fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-01-02 18:10:17 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

broadcom/compiler: use ldunifa with unaligned constant offset

Browse source 
If we know we have a load with a constant offset, then even if it
is not aligned to 32-bit we can still produce an aligned offset
and then skip over the bytes we don't need.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/14648>
This commit is contained in:
Iago Toral Quiroga 2022-01-18 12:02:36 +01:00 • committed by Marge Bot
parent 2eb6910d96
commit bdb6201ea1
1 changed files with 31 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

38
src/broadcom/compiler/nir_to_vir.c
View file

@ -2785,8 +2785,22 @@ ntq_emit_load_ubo_unifa(struct v3d_compile *c, nir_intrinsic_instr *instr)
bool dynamic_src = !nir_src_is_const(instr->src[1]);
uint32_t const_offset =
dynamic_src ? 0 : nir_src_as_uint(instr->src[1]);
if (bit_size < 32 && (dynamic_src || (const_offset % 4) != 0))
return false;
uint32_t value_skips = 0;
if (bit_size < 32) {
if (dynamic_src) {
return false;
} else if (const_offset % 4 != 0) {
/* If we are loading from an unaligned offset, fix
* alignment and skip over unused elements in result.
*/
value_skips = (const_offset % 4) / (bit_size / 8);
const_offset &= ~0xf;
}
}
assert((bit_size == 32 && value_skips == 0) ||
(bit_size == 16 && value_skips <= 1) ||
(bit_size == 8 && value_skips <= 3));
/* On OpenGL QUNIFORM_UBO_ADDR takes a UBO index
* shifted up by 1 (0 is gallium's constant buffer 0).
@ -2835,19 +2849,29 @@ ntq_emit_load_ubo_unifa(struct v3d_compile *c, nir_intrinsic_instr *instr)
}
uint32_t num_components = nir_intrinsic_dest_components(instr);
for (uint32_t i = 0; i < num_components; i++) {
for (uint32_t i = 0; i < num_components; ) {
struct qreg data;
emit_ldunifa(c, &data);
if (bit_size == 32) {
assert(value_skips == 0);
ntq_store_dest(c, &instr->dest, i, vir_MOV(c, data));
i++;
} else {
assert(bit_size == 16);
struct qreg tmp = vir_AND(c, vir_MOV(c, data),
vir_uniform_ui(c, 0xffff));
ntq_store_dest(c, &instr->dest, i, vir_MOV(c, tmp));
i++;
assert(value_skips <= 1);
struct qreg tmp;
if (value_skips == 0) {
tmp = vir_AND(c, vir_MOV(c, data),
vir_uniform_ui(c, 0xffff));
ntq_store_dest(c, &instr->dest, i, vir_MOV(c, tmp));
i++;
} else {
value_skips--;
}
assert(value_skips == 0);
if (i < num_components) {
tmp = vir_SHR(c, data, vir_uniform_ui(c, 16));