AlexIndustrial/mesa
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

v3d: Add a fallthrough path for utile load/store of 32 byte lines.

Browse Source 
Now that V3D has 8 byte per pixel formats exposed, we've got stride==32
utiles to load and store.  Just handle them through the non-NEON paths for
now.
This commit is contained in:
Eric Anholt
2018-12-17 12:20:41 -08:00
parent f6a0f4f41e
commit 7c56b7a6ea
1 changed files with 16 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/broadcom/common/v3d_cpu_tiling.h
+16 -12
View File
@@ -52,8 +52,8 @@ v3d_load_utile(void *cpu, uint32_t cpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu_stride)
: "q0", "q1", "q2", "q3");
} else {
assert(gpu_stride == 16);
return;
} else if (gpu_stride == 16) {
__asm__ volatile (
/* Load from the GPU in one shot, no interleave, to
* d0-d7.
@@ -74,6 +74,7 @@ v3d_load_utile(void *cpu, uint32_t cpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu + 8), "r"(cpu_stride)
: "q0", "q1", "q2", "q3");
return;
}
#elif defined (PIPE_ARCH_AARCH64)
if (gpu_stride == 8) {
@@ -96,8 +97,8 @@ v3d_load_utile(void *cpu, uint32_t cpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu_stride)
: "v0", "v1", "v2", "v3");
} else {
assert(gpu_stride == 16);
return;
} else if (gpu_stride == 16) {
__asm__ volatile (
/* Load from the GPU in one shot, no interleave, to
* d0-d7.
@@ -118,13 +119,14 @@ v3d_load_utile(void *cpu, uint32_t cpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu + 8), "r"(cpu_stride)
: "v0", "v1", "v2", "v3");
return;
}
#else
#endif
for (uint32_t gpu_offset = 0; gpu_offset < 64; gpu_offset += gpu_stride) {
memcpy(cpu, gpu + gpu_offset, gpu_stride);
cpu += cpu_stride;
}
#endif
}
static inline void
@@ -152,8 +154,8 @@ v3d_store_utile(void *gpu, uint32_t gpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu_stride)
: "q0", "q1", "q2", "q3");
} else {
assert(gpu_stride == 16);
return;
} else if (gpu_stride == 16) {
__asm__ volatile (
/* Load each 16-byte line in 2 parts from the cpu-side
* destination. (vld1 can only store one d-register
@@ -172,6 +174,7 @@ v3d_store_utile(void *gpu, uint32_t gpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu + 8), "r"(cpu_stride)
: "q0", "q1", "q2", "q3");
return;
}
#elif defined (PIPE_ARCH_AARCH64)
if (gpu_stride == 8) {
@@ -192,8 +195,8 @@ v3d_store_utile(void *gpu, uint32_t gpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu_stride)
: "v0", "v1", "v2", "v3");
} else {
assert(gpu_stride == 16);
return;
} else if (gpu_stride == 16) {
__asm__ volatile (
/* Load each 16-byte line in 2 parts from the cpu-side
* destination. (vld1 can only store one d-register
@@ -212,11 +215,12 @@ v3d_store_utile(void *gpu, uint32_t gpu_stride,
:
: "r"(gpu), "r"(cpu), "r"(cpu + 8), "r"(cpu_stride)
: "v0", "v1", "v2", "v3");
return;
}
#else
#endif
for (uint32_t gpu_offset = 0; gpu_offset < 64; gpu_offset += gpu_stride) {
memcpy(gpu + gpu_offset, cpu, gpu_stride);
cpu += cpu_stride;
}
#endif
}