nir/builder: Add a nir_extract_bits helper

This new helper is better than nir_bitcast_vector because it's able to
take a (mostly) arbitrary range from the source vector.  The only
requirement is that first_bit has to be aligned to the smaller of the
two bit sizes.  It wouldn't be hard to lift that requirement but it's
reasonable for now.

Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
This commit is contained in:
Jason Ekstrand
2019-11-06 12:09:56 -06:00
parent 86d3a346f1
commit d0bbf98c96

View File

@@ -757,6 +757,85 @@ nir_unpack_bits(nir_builder *b, nir_ssa_def *src, unsigned dest_bit_size)
return nir_vec(b, dest_comps, dest_num_components);
}
/**
* Treats srcs as if it's one big blob of bits and extracts the range of bits
* given by
*
* [first_bit, first_bit + dest_num_components * dest_bit_size)
*
* The range can have any alignment or size as long as it's an integer number
* of destination components and fits inside the concatenated sources.
*
* TODO: The one caveat here is that we can't handle byte alignment if 64-bit
* values are involved because that would require pack/unpack to/from a vec8
* which NIR currently does not support.
*/
static inline nir_ssa_def *
nir_extract_bits(nir_builder *b, nir_ssa_def **srcs, unsigned num_srcs,
unsigned first_bit,
unsigned dest_num_components, unsigned dest_bit_size)
{
const unsigned num_bits = dest_num_components * dest_bit_size;
/* Figure out the common bit size */
unsigned common_bit_size = dest_bit_size;
for (unsigned i = 0; i < num_srcs; i++)
common_bit_size = MIN2(common_bit_size, srcs[i]->bit_size);
if (first_bit > 0)
common_bit_size = MIN2(common_bit_size, (1 << (ffs(first_bit) - 1)));
/* We don't want to have to deal with 1-bit values */
assert(common_bit_size >= 8);
nir_ssa_def *common_comps[NIR_MAX_VEC_COMPONENTS * sizeof(uint64_t)];
assert(num_bits / common_bit_size <= ARRAY_SIZE(common_comps));
/* First, unpack to the common bit size and select the components from the
* source.
*/
int src_idx = -1;
unsigned src_start_bit = 0;
unsigned src_end_bit = 0;
for (unsigned i = 0; i < num_bits / common_bit_size; i++) {
const unsigned bit = first_bit + (i * common_bit_size);
while (bit >= src_end_bit) {
src_idx++;
assert(src_idx < num_srcs);
src_start_bit = src_end_bit;
src_end_bit += srcs[src_idx]->bit_size *
srcs[src_idx]->num_components;
}
assert(bit >= src_start_bit);
assert(bit + common_bit_size <= src_end_bit);
const unsigned rel_bit = bit - src_start_bit;
const unsigned src_bit_size = srcs[src_idx]->bit_size;
nir_ssa_def *comp = nir_channel(b, srcs[src_idx],
rel_bit / src_bit_size);
if (srcs[src_idx]->bit_size > common_bit_size) {
nir_ssa_def *unpacked = nir_unpack_bits(b, comp, common_bit_size);
comp = nir_channel(b, unpacked, (rel_bit % src_bit_size) /
common_bit_size);
}
common_comps[i] = comp;
}
/* Now, re-pack the destination if we have to */
if (dest_bit_size > common_bit_size) {
unsigned common_per_dest = dest_bit_size / common_bit_size;
nir_ssa_def *dest_comps[NIR_MAX_VEC_COMPONENTS];
for (unsigned i = 0; i < dest_num_components; i++) {
nir_ssa_def *unpacked = nir_vec(b, common_comps + i * common_per_dest,
common_per_dest);
dest_comps[i] = nir_pack_bits(b, unpacked, dest_bit_size);
}
return nir_vec(b, dest_comps, dest_num_components);
} else {
assert(dest_bit_size == common_bit_size);
return nir_vec(b, common_comps, dest_num_components);
}
}
static inline nir_ssa_def *
nir_bitcast_vector(nir_builder *b, nir_ssa_def *src, unsigned dest_bit_size)
{
@@ -765,43 +844,7 @@ nir_bitcast_vector(nir_builder *b, nir_ssa_def *src, unsigned dest_bit_size)
(src->bit_size * src->num_components) / dest_bit_size;
assert(dest_num_components <= NIR_MAX_VEC_COMPONENTS);
if (src->bit_size > dest_bit_size) {
assert(src->bit_size % dest_bit_size == 0);
if (src->num_components == 1) {
return nir_unpack_bits(b, src, dest_bit_size);
} else {
const unsigned divisor = src->bit_size / dest_bit_size;
assert(src->num_components * divisor == dest_num_components);
nir_ssa_def *dest[NIR_MAX_VEC_COMPONENTS];
for (unsigned i = 0; i < src->num_components; i++) {
nir_ssa_def *unpacked =
nir_unpack_bits(b, nir_channel(b, src, i), dest_bit_size);
assert(unpacked->num_components == divisor);
for (unsigned j = 0; j < divisor; j++)
dest[i * divisor + j] = nir_channel(b, unpacked, j);
}
return nir_vec(b, dest, dest_num_components);
}
} else if (src->bit_size < dest_bit_size) {
assert(dest_bit_size % src->bit_size == 0);
if (dest_num_components == 1) {
return nir_pack_bits(b, src, dest_bit_size);
} else {
const unsigned divisor = dest_bit_size / src->bit_size;
assert(src->num_components == dest_num_components * divisor);
nir_ssa_def *dest[NIR_MAX_VEC_COMPONENTS];
for (unsigned i = 0; i < dest_num_components; i++) {
nir_component_mask_t src_mask =
((1 << divisor) - 1) << (i * divisor);
dest[i] = nir_pack_bits(b, nir_channels(b, src, src_mask),
dest_bit_size);
}
return nir_vec(b, dest, dest_num_components);
}
} else {
assert(src->bit_size == dest_bit_size);
return src;
}
return nir_extract_bits(b, &src, 1, 0, dest_num_components, dest_bit_size);
}
/**