mirror of
https://github.com/pytorch/pytorch.git
synced 2025-10-21 05:34:18 +08:00
5b326d6b61
`pytorch-lldb` support pretty printing size and key_set of tensor via #97101 Add same pretty printing for gdb debugging. **Test Result** ```bash $ gdb python (gdb) break at::native::negative (gdb) r >>> import torch >>> t = torch.tensor([1, 2, 3, 4], dtype=torch.float64) >>> t.negative() Thread 1 "python" hit Breakpoint 1, at::native::negative (self=...) at /home/zong/code/pytorch/aten/src/ATen/native/UnaryOps.cpp:854 854 Tensor negative(const Tensor& self) { return self.neg(); } ``` **Before** ```bash (gdb) p self.key_set() $2 = {repr_ = 1271310352385} (gdb) p self.sizes() $3 = {Data = 0x9cb488, Length = 1} ``` **After** ```bash (gdb) torch-int-array-ref-repr self.sizes() [4] (gdb) torch-dispatch-keyset-repr self.key_set() DispatchKeySet(CPU, ADInplaceOrView, AutogradCPU, AutocastCPU) ``` ```bash $ lintrunner ```  Pull Request resolved: https://github.com/pytorch/pytorch/pull/140935 Approved by: https://github.com/drisspg
108 lines
3.3 KiB
Python
108 lines
3.3 KiB
Python
import textwrap
|
|
from typing import Any
|
|
|
|
import gdb # type: ignore[import]
|
|
|
|
|
|
class DisableBreakpoints:
|
|
"""
|
|
Context-manager to temporarily disable all gdb breakpoints, useful if
|
|
there is a risk to hit one during the evaluation of one of our custom
|
|
commands
|
|
"""
|
|
|
|
def __enter__(self) -> None:
|
|
self.disabled_breakpoints = []
|
|
for b in gdb.breakpoints():
|
|
if b.enabled:
|
|
b.enabled = False
|
|
self.disabled_breakpoints.append(b)
|
|
|
|
def __exit__(self, etype: Any, evalue: Any, tb: Any) -> None:
|
|
for b in self.disabled_breakpoints:
|
|
b.enabled = True
|
|
|
|
|
|
class TensorRepr(gdb.Command): # type: ignore[misc, no-any-unimported]
|
|
"""
|
|
Print a human readable representation of the given at::Tensor.
|
|
Usage: torch-tensor-repr EXP
|
|
|
|
at::Tensor instances do not have a C++ implementation of a repr method: in
|
|
pytorch, this is done by pure-Python code. As such, torch-tensor-repr
|
|
internally creates a Python wrapper for the given tensor and call repr()
|
|
on it.
|
|
"""
|
|
|
|
__doc__ = textwrap.dedent(__doc__).strip()
|
|
|
|
def __init__(self) -> None:
|
|
gdb.Command.__init__(
|
|
self, "torch-tensor-repr", gdb.COMMAND_USER, gdb.COMPLETE_EXPRESSION
|
|
)
|
|
|
|
def invoke(self, args: str, from_tty: bool) -> None:
|
|
args = gdb.string_to_argv(args)
|
|
if len(args) != 1:
|
|
print("Usage: torch-tensor-repr EXP")
|
|
return
|
|
name = args[0]
|
|
with DisableBreakpoints():
|
|
res = gdb.parse_and_eval(f"torch::gdb::tensor_repr({name})")
|
|
print(f"Python-level repr of {name}:")
|
|
print(res.string())
|
|
# torch::gdb::tensor_repr returns a malloc()ed buffer, let's free it
|
|
gdb.parse_and_eval(f"(void)free({int(res)})")
|
|
|
|
|
|
class IntArrayRefRepr(gdb.Command): # type: ignore[misc, no-any-unimported]
|
|
"""
|
|
Print human readable representation of c10::IntArrayRef
|
|
"""
|
|
|
|
def __init__(self) -> None:
|
|
gdb.Command.__init__(
|
|
self, "torch-int-array-ref-repr", gdb.COMMAND_USER, gdb.COMPLETE_EXPRESSION
|
|
)
|
|
|
|
def invoke(self, args: str, from_tty: bool) -> None:
|
|
args = gdb.string_to_argv(args)
|
|
if len(args) != 1:
|
|
print("Usage: torch-int-array-ref-repr EXP")
|
|
return
|
|
name = args[0]
|
|
with DisableBreakpoints():
|
|
res = gdb.parse_and_eval(f"torch::gdb::int_array_ref_string({name})")
|
|
res = str(res)
|
|
print(res[res.find('"') + 1 : -1])
|
|
|
|
|
|
class DispatchKeysetRepr(gdb.Command): # type: ignore[misc, no-any-unimported]
|
|
"""
|
|
Print human readable representation of c10::DispatchKeyset
|
|
"""
|
|
|
|
def __init__(self) -> None:
|
|
gdb.Command.__init__(
|
|
self,
|
|
"torch-dispatch-keyset-repr",
|
|
gdb.COMMAND_USER,
|
|
gdb.COMPLETE_EXPRESSION,
|
|
)
|
|
|
|
def invoke(self, args: str, from_tty: bool) -> None:
|
|
args = gdb.string_to_argv(args)
|
|
if len(args) != 1:
|
|
print("Usage: torch-dispatch-keyset-repr EXP")
|
|
return
|
|
keyset = args[0]
|
|
with DisableBreakpoints():
|
|
res = gdb.parse_and_eval(f"torch::gdb::dispatch_keyset_string({keyset})")
|
|
res = str(res)
|
|
print(res[res.find('"') + 1 : -1])
|
|
|
|
|
|
TensorRepr()
|
|
IntArrayRefRepr()
|
|
DispatchKeysetRepr()
|