mirrors/ZLUDA
1
0
Fork 0
mirror of https://github.com/vosen/ZLUDA.git synced 2025-04-20 00:14:45 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Throw away useless stuff

Browse source
This commit is contained in:
Andrzej Janik 2020-11-23 19:44:57 +01:00
parent cd141590be
commit 0415f873ae
4 changed files with 0 additions and 181 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

83
doc/NOTES.md
View file

@ -1,83 +0,0 @@
Parser generators in Rust:
--------------------------
I'm convinced nobody actually uses parser generators in Rust:
* pomelo can't generate lexer (understandable, as it is a port of lemon and lemon can't do this either)
* pest can't do parse actions, you have to convert your parse tree to ast manually
* lalrpop can't do comments
* and the day I wrote the line above it can
* reports parsing errors as byte offsets
* if you want to skip parsing one of the alternatives, functional design gets quite awkward
* antlr4rust is untried and requires java to build
* no library supports island grammars
What to emit?
-------------
* SPIR-V
* Better library support, easier to emit
* Can by optimized by IGC
* Can't do some things (not sure what exactly yet)
* But we can work around with inline VISA
* VISA
* Quicker compilation
A64 vs BTS
----------
* How to force A64: -cl-intel-greater-than-4GB-buffer-required
* PTX made a baffling desing choice: global pointers are represented as untyped 64bit integers
* Consequently, there's no 100% certain way to know which argument is a surface and which is a scalar
* It seems that NVidia guys realized what a horrible idea that was and emit `cvta.to.global` as a marker for global pointers?
* But it's only emitted in a recent release build, can't rely on it
* Maybe debug builds emit debug metadata to detect surfaces?
* Might add this as an optimization later
* `cuLaunchKernel` docs say this: "The number of kernel parameters and their offsets and sizes do not need to be specified as that information is retrieved directly from the kernel's image", note the wording: _offsets_ and _sizes_ and not _types_
* Wait, you can mark an argument as a pointer with `.ptr`: https://docs.nvidia.com/cuda/parallel-thread-execution/index.html#kernel-parameter-attribute-ptr, but it's useless with NV compiler not emitting it
* Potential solution: compile only during the dispatch, when type of arguments is known?
* Can't do, the set of arguments passed to cuLaunchKernel is untyped
* Solution: treat all arguments as untyped integers and say goodbye to BTS access
Implicit conversions
--------------------
* PTX support for implicit conversions is completely degenerate, docs say:
_For convenience, ld, st, and cvt instructions permit source and destination data operands to be wider than the instruction-type size, so that narrow values may be loaded, stored, and converted using regular-width registers. For example, 8-bit or 16-bit values may be held directly in 32-bit or 64-bit registers when being loaded, stored, or converted to other types and sizes_
Which is sensible, but completely untrue. In reality ptxas compiles silly code like this:
```
param.f32 param_1
...
.reg.s32 %r1
ld.param.b16 %r1, [param_1];
```
* Surprise, surprise, there's two kind of implicit conversions at play in the example above:
* "Relaxed type-checking rules": this is the conversion of b16 operation type to s32 dst register
* Undocumented type coercion when dereferencing param_1. The PTX behaviour is to coerce **every** type. It's something to the effect of `[param_1] = *(b16*)param_1`
PTX grammar
-----------
* PTX grammar rules are atrocious, keywords can be freely reused as ids without escaping
* Modifiers can be applied to instructions in any arbitrary order. We don't support it and hope we will never have to
Rust debugging
--------------
* Nothing works 100% well on vscode/Windows:
* MSVC/lldb - always garbage (simple enums are fubar)
* MSVC/cppvsdbg - sometimes garbage (nested enums are fubar)
* GNU/lldb - mostly fine, but can't follow child processes
* GNU/gdb - always garbage (I don't have the patience to manually QA rust-gdb on Windows) and doesn't quite understand file paths for break points
* Neither on vscode/Linux:
* lldb - mostly fine, but can't follow child processes
* gdb - visualizes variables somewhat awkardly (shows all possible variants of an enum)
* CLion could be the solution, but intellij-rust can't load this project
CUDA <-> L0
-----------
* device ~= device
* stream ~= command queue
* context ~= context (1.0+)
* graph ~= command list
* module ~= module
IGC
---
* IGC is extremely brittle and segfaults on fairly innocent code:
* OpBitcast of pointer to uint
* OpCopyMemory of alloca'd variable

36
ptx/tools/cvt.py
View file

@ -1,36 +0,0 @@
import os
import subprocess
import tempfile
types = ["u8", "u16", "u32", "u64", "s8", "s16", "s32", "s64", "f16", "f32", "f64"]
rnd = ["", ".rn", ".rni"]
ftz_all = ["", ".ftz"]
sat = ["", ".sat"]
for in_type in types:
for out_type in types:
for r in rnd:
for ftz in ftz_all:
for s in sat:
with tempfile.TemporaryDirectory() as dir:
f_name = os.path.join(dir, 'ptx')
out_name = os.path.join(dir, 'out')
with open(f_name, 'w') as f:
f.write(
f"""
.version 6.5
.target sm_30
.address_size 64
.visible .entry VecAdd_kernel()
{{
.reg.{in_type} r1;
.reg.{out_type} r2;
cvt{r}{ftz}{s}.{out_type}.{in_type} r2, r1;
ret;
}}
""")
err = subprocess.run(f"ptxas {f_name} -o {out_name}", capture_output = True)
if err.returncode == 0:
print(f"cvt{r}{ftz}{s}.{out_type}.{in_type}")
#else:
# print(f"[INVALID] cvt{r}{ftz}{s}.{out_type}.{in_type}")

31
ptx/tools/implicit_ld_dst.py
View file

@ -1,31 +0,0 @@
import os
import subprocess
import tempfile
types = ["b8", "b16", "b32", "b64", "u8", "u16", "u32", "u64", "s8", "s16", "s32", "s64", "f32", "f64"]
for op_type in types:
for output_type in types:
with tempfile.TemporaryDirectory() as dir:
f_name = os.path.join(dir, 'ptx')
out_name = os.path.join(dir, 'out')
with open(f_name, 'w') as f:
f.write(
f"""
.version 6.5
.target sm_30
.address_size 64
.visible .entry VecAdd_kernel(
.param .{op_type} input
)
{{
.reg.{output_type} r1;
ld.param.{op_type} r1, [input];
ret;
}}
""")
err = subprocess.run(f"ptxas {f_name} -o {out_name}", capture_output = True)
if err.returncode == 0:
print(f"{op_type} {output_type}")
else:
print(f"[INVALID] {op_type} {output_type}")

31
ptx/tools/implicit_ld_src.py
View file

@ -1,31 +0,0 @@
import os
import subprocess
import tempfile
types = ["b8", "b16", "b32", "b64", "u8", "u16", "u32", "u64", "s8", "s16", "s32", "s64", "f32", "f64"]
for input_type in types:
for op_type in types:
with tempfile.TemporaryDirectory() as dir:
f_name = os.path.join(dir, 'ptx')
out_name = os.path.join(dir, 'out')
with open(f_name, 'w') as f:
f.write(
f"""
.version 6.5
.target sm_30
.address_size 64
.visible .entry VecAdd_kernel(
.param .{input_type} input
)
{{
.reg.{op_type} r1;
ld.param.{op_type} r1, [input];
ret;
}}
""")
err = subprocess.run(f"ptxas {f_name} -o {out_name}")
if err.returncode == 0:
print(f"{op_type} {input_type}")
else:
print(f"[INVALID] {op_type} {input_type}")