Signed-off-by: Paul B Mahol <onemda(a)gmail.com>
---
Please test it and reports benchmark results on non toyish GPUs.
---
configure | 1 +
libavfilter/Makefile | 1 +
libavfilter/allfilters.c | 1 +
libavfilter/vf_v360_vulkan.c | 668 +++++++++++++++++++++++++++++++++++
4 files changed, 671 insertions(+)
create mode 100644 libavfilter/vf_v360_vulkan.c
diff --git a/configure b/configure
index 7495f35faa..2a3ac563db 100755
--- a/configure
+++ b/configure
@@ -3622,6 +3622,7 @@ transpose_opencl_filter_deps="opencl"
transpose_vaapi_filter_deps="vaapi VAProcPipelineCaps_rotation_flags"
unsharp_opencl_filter_deps="opencl"
uspp_filter_deps="gpl avcodec"
+v360_vulkan_filter_deps="vulkan libglslang"
vaguedenoiser_filter_deps="gpl"
vidstabdetect_filter_deps="libvidstab"
vidstabtransform_filter_deps="libvidstab"
diff --git a/libavfilter/Makefile b/libavfilter/Makefile
index 191826a622..6440794bcf 100644
--- a/libavfilter/Makefile
+++ b/libavfilter/Makefile
@@ -441,6 +441,7 @@ OBJS-$(CONFIG_UNSHARP_OPENCL_FILTER) += vf_unsharp_opencl.o opencl.o \
OBJS-$(CONFIG_UNTILE_FILTER) += vf_untile.o
OBJS-$(CONFIG_USPP_FILTER) += vf_uspp.o
OBJS-$(CONFIG_V360_FILTER) += vf_v360.o
+OBJS-$(CONFIG_V360_VULKAN_FILTER) += vf_v360_vulkan.o vulkan.o
OBJS-$(CONFIG_VAGUEDENOISER_FILTER) += vf_vaguedenoiser.o
OBJS-$(CONFIG_VECTORSCOPE_FILTER) += vf_vectorscope.o
OBJS-$(CONFIG_VFLIP_FILTER) += vf_vflip.o
diff --git a/libavfilter/allfilters.c b/libavfilter/allfilters.c
index de5884529c..65669c4422 100644
--- a/libavfilter/allfilters.c
+++ b/libavfilter/allfilters.c
@@ -420,6 +420,7 @@ extern AVFilter ff_vf_unsharp_opencl;
extern AVFilter ff_vf_untile;
extern AVFilter ff_vf_uspp;
extern AVFilter ff_vf_v360;
+extern AVFilter ff_vf_v360_vulkan;
extern AVFilter ff_vf_vaguedenoiser;
extern AVFilter ff_vf_vectorscope;
extern AVFilter ff_vf_vflip;
diff --git a/libavfilter/vf_v360_vulkan.c b/libavfilter/vf_v360_vulkan.c
new file mode 100644
index 0000000000..7db200f1c6
--- /dev/null
+++ b/libavfilter/vf_v360_vulkan.c
@@ -0,0 +1,668 @@
+/*
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavutil/random_seed.h"
+#include "libavutil/opt.h"
+#include "vulkan.h"
+#include "internal.h"
+
+enum RotationOrder {
+ YAW,
+ PITCH,
+ ROLL,
+ NB_RORDERS,
+};
+
+enum Projections {
+ EQUIRECTANGULAR,
+ CUBEMAP_3_2,
+ CUBEMAP_6_1,
+ EQUIANGULAR,
+ FLAT,
+ DUAL_FISHEYE,
+ BARREL,
+ CUBEMAP_1_6,
+ STEREOGRAPHIC,
+ MERCATOR,
+ BALL,
+ HAMMER,
+ SINUSOIDAL,
+ FISHEYE,
+ PANNINI,
+ CYLINDRICAL,
+ PERSPECTIVE,
+ TETRAHEDRON,
+ BARREL_SPLIT,
+ TSPYRAMID,
+ HEQUIRECTANGULAR,
+ NB_PROJECTIONS,
+};
+
+#define CGROUPS (int [3]){ 32, 32, 1 }
+
+typedef struct V360VulkanContext {
+ VulkanFilterContext vkctx;
+
+ int initialized;
+ FFVkExecContext *exec;
+ VulkanPipeline *pl;
+
+ /* Shader updators, must be in the main filter struct */
+ VkDescriptorImageInfo input_images[3];
+ VkDescriptorImageInfo output_images[3];
+
+ int planewidth[4], planeheight[4];
+ int inplanewidth[4], inplaneheight[4];
+ int in, out;
+ int width, height;
+ float h_fov, v_fov;
+ float ih_fov, iv_fov;
+ float yaw, pitch, roll;
+ char *rorder;
+ int rotation_order[3];
+
+ /* Push constants / options */
+ struct {
+ float flat_range[2];
+ float iflat_range[2];
+ float rot_mat[4][4];
+ } opts;
+} V360VulkanContext;
+
+static const char flat_to_xyz[] = {
+ C(0, void out_transform(out vec3 v, in ivec2 out_size, in ivec2 pos) )
+ C(0, { )
+ C(1, vec2 fpos = vec2(pos) + vec2(0.5f, 0.5f); )
+ C(1, vec2 p = ((fpos / vec2(out_size)) - 0.5f)*2.0f; )
+ C(1, v = vec3(p[0], p[1], 1.f) * vec3(flat_range, 1.f); )
+ C(1, v = normalize(v); )
+ C(0, } )
+};
+
+static const char xyz_to_flat[] = {
+ C(0, void in_transform(int idx, in vec3 v, in ivec2 pos, in ivec2 in_size) )
+ C(0, { )
+ C(1, const float r = tan(acos(v[2])); )
+ C(1, const float rr = abs(r) < 1e+6f ? r : length(in_size); )
+ C(1, const float h = length(vec2(v[0], v[1])); )
+ C(1, const float c = h <= 1e-6f ? 1.f : rr / h; )
+ C(1, vec2 p = vec2(v[0], v[1]) / iflat_range * c; )
+ C(1, p = IS_WITHIN(abs(p), vec2(1.f)) ? (p/2.0f)+0.5f:vec2(0.f); )
+ C(1, p = v[2] >= 0.f ? p : vec2(0.f); )
+ C(1, vec4 res = texture(input_img[idx], p); )
+ C(1, imageStore(output_img[idx], pos, res); )
+ C(0, } )
+};
+
+static const char equirect_to_xyz[] = {
+ C(0, void out_transform(out vec3 v, in ivec2 out_size, in ivec2 pos) )
+ C(0, { )
+ C(1, vec2 fpos = 2.f * vec2(pos) + 0.5f; )
+ C(1, vec2 p = fpos / vec2(out_size) - 1.f; )
+ C(1, p = vec2(p[0] * PI, p[1] * PI_2); )
+ C(1, v = vec3(cos(p[1]) * sin(p[0]), sin(p[1]), cos(p[1])*cos(p[0])); )
+ C(0, } )
+};
+
+static const char xyz_to_equirect[] = {
+ C(0, void in_transform(int idx, in vec3 v, in ivec2 pos, in ivec2 in_size) )
+ C(0, { )
+ C(1, vec2 p = vec2(atan(v[0], v[2]) / PI, asin(v[1]) / PI_2); )
+ C(1, vec4 res = texture(input_img[idx], (p/2.0f) + 0.5f); )
+ C(1, imageStore(output_img[idx], pos, res); )
+ C(0, } )
+};
+
+static const char stereographic_to_xyz[] = {
+ C(0, void out_transform(out vec3 v, in ivec2 out_size, in ivec2 pos) )
+ C(0, { )
+ C(1, vec2 fpos = vec2(pos) + vec2(0.5f, 0.5f); )
+ C(1, vec2 p = (fpos / vec2(out_size) - 0.5f) * 2.0f * flat_range; )
+ C(1, const float r = length(p); )
+ C(1, const float theta = atan(r) * 2.0f; )
+ C(1, v = vec3(p[0] / r*sin(theta), p[1] / r*sin(theta), cos(theta)); )
+ C(1, v = normalize(v); )
+ C(0, } )
+};
+
+static const char xyz_to_stereographic[] = {
+ C(0, void in_transform(int idx, in vec3 v, in ivec2 pos, in ivec2 in_size) )
+ C(0, { )
+ C(1, const float theta = acos(v[2]); )
+ C(1, const float r = tan(theta * 0.5f); )
+ C(1, const vec2 c = (r / length(vec2(v[0], v[1]))) / iflat_range; )
+ C(1, vec2 p = vec2(v[0], v[1]) * c; )
+ C(1, p = IS_WITHIN(abs(p), vec2(1.f)) ? (p/2.0f)+0.5f:vec2(0.f); )
+ C(1, vec4 res = texture(input_img[idx], p); )
+ C(1, imageStore(output_img[idx], pos, res); )
+ C(0, } )
+};
+
+static const char fisheye_to_xyz[] = {
+ C(0, void out_transform(out vec3 v, in ivec2 out_size, in ivec2 pos) )
+ C(0, { )
+ C(1, vec2 fpos = vec2(pos) + vec2(0.5f, 0.5f); )
+ C(1, vec2 p = (fpos / vec2(out_size) - 0.5f) * 2.0f * flat_range; )
+ C(1, const float r = length(p); )
+ C(1, const float phi = atan(p[1], p[0]); )
+ C(1, const float theta = (1.f - r) * PI_2; )
+ C(1, v = vec3(cos(theta)*cos(phi), cos(theta)*sin(phi), sin(theta)); )
+ C(1, v = normalize(v); )
+ C(0, } )
+};
+
+static const char xyz_to_fisheye[] = {
+ C(0, void in_transform(int idx, in vec3 v, in ivec2 pos, in ivec2 in_size) )
+ C(0, { )
+ C(1, const float h = length(vec2(v[0], v[1])); )
+ C(1, const float lh = h > 0.f ? h / 2.f : 1.f; )
+ C(1, const float phi = atan(h, v[2]) / PI; )
+ C(1, vec2 p = vec2(v[0], v[1]) * phi / lh / iflat_range; )
+ C(1, p = (length(p) <= 1.f) ? (p/2.0f)+0.5f:vec2(0.f); )
+ C(1, vec4 res = texture(input_img[idx], p); )
+ C(1, imageStore(output_img[idx], pos, res); )
+ C(0, } )
+};
+
+static const char dfisheye_to_xyz[] = {
+ C(0, void out_transform(out vec3 v, in ivec2 out_size, in ivec2 pos) )
+ C(0, { )
+ C(1, const float m = pos[0] >= out_size[0] / 2 ? 1.f : -1.f; )
+ C(1, vec2 npos = m == 1.f ? vec2(out_size[0] / 2, 0.f) : vec2(0.f); )
+ C(1, vec2 fpos = vec2(pos) - npos + vec2(0.5f, 0.5f); )
+ C(1, vec2 osize = vec2(out_size) * vec2(0.5f, 1.f); )
+ C(1, vec2 p = (fpos / osize - 0.5f) * 2.0f * flat_range; )
+ C(1, const float h = length(p); )
+ C(1, const float lh = h > 0.f ? h : 1.f; )
+ C(1, const float theta = m * PI_2 * (1.f - h); )
+ C(1, p = p / lh; )
+ C(1, v = vec3(cos(theta)*m*p[0], cos(theta)*p[1], sin(theta)); )
+ C(1, v = normalize(v); )
+ C(0, } )
+};
+
+static const char xyz_to_dfisheye[] = {
+ C(0, void in_transform(int idx, in vec3 v, in ivec2 pos, in ivec2 in_size) )
+ C(0, { )
+ C(1, const float h = length(vec2(v[0], v[1])); )
+ C(1, const float lh = h > 0.f ? h : 1.f; )
+ C(1, const float theta = acos(abs(v[2])) / PI; )
+ C(1, vec2 p = (vec2(v[0], v[1]) * theta)/lh/iflat_range + 0.5f; )
+ C(1, p = p * vec2(0.5f, 1.f); )
+ C(1, p = v[2] >= 0.f ? vec2(p[0]+0.5f, p[1]) : vec2(0.5f-p[0], p[1]); )
+ C(1, vec4 res = texture(input_img[idx], p); )
+ C(1, imageStore(output_img[idx], pos, res); )
+ C(0, } )
+};
+
+static void multiply_matrix(float c[4][4], const float a[4][4], const float b[4][4])
+{
+ for (int i = 0; i < 3; i++) {
+ for (int j = 0; j < 3; j++) {
+ float sum = 0.f;
+
+ for (int k = 0; k < 3; k++)
+ sum += a[i][k] * b[k][j];
+
+ c[i][j] = sum;
+ }
+ }
+}
+
+static inline void calculate_rotation_matrix(float yaw, float pitch, float roll,
+ float rot_mat[4][4],
+ const int rotation_order[3])
+{
+ const float yaw_rad = yaw * M_PI / 180.f;
+ const float pitch_rad = pitch * M_PI / 180.f;
+ const float roll_rad = roll * M_PI / 180.f;
+
+ const float sin_yaw = sinf(yaw_rad);
+ const float cos_yaw = cosf(yaw_rad);
+ const float sin_pitch = sinf(pitch_rad);
+ const float cos_pitch = cosf(pitch_rad);
+ const float sin_roll = sinf(roll_rad);
+ const float cos_roll = cosf(roll_rad);
+
+ float m[3][4][4];
+ float temp[4][4];
+
+ m[0][0][0] = cos_yaw; m[0][0][1] = 0; m[0][0][2] = sin_yaw;
+ m[0][1][0] = 0; m[0][1][1] = 1; m[0][1][2] = 0;
+ m[0][2][0] = -sin_yaw; m[0][2][1] = 0; m[0][2][2] = cos_yaw;
+
+ m[1][0][0] = 1; m[1][0][1] = 0; m[1][0][2] = 0;
+ m[1][1][0] = 0; m[1][1][1] = cos_pitch; m[1][1][2] = -sin_pitch;
+ m[1][2][0] = 0; m[1][2][1] = sin_pitch; m[1][2][2] = cos_pitch;
+
+ m[2][0][0] = cos_roll; m[2][0][1] = -sin_roll; m[2][0][2] = 0;
+ m[2][1][0] = sin_roll; m[2][1][1] = cos_roll; m[2][1][2] = 0;
+ m[2][2][0] = 0; m[2][2][1] = 0; m[2][2][2] = 1;
+
+ multiply_matrix(temp, m[rotation_order[0]], m[rotation_order[1]]);
+ multiply_matrix(rot_mat, temp, m[rotation_order[2]]);
+}
+
+static void set_dimensions(int *outw, int *outh, int w, int h, const AVPixFmtDescriptor *desc)
+{
+ outw[1] = outw[2] = FF_CEIL_RSHIFT(w, desc->log2_chroma_w);
+ outw[0] = outw[3] = w;
+ outh[1] = outh[2] = FF_CEIL_RSHIFT(h, desc->log2_chroma_h);
+ outh[0] = outh[3] = h;
+}
+
+static av_cold int init_filter(AVFilterContext *ctx, AVFrame *in)
+{
+ AVFilterLink *outlink = ctx->outputs[0];
+ V360VulkanContext *s = ctx->priv;
+ int err;
+
+ /* Create a sampler */
+ VkSampler *sampler = ff_vk_init_sampler(ctx, 0, VK_FILTER_LINEAR);
+ if (!sampler)
+ return AVERROR_EXTERNAL;
+
+ s->vkctx.queue_family_idx = s->vkctx.hwctx->queue_family_comp_index;
+ s->vkctx.queue_count = GET_QUEUE_COUNT(s->vkctx.hwctx, 0, 1, 0);
+ s->vkctx.cur_queue_idx = av_get_random_seed() % s->vkctx.queue_count;
+
+ s->pl = ff_vk_create_pipeline(ctx);
+ if (!s->pl)
+ return AVERROR(ENOMEM);
+
+ switch (s->out) {
+ case FLAT:
+ s->opts.flat_range[0] = tanf(0.5f * s->h_fov * M_PI / 180.f);
+ s->opts.flat_range[1] = tanf(0.5f * s->v_fov * M_PI / 180.f);
+ break;
+ case STEREOGRAPHIC:
+ s->opts.flat_range[0] = tanf(FFMIN(s->h_fov, 359.f) * M_PI / 720.f);
+ s->opts.flat_range[1] = tanf(FFMIN(s->v_fov, 359.f) * M_PI / 720.f);
+ break;
+ case DUAL_FISHEYE:
+ case FISHEYE:
+ s->opts.flat_range[0] = s->h_fov / 180.f;
+ s->opts.flat_range[1] = s->v_fov / 180.f;
+ break;
+ }
+
+ switch (s->in) {
+ case FLAT:
+ s->opts.iflat_range[0] = tanf(0.5f * s->ih_fov * M_PI / 180.f);
+ s->opts.iflat_range[1] = tanf(0.5f * s->iv_fov * M_PI / 180.f);
+ break;
+ case STEREOGRAPHIC:
+ s->opts.iflat_range[0] = tanf(FFMIN(s->ih_fov, 359.f) * M_PI / 720.f);
+ s->opts.iflat_range[1] = tanf(FFMIN(s->iv_fov, 359.f) * M_PI / 720.f);
+ break;
+ case DUAL_FISHEYE:
+ case FISHEYE:
+ s->opts.iflat_range[0] = s->ih_fov / 180.f;
+ s->opts.iflat_range[1] = s->iv_fov / 180.f;
+ break;
+ }
+
+ s->rotation_order[0] = YAW;
+ s->rotation_order[1] = PITCH;
+ s->rotation_order[2] = ROLL;
+
+ { /* Create the shader */
+ const int planes = av_pix_fmt_count_planes(s->vkctx.output_format);
+ const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(s->vkctx.output_format);
+
+ set_dimensions(s->inplanewidth, s->inplaneheight, in->width, in->height, desc);
+ set_dimensions(s->planewidth, s->planeheight, outlink->w, outlink->h, desc);
+ calculate_rotation_matrix(s->yaw, s->pitch, s->roll, s->opts.rot_mat, s->rotation_order);
+
+ VulkanDescriptorSetBinding desc_i[2] = {
+ {
+ .name = "input_img",
+ .type = VK_DESCRIPTOR_TYPE_COMBINED_IMAGE_SAMPLER,
+ .dimensions = 2,
+ .elems = planes,
+ .stages = VK_SHADER_STAGE_COMPUTE_BIT,
+ .updater = s->input_images,
+ .samplers = DUP_SAMPLER_ARRAY4(*sampler),
+ },
+ {
+ .name = "output_img",
+ .type = VK_DESCRIPTOR_TYPE_STORAGE_IMAGE,
+ .mem_layout = ff_vk_shader_rep_fmt(s->vkctx.output_format),
+ .mem_quali = "writeonly",
+ .dimensions = 2,
+ .elems = planes,
+ .stages = VK_SHADER_STAGE_COMPUTE_BIT,
+ .updater = s->output_images,
+ },
+ };
+
+ SPIRVShader *shd = ff_vk_init_shader(ctx, s->pl, "v360_compute",
+ VK_SHADER_STAGE_COMPUTE_BIT);
+ if (!shd)
+ return AVERROR(ENOMEM);
+
+ ff_vk_set_compute_shader_sizes(ctx, shd, CGROUPS);
+
+ GLSLC(0, layout(push_constant, std430) uniform pushConstants { );
+ GLSLC(1, vec2 flat_range; );
+ GLSLC(1, vec2 iflat_range; );
+ GLSLC(1, mat4 rot_mat; );
+ GLSLC(0, }; );
+ GLSLC(0, );
+
+ ff_vk_add_push_constant(ctx, s->pl, 0, sizeof(s->opts),
+ VK_SHADER_STAGE_COMPUTE_BIT);
+
+ RET(ff_vk_add_descriptor_set(ctx, s->pl, shd, desc_i, 2, 0)); /* set 0 */
+
+ GLSLF(0, #define PI (%f) ,M_PI);
+ GLSLF(0, #define PI_2 (%f) ,M_PI_2);
+ GLSLF(0, #define SQRT2 (%f) ,M_SQRT2);
+ GLSLF(0, #define inplanewidth ivec4(%i, %i, %i, %i), s->inplanewidth[0],
+ s->inplanewidth[1],
+ s->inplanewidth[2],
+ s->inplanewidth[3]);
+ GLSLF(0, #define inplaneheight ivec4(%i, %i, %i, %i), s->inplaneheight[0],
+ s->inplaneheight[1],
+ s->inplaneheight[2],
+ s->inplaneheight[3]);
+
+ switch (s->out) {
+ case FLAT:
+ GLSLD(flat_to_xyz);
+ break;
+ case EQUIRECTANGULAR:
+ GLSLD(equirect_to_xyz);
+ break;
+ case STEREOGRAPHIC:
+ GLSLD(stereographic_to_xyz);
+ break;
+ case FISHEYE:
+ GLSLD(fisheye_to_xyz);
+ break;
+ case DUAL_FISHEYE:
+ GLSLD(dfisheye_to_xyz);
+ break;
+ }
+
+ switch (s->in) {
+ case FLAT:
+ GLSLD(xyz_to_flat);
+ break;
+ case EQUIRECTANGULAR:
+ GLSLD(xyz_to_equirect);
+ break;
+ case STEREOGRAPHIC:
+ GLSLD(xyz_to_stereographic);
+ break;
+ case FISHEYE:
+ GLSLD(xyz_to_fisheye);
+ break;
+ case DUAL_FISHEYE:
+ GLSLD(xyz_to_dfisheye);
+ break;
+ }
+
+ GLSLC(0, void main() );
+ GLSLC(0, { );
+ GLSLC(1, ivec2 pos = ivec2(gl_GlobalInvocationID.xy); );
+ GLSLC(1, vec3 vector; );
+ GLSLF(1, int planes = %i; ,planes);
+ GLSLC(1, for (int i = 0; i < planes; i++) { );
+ GLSLC(2, ivec2 out_size = imageSize(output_img[i]); );
+ GLSLC(2, ivec2 in_size = ivec2(inplanewidth[i],inplaneheight[i]););
+ GLSLC(2, out_transform(vector, out_size, pos); );
+ GLSLC(2, vector = normalize((rot_mat * vec4(vector, 1.f)).xyz); );
+ GLSLC(2, in_transform(i, vector, pos, in_size); );
+ GLSLC(1, } );
+ GLSLC(0, } );
+
+ RET(ff_vk_compile_shader(ctx, shd, "main"));
+ }
+
+ RET(ff_vk_init_pipeline_layout(ctx, s->pl));
+ RET(ff_vk_init_compute_pipeline(ctx, s->pl));
+
+ /* Execution context */
+ RET(ff_vk_create_exec_ctx(ctx, &s->exec));
+
+ s->initialized = 1;
+
+ return 0;
+
+fail:
+ return err;
+}
+
+static int process_frames(AVFilterContext *avctx, AVFrame *out_f, AVFrame *in_f)
+{
+ int err = 0;
+ VkCommandBuffer cmd_buf;
+ V360VulkanContext *s = avctx->priv;
+ AVVkFrame *in = (AVVkFrame *)in_f->data[0];
+ AVVkFrame *out = (AVVkFrame *)out_f->data[0];
+ int planes = av_pix_fmt_count_planes(s->vkctx.output_format);
+
+ /* Update descriptors and init the exec context */
+ ff_vk_start_exec_recording(avctx, s->exec);
+ cmd_buf = ff_vk_get_exec_buf(avctx, s->exec);
+
+ for (int i = 0; i < planes; i++) {
+ RET(ff_vk_create_imageview(avctx, s->exec, &s->input_images[i].imageView, in->img[i],
+ av_vkfmt_from_pixfmt(s->vkctx.input_format)[i],
+ ff_comp_identity_map));
+
+ RET(ff_vk_create_imageview(avctx, s->exec, &s->output_images[i].imageView, out->img[i],
+ av_vkfmt_from_pixfmt(s->vkctx.output_format)[i],
+ ff_comp_identity_map));
+
+ s->input_images[i].imageLayout = VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL;
+ s->output_images[i].imageLayout = VK_IMAGE_LAYOUT_GENERAL;
+ }
+
+ ff_vk_update_descriptor_set(avctx, s->pl, 0);
+
+ for (int i = 0; i < planes; i++) {
+ VkImageMemoryBarrier bar[2] = {
+ {
+ .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+ .srcAccessMask = 0,
+ .dstAccessMask = VK_ACCESS_SHADER_READ_BIT,
+ .oldLayout = in->layout[i],
+ .newLayout = s->input_images[i].imageLayout,
+ .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+ .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+ .image = in->img[i],
+ .subresourceRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT,
+ .subresourceRange.levelCount = 1,
+ .subresourceRange.layerCount = 1,
+ },
+ {
+ .sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER,
+ .srcAccessMask = 0,
+ .dstAccessMask = VK_ACCESS_SHADER_WRITE_BIT,
+ .oldLayout = out->layout[i],
+ .newLayout = s->output_images[i].imageLayout,
+ .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+ .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
+ .image = out->img[i],
+ .subresourceRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT,
+ .subresourceRange.levelCount = 1,
+ .subresourceRange.layerCount = 1,
+ },
+ };
+
+ vkCmdPipelineBarrier(cmd_buf, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT,
+ VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, 0,
+ 0, NULL, 0, NULL, FF_ARRAY_ELEMS(bar), bar);
+
+ in->layout[i] = bar[0].newLayout;
+ in->access[i] = bar[0].dstAccessMask;
+
+ out->layout[i] = bar[1].newLayout;
+ out->access[i] = bar[1].dstAccessMask;
+ }
+
+ ff_vk_bind_pipeline_exec(avctx, s->exec, s->pl);
+
+ ff_vk_update_push_exec(avctx, s->exec, VK_SHADER_STAGE_COMPUTE_BIT,
+ 0, sizeof(s->opts), &s->opts);
+
+ vkCmdDispatch(cmd_buf,
+ FFALIGN(s->vkctx.output_width, CGROUPS[0])/CGROUPS[0],
+ FFALIGN(s->vkctx.output_height, CGROUPS[1])/CGROUPS[1], 1);
+
+ ff_vk_add_exec_dep(avctx, s->exec, in_f, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT);
+ ff_vk_add_exec_dep(avctx, s->exec, out_f, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT);
+
+ err = ff_vk_submit_exec_queue(avctx, s->exec);
+ if (err)
+ return err;
+
+fail:
+ ff_vk_discard_exec_deps(avctx, s->exec);
+ return err;
+}
+
+static int v360_vulkan_filter_frame(AVFilterLink *link, AVFrame *in)
+{
+ int err;
+ AVFilterContext *ctx = link->dst;
+ V360VulkanContext *s = ctx->priv;
+ AVFilterLink *outlink = ctx->outputs[0];
+
+ AVFrame *out = ff_get_video_buffer(outlink, outlink->w, outlink->h);
+ if (!out) {
+ err = AVERROR(ENOMEM);
+ goto fail;
+ }
+
+ if (!s->initialized)
+ RET(init_filter(ctx, in));
+
+ RET(process_frames(ctx, out, in));
+
+ err = av_frame_copy_props(out, in);
+ if (err < 0)
+ goto fail;
+
+ av_frame_free(&in);
+
+ return ff_filter_frame(outlink, out);
+
+fail:
+ av_frame_free(&in);
+ av_frame_free(&out);
+ return err;
+}
+
+static int v360_vulkan_config_output(AVFilterLink *outlink)
+{
+ AVFilterContext *avctx = outlink->src;
+ V360VulkanContext *s = avctx->priv;
+ AVFilterLink *inlink = avctx->inputs[0];
+ int err;
+
+ if (s->width > 0 && s->height > 0) {
+ s->vkctx.output_width = s->width;
+ s->vkctx.output_height = s->height;
+ }
+
+ s->vkctx.output_format = s->vkctx.input_format;
+
+ err = ff_vk_filter_config_output(outlink);
+ if (err < 0)
+ return err;
+
+ outlink->sample_aspect_ratio = inlink->sample_aspect_ratio;
+
+ return 0;
+}
+
+static void v360_vulkan_uninit(AVFilterContext *avctx)
+{
+ V360VulkanContext *s = avctx->priv;
+
+ ff_vk_filter_uninit(avctx);
+
+ s->initialized = 0;
+}
+
+#define OFFSET(x) offsetof(V360VulkanContext, x)
+#define FLAGS (AV_OPT_FLAG_FILTERING_PARAM | AV_OPT_FLAG_VIDEO_PARAM)
+static const AVOption v360_vulkan_options[] = {
+ { "input", "set input projection", OFFSET(in), AV_OPT_TYPE_INT, {.i64=EQUIRECTANGULAR}, 0, NB_PROJECTIONS-1, FLAGS, "in" },
+ { "e", "equirectangular", 0, AV_OPT_TYPE_CONST, {.i64=EQUIRECTANGULAR}, 0, 0, FLAGS, "in" },
+ { "equirect", "equirectangular", 0, AV_OPT_TYPE_CONST, {.i64=EQUIRECTANGULAR}, 0, 0, FLAGS, "in" },
+ { "flat", "regular video", 0, AV_OPT_TYPE_CONST, {.i64=FLAT}, 0, 0, FLAGS, "in" },
+ { "dfisheye", "dual fisheye", 0, AV_OPT_TYPE_CONST, {.i64=DUAL_FISHEYE}, 0, 0, FLAGS, "in" },
+ { "sg", "stereographic", 0, AV_OPT_TYPE_CONST, {.i64=STEREOGRAPHIC}, 0, 0, FLAGS, "in" },
+ { "fisheye", "fisheye", 0, AV_OPT_TYPE_CONST, {.i64=FISHEYE} , 0, 0, FLAGS, "in" },
+ { "output", "set output projection", OFFSET(out), AV_OPT_TYPE_INT, {.i64=FLAT}, 0, NB_PROJECTIONS-1, FLAGS, "out" },
+ { "e", "equirectangular", 0, AV_OPT_TYPE_CONST, {.i64=EQUIRECTANGULAR}, 0, 0, FLAGS, "out" },
+ { "equirect", "equirectangular", 0, AV_OPT_TYPE_CONST, {.i64=EQUIRECTANGULAR}, 0, 0, FLAGS, "out" },
+ { "flat", "regular video", 0, AV_OPT_TYPE_CONST, {.i64=FLAT}, 0, 0, FLAGS, "out" },
+ { "dfisheye", "dual fisheye", 0, AV_OPT_TYPE_CONST, {.i64=DUAL_FISHEYE}, 0, 0, FLAGS, "out" },
+ { "sg", "stereographic", 0, AV_OPT_TYPE_CONST, {.i64=STEREOGRAPHIC}, 0, 0, FLAGS, "out" },
+ { "fisheye", "fisheye", 0, AV_OPT_TYPE_CONST, {.i64=FISHEYE} , 0, 0, FLAGS, "out" },
+ { "w", "output width", OFFSET(width), AV_OPT_TYPE_INT, {.i64=0}, 0, INT16_MAX, FLAGS, "w"},
+ { "h", "output height", OFFSET(height), AV_OPT_TYPE_INT, {.i64=0}, 0, INT16_MAX, FLAGS, "h"},
+ { "yaw", "yaw rotation", OFFSET(yaw), AV_OPT_TYPE_FLOAT, {.dbl=0.f}, -180.f, 180.f, FLAGS, "yaw"},
+ { "pitch", "pitch rotation", OFFSET(pitch), AV_OPT_TYPE_FLOAT, {.dbl=0.f}, -180.f, 180.f, FLAGS, "pitch"},
+ { "roll", "roll rotation", OFFSET(roll), AV_OPT_TYPE_FLOAT, {.dbl=0.f}, -180.f, 180.f, FLAGS, "roll"},
+ { "rorder", "rotation order", OFFSET(rorder), AV_OPT_TYPE_STRING, {.str="ypr"}, 0, 0, FLAGS, "rorder"},
+ { "h_fov", "set output horizontal FOV angle", OFFSET(h_fov), AV_OPT_TYPE_FLOAT, {.dbl = 90.0f}, 0.00001f, 360.0f, .flags = FLAGS },
+ { "v_fov", "set output vertical FOV angle", OFFSET(v_fov), AV_OPT_TYPE_FLOAT, {.dbl = 45.0f}, 0.00001f, 360.0f, .flags = FLAGS },
+ { "ih_fov", "set input horizontal FOV angle", OFFSET(ih_fov), AV_OPT_TYPE_FLOAT, {.dbl = 90.0f}, 0.00001f, 360.0f, .flags = FLAGS },
+ { "iv_fov", "set input vertical FOV angle", OFFSET(iv_fov), AV_OPT_TYPE_FLOAT, {.dbl = 45.0f}, 0.00001f, 360.0f, .flags = FLAGS },
+ { NULL },
+};
+
+AVFILTER_DEFINE_CLASS(v360_vulkan);
+
+static const AVFilterPad v360_vulkan_inputs[] = {
+ {
+ .name = "default",
+ .type = AVMEDIA_TYPE_VIDEO,
+ .filter_frame = &v360_vulkan_filter_frame,
+ .config_props = &ff_vk_filter_config_input,
+ },
+ { NULL }
+};
+
+static const AVFilterPad v360_vulkan_outputs[] = {
+ {
+ .name = "default",
+ .type = AVMEDIA_TYPE_VIDEO,
+ .config_props = &v360_vulkan_config_output,
+ },
+ { NULL }
+};
+
+AVFilter ff_vf_v360_vulkan = {
+ .name = "v360_vulkan",
+ .description = NULL_IF_CONFIG_SMALL("Convert 360 projection of video."),
+ .priv_size = sizeof(V360VulkanContext),
+ .init = &ff_vk_filter_init,
+ .uninit = &v360_vulkan_uninit,
+ .query_formats = &ff_vk_filter_query_formats,
+ .inputs = v360_vulkan_inputs,
+ .outputs = v360_vulkan_outputs,
+ .priv_class = &v360_vulkan_class,
+ .flags_internal = FF_FILTER_FLAG_HWFRAME_AWARE,
+};
--
2.17.1
