[rocprofiler-compute] fix parser to prevent missing metrics in analysis mode (#1613)
* fix parser * fix parser * fix parser --------- Co-authored-by: fei.zheng <fei.zheng@amd.com> Co-authored-by: ywang103-amd <ywang103@amd.com>
Dieser Commit ist enthalten in:
vedithal-amd
GitHub
@@ -315,33 +315,13 @@ class MetricEvaluator:
|
||||
self.raw_pmc_df = raw_pmc_df
|
||||
self.sys_vars = sys_vars
|
||||
self.empirical_peaks = empirical_peaks
|
||||
self._prepare_df_cache()
|
||||
|
||||
def _prepare_df_cache(self) -> None:
|
||||
"""Prepare cached dataframe access for performance."""
|
||||
if isinstance(self.raw_pmc_df, dict):
|
||||
self.df_cache = {
|
||||
f"raw_pmc_df_{key}": self.raw_pmc_df[key]
|
||||
for key in self.raw_pmc_df.keys()
|
||||
}
|
||||
elif isinstance(self.raw_pmc_df, pd.DataFrame):
|
||||
raw_pmc_df_keys = set(self.raw_pmc_df.columns.get_level_values(0))
|
||||
self.df_cache = {
|
||||
f"raw_pmc_df_{key}": self.raw_pmc_df[key] for key in raw_pmc_df_keys
|
||||
}
|
||||
else:
|
||||
raise ValueError(f'Unknown `raw_pmc_df` type: "{type(self.raw_pmc_df)}".')
|
||||
|
||||
def eval_expression(self, expr: str) -> Union[str, float, int]:
|
||||
"""Evaluate a single expression with proper local context."""
|
||||
try:
|
||||
# Optimize dataframe access by replacing dict notation with dir_path
|
||||
# variable access
|
||||
opt_expr = re.sub(r"raw_pmc_df\['(.*?)'\]", r"raw_pmc_df_\1", expr)
|
||||
|
||||
# Create comprehensive local context
|
||||
local_expr_context = {}
|
||||
local_expr_context.update(self.df_cache)
|
||||
local_expr_context.update({"raw_pmc_df": self.raw_pmc_df})
|
||||
local_expr_context.update(self.sys_vars)
|
||||
local_expr_context.update(self.empirical_peaks)
|
||||
|
||||
@@ -361,12 +341,12 @@ class MetricEvaluator:
|
||||
})
|
||||
|
||||
eval_result = eval(
|
||||
compile(opt_expr, "<string>", "eval"),
|
||||
compile(expr, "<string>", "eval"),
|
||||
{},
|
||||
local_expr_context,
|
||||
)
|
||||
|
||||
if np.isnan(eval_result):
|
||||
if np.isnan(eval_result).any():
|
||||
return ""
|
||||
else:
|
||||
return eval_result
|
||||
@@ -378,10 +358,14 @@ class MetricEvaluator:
|
||||
)
|
||||
return ""
|
||||
else:
|
||||
console_warning(f"Failed to evaluate expression '{expr}': {exception}.")
|
||||
return ""
|
||||
|
||||
except AttributeError as attribute_error:
|
||||
if str(attribute_error) == "'NoneType' object has no attribute 'get'":
|
||||
console_warning(
|
||||
f"Failed to evaluate expression '{expr}': {attribute_error}."
|
||||
)
|
||||
return ""
|
||||
else:
|
||||
console_error("analysis", str(attribute_error))
|
||||
@@ -477,8 +461,17 @@ def build_eval_string(equation: str, coll_level: str, config: dict) -> str:
|
||||
equation_string,
|
||||
)
|
||||
else:
|
||||
# Use pmc_perf.csv for all counters
|
||||
equation_string = re.sub(
|
||||
r"raw_pmc_df", f"raw_pmc_df['{coll_level}']", equation_string
|
||||
r"raw_pmc_df",
|
||||
f"raw_pmc_df['{schema.PMC_PERF_FILE_PREFIX}']",
|
||||
equation_string,
|
||||
)
|
||||
# Use coll_level csv for SQ_ACCUM_PREV_HIRES counter only
|
||||
equation_string = re.sub(
|
||||
rf"raw_pmc_df['{schema.PMC_PERF_FILE_PREFIX}']['SQ_ACCUM_PREV_HIRES']",
|
||||
f"raw_pmc_df['{coll_level}']['SQ_ACCUM_PREV_HIRES']",
|
||||
equation_string,
|
||||
)
|
||||
return equation_string
|
||||
|
||||
@@ -911,7 +904,9 @@ def create_sys_vars(sys_info: pd.Series) -> dict[str, Union[int, float]]:
|
||||
|
||||
|
||||
def calc_builtin_vars(
|
||||
raw_pmc_df: Union[pd.DataFrame, dict], config: dict
|
||||
raw_pmc_df: Union[pd.DataFrame, dict],
|
||||
config: dict,
|
||||
sys_vars: dict[str, Union[int, float]],
|
||||
) -> dict[str, Optional[Union[str, float, int]]]:
|
||||
"""Calculate built-in variables"""
|
||||
# TODO: fix all $normUnit in Unit column or title
|
||||
@@ -929,7 +924,8 @@ def calc_builtin_vars(
|
||||
)
|
||||
try:
|
||||
# Create temporary evaluator for this calculation
|
||||
temporary_evaluator = MetricEvaluator(raw_pmc_df, {}, {})
|
||||
# Pass sys_vars so that $num_xcd and other system variables are available
|
||||
temporary_evaluator = MetricEvaluator(raw_pmc_df, sys_vars, {})
|
||||
calculation_result = temporary_evaluator.eval_expression(eval_string)
|
||||
builtin_vars_collection[f"ammolite__{variable_key}"] = calculation_result
|
||||
except (TypeError, NameError, KeyError, AttributeError):
|
||||
@@ -944,9 +940,9 @@ def calc_builtin_vars(
|
||||
variable_value, schema.PMC_PERF_FILE_PREFIX, config
|
||||
)
|
||||
try:
|
||||
temporary_evaluator = MetricEvaluator(
|
||||
raw_pmc_df, builtin_vars_collection, {}
|
||||
)
|
||||
# Merge sys_vars with builtin_vars_collection for second pass
|
||||
combined_vars = {**sys_vars, **builtin_vars_collection}
|
||||
temporary_evaluator = MetricEvaluator(raw_pmc_df, combined_vars, {})
|
||||
calculation_result = temporary_evaluator.eval_expression(eval_string)
|
||||
builtin_vars_collection[f"ammolite__{variable_key}"] = calculation_result
|
||||
except (TypeError, NameError, KeyError, AttributeError):
|
||||
@@ -981,7 +977,7 @@ def eval_metric(
|
||||
|
||||
sys_vars = create_sys_vars(sys_info)
|
||||
empirical_peaks = create_empirical_peaks_dict(empirical_peaks_df)
|
||||
builtin_vars = calc_builtin_vars(raw_pmc_df, config)
|
||||
builtin_vars = calc_builtin_vars(raw_pmc_df, config, sys_vars)
|
||||
sys_vars.update(builtin_vars)
|
||||
|
||||
# Create metric evaluator
|
||||
@@ -1323,28 +1319,28 @@ def search_pc_sampling_record(
|
||||
console_warning("PC sampling: no pc sampling record found!")
|
||||
return None
|
||||
|
||||
# Prepare sorted output list
|
||||
# Convert to sorted list of tuples:
|
||||
# (code_object_id, inst_index, code_object_offset, count, count_issued,
|
||||
# count_stalled, stall_reason)
|
||||
sorted_counts = sorted(
|
||||
[
|
||||
(
|
||||
code_object_id,
|
||||
code_object_offset,
|
||||
inst_index,
|
||||
info[0], # total_count
|
||||
info[3], # inst_index
|
||||
offset,
|
||||
info[0], # count
|
||||
info[1], # count_issued
|
||||
info[2], # count_stalled
|
||||
# For info[4] (stall_reason dict), remove the zero entries,
|
||||
# sorting the remaining items by their values in descending order
|
||||
sorted(
|
||||
((k, v) for k, v in info[3].items() if v > 0),
|
||||
((k, v) for k, v in info[4].items() if v > 0),
|
||||
key=lambda item: item[1],
|
||||
reverse=True,
|
||||
), # sorted stall reasons
|
||||
sorted(info[4]), # sorted dispatch_ids list
|
||||
)
|
||||
for (
|
||||
code_object_id,
|
||||
code_object_offset,
|
||||
inst_index,
|
||||
), info in grouped_data.items()
|
||||
for (code_object_id, offset), info in grouped_data.items()
|
||||
],
|
||||
key=lambda x: (x[0], x[1], x[2]),
|
||||
)
|
||||
|
||||
In neuem Issue referenzieren
Einen Benutzer sperren