sse2 integer IDCT. not the fastest possible implementation but it produces bit-identical results to the generic C version so it's fully "transparent".

* @brief Store a vector to an unaligned memory address. */

* @brief Store a vector to an unaligned memory address. */

* @brief Save a BMP image using STBImageWrite to provide the store routine. * * @param img The source data for the image. * @param filename

* @brief Save a DDS uncompressed image using a local store routine. * * @param img The source data for the image. * @param filename The na

* @brief Save an EXR image using TinyExr to provide the store routine. * * @param img The source data for the image. * @param filename The

* @brief Save a HDR image using STBImageWrite to provide the store routine. * * @param img The source data for the image. * @param filename

* @brief Save a KTX uncompressed image using a local store routine. * * @param img The source data for the image. * @param filename The na

* @brief Store a vector, skipping masked lanes. * * All masked lanes must be at the end of vector, after all non-masked lanes. * Input is a byte ar

* @brief Store a vector, skipping masked lanes. * * All masked lanes must be at the end of vector, after all non-masked lanes. */

* @brief Store a vector, skipping masked lanes. * * All masked lanes must be at the end of vector, after all non-masked lanes. */

* @brief Store a vector, skipping masked lanes. * * All masked lanes must be at the end of vector, after all non-masked lanes. */

* @brief Store a vector, skipping masked lanes. * * All masked lanes must be at the end of vector, after all non-masked lanes. */

* @brief Store lowest N (vector width) bytes into an unaligned address. */

* @brief Save a PNG image using STBImageWrite to provide the store routine. * * @param img The source data for the image. * @param filename

* @brief Save a TGA image using STBImageWrite to provide the store routine. * * @param img The source data for the image. * @param filename

* @brief Store a vector to an aligned memory address. */

* @brief Store a vector to a 16B aligned memory address. */

* @brief Swizzle a texel into a new arrangement. * * @param data The source RGBA vector to swizzle. * @param swz The swizzle to use. */

* @brief Dummy no-op swizzle function. * * @param data The source RGBA vector to swizzle. * @param swz The swizzle to use. */

Tear down a test case. Clean up the temporary directory created for output files.

Get current position

Test 2 partition partitioning candidate limit.

Test 2 partition early limit.

Test partition index limit.

Test 2 plane correlation limit.

Test 3 partition partitioning candidate limit.

Test 3 partition early limit.

Test partition index limit.

Test 4 partition partitioning candidate limit.

Test partition index limit.

Test block mode limit.

Test candidate limit.

Test -ch with -mpsnr and missing arguments.

Test channel weighting.

Test -cl with a 2D block size and 3D input data.

Test -cl with -2partitioncandidatelimit and missing arguments.

Test -cl with -2partitionlimitfactor and missing arguments.

Test -cl with -2partitionindexlimit and missing arguments.

Test -cl with -2planelimitcorrelation and missing arguments.

Test -cl with -3partitioncandidatelimit and missing arguments.

Test -cl with -3partitionlimitfactor and missing arguments.

Test -cl with -3partitionindexlimit and missing arguments.

Test -cl with -4partitioncandidatelimit and missing arguments.

Test -cl with -4partitionindexlimit and missing arguments.

Test -cl with -a and missing arguments.

Test -cl with a 2D block size and 3D input data.