nimlgen/tinygrad
1
0
Fork 0
forked from tinygrad/tinygrad
Code Pull Requests Activity

hcq2: cleanup (#16767)

Browse Source
This commit is contained in:
nimlgen
2026-06-26 20:26:35 +03:00
committed by GitHub
parent 8c9e8d38a9
commit fe9cd76e8b
2 changed files with 16 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files
extra/hcq2/hcq2.py
+13 -12
View File
@@ -27,7 +27,7 @@ class HCQ2Compiled(Compiled):
(UPat(Ops.BUFFER, tag="timeline_value"), lambda ctx: ctx.timeline_value()),
(UPat(Ops.BUFFER, tag="sentinel_signal"), lambda ctx: ctx.timeline_signal("sentinel", (1 << 64) - 1)),
(UPat(Ops.BUFFER, name="b"), lambda ctx, b:
Buffer(ctx.device, b.arg, b.dtype, options=BufferSpec(host=False, uncached=True, cpu_access=True, nolru=True))), # TODO: remove nolru
Buffer(ctx.device, b.max_numel(), b.dtype, options=BufferSpec(host=False, uncached=True, cpu_access=True, nolru=True))), # TODO: remove nolru
])
super().__init__(device, allocator, compilers, lambda *a, **kw: None, None, arch=arch)
@@ -137,7 +137,7 @@ def unwrap_after(uop):
return uop
def make_getaddr(u, device=None):
if unwrap_after(u).op not in (Ops.BUFFER, Ops.SLICE, Ops.BINARY, Ops.MSTACK, Ops.MSELECT): return u
if unwrap_after(u).op not in (Ops.BUFFER, Ops.SLICE, Ops.BINARY, Ops.MSTACK, Ops.MSELECT, Ops.PARAM): return u
return UOp(Ops.GETADDR, dtypes.uint64, src=(u, UOp(Ops.DEVICE, arg=device or to_tuple(u.device)[0])))
def make_ins(op, *srcs):
@@ -183,19 +183,19 @@ class HCQInfo:
def from_call(call:UOp) -> HCQInfo: return HCQInfo(get_call_name(call, get_call_arg_uops(call)), estimate_uop(call), get_call_outs_ins(call)[0])
# *****************
# 1.1. prep runtimes: staging copies
# 1.1. prep: staging copies
def _need_staging(a, b): return all_devices_in(a.device, HCQ_DEVS) and not all_devices_in(b.device, HCQ_P2P_DEVS)
def stage_copy(dst:UOp, src:UOp) -> UOp|None:
if not (_need_staging(src, dst) or _need_staging(dst, src)): return None
stage = UOp.new_buffer("CPU", src.buffer.nbytes, dtypes.uint8)
stage = UOp.new_buffer("CPU", src.nbytes(), dtypes.uint8)
return UOp(Ops.LINEAR, dtypes.void, (src.copy_to_device("CPU").call(stage, src), stage.copy_to_device(dst.device).call(dst, stage)))
pm_insert_copy_staging = PatternMatcher([(UPat(Ops.CALL, src=(UPat(Ops.COPY), UPat(name="dst"), UPat(name="src"))), stage_copy)])
# *****************
# 1.2. prep runtimes: programs/kernargs
# 2.1. hcq lowering: programs/kernargs
@functools.cache
def get_pm_prep_program(name:str) -> PatternMatcher|None:
@@ -216,7 +216,7 @@ def prep_program(call:UOp, prg:UOp) -> UOp|None:
def prep_kernargs(call:UOp, prg:UOp) -> UOp:
(data, info), dev_uop = prg.arg, UOp(Ops.DEVICE, arg=call.src[1].device)
buf = UOp.new_buffer(dev_uop.arg, data.kernargs_alloc_size, dtypes.uint8).rtag("kernargs")
patches = [make_patch(buf, i*8, UOp(Ops.GETADDR, dtypes.uint64, src=(call.src[1+gi], dev_uop))) for i,gi in enumerate(info.globals)] \
patches = [make_patch(buf, i*8, make_getaddr(call.src[1+gi], dev_uop.arg)) for i,gi in enumerate(info.globals)] \
+ [make_patch(buf, len(info.globals)*8 + i*4, v, dtypes.uint32) for i,v in enumerate(info.vars)]
return call.replace(src=(prg.replace(src=prg.src + (buf.after(*patches),), arg=(data, info)),) + call.src[1:])
@@ -230,7 +230,7 @@ pm_prep_runtime = PatternMatcher([
])
# *****************
# 2. lowering to hcq ir
# 2.2. hcq lowering: ops to ir
def make_submit(*cmds, devs:str|tuple[str, ...], queue:str) -> UOp:
devs:tuple[str, ...] = to_tuple(devs)
@@ -470,15 +470,16 @@ def pack_hcq_placeholders(call:UOp) -> UOp|None:
off_per_buf:dict[UOp, int] = {}
size_per_tag:dict[str, int] = {}
for b in bufs:
if b.tag in maxtags: size_per_tag[b.tag] = max(size_per_tag.get(b.tag, 0), b.arg)
bsz = b.max_numel()
if b.tag in maxtags: size_per_tag[b.tag] = max(size_per_tag.get(b.tag, 0), bsz)
elif b.tag in sumtags:
off_per_buf[b] = round_up(size_per_tag.get(b.tag, 0), {"program": 0x1000}.get(b.tag, 128))
size_per_tag[b.tag] = off_per_buf[b] + b.arg
size_per_tag[b.tag] = off_per_buf[b] + bsz
count_per_tag = collections.Counter(b.tag for b in bufs)
ref_bufs = {b.tag:b for b in bufs if count_per_tag[b.tag] > 1}
bases = {tag:UOp.new_buffer(b.src[1].arg, size_per_tag[tag], b.dtype).rtag(tag) for tag,b in ref_bufs.items()}
subs = {b:UOp(Ops.SLICE, b.dtype, (bases[b.tag], UOp.const(dtypes.weakint, off_per_buf.get(b, 0))), b.arg) for b in bufs if b.tag in bases}
bases = {tag:UOp.new_buffer(b.device, size_per_tag[tag], b.dtype).rtag(tag) for tag,b in ref_bufs.items()}
subs = {b:UOp(Ops.SLICE, b.dtype, (bases[b.tag], UOp.const(dtypes.weakint, off_per_buf.get(b, 0))), b.max_numel()) for b in bufs if b.tag in bases}
return call.replace(src=(call.src[0].substitute(subs, walk=True), *call.src[1:])) if subs else None
pm_pack_placeholders = PatternMatcher([(UPat(Ops.CALL, tag="hcq", name="call"), pack_hcq_placeholders)])
@@ -495,7 +496,7 @@ pm_hold_call_buffers = PatternMatcher([(UPat(Ops.CALL, tag="hcq", name="call"),
def bufferize_buf(buf:UOp) -> UOp|None:
if buf.tag is None: return None
uops = tuple(UOp.from_buffer((dv:=Device[dev]).pm_bufferize.rewrite(buf, ctx=dv), "CPU") for dev in to_tuple(buf.src[1].arg))
uops = tuple(UOp.from_buffer((dv:=Device[dev]).pm_bufferize.rewrite(buf, ctx=dv), "CPU") for dev in to_tuple(buf.device))
return make_mstack(uops)
pm_bufferize = PatternMatcher([(UPat(Ops.BUFFER, name="buf"), bufferize_buf)])
extra/hcq2/ops_amd2.py
+3 -3
View File
@@ -142,7 +142,7 @@ pm_pm4_opsel = PatternMatcher([
])
def pm4_submit(cmdbuf, devs):
size, zero = UOp.const(dtypes.uint32, cmdbuf.src[0].arg // dtypes.uint32.itemsize), UOp.const(dtypes.int, 0)
size, zero = UOp.const(dtypes.uint32, cmdbuf.nbytes() // dtypes.uint32.itemsize), UOp.const(dtypes.int, 0)
# the compute queue's ring and its host-side ring/write/put pointers (placeholders, resolved in pm_bufferize)
for d in devs: q = Device[d].compute_queue
@@ -205,7 +205,7 @@ pm_sdma_opsel = PatternMatcher([
def sdma_submit(cmdbuf, devs):
# the cmdbuf to submit + the patch writes that fill it
size_dw, zero = cmdbuf.src[0].arg // dtypes.uint32.itemsize, UOp.const(dtypes.int, 0)
size_dw, zero = cmdbuf.nbytes() // dtypes.uint32.itemsize, UOp.const(dtypes.int, 0)
# the sdma queue's ring and its host-side ring/write/put pointers
for d in devs: q = Device[d].sdma_queue(0)
@@ -579,7 +579,7 @@ class AMDDevice(HCQ2Compiled):
# Scratch setup
self.max_private_segment_size = 0
self.pm_bufferize = PatternMatcher([(UPat(Ops.BUFFER, tag="scratch", name="b"), lambda ctx, b: ctx.scratch_buffer(b.arg))]) + self.pm_bufferize
self.pm_bufferize = PatternMatcher([(UPat(Ops.BUFFER, tag="scratch", name="b"), lambda ctx, b: ctx.scratch_buffer(b.max_numel()))]) + self.pm_bufferize
self.pmc_enabled:bool = PROFILE > 0 and PMC > 0
if self.pmc_enabled: