mirror-github-bspeice/cuburn
1
0
Fork 0
mirror of https://github.com/stevenrobertson/cuburn.git synced 2025-02-05 11:40:04 -05:00
Code Issues Actions Packages Projects Releases Wiki Activity

Deferred works again. Time to break it.

Browse Source
This commit is contained in:
Steven Robertson 2011-12-08 15:28:10 -05:00
parent e106524701
commit b76208078f
3 changed files with 118 additions and 60 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

154
cuburn/code/iter.py
View File

@ -363,9 +363,9 @@ void iter(
pix.w += 1.0f; pix.w += 1.0f;
*accbuf = pix; *accbuf = pix;
{{elif info.acc_mode == 'deferred'}} {{elif info.acc_mode == 'deferred'}}
// 'color' gets the top 9 bits. TODO: add dithering via precalc. // 'color' gets the top 8 bits. TODO: add dithering via precalc.
uint32_t icolor = fminf(1.0f, cc) * 511.0f + color_dither; uint32_t icolor = fminf(1.0f, cc) * 255.0f + color_dither;
asm("bfi.b32 %0, %1, %0, 23, 9;" : "+r"(i) : "r"(icolor)); asm("bfi.b32 %0, %1, %0, 24, 8;" : "+r"(i) : "r"(icolor));
*log = i; *log = i;
{{endif}} {{endif}}
} }
@ -394,10 +394,10 @@ void write_shmem_helper(
const uint32_t gb const uint32_t gb
) { ) {
float4 pix = acc[glo_idx]; float4 pix = acc[glo_idx];
pix.x += (dr & 0xffff) / 255.0f; pix.x += (dr & 0xffff) / 127.0f;
pix.w += dr >> 16; pix.w += dr >> 16;
pix.y += (gb & 0xffff) / 255.0f; pix.y += (gb & 0xffff) / 127.0f;
pix.z += (gb >> 16) / 255.0f; pix.z += (gb >> 16) / 127.0f;
acc[glo_idx] = pix; acc[glo_idx] = pix;
} }
@ -417,8 +417,7 @@ __launch_bounds__(BS, 1)
write_shmem( write_shmem(
float4 *acc, float4 *acc,
const uint32_t *log, const uint32_t *log,
const uint32_t *log_bounds, const uint32_t *log_bounds
const int log_bounds_shift
) { ) {
const int tid = threadIdx.x; const int tid = threadIdx.x;
const int bid = blockIdx.x; const int bid = blockIdx.x;
@ -441,62 +440,94 @@ write_shmem(
s_acc_gb[tid+3*BS] = 0; s_acc_gb[tid+3*BS] = 0;
__syncthreads(); __syncthreads();
// TODO: share across threads - discernable performance impact? // This predicate is used for the horrible monkey-patching magic.
int lb_idx_lo, lb_idx_hi; asm volatile(".reg .pred p; setp.lt.u32 p, %0, 42;" :: "r"(s_acc_dr[0]));
if (log_bounds_shift > 0) {
lb_idx_hi = ((bid + 1) << log_bounds_shift) - 1;
lb_idx_lo = (bid << log_bounds_shift) - 1;
} else {
lb_idx_hi = bid >> (-log_bounds_shift);
lb_idx_lo = lb_idx_hi - 1;
}
int idx_lo, idx_hi; // log_bounds[] holds inclusive prefix sums, so that log_bounds[0] is the
if (lb_idx_lo < 0) idx_lo = 0; // largest index with radix 0, and so on.
else idx_lo = log_bounds[lb_idx_lo] & ~(BS-1); int lb_idx_hi = bid & 0xff;
idx_hi = (log_bounds[lb_idx_hi] & ~(BS - 1)) + BS; int lb_idx_lo = lb_idx_hi - 1;
int idx_lo;
if (lb_idx_lo > 0) idx_lo = log_bounds[lb_idx_lo] & ~(BS - 1);
else idx_lo = 0;
int idx_hi = (log_bounds[lb_idx_hi] & ~(BS - 1)) + BS;
float rnrounds = 1.0f / (idx_hi - idx_lo); float rnrounds = 1.0f / (idx_hi - idx_lo);
float time = tid * rnrounds; float time = tid * rnrounds;
float time_step = BS * rnrounds; float time_step = BS * rnrounds;
int glo_base = bid << SHAB; int glo_base = bid << SHAB;
float4* glo_ptr = &acc[glo_base];
for (int i = idx_lo + tid; i < idx_hi; i += BS) { for (int i = idx_lo + tid; i < idx_hi; i += BS) {
int entry = log[i]; int entry = log[i];
// Constant '12' is 32 - 8 - SHAB, where 8 is the
// TODO: constant '11' is really just 32 - 9 - SHAB, where 9 is the // number of bits assigned to color. TODO: This ignores opacity.
// number of bits assigned to color. This ignores opacity. bfe_decl(glob_addr, entry, SHAB, 12);
bfe_decl(glob_addr, entry, SHAB, 11);
if (glob_addr != bid) continue; if (glob_addr != bid) continue;
bfe_decl(shr_addr, entry, 0, SHAB); // Shared memory address, pre-shifted
bfe_decl(color, entry, 23, 9); int shr_addr;
asm("bfi.b32 %0, %1, 0, 2, 12;" : "=r"(shr_addr) : "r"(entry));
bfe_decl(color, entry, 24, 8);
float colorf = color / 511.0f; float colorf = color / 255.0f;
float4 outcol = tex2D(palTex, colorf, time); float4 outcol = tex2D(palTex, colorf, time);
// TODO: change texture sampler to return shorts and avoid this // TODO: change texture sampler to return shorts and avoid this
uint32_t r = 255.0f * outcol.x; uint32_t r = 127.0f * outcol.x;
uint32_t g = 255.0f * outcol.y; uint32_t g = 127.0f * outcol.y;
uint32_t b = 255.0f * outcol.z; uint32_t b = 127.0f * outcol.z;
uint32_t dr = atomicAdd(s_acc_dr + shr_addr, r + 0x10000); uint32_t dr = r + 0x10000, gb = g + (b << 16);
uint32_t gb = atomicAdd(s_acc_gb + shr_addr, g + (b << 16));
uint32_t d = dr >> 16;
// Neat trick: if overflow is about to happen, write the accumulator, asm volatile ({{crep("""
// and subtract the last known values from the accumulator again. {
// Even if the ints end up wrapping around once before the subtraction .reg .pred q;
// can occur, the results after the subtraction will be correct. .reg .u32 d, r, g, b, dr, gb, drw, gbw, off;
// (Wrapping twice will mess up the intermediate write, but is pretty .reg .u64 ptr;
// unlikely.) .reg .f32 rf, gf, bf, df;
if (d == 250) {
atomicSub(s_acc_dr + shr_addr, dr); acc_write_start:
atomicSub(s_acc_gb + shr_addr, gb); // This instruction will get replaced with a LDSLK that sets 'p'.
write_shmem_helper(acc, glo_base + shr_addr, dr, gb); // The 0xffff is a signature to make sure we get the right instruction,
} // and will get replaced with a 0-offset when patching.
ld.shared.volatile.u32 dr, [%2+0xffff];
@p ld.shared.volatile.u32 gb, [%2+0x4000];
@p add.u32 %0, dr, %0;
@p add.u32 %1, gb, %1;
// TODO: clever use of slct could remove an instruction?
@p setp.lo.u32 q, %0, 32768000;
@p selp.b32 drw, %0, 0, q;
@p selp.b32 gbw, %1, 0, q;
@p st.shared.volatile.u32 [%2+0x4000], gbw;
// This instruction will get replaced with an STSUL
@p st.shared.volatile.u32 [%2+0xffff], drw;
@!p bra acc_write_start;
@q bra oflow_write_end;
shl.b32 %2, %2, 2;
cvt.u64.u32 ptr, %2;
add.u64 ptr, ptr, %3;
and.b32 r, %0, 0xffff;
shr.b32 g, %1, 16;
and.b32 b, %1, 0xffff;
cvt.rn.f32.u32 rf, r;
cvt.rn.f32.u32 gf, g;
cvt.rn.f32.u32 bf, b;
mul.ftz.f32 rf, rf, 0.007874015748031496;
mul.ftz.f32 gf, gf, 0.007874015748031496;
mul.ftz.f32 bf, bf, 0.007874015748031496;
red.add.f32 [ptr], 500.0;
red.add.f32 [ptr+4], bf;
red.add.f32 [ptr+8], gf;
red.add.f32 [ptr+12], rf;
oflow_write_end:
}
""")}} :: "r"(dr), "r"(gb), "r"(shr_addr), "l"(glo_ptr));
// TODO: go through the pain of manual address calculation for global ptr
time += time_step; time += time_step;
} }
@ -519,3 +550,36 @@ write_shmem(
if n != 'final'], if n != 'final'],
**globals()) **globals())
@staticmethod
def monkey_patch(cubin):
LD = np.uint64(0x851c00fcff0300c1)
LDSLK = np.uint64(0x851c0000000000c4)
ST = np.uint64(0x850000fcff0300c9)
STSUL = np.uint64(0x85000000000000cc)
regmask = np.uint64(0x00c0ff0300000000)
prdmask = np.uint64(0x003c000000000000)
O = 64 # Expected offset to last instruction
offset = cubin.find('\x85')
while offset >= 0:
# Using these fixed offsets makes this code intentionally
# intolerant of compiler instruction reordering
if cubin[offset+7] == '\xc1' and cubin[offset+O] == '\x85':
ld = np.frombuffer(cubin[offset:offset+8], dtype='>u8')
st = np.frombuffer(cubin[offset+O:offset+8+O], dtype='>u8')
if ((ld & (~regmask)) == LD and
((st & (~regmask)) & (~prdmask)) == ST):
break
offset = cubin.find('\x85', offset+1)
assert offset > 0, 'Could not find patch point!'
# Note that these bits are still reversed, and we ignore the
# (im)possibility of a negative predicate in this case
pred = (st & prdmask) >> 50
ld = LDSLK | (ld & regmask) | (pred << 10)
st = STSUL | (st & regmask) | (st & prdmask)
return ( cubin[:offset] + ld.byteswap().tostring()
+ cubin[offset+8:offset+O]
+ st.byteswap().tostring() + cubin[offset+8+O:] )

2
cuburn/genome.py
View File

@ -131,7 +131,7 @@ class RenderInfo(object):
# position into a single 32-bit int for now, which limits resolution to # position into a single 32-bit int for now, which limits resolution to
# 1080p when xform opacity is respected, so the other two modes will hang # 1080p when xform opacity is respected, so the other two modes will hang
# around until that can be extended to be memory-limited again. # around until that can be extended to be memory-limited again.
acc_mode = 'global' acc_mode = 'deferred'
# TODO: fix this # TODO: fix this
chaos_used = False chaos_used = False

22
cuburn/render.py
View File

@ -81,6 +81,11 @@ class Renderer(object):
self.cubin = pycuda.compiler.compile( self.cubin = pycuda.compiler.compile(
self.src, keep=keep, options=cmp_options, self.src, keep=keep, options=cmp_options,
cache_dir=False if keep else None) cache_dir=False if keep else None)
with open('/tmp/iter_kern.cubin', 'wb') as fp:
fp.write(self.cubin)
# For now, we apply the monkey-patch manually. May eventually make
# this more of a framework if I do it in more than one code segment.
self.cubin = self._iter.monkey_patch(self.cubin)
self.mod = cuda.module_from_buffer(self.cubin, jit_options) self.mod = cuda.module_from_buffer(self.cubin, jit_options)
with open('/tmp/iter_kern.cubin', 'wb') as fp: with open('/tmp/iter_kern.cubin', 'wb') as fp:
fp.write(self.cubin) fp.write(self.cubin)
@ -142,18 +147,7 @@ class Renderer(object):
d_log = cuda.mem_alloc(log_size * 4) d_log = cuda.mem_alloc(log_size * 4)
d_log_sorted = cuda.mem_alloc(log_size * 4) d_log_sorted = cuda.mem_alloc(log_size * 4)
sorter = sort.Sorter(log_size) sorter = sort.Sorter(log_size)
nwriteblocks = int(np.ceil(nbins / float(1<<12)))
# Shared accumulators take care of the lowest 12 bits, but due to
# a quirk of the sort implementation, asking the sort to handle
# fewer bits than it is compiled for will make it considerably
# slower (and it can't be compiled for <7b), so we actually dig in
# to the accumulator's SHAB window for those cases.
SHAB = np.int32(12)
address_bits = np.int32(np.ceil(np.log2(nbins+1)))
start_bit = address_bits - sorter.radix_bits
log_shift = np.int32(SHAB - start_bit)
nwriteblocks = int(np.ceil(nbins / (1<<SHAB)))
print start_bit, log_shift, nwriteblocks
# Calculate 'nslots', the number of simultaneous running threads that # Calculate 'nslots', the number of simultaneous running threads that
# can be active on the GPU during iteration (and thus the number of # can be active on the GPU during iteration (and thus the number of
@ -252,10 +246,10 @@ class Renderer(object):
block=(32, self._iter.NTHREADS/32, 1), block=(32, self._iter.NTHREADS/32, 1),
grid=(ntemporal_samples, 1), stream=iter_stream) grid=(ntemporal_samples, 1), stream=iter_stream)
_sync_stream(write_stream, iter_stream) _sync_stream(write_stream, iter_stream)
sorter.sort(d_log_sorted, d_log, log_size, start_bit, True, sorter.sort(d_log_sorted, d_log, log_size, 12, True,
stream=write_stream) stream=write_stream)
_sync_stream(iter_stream, write_stream) _sync_stream(iter_stream, write_stream)
write_fun(d_accum, d_log_sorted, sorter.dglobal, log_shift, write_fun(d_accum, d_log_sorted, sorter.dglobal,
block=(1024, 1, 1), grid=(nwriteblocks, 1), block=(1024, 1, 1), grid=(nwriteblocks, 1),
texrefs=[tref], stream=write_stream) texrefs=[tref], stream=write_stream)
else: else: