Pad etcpak input to 4x4 blocks. Fixes #49981

This commit is contained in:
Lyuma 2022-01-18 20:29:39 -08:00
parent f8b8d0d4b8
commit b1b20f5f91
2 changed files with 59 additions and 9 deletions

View File

@ -1434,12 +1434,11 @@ int Image::_get_dst_image_size(int p_width, int p_height, Format p_format, int &
}
// Set mipmap size.
// It might be necessary to put this after the minimum mipmap size check because of the possible occurrence of "1 >> 1".
if (r_mm_width) {
*r_mm_width = bw >> 1;
*r_mm_width = w;
}
if (r_mm_height) {
*r_mm_height = bh >> 1;
*r_mm_height = h;
}
// Reach target mipmap.

View File

@ -132,8 +132,35 @@ void _compress_etcpak(EtcpakType p_compresstype, Image *r_img, float p_lossy_qua
// Compress image data and (if required) mipmaps.
const bool mipmaps = r_img->has_mipmaps();
const int width = r_img->get_width();
const int height = r_img->get_height();
int width = r_img->get_width();
int height = r_img->get_height();
/*
The first mipmap level of a compressed texture must be a multiple of 4. Quote from D3D11.3 spec:
BC format surfaces are always multiples of full blocks, each block representing 4x4 pixels.
For mipmaps, the top level map is required to be a multiple of 4 size in all dimensions.
The sizes for the lower level maps are computed as they are for all mipmapped surfaces,
and thus may not be a multiple of 4, for example a top level map of 20 results in a second level
map size of 10. For these cases, there is a differing 'physical' size and a 'virtual' size.
The virtual size is that computed for each mip level without adjustment, which is 10 for the example.
The physical size is the virtual size rounded up to the next multiple of 4, which is 12 for the example,
and this represents the actual memory size. The sampling hardware will apply texture address
processing based on the virtual size (using, for example, border color if specified for accesses
beyond 10), and thus for the example case will not access the 11th and 12th row of the resource.
So for mipmap chains when an axis becomes < 4 in size, only texels 'a','b','e','f'
are used for a 2x2 map, and texel 'a' is used for 1x1. Note that this is similar to, but distinct from,
the surface pitch, which can encompass additional padding beyond the physical surface size.
*/
int next_width = (width + 3) & ~3;
int next_height = (height + 3) & ~3;
if (next_width != width || next_height != height) {
r_img->resize(next_width, next_height, Image::INTERPOLATE_LANCZOS);
width = r_img->get_width();
height = r_img->get_height();
}
ERR_FAIL_COND(width % 4 != 0 || height % 4 != 0); // Should be guaranteed by above
const uint8_t *src_read = r_img->get_data().ptr();
print_verbose(vformat("ETCPAK: Encoding image size %dx%d to format %s.", width, height, Image::get_format_name(target_format)));
@ -144,24 +171,48 @@ void _compress_etcpak(EtcpakType p_compresstype, Image *r_img, float p_lossy_qua
uint8_t *dest_write = dest_data.ptrw();
int mip_count = mipmaps ? Image::get_image_required_mipmaps(width, height, target_format) : 0;
Vector<uint32_t> padded_src;
for (int i = 0; i < mip_count + 1; i++) {
// Get write mip metrics for target image.
int mip_w, mip_h;
int mip_ofs = Image::get_image_mipmap_offset_and_dimensions(width, height, target_format, i, mip_w, mip_h);
int orig_mip_w, orig_mip_h;
int mip_ofs = Image::get_image_mipmap_offset_and_dimensions(width, height, target_format, i, orig_mip_w, orig_mip_h);
// Ensure that mip offset is a multiple of 8 (etcpak expects uint64_t pointer).
ERR_FAIL_COND(mip_ofs % 8 != 0);
uint64_t *dest_mip_write = (uint64_t *)&dest_write[mip_ofs];
// Block size. Align stride to multiple of 4 (RGBA8).
mip_w = (mip_w + 3) & ~3;
mip_h = (mip_h + 3) & ~3;
int mip_w = (orig_mip_w + 3) & ~3;
int mip_h = (orig_mip_h + 3) & ~3;
const uint32_t blocks = mip_w * mip_h / 16;
// Get mip data from source image for reading.
int src_mip_ofs = r_img->get_mipmap_offset(i);
const uint32_t *src_mip_read = (const uint32_t *)&src_read[src_mip_ofs];
// Pad textures to nearest block by smearing.
if (mip_w != orig_mip_w || mip_h != orig_mip_h) {
padded_src.resize(mip_w * mip_h);
uint32_t *ptrw = padded_src.ptrw();
int x = 0, y = 0;
for (y = 0; y < orig_mip_h; y++) {
for (x = 0; x < orig_mip_w; x++) {
ptrw[mip_w * y + x] = src_mip_read[orig_mip_w * y + x];
}
// First, smear in x.
for (; x < mip_w; x++) {
ptrw[mip_w * y + x] = ptrw[mip_w * y + x - 1];
}
}
// Then, smear in y.
for (; y < mip_h; y++) {
for (x = 0; x < mip_w; x++) {
ptrw[mip_w * y + x] = ptrw[mip_w * y + x - mip_w];
}
}
// Override the src_mip_read pointer to our temporary Vector.
src_mip_read = padded_src.ptr();
}
if (p_compresstype == EtcpakType::ETCPAK_TYPE_ETC1) {
CompressEtc1RgbDither(src_mip_read, dest_mip_write, blocks, mip_w);
} else if (p_compresstype == EtcpakType::ETCPAK_TYPE_ETC2 || p_compresstype == EtcpakType::ETCPAK_TYPE_ETC2_RA_AS_RG) {