +             }
> +             dpu_run(ctx);
> +             pr_info("enhance gamma get\n");
> +             break;
> +     default:
> +             break;
> +     }
> +}
> +
> +static void dpu_enhance_reload(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +     struct scale_cfg *scale;
> +     struct cm_cfg *cm;
> +     struct slp_cfg *slp;
> +     struct gamma_lut *gamma;
> +     struct hsv_lut *hsv;
> +     struct epf_cfg *epf;
> +     int i;
> +
> +     if (enhance_en & BIT(0)) {
> +             scale = &scale_copy;
> +             reg->blend_size = (scale->in_h << 16) | scale->in_w;
> +             pr_info("enhance scaling from %ux%u to %ux%u\n", scale->in_w,
> +                     scale->in_h, ctx->vm.hactive, ctx->vm.vactive);
> +     }
> +
> +     if (enhance_en & BIT(1)) {
> +             epf = &epf_copy;
> +             dpu_epf_set(reg, epf);
> +             pr_info("enhance epf reload\n");
> +     }
> +
> +     if (enhance_en & BIT(2)) {
> +             hsv = &hsv_copy;
> +             for (i = 0; i < 360; i++) {
> +                     reg->hsv_lut_addr = i;
> +                     udelay(1);
> +                     reg->hsv_lut_wdata = (hsv->table[i].sat << 16) |
> +                                             hsv->table[i].hue;
> +             }
> +             pr_info("enhance hsv reload\n");
> +     }
> +
> +     if (enhance_en & BIT(3)) {
> +             cm = &cm_copy;
> +             reg->cm_coef01_00 = (cm->coef01 << 16) | cm->coef00;
> +             reg->cm_coef03_02 = (cm->coef03 << 16) | cm->coef02;
> +             reg->cm_coef11_10 = (cm->coef11 << 16) | cm->coef10;
> +             reg->cm_coef13_12 = (cm->coef13 << 16) | cm->coef12;
> +             reg->cm_coef21_20 = (cm->coef21 << 16) | cm->coef20;
> +             reg->cm_coef23_22 = (cm->coef23 << 16) | cm->coef22;
> +             pr_info("enhance cm reload\n");
> +     }
> +
> +     if (enhance_en & BIT(4)) {
> +             slp = &slp_copy;
> +             reg->slp_cfg0 = (slp->second_bright_factor << 24) |
> +                             (slp->brightness_step << 16) |
> +                             (slp->conversion_matrix << 8) |
> +                             slp->brightness;
> +             reg->slp_cfg1 = (slp->first_max_bright_th << 8) |
> +                             slp->first_percent_th;
> +             pr_info("enhance slp reload\n");
> +     }
> +
> +     if (enhance_en & BIT(5)) {
> +             gamma = &gamma_copy;
> +             for (i = 0; i < 256; i++) {
> +                     reg->gamma_lut_addr = i;
> +                     udelay(1);
> +                     reg->gamma_lut_wdata = (gamma->r[i] << 20) |
> +                                             (gamma->g[i] << 10) |
> +                                             gamma->b[i];
> +                     pr_debug("0x%02x: r=%u, g=%u, b=%u\n", i,
> +                             gamma->r[i], gamma->g[i], gamma->b[i]);
> +             }
> +             pr_info("enhance gamma reload\n");
> +     }
> +
> +     reg->dpu_enhance_cfg = enhance_en;
> +}
> +
> +static void dpu_sr_config(struct dpu_context *ctx)
> +{
> +     struct dpu_reg *reg = (struct dpu_reg *)ctx->base;
> +
> +     reg->blend_size = (scale_copy.in_h << 16) | scale_copy.in_w;
> +     if (need_scale) {
> +             /* SLP is disabled mode or bypass mode */
> +             if (slp_copy.brightness <= SLP_BRIGHTNESS_THRESHOLD) {
> +
> +             /*
> +              * valid range of gain3 is [128,255];dpu_scaling maybe
> +              * called before epf_copy is assinged a value
> +              */
> +                     if (epf_copy.gain3 > 0) {
> +                             dpu_epf_set(reg, &epf_copy);
> +                             enhance_en |= BIT(1);
> +                     }
> +             }
> +             enhance_en |= BIT(0);
> +             reg->dpu_enhance_cfg = enhance_en;
> +     } else {
> +             if (slp_copy.brightness <= SLP_BRIGHTNESS_THRESHOLD)
> +                     enhance_en &= ~(BIT(1));
> +
> +             enhance_en &= ~(BIT(0));
> +             reg->dpu_enhance_cfg = enhance_en;
> +     }
> +}
> +
> +static int dpu_modeset(struct dpu_context *ctx,
> +             struct drm_mode_modeinfo *mode)
> +{
> +     scale_copy.in_w = mode->hdisplay;
> +     scale_copy.in_h = mode->vdisplay;
> +
> +     if ((mode->hdisplay != ctx->vm.hactive) ||
> +             (mode->vdisplay != ctx->vm.vactive))
> +             need_scale = true;
> +     else
> +             need_scale = false;
> +
> +     mode_changed = true;
> +     pr_info("begin switch to %u x %u\n", mode->hdisplay, mode->vdisplay);
> +
> +     return 0;
> +}
> +
> +static const u32 primary_fmts[] = {
> +     DRM_FORMAT_XRGB8888, DRM_FORMAT_XBGR8888,
> +     DRM_FORMAT_ARGB8888, DRM_FORMAT_ABGR8888,
> +     DRM_FORMAT_RGBA8888, DRM_FORMAT_BGRA8888,
> +     DRM_FORMAT_RGBX8888, DRM_FORMAT_BGRX8888,
> +     DRM_FORMAT_RGB565, DRM_FORMAT_BGR565,
> +     DRM_FORMAT_NV12, DRM_FORMAT_NV21,
> +     DRM_FORMAT_NV16, DRM_FORMAT_NV61,
> +     DRM_FORMAT_YUV420, DRM_FORMAT_YVU420,
> +};
> +
> +static int dpu_capability(struct dpu_context *ctx,
> +                     struct dpu_capability *cap)
> +{
> +     if (!cap)
> +             return -EINVAL;
> +
> +     cap->max_layers = 6;
> +     cap->fmts_ptr = primary_fmts;
> +     cap->fmts_cnt = ARRAY_SIZE(primary_fmts);
> +
> +     return 0;
> +}
> +
> +static struct dpu_core_ops dpu_r2p0_ops = {
> +     .parse_dt = dpu_parse_dt,
> +     .version = dpu_get_version,
> +     .init = dpu_init,
> +     .uninit = dpu_uninit,
> +     .run = dpu_run,
> +     .stop = dpu_stop,
> +     .isr = dpu_isr,
> +     .ifconfig = dpu_dpi_init,
> +     .capability = dpu_capability,
> +     .flip = dpu_flip,
> +     .bg_color = dpu_bgcolor,
> +     .enable_vsync = enable_vsync,
> +     .disable_vsync = disable_vsync,
> +     .enhance_set = dpu_enhance_set,
> +     .enhance_get = dpu_enhance_get,
> +     .modeset = dpu_modeset,
> +     .check_raw_int = dpu_check_raw_int,
> +};
> +
> +static struct ops_entry entry = {
> +     .ver = "dpu-r2p0",
> +     .ops = &dpu_r2p0_ops,
> +};
> +
> +static int __init dpu_core_register(void)
> +{
> +     return dpu_core_ops_register(&entry);
> +}
> +
> +subsys_initcall(dpu_core_register);
> diff --git a/drivers/gpu/drm/sprd/sprd_dpu.c b/drivers/gpu/drm/sprd/sprd_dpu.c
> new file mode 100644
> index 0000000..43142b3
> --- /dev/null
> +++ b/drivers/gpu/drm/sprd/sprd_dpu.c
> @@ -0,0 +1,1152 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (C) 2019 Unisoc Inc.
> + */
> +
> +#include <linux/component.h>
> +#include <linux/dma-buf.h>
> +#include <linux/module.h>
> +#include <linux/of.h>
> +#include <linux/of_address.h>
> +#include <linux/of_irq.h>
> +#include <linux/pm_runtime.h>
> +
> +#include <drm/drm_atomic_helper.h>
> +#include <drm/drm_crtc_helper.h>
> +#include <drm/drm_gem_framebuffer_helper.h>
> +#include <drm/drm_plane_helper.h>
> +
> +#include "sprd_drm.h"
> +#include "sprd_dpu.h"
> +#include "sprd_gem.h"
> +
> +struct sprd_plane {
> +     struct drm_plane plane;
> +     struct drm_property *alpha_property;
> +     struct drm_property *blend_mode_property;
> +     struct drm_property *fbc_hsize_r_property;
> +     struct drm_property *fbc_hsize_y_property;
> +     struct drm_property *fbc_hsize_uv_property;
> +     struct drm_property *y2r_coef_property;
> +     struct drm_property *pallete_en_property;
> +     struct drm_property *pallete_color_property;
> +     u32 index;
> +};
> +
> +struct sprd_plane_state {
> +     struct drm_plane_state state;
> +     u8 alpha;
> +     u8 blend_mode;
> +     u32 fbc_hsize_r;
> +     u32 fbc_hsize_y;
> +     u32 fbc_hsize_uv;
> +     u32 y2r_coef;
> +     u32 pallete_en;
> +     u32 pallete_color;
> +};
> +
> +LIST_HEAD(dpu_core_head);
> +LIST_HEAD(dpu_clk_head);
> +LIST_HEAD(dpu_glb_head);
> +
> +static int sprd_dpu_init(struct sprd_dpu *dpu);
> +static int sprd_dpu_uninit(struct sprd_dpu *dpu);
> +
> +static inline struct sprd_plane *to_sprd_plane(struct drm_plane *plane)
> +{
> +     return container_of(plane, struct sprd_plane, plane);
> +}
> +
> +static inline struct
> +sprd_plane_state *to_sprd_plane_state(const struct drm_plane_state *state)
> +{
> +     return container_of(state, struct sprd_plane_state, state);
> +}
> +
> +static int sprd_dpu_iommu_map(struct device *dev,
> +                             struct sprd_gem_obj *sprd_gem)
> +{
> +     /*
> +      * TODO:
> +      * Currently only supports dumb buffer,
> +      * will support iommu in the future.
> +      */
> +
> +     return 0;
> +}
> +
> +static void sprd_dpu_iommu_unmap(struct device *dev,
> +                             struct sprd_gem_obj *sprd_gem)
> +{
> +     /*
> +      * TODO:
> +      * Currently only supports dumb buffer,
> +      * will support iommu in the future.
> +      */
> +}
> +
> +static int of_get_logo_memory_info(struct sprd_dpu *dpu,
> +     struct device_node *np)
> +{
> +     struct device_node *node;
> +     struct resource r;
> +     int ret;
> +     struct dpu_context *ctx = &dpu->ctx;
> +
> +     node = of_parse_phandle(np, "sprd,logo-memory", 0);
> +     if (!node) {
> +             DRM_INFO("no sprd,logo-memory specified\n");
> +             return 0;
> +     }
> +
> +     ret = of_address_to_resource(node, 0, &r);
> +     of_node_put(node);
> +     if (ret) {
> +             DRM_ERROR("invalid logo reserved memory node!\n");
> +             return -EINVAL;
> +     }
> +
> +     ctx->logo_addr = r.start;
> +     ctx->logo_size = resource_size(&r);
> +
> +     return 0;
> +}
> +
> +static int sprd_plane_prepare_fb(struct drm_plane *plane,
> +                             struct drm_plane_state *new_state)
> +{
> +     struct drm_plane_state *curr_state = plane->state;
> +     struct drm_framebuffer *fb;
> +     struct drm_gem_object *obj;
> +     struct sprd_gem_obj *sprd_gem;
> +     struct sprd_dpu *dpu;
> +     int i;
> +
> +     if ((curr_state->fb == new_state->fb) || !new_state->fb)
> +             return 0;
> +
> +     fb = new_state->fb;
> +     dpu = crtc_to_dpu(new_state->crtc);
> +
> +     if (!dpu->ctx.is_inited) {
> +             DRM_WARN("dpu has already powered off\n");
> +             return 0;
> +     }
> +
> +     for (i = 0; i < fb->format->num_planes; i++) {
> +             obj = drm_gem_fb_get_obj(fb, i);
> +             sprd_gem = to_sprd_gem_obj(obj);
> +             if (sprd_gem->sgtb && sprd_gem->sgtb->nents > 1)
> +                     sprd_dpu_iommu_map(&dpu->dev, sprd_gem);
> +     }
> +
> +     return 0;
> +}
> +
> +static void sprd_plane_cleanup_fb(struct drm_plane *plane,
> +                             struct drm_plane_state *old_state)
> +{
> +     struct drm_plane_state *curr_state = plane->state;
> +     struct drm_framebuffer *fb;
> +     struct drm_gem_object *obj;
> +     struct sprd_gem_obj *sprd_gem;
> +     struct sprd_dpu *dpu;
> +     int i;
> +     static atomic_t logo2animation = { -1 };
> +
> +     if ((curr_state->fb == old_state->fb) || !old_state->fb)
> +             return;
> +
> +     fb = old_state->fb;
> +     dpu = crtc_to_dpu(old_state->crtc);
> +
> +     if (!dpu->ctx.is_inited) {
> +             DRM_WARN("dpu has already powered off\n");
> +             return;
> +     }
> +
> +     for (i = 0; i < fb->format->num_planes; i++) {
> +             obj = drm_gem_fb_get_obj(fb, i);
> +             sprd_gem = to_sprd_gem_obj(obj);
> +             if (sprd_gem->sgtb && sprd_gem->sgtb->nents > 1)
> +                     sprd_dpu_iommu_unmap(&dpu->dev, sprd_gem);
> +     }
> +
> +     if (unlikely(atomic_inc_not_zero(&logo2animation)) &&
> +             dpu->ctx.logo_addr) {
> +             DRM_INFO("free logo memory addr:0x%lx size:0x%lx\n",
> +                     dpu->ctx.logo_addr, dpu->ctx.logo_size);
> +             free_reserved_area(phys_to_virt(dpu->ctx.logo_addr),
> +                     phys_to_virt(dpu->ctx.logo_addr + dpu->ctx.logo_size),
> +                     -1, "logo");
> +     }
> +}
> +
> +static int sprd_plane_atomic_check(struct drm_plane *plane,
> +                               struct drm_plane_state *state)
> +{
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     return 0;
> +}
> +
> +static void sprd_plane_atomic_update(struct drm_plane *plane,
> +                                 struct drm_plane_state *old_state)
> +{
> +     struct drm_plane_state *state = plane->state;
> +     struct drm_framebuffer *fb = plane->state->fb;
> +     struct drm_gem_object *obj;
> +     struct sprd_gem_obj *sprd_gem;
> +     struct sprd_plane *p = to_sprd_plane(plane);
> +     struct sprd_plane_state *s = to_sprd_plane_state(state);
> +     struct sprd_dpu *dpu = crtc_to_dpu(plane->state->crtc);
> +     struct sprd_dpu_layer *layer = &dpu->layers[p->index];
> +     int i;
> +
> +     if (plane->state->crtc->state->active_changed) {
> +             DRM_DEBUG("resume or suspend, no need to update plane\n");
> +             return;
> +     }
> +
> +     if (s->pallete_en) {
> +             layer->index = p->index;
> +             layer->dst_x = state->crtc_x;
> +             layer->dst_y = state->crtc_y;
> +             layer->dst_w = state->crtc_w;
> +             layer->dst_h = state->crtc_h;
> +             layer->alpha = s->alpha;
> +             layer->blending = s->blend_mode;
> +             layer->pallete_en = s->pallete_en;
> +             layer->pallete_color = s->pallete_color;
> +             dpu->pending_planes++;
> +             DRM_DEBUG("%s() pallete_color = %u, index = %u\n",
> +                     __func__, layer->pallete_color, layer->index);
> +             return;
> +     }
> +
> +     layer->index = p->index;
> +     layer->src_x = state->src_x >> 16;
> +     layer->src_y = state->src_y >> 16;
> +     layer->src_w = state->src_w >> 16;
> +     layer->src_h = state->src_h >> 16;
> +     layer->dst_x = state->crtc_x;
> +     layer->dst_y = state->crtc_y;
> +     layer->dst_w = state->crtc_w;
> +     layer->dst_h = state->crtc_h;
> +     layer->rotation = state->rotation;
> +     layer->planes = fb->format->num_planes;
> +     layer->format = fb->format->format;
> +     layer->alpha = s->alpha;
> +     layer->blending = s->blend_mode;
> +     layer->xfbc = fb->modifier;
> +     layer->header_size_r = s->fbc_hsize_r;
> +     layer->header_size_y = s->fbc_hsize_y;
> +     layer->header_size_uv = s->fbc_hsize_uv;
> +     layer->y2r_coef = s->y2r_coef;
> +     layer->pallete_en = s->pallete_en;
> +     layer->pallete_color = s->pallete_color;
> +
> +     DRM_DEBUG("%s() alpha = %u, blending = %u, rotation = %u, y2r_coef = %u\n",
> +               __func__, layer->alpha, layer->blending, layer->rotation, s->y2r_coef);
> +
> +     DRM_DEBUG("%s() xfbc = %u, hsize_r = %u, hsize_y = %u, hsize_uv = %u\n",
> +               __func__, layer->xfbc, layer->header_size_r,
> +               layer->header_size_y, layer->header_size_uv);
> +
> +     for (i = 0; i < layer->planes; i++) {
> +             obj = drm_gem_fb_get_obj(fb, i);
> +             sprd_gem = to_sprd_gem_obj(obj);
> +             layer->addr[i] = sprd_gem->dma_addr + fb->offsets[i];
> +             layer->pitch[i] = fb->pitches[i];
> +     }
> +
> +     dpu->pending_planes++;
> +}
> +
> +static void sprd_plane_atomic_disable(struct drm_plane *plane,
> +                                  struct drm_plane_state *old_state)
> +{
> +     struct sprd_plane *p = to_sprd_plane(plane);
> +
> +     /*
> +      * NOTE:
> +      * The dpu->core->flip() will disable all the planes each time.
> +      * So there is no need to impliment the atomic_disable() function.
> +      * But this function can not be removed, because it will change
> +      * to call atomic_update() callback instead. Which will cause
> +      * kernel panic in sprd_plane_atomic_update().
> +      *
> +      * We do nothing here but just print a debug log.
> +      */
> +     DRM_DEBUG("%s() layer_id = %u\n", __func__, p->index);
> +}
> +
> +static void sprd_plane_reset(struct drm_plane *plane)
> +{
> +     struct sprd_plane *p = to_sprd_plane(plane);
> +     struct sprd_plane_state *s;
> +
> +     DRM_INFO("%s()\n", __func__);
> +
> +     if (plane->state) {
> +             __drm_atomic_helper_plane_destroy_state(plane->state);
> +
> +             s = to_sprd_plane_state(plane->state);
> +             memset(s, 0, sizeof(*s));
> +     } else {
> +             s = kzalloc(sizeof(*s), GFP_KERNEL);
> +             if (!s)
> +                     return;
> +             plane->state = &s->state;
> +     }
> +
> +     s->state.plane = plane;
> +     s->state.zpos = p->index;
> +     s->alpha = 255;
> +     s->blend_mode = DRM_MODE_BLEND_PIXEL_NONE;
> +}
> +
> +static struct drm_plane_state *
> +sprd_plane_atomic_duplicate_state(struct drm_plane *plane)
> +{
> +     struct sprd_plane_state *s;
> +     struct sprd_plane_state *old_state = to_sprd_plane_state(plane->state);
> +
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     s = kzalloc(sizeof(*s), GFP_KERNEL);
> +     if (!s)
> +             return NULL;
> +
> +     __drm_atomic_helper_plane_duplicate_state(plane, &s->state);
> +
> +     WARN_ON(s->state.plane != plane);
> +
> +     s->alpha = old_state->alpha;
> +     s->blend_mode = old_state->blend_mode;
> +     s->fbc_hsize_r = old_state->fbc_hsize_r;
> +     s->fbc_hsize_y = old_state->fbc_hsize_y;
> +     s->fbc_hsize_uv = old_state->fbc_hsize_uv;
> +     s->y2r_coef = old_state->y2r_coef;
> +     s->pallete_en = old_state->pallete_en;
> +     s->pallete_color = old_state->pallete_color;
> +
> +     return &s->state;
> +}
> +
> +static void sprd_plane_atomic_destroy_state(struct drm_plane *plane,
> +                                         struct drm_plane_state *state)
> +{
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     __drm_atomic_helper_plane_destroy_state(state);
> +     kfree(to_sprd_plane_state(state));
> +}
> +
> +static int sprd_plane_atomic_set_property(struct drm_plane *plane,
> +                                       struct drm_plane_state *state,
> +                                       struct drm_property *property,
> +                                       u64 val)
> +{
> +     struct sprd_plane *p = to_sprd_plane(plane);
> +     struct sprd_plane_state *s = to_sprd_plane_state(state);
> +
> +     DRM_DEBUG("%s() name = %s, val = %llu\n",
> +               __func__, property->name, val);
> +
> +     if (property == p->alpha_property)
> +             s->alpha = val;
> +     else if (property == p->blend_mode_property)
> +             s->blend_mode = val;
> +     else if (property == p->fbc_hsize_r_property)
> +             s->fbc_hsize_r = val;
> +     else if (property == p->fbc_hsize_y_property)
> +             s->fbc_hsize_y = val;
> +     else if (property == p->fbc_hsize_uv_property)
> +             s->fbc_hsize_uv = val;
> +     else if (property == p->y2r_coef_property)
> +             s->y2r_coef = val;
> +     else if (property == p->pallete_en_property)
> +             s->pallete_en = val;
> +     else if (property == p->pallete_color_property)
> +             s->pallete_color = val;
> +     else {
> +             DRM_ERROR("property %s is invalid\n", property->name);
> +             return -EINVAL;
> +     }
> +
> +     return 0;
> +}
> +
> +static int sprd_plane_atomic_get_property(struct drm_plane *plane,
> +                                       const struct drm_plane_state *state,
> +                                       struct drm_property *property,
> +                                       u64 *val)
> +{
> +     struct sprd_plane *p = to_sprd_plane(plane);
> +     const struct sprd_plane_state *s = to_sprd_plane_state(state);
> +
> +     DRM_DEBUG("%s() name = %s\n", __func__, property->name);
> +
> +     if (property == p->alpha_property)
> +             *val = s->alpha;
> +     else if (property == p->blend_mode_property)
> +             *val = s->blend_mode;
> +     else if (property == p->fbc_hsize_r_property)
> +             *val = s->fbc_hsize_r;
> +     else if (property == p->fbc_hsize_y_property)
> +             *val = s->fbc_hsize_y;
> +     else if (property == p->fbc_hsize_uv_property)
> +             *val = s->fbc_hsize_uv;
> +     else if (property == p->y2r_coef_property)
> +             *val = s->y2r_coef;
> +     else if (property == p->pallete_en_property)
> +             *val = s->pallete_en;
> +     else if (property == p->pallete_color_property)
> +             *val = s->pallete_color;
> +     else {
> +             DRM_ERROR("property %s is invalid\n", property->name);
> +             return -EINVAL;
> +     }
> +
> +     return 0;
> +}
> +
> +static int sprd_plane_create_properties(struct sprd_plane *p, int index)
> +{
> +     struct drm_property *prop;
> +     static const struct drm_prop_enum_list blend_mode_enum_list[] = {
> +             { DRM_MODE_BLEND_PIXEL_NONE, "None" },
> +             { DRM_MODE_BLEND_PREMULTI, "Pre-multiplied" },
> +             { DRM_MODE_BLEND_COVERAGE, "Coverage" },
> +     };
> +
> +     /* create rotation property */
> +     drm_plane_create_rotation_property(&p->plane,
> +                                        DRM_MODE_ROTATE_0,
> +                                        DRM_MODE_ROTATE_MASK |
> +                                        DRM_MODE_REFLECT_MASK);
> +
> +     /* create zpos property */
> +     drm_plane_create_zpos_immutable_property(&p->plane, index);
> +
> +     /* create layer alpha property */
> +     prop = drm_property_create_range(p->plane.dev, 0, "alpha", 0, 255);
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop, 255);
> +     p->alpha_property = prop;
> +
> +     /* create blend mode property */
> +     prop = drm_property_create_enum(p->plane.dev, DRM_MODE_PROP_ENUM,
> +                                     "pixel blend mode",
> +                                     blend_mode_enum_list,
> +                                     ARRAY_SIZE(blend_mode_enum_list));
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop,
> +                                DRM_MODE_BLEND_PIXEL_NONE);
> +     p->blend_mode_property = prop;
> +
> +     /* create fbc header size property */
> +     prop = drm_property_create_range(p->plane.dev, 0,
> +                     "FBC header size RGB", 0, UINT_MAX);
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop, 0);
> +     p->fbc_hsize_r_property = prop;
> +
> +     prop = drm_property_create_range(p->plane.dev, 0,
> +                     "FBC header size Y", 0, UINT_MAX);
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop, 0);
> +     p->fbc_hsize_y_property = prop;
> +
> +     prop = drm_property_create_range(p->plane.dev, 0,
> +                     "FBC header size UV", 0, UINT_MAX);
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop, 0);
> +     p->fbc_hsize_uv_property = prop;
> +
> +     /* create y2r coef property */
> +     prop = drm_property_create_range(p->plane.dev, 0,
> +                     "YUV2RGB coef", 0, UINT_MAX);
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop, 0);
> +     p->y2r_coef_property = prop;
> +
> +     /* create pallete enable property */
> +     prop = drm_property_create_range(p->plane.dev, 0,
> +                     "pallete enable", 0, UINT_MAX);
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop, 0);
> +     p->pallete_en_property = prop;
> +
> +     /* create pallete color property */
> +     prop = drm_property_create_range(p->plane.dev, 0,
> +                     "pallete color", 0, UINT_MAX);
> +     if (!prop)
> +             return -ENOMEM;
> +     drm_object_attach_property(&p->plane.base, prop, 0);
> +     p->pallete_color_property = prop;
> +
> +     return 0;
> +}
> +
> +static const struct drm_plane_helper_funcs sprd_plane_helper_funcs = {
> +     .prepare_fb = sprd_plane_prepare_fb,
> +     .cleanup_fb = sprd_plane_cleanup_fb,
> +     .atomic_check = sprd_plane_atomic_check,
> +     .atomic_update = sprd_plane_atomic_update,
> +     .atomic_disable = sprd_plane_atomic_disable,
> +};
> +
> +static const struct drm_plane_funcs sprd_plane_funcs = {
> +     .update_plane = drm_atomic_helper_update_plane,
> +     .disable_plane  = drm_atomic_helper_disable_plane,
> +     .destroy = drm_plane_cleanup,
> +     .reset = sprd_plane_reset,
> +     .atomic_duplicate_state = sprd_plane_atomic_duplicate_state,
> +     .atomic_destroy_state = sprd_plane_atomic_destroy_state,
> +     .atomic_set_property = sprd_plane_atomic_set_property,
> +     .atomic_get_property = sprd_plane_atomic_get_property,
> +};
> +
> +static struct drm_plane *sprd_plane_init(struct drm_device *drm,
> +                                     struct sprd_dpu *dpu)
> +{
> +     struct drm_plane *primary = NULL;
> +     struct sprd_plane *p = NULL;
> +     struct dpu_capability cap = {};
> +     int err, i;
> +
> +     if (dpu->core && dpu->core->capability)
> +             dpu->core->capability(&dpu->ctx, &cap);
> +
> +     dpu->layers = devm_kcalloc(drm->dev, cap.max_layers,
> +                               sizeof(struct sprd_dpu_layer), GFP_KERNEL);
> +     if (!dpu->layers)
> +             return ERR_PTR(-ENOMEM);
> +
> +     for (i = 0; i < cap.max_layers; i++) {
> +
> +             p = devm_kzalloc(drm->dev, sizeof(*p), GFP_KERNEL);
> +             if (!p)
> +                     return ERR_PTR(-ENOMEM);
> +
> +             err = drm_universal_plane_init(drm, &p->plane, 1,
> +                                            &sprd_plane_funcs, cap.fmts_ptr,
> +                                            cap.fmts_cnt, NULL,
> +                                            DRM_PLANE_TYPE_PRIMARY, NULL);
> +             if (err) {
> +                     DRM_ERROR("fail to init primary plane\n");
> +                     return ERR_PTR(err);
> +             }
> +
> +             drm_plane_helper_add(&p->plane, &sprd_plane_helper_funcs);
> +
> +             sprd_plane_create_properties(p, i);
> +
> +             p->index = i;
> +             if (i == 0)
> +                     primary = &p->plane;
> +     }
> +
> +     if (p)
> +             DRM_INFO("dpu plane init ok\n");
> +
> +     return primary;
> +}
> +
> +static void sprd_crtc_mode_set_nofb(struct drm_crtc *crtc)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +     struct drm_display_mode *mode = &crtc->state->adjusted_mode;
> +
> +     DRM_INFO("%s() set mode: %s\n", __func__, dpu->mode->name);
> +
> +     /*
> +      * TODO:
> +      * Currently, low simulator resolution only support
> +      * DPI mode, support for EDPI in the future.
> +      */
> +     if (mode->type & DRM_MODE_TYPE_BUILTIN) {
> +             dpu->ctx.if_type = SPRD_DISPC_IF_DPI;
> +             return;
> +     }
> +
> +     if ((dpu->mode->hdisplay == dpu->mode->htotal) ||
> +         (dpu->mode->vdisplay == dpu->mode->vtotal))
> +             dpu->ctx.if_type = SPRD_DISPC_IF_EDPI;
> +     else
> +             dpu->ctx.if_type = SPRD_DISPC_IF_DPI;
> +
> +     if (dpu->core && dpu->core->modeset) {
> +             if (crtc->state->mode_changed) {
> +                     struct drm_mode_modeinfo umode;
> +
> +                     drm_mode_convert_to_umode(&umode, mode);
> +                     dpu->core->modeset(&dpu->ctx, &umode);
> +             }
> +     }
> +}
> +
> +static enum drm_mode_status sprd_crtc_mode_valid(struct drm_crtc *crtc,
> +                                     const struct drm_display_mode *mode)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +
> +     DRM_INFO("%s() mode: "DRM_MODE_FMT"\n", __func__, DRM_MODE_ARG(mode));
> +
> +     if (mode->type & DRM_MODE_TYPE_DEFAULT)
> +             dpu->mode = (struct drm_display_mode *)mode;
> +
> +     if (mode->type & DRM_MODE_TYPE_PREFERRED) {
> +             dpu->mode = (struct drm_display_mode *)mode;
> +             drm_display_mode_to_videomode(dpu->mode, &dpu->ctx.vm);
> +     }
> +
> +     if (mode->type & DRM_MODE_TYPE_BUILTIN)
> +             dpu->mode = (struct drm_display_mode *)mode;
> +
> +     return MODE_OK;
> +}
> +
> +static void sprd_crtc_atomic_enable(struct drm_crtc *crtc,
> +                                struct drm_crtc_state *old_state)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +     static bool is_enabled = true;
> +
> +     DRM_INFO("%s()\n", __func__);
> +
> +     /*
> +      * add if condition to avoid resume dpu for SR feature.
> +      */
> +     if (crtc->state->mode_changed && !crtc->state->active_changed)
> +             return;
> +
> +     if (is_enabled)
> +             is_enabled = false;
> +     else
> +             pm_runtime_get_sync(dpu->dev.parent);
> +
> +     sprd_dpu_init(dpu);
> +
> +     enable_irq(dpu->ctx.irq);
> +}
> +
> +static void sprd_crtc_wait_last_commit_complete(struct drm_crtc *crtc)
> +{
> +     struct drm_crtc_commit *commit;
> +     int ret, i = 0;
> +
> +     spin_lock(&crtc->commit_lock);
> +     list_for_each_entry(commit, &crtc->commit_list, commit_entry) {
> +             i++;
> +             /* skip the first entry, that's the current commit */
> +             if (i == 2)
> +                     break;
> +     }
> +     if (i == 2)
> +             drm_crtc_commit_get(commit);
> +     spin_unlock(&crtc->commit_lock);
> +
> +     if (i != 2)
> +             return;
> +
> +     ret = wait_for_completion_interruptible_timeout(&commit->cleanup_done,
> +                                                     HZ);
> +     if (ret == 0)
> +             DRM_WARN("wait last commit completion timed out\n");
> +
> +     drm_crtc_commit_put(commit);
> +}
> +
> +static void sprd_crtc_atomic_disable(struct drm_crtc *crtc,
> +                                 struct drm_crtc_state *old_state)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +     struct drm_device *drm = dpu->crtc.dev;
> +
> +     DRM_INFO("%s()\n", __func__);
> +
> +     /* add if condition to avoid suspend dpu for SR feature */
> +     if (crtc->state->mode_changed && !crtc->state->active_changed)
> +             return;
> +
> +     sprd_crtc_wait_last_commit_complete(crtc);
> +
> +     disable_irq(dpu->ctx.irq);
> +
> +     sprd_dpu_uninit(dpu);
> +
> +     pm_runtime_put(dpu->dev.parent);
> +
> +     spin_lock_irq(&drm->event_lock);
> +     if (crtc->state->event) {
> +             drm_crtc_send_vblank_event(crtc, crtc->state->event);
> +             crtc->state->event = NULL;
> +     }
> +     spin_unlock_irq(&drm->event_lock);
> +}
> +
> +static int sprd_crtc_atomic_check(struct drm_crtc *crtc,
> +                              struct drm_crtc_state *state)
> +{
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     return 0;
> +}
> +
> +static void sprd_crtc_atomic_begin(struct drm_crtc *crtc,
> +                               struct drm_crtc_state *old_state)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     down(&dpu->ctx.refresh_lock);
> +
> +     memset(dpu->layers, 0, sizeof(*dpu->layers) * dpu->pending_planes);
> +
> +     dpu->pending_planes = 0;
> +}
> +
> +static void sprd_crtc_atomic_flush(struct drm_crtc *crtc,
> +                               struct drm_crtc_state *old_state)
> +
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +     struct drm_device *drm = dpu->crtc.dev;
> +
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     if (dpu->core && dpu->core->flip &&
> +         dpu->pending_planes && !dpu->ctx.disable_flip)
> +             dpu->core->flip(&dpu->ctx, dpu->layers, dpu->pending_planes);
> +
> +     up(&dpu->ctx.refresh_lock);
> +
> +     spin_lock_irq(&drm->event_lock);
> +     if (crtc->state->event) {
> +             drm_crtc_send_vblank_event(crtc, crtc->state->event);
> +             crtc->state->event = NULL;
> +     }
> +     spin_unlock_irq(&drm->event_lock);
> +}
> +
> +static int sprd_crtc_enable_vblank(struct drm_crtc *crtc)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     if (dpu->core && dpu->core->enable_vsync)
> +             dpu->core->enable_vsync(&dpu->ctx);
> +
> +     return 0;
> +}
> +
> +static void sprd_crtc_disable_vblank(struct drm_crtc *crtc)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +
> +     DRM_DEBUG("%s()\n", __func__);
> +
> +     if (dpu->core && dpu->core->disable_vsync)
> +             dpu->core->disable_vsync(&dpu->ctx);
> +}
> +
> +static int sprd_crtc_create_properties(struct drm_crtc *crtc)
> +{
> +     struct sprd_dpu *dpu = crtc_to_dpu(crtc);
> +     struct drm_device *drm = dpu->crtc.dev;
> +     struct drm_property *prop;
> +     struct drm_property_blob *blob;
> +     size_t blob_size;
> +
> +     blob_size = strlen(dpu->ctx.version) + 1;
> +
> +     blob = drm_property_create_blob(dpu->crtc.dev, blob_size,
> +                     dpu->ctx.version);
> +     if (IS_ERR(blob)) {
> +             DRM_ERROR("drm_property_create_blob dpu version failed\n");
> +             return PTR_ERR(blob);
> +     }
> +
> +     /* create dpu version property */
> +     prop = drm_property_create(drm,
> +             DRM_MODE_PROP_IMMUTABLE | DRM_MODE_PROP_BLOB,
> +             "dpu version", 0);
> +     if (!prop) {
> +             DRM_ERROR("drm_property_create dpu version failed\n");
> +             return -ENOMEM;
> +     }
> +     drm_object_attach_property(&crtc->base, prop, blob->base.id);
> +
> +     /* create corner size property */
> +     prop = drm_property_create(drm,
> +             DRM_MODE_PROP_IMMUTABLE | DRM_MODE_PROP_RANGE,
> +             "corner size", 0);
> +     if (!prop) {
> +             DRM_ERROR("drm_property_create corner size failed\n");
> +             return -ENOMEM;
> +     }
> +     drm_object_attach_property(&crtc->base, prop, dpu->ctx.corner_size);
> +
> +     return 0;
> +}
> +
> +static const struct drm_crtc_helper_funcs sprd_crtc_helper_funcs = {
> +     .mode_set_nofb  = sprd_crtc_mode_set_nofb,
> +     .mode_valid     = sprd_crtc_mode_valid,
> +     .atomic_check   = sprd_crtc_atomic_check,
> +     .atomic_begin   = sprd_crtc_atomic_begin,
> +     .atomic_flush   = sprd_crtc_atomic_flush,
> +     .atomic_enable  = sprd_crtc_atomic_enable,
> +     .atomic_disable = sprd_crtc_atomic_disable,
> +};
> +
> +static const struct drm_crtc_funcs sprd_crtc_funcs = {
> +     .destroy        = drm_crtc_cleanup,
> +     .set_config     = drm_atomic_helper_set_config,
> +     .page_flip      = drm_atomic_helper_page_flip,
> +     .reset          = drm_atomic_helper_crtc_reset,
> +     .atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state,
> +     .atomic_destroy_state   = drm_atomic_helper_crtc_destroy_state,
> +     .enable_vblank  = sprd_crtc_enable_vblank,
> +     .disable_vblank = sprd_crtc_disable_vblank,
> +};
> +
> +static int sprd_crtc_init(struct drm_device *drm, struct drm_crtc *crtc,
> +                      struct drm_plane *primary)
> +{
> +     struct device_node *port;
> +     int err;
> +
> +     /*
> +      * set crtc port so that drm_of_find_possible_crtcs call works
> +      */
> +     port = of_parse_phandle(drm->dev->of_node, "ports", 0);
> +     if (!port) {
> +             DRM_ERROR("find 'ports' phandle of %s failed\n",
> +                       drm->dev->of_node->full_name);
> +             return -EINVAL;
> +     }
> +     of_node_put(port);
> +     crtc->port = port;
> +
> +     err = drm_crtc_init_with_planes(drm, crtc, primary, NULL,
> +                                     &sprd_crtc_funcs, NULL);
> +     if (err) {
> +             DRM_ERROR("failed to init crtc.\n");
> +             return err;
> +     }
> +
> +     drm_mode_crtc_set_gamma_size(crtc, 256);
> +
> +     drm_crtc_helper_add(crtc, &sprd_crtc_helper_funcs);
> +
> +     sprd_crtc_create_properties(crtc);
> +
> +     DRM_INFO("%s() ok\n", __func__);
> +     return 0;
> +}
> +
> +int sprd_dpu_run(struct sprd_dpu *dpu)
> +{
> +     struct dpu_context *ctx = &dpu->ctx;
> +
> +     down(&ctx->refresh_lock);
> +
> +     if (!ctx->is_inited) {
> +             DRM_ERROR("dpu is not initialized\n");
> +             up(&ctx->refresh_lock);
> +             return -EINVAL;
> +     }
> +
> +     if (!ctx->is_stopped) {
> +             up(&ctx->refresh_lock);
> +             return 0;
> +     }
> +
> +     if (dpu->core && dpu->core->run)
> +             dpu->core->run(ctx);
> +
> +     up(&ctx->refresh_lock);
> +
> +     drm_crtc_vblank_on(&dpu->crtc);
> +
> +     return 0;
> +}
> +
> +int sprd_dpu_stop(struct sprd_dpu *dpu)
> +{
> +     struct dpu_context *ctx = &dpu->ctx;
> +
> +     down(&ctx->refresh_lock);
> +
> +     if (!ctx->is_inited) {
> +             DRM_ERROR("dpu is not initialized\n");
> +             up(&ctx->refresh_lock);
> +             return -EINVAL;
> +     }
> +
> +     if (ctx->is_stopped) {
> +             up(&ctx->refresh_lock);
> +             return 0;
> +     }
> +
> +     if (dpu->core && dpu->core->stop)
> +             dpu->core->stop(ctx);
> +
> +     up(&ctx->refresh_lock);
> +
> +     drm_crtc_handle_vblank(&dpu->crtc);
> +     drm_crtc_vblank_off(&dpu->crtc);
> +
> +     return 0;
> +}
> +
> +static int sprd_dpu_init(struct sprd_dpu *dpu)
> +{
> +     struct dpu_context *ctx = &dpu->ctx;
> +
> +     down(&ctx->refresh_lock);
> +
> +     if (dpu->ctx.is_inited) {
> +             up(&ctx->refresh_lock);
> +             return 0;
> +     }
> +
> +     if (dpu->glb && dpu->glb->power)
> +             dpu->glb->power(ctx, true);
> +     if (dpu->glb && dpu->glb->enable)
> +             dpu->glb->enable(ctx);
> +
> +     if (ctx->is_stopped && dpu->glb && dpu->glb->reset)
> +             dpu->glb->reset(ctx);
> +
> +     if (dpu->clk && dpu->clk->init)
> +             dpu->clk->init(ctx);
> +     if (dpu->clk && dpu->clk->enable)
> +             dpu->clk->enable(ctx);
> +
> +     if (dpu->core && dpu->core->init)
> +             dpu->core->init(ctx);
> +     if (dpu->core && dpu->core->ifconfig)
> +             dpu->core->ifconfig(ctx);
> +
> +     ctx->is_inited = true;
> +
> +     up(&ctx->refresh_lock);
> +
> +     return 0;
> +}
> +
> +static int sprd_dpu_uninit(struct sprd_dpu *dpu)
> +{
> +     struct dpu_context *ctx = &dpu->ctx;
> +
> +     down(&ctx->refresh_lock);
> +
> +     if (!dpu->ctx.is_inited) {
> +             up(&ctx->refresh_lock);
> +             return 0;
> +     }
> +
> +     if (dpu->core && dpu->core->uninit)
> +             dpu->core->uninit(ctx);
> +     if (dpu->clk && dpu->clk->disable)
> +             dpu->clk->disable(ctx);
> +     if (dpu->glb && dpu->glb->disable)
> +             dpu->glb->disable(ctx);
> +     if (dpu->glb && dpu->glb->power)
> +             dpu->glb->power(ctx, false);
> +
> +     ctx->is_inited = false;
> +
> +     up(&ctx->refresh_lock);
> +
> +     return 0;
> +}
> +
> +static irqreturn_t sprd_dpu_isr(int irq, void *data)
> +{
> +     struct sprd_dpu *dpu = data;
> +     struct dpu_context *ctx = &dpu->ctx;
> +     u32 int_mask = 0;
> +
> +     if (dpu->core && dpu->core->isr)
> +             int_mask = dpu->core->isr(ctx);
> +
> +     if (int_mask & DISPC_INT_TE_MASK) {
> +             if (ctx->te_check_en) {
> +                     ctx->evt_te = true;
> +                     wake_up_interruptible_all(&ctx->te_wq);
> +             }
> +     }
> +
> +     if (int_mask & DISPC_INT_ERR_MASK)
> +             DRM_WARN("Warning: dpu underflow!\n");
> +
> +     if ((int_mask & DISPC_INT_DPI_VSYNC_MASK) && ctx->is_inited)
> +             drm_crtc_handle_vblank(&dpu->crtc);
> +
> +     return IRQ_HANDLED;
> +}
> +
> +static int sprd_dpu_irq_request(struct sprd_dpu *dpu)
> +{
> +     int err;
> +     int irq_num;
> +
> +     irq_num = irq_of_parse_and_map(dpu->dev.of_node, 0);
> +     if (!irq_num) {
> +             DRM_ERROR("error: dpu parse irq num failed\n");
> +             return -EINVAL;
> +     }
> +     DRM_INFO("dpu irq_num = %d\n", irq_num);
> +
> +     irq_set_status_flags(irq_num, IRQ_NOAUTOEN);
> +     err = devm_request_irq(&dpu->dev, irq_num, sprd_dpu_isr,
> +                                     0, "DISPC", dpu);
> +     if (err) {
> +             DRM_ERROR("error: dpu request irq failed\n");
> +             return -EINVAL;
> +     }
> +     dpu->ctx.irq = irq_num;
> +     dpu->ctx.dpu_isr = sprd_dpu_isr;
> +
> +     return 0;
> +}
> +
> +static int sprd_dpu_bind(struct device *dev, struct device *master, void *data)
> +{
> +     struct drm_device *drm = data;
> +     struct sprd_drm *sprd = drm->dev_private;
> +     struct sprd_dpu *dpu = dev_get_drvdata(dev);
> +     struct drm_plane *plane;
> +     int err;
> +
> +     DRM_INFO("%s()\n", __func__);
> +
> +     plane = sprd_plane_init(drm, dpu);
> +     if (IS_ERR_OR_NULL(plane)) {
> +             err = PTR_ERR(plane);
> +             return err;
> +     }
> +
> +     err = sprd_crtc_init(drm, &dpu->crtc, plane);
> +     if (err)
> +             return err;
> +
> +     sprd_dpu_irq_request(dpu);
> +
> +     sprd->dpu_dev = dev;
> +
> +     return 0;
> +}
> +
> +static void sprd_dpu_unbind(struct device *dev, struct device *master,
> +     void *data)
> +{
> +     struct sprd_dpu *dpu = dev_get_drvdata(dev);
> +
> +     DRM_INFO("%s()\n", __func__);
> +
> +     drm_crtc_cleanup(&dpu->crtc);
> +}
> +
> +static const struct component_ops dpu_component_ops = {
> +     .bind = sprd_dpu_bind,
> +     .unbind = sprd_dpu_unbind,
> +};
> +
> +static int sprd_dpu_context_init(struct sprd_dpu *dpu,
> +                             struct device_node *np)
> +{
> +     u32 temp;
> +     struct resource r;
> +     struct dpu_context *ctx = &dpu->ctx;
> +
> +     if (dpu->core && dpu->core->parse_dt)
> +             dpu->core->parse_dt(&dpu->ctx, np);
> +     if (dpu->clk && dpu->clk->parse_dt)
> +             dpu->clk->parse_dt(&dpu->ctx, np);
> +     if (dpu->glb && dpu->glb->parse_dt)
> +             dpu->glb->parse_dt(&dpu->ctx, np);
> +
> +     if (!of_property_read_u32(np, "sprd,dev-id", &temp))
> +             ctx->id = temp;
> +
> +     if (of_address_to_resource(np, 0, &r)) {
> +             DRM_ERROR("parse dt base address failed\n");
> +             return -ENODEV;
> +     }
> +     ctx->base = (unsigned long)ioremap_nocache(r.start,
> +                                     resource_size(&r));
> +     if (ctx->base == 0) {
> +             DRM_ERROR("ioremap base address failed\n");
> +             return -EFAULT;
> +     }
> +
> +     of_get_logo_memory_info(dpu, np);
> +
> +     sema_init(&ctx->refresh_lock, 1);
> +
> +     return 0;
> +}
> +
> +static int sprd_dpu_probe(struct platform_device *pdev)
> +{
> +     struct device_node *np = pdev->dev.of_node;
> +     struct sprd_dpu *dpu;
> +     int ret;
> +
> +     dpu = devm_kzalloc(&pdev->dev, sizeof(*dpu), GFP_KERNEL);
> +     if (!dpu)
> +             return -ENOMEM;
> +
> +     ret = sprd_dpu_context_init(dpu, np);
> +     if (ret)
> +             return ret;
> +
> +     platform_set_drvdata(pdev, dpu);
> +     pm_runtime_set_active(&pdev->dev);
> +     pm_runtime_get_noresume(&pdev->dev);
> +     pm_runtime_enable(&pdev->dev);
> +
> +     return component_add(&pdev->dev, &dpu_component_ops);
> +}
> +
> +static int sprd_dpu_remove(struct platform_device *pdev)
> +{
> +     component_del(&pdev->dev, &dpu_component_ops);
> +     return 0;
> +}
> +
> +static const struct of_device_id dpu_match_table[] = {
> +     { .compatible = "sprd,display-processor",},
> +     {},
> +};
> +
> +static struct platform_driver sprd_dpu_driver = {
> +     .probe = sprd_dpu_probe,
> +     .remove = sprd_dpu_remove,
> +     .driver = {
> +             .name = "sprd-dpu-drv",
> +             .of_match_table = dpu_match_table,
> +     },
> +};
> +module_platform_driver(sprd_dpu_driver);
> +
> +MODULE_AUTHOR("Leon He <leon.he@xxxxxxxxxx>");
> +MODULE_AUTHOR("Kevin Tang <kevin.tang@xxxxxxxxxx>");
> +MODULE_DESCRIPTION("Unisoc Display Controller Driver");
> +MODULE_LICENSE("GPL v2");
> diff --git a/drivers/gpu/drm/sprd/sprd_dpu.h b/drivers/gpu/drm/sprd/sprd_dpu.h
> new file mode 100644
> index 0000000..998ebc7
> --- /dev/null
> +++ b/drivers/gpu/drm/sprd/sprd_dpu.h
> @@ -0,0 +1,217 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +/*
> + * Copyright (C) 2019 Unisoc Inc.
> + */
> +
> +#ifndef __SPRD_DPU_H__
> +#define __SPRD_DPU_H__
> +
> +#include <linux/bug.h>
> +#include <linux/delay.h>
> +#include <linux/device.h>
> +#include <linux/kernel.h>
> +#include <linux/platform_device.h>
> +#include <linux/string.h>
> +#include <video/videomode.h>
> +
> +#include <drm/drm_crtc.h>
> +#include <drm/drm_fourcc.h>
> +#include <drm/drm_print.h>
> +#include <drm/drm_vblank.h>
> +#include <uapi/drm/drm_mode.h>
> +#include "disp_lib.h"
> +
> +#define DISPC_INT_DONE_MASK          BIT(0)
> +#define DISPC_INT_TE_MASK            BIT(1)
> +#define DISPC_INT_ERR_MASK           BIT(2)
> +#define DISPC_INT_EDPI_TE_MASK               BIT(3)
> +#define DISPC_INT_UPDATE_DONE_MASK   BIT(4)
> +#define DISPC_INT_DPI_VSYNC_MASK     BIT(5)
> +#define DISPC_INT_WB_DONE_MASK               BIT(6)
> +#define DISPC_INT_WB_FAIL_MASK               BIT(7)
> +
> +/* NOTE: this mask is not a realy dpu interrupt mask */
> +#define DISPC_INT_FENCE_SIGNAL_REQUEST       BIT(31)
> +
> +enum {
> +     SPRD_DISPC_IF_DBI = 0,
> +     SPRD_DISPC_IF_DPI,
> +     SPRD_DISPC_IF_EDPI,
> +     SPRD_DISPC_IF_LIMIT
> +};
> +
> +enum {
> +     SPRD_IMG_DATA_ENDIAN_B0B1B2B3 = 0,
> +     SPRD_IMG_DATA_ENDIAN_B3B2B1B0,
> +     SPRD_IMG_DATA_ENDIAN_B2B3B0B1,
> +     SPRD_IMG_DATA_ENDIAN_B1B0B3B2,
> +     SPRD_IMG_DATA_ENDIAN_LIMIT
> +};
> +
> +enum {
> +     DISPC_CLK_ID_CORE = 0,
> +     DISPC_CLK_ID_DBI,
> +     DISPC_CLK_ID_DPI,
> +     DISPC_CLK_ID_MAX
> +};
> +
> +enum {
> +     ENHANCE_CFG_ID_ENABLE,
> +     ENHANCE_CFG_ID_DISABLE,
> +     ENHANCE_CFG_ID_SCL,
> +     ENHANCE_CFG_ID_EPF,
> +     ENHANCE_CFG_ID_HSV,
> +     ENHANCE_CFG_ID_CM,
> +     ENHANCE_CFG_ID_SLP,
> +     ENHANCE_CFG_ID_GAMMA,
> +     ENHANCE_CFG_ID_LTM,
> +     ENHANCE_CFG_ID_CABC,
> +     ENHANCE_CFG_ID_SLP_LUT,
> +     ENHANCE_CFG_ID_LUT3D,
> +     ENHANCE_CFG_ID_MAX
> +};
> +
> +struct sprd_dpu_layer {
> +     u8 index;
> +     u8 planes;
> +     u32 addr[4];
> +     u32 pitch[4];
> +     s16 src_x;
> +     s16 src_y;
> +     s16 src_w;
> +     s16 src_h;
> +     s16 dst_x;
> +     s16 dst_y;
> +     u16 dst_w;
> +     u16 dst_h;
> +     u32 format;
> +     u32 alpha;
> +     u32 blending;
> +     u32 rotation;
> +     u32 xfbc;
> +     u32 height;
> +     u32 header_size_r;
> +     u32 header_size_y;
> +     u32 header_size_uv;
> +     u32 y2r_coef;
> +     u8 pallete_en;
> +     u32 pallete_color;
> +};
> +
> +struct dpu_capability {
> +     u32 max_layers;
> +     const u32 *fmts_ptr;
> +     u32 fmts_cnt;
> +};
> +
> +struct dpu_context;
> +
> +struct dpu_core_ops {
> +     int (*parse_dt)(struct dpu_context *ctx,
> +                     struct device_node *np);
> +     u32 (*version)(struct dpu_context *ctx);
> +     int (*init)(struct dpu_context *ctx);
> +     void (*uninit)(struct dpu_context *ctx);
> +     void (*run)(struct dpu_context *ctx);
> +     void (*stop)(struct dpu_context *ctx);
> +     void (*disable_vsync)(struct dpu_context *ctx);
> +     void (*enable_vsync)(struct dpu_context *ctx);
> +     u32 (*isr)(struct dpu_context *ctx);
> +     void (*ifconfig)(struct dpu_context *ctx);
> +     void (*write_back)(struct dpu_context *ctx, u8 count, bool debug);
> +     void (*flip)(struct dpu_context *ctx,
> +                  struct sprd_dpu_layer layers[], u8 count);
> +     int (*capability)(struct dpu_context *ctx,
> +                     struct dpu_capability *cap);
> +     void (*bg_color)(struct dpu_context *ctx, u32 color);
> +     void (*enhance_set)(struct dpu_context *ctx, u32 id, void *param);
> +     void (*enhance_get)(struct dpu_context *ctx, u32 id, void *param);
> +     int (*modeset)(struct dpu_context *ctx,
> +                     struct drm_mode_modeinfo *mode);
> +     bool (*check_raw_int)(struct dpu_context *ctx, u32 mask);
> +};
> +
> +struct dpu_clk_ops {
> +     int (*parse_dt)(struct dpu_context *ctx,
> +                     struct device_node *np);
> +     int (*init)(struct dpu_context *ctx);
> +     int (*uinit)(struct dpu_context *ctx);
> +     int (*enable)(struct dpu_context *ctx);
> +     int (*disable)(struct dpu_context *ctx);
> +     int (*update)(struct dpu_context *ctx, int clk_id, int val);
> +};
> +
> +struct dpu_glb_ops {
> +     int (*parse_dt)(struct dpu_context *ctx,
> +                     struct device_node *np);
> +     void (*enable)(struct dpu_context *ctx);
> +     void (*disable)(struct dpu_context *ctx);
> +     void (*reset)(struct dpu_context *ctx);
> +     void (*power)(struct dpu_context *ctx, int enable);
> +};
> +
> +struct dpu_context {
> +     unsigned long base;
> +     u32 base_offset[2];
> +     const char *version;
> +     u32 corner_size;
> +     int irq;
> +     u8 if_type;
> +     u8 id;
> +     bool is_inited;
> +     bool is_stopped;
> +     bool disable_flip;
> +     struct videomode vm;
> +     struct semaphore refresh_lock;
> +     struct work_struct wb_work;
> +     struct tasklet_struct dvfs_task;
> +     u32 wb_addr_p;
> +     irqreturn_t (*dpu_isr)(int irq, void *data);
> +     wait_queue_head_t te_wq;
> +     bool te_check_en;
> +     bool evt_te;
> +     unsigned long logo_addr;
> +     unsigned long logo_size;
> +     struct work_struct cabc_work;
> +     struct work_struct cabc_bl_update;
> +};
> +
> +struct sprd_dpu {
> +     struct device dev;
> +     struct drm_crtc crtc;
> +     struct dpu_context ctx;
> +     struct dpu_core_ops *core;
> +     struct dpu_clk_ops *clk;
> +     struct dpu_glb_ops *glb;
> +     struct drm_display_mode *mode;
> +     struct sprd_dpu_layer *layers;
> +     u8 pending_planes;
> +};
> +
> +extern struct list_head dpu_core_head;
> +extern struct list_head dpu_clk_head;
> +extern struct list_head dpu_glb_head;
> +
> +static inline struct sprd_dpu *crtc_to_dpu(struct drm_crtc *crtc)
> +{
> +     return crtc ? container_of(crtc, struct sprd_dpu, crtc) : NULL;
> +}
> +
> +#define dpu_core_ops_register(entry) \
> +     disp_ops_register(entry, &dpu_core_head)
> +#define dpu_clk_ops_register(entry) \
> +     disp_ops_register(entry, &dpu_clk_head)
> +#define dpu_glb_ops_register(entry) \
> +     disp_ops_register(entry, &dpu_glb_head)
> +
> +#define dpu_core_ops_attach(str) \
> +     disp_ops_attach(str, &dpu_core_head)
> +#define dpu_clk_ops_attach(str) \
> +     disp_ops_attach(str, &dpu_clk_head)
> +#define dpu_glb_ops_attach(str) \
> +     disp_ops_attach(str, &dpu_glb_head)
> +
> +int sprd_dpu_run(struct sprd_dpu *dpu);
> +int sprd_dpu_stop(struct sprd_dpu *dpu);
> +
> +#endif
>

--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Maxfeldstr. 5, 90409 Nürnberg, Germany
(HRB 36809, AG Nürnberg)
Geschäftsführer: Felix Imendörffer

_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[Index of Archives]     [Linux DRI Users]     [Linux Intel Graphics]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]
  Powered by Linux