pytorch/torch/csrc/jit/testing/file_check.cpp
Han Qi b34b192d6b Reland "Make debug_pkl smaller by only emitting unique traces." (#73368)
Summary:
## Original commit message:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/73368

debug_pkl file inside of pytorch's .pt file consists of a list of SourceRanges. Each SourceRange points to a Source which is a stack track, filename, and start, end numbers. Those are emitted in debug_pkl file as strings.
Since many SourceRange shares the same source, the string for trace can be deduped.
The newer format saves a set of unique traces in a tuple, then each SourceRange will save the offset of it's trace w.r.t. position in that tuple. (i.e. manually applying dictionary compression).
The above helps with smaller file size. On loading, if we copy each trace to Source as string the runtime memory would still blowup.
To mitigate this, we use SourceView directly instead of source which will take the reference of string inside of Deserializer and make that into string_view. This is safe because Deserializer is hold by Unpickler by shared_ptr, and Unpickler is also hold by shared_ptr by another Source object. That Source object will be alive during the model construction.

Test Plan:
## Original Test plan
unit test

Took original file (312271638_930.predictor.disagg.local); loaded with `torch.jit.load` save again with `torch.jit.save`. Unzip both, look at contents:
```
[qihan@devvm5585.vll0 ~]$ du archive -h
4.0K    archive/xl_model_weights
3.7M    archive/extra
8.0K    archive/code/__torch__/caffe2/torch/fb/model_transform/splitting
8.0K    archive/code/__torch__/caffe2/torch/fb/model_transform
8.0K    archive/code/__torch__/caffe2/torch/fb
8.0K    archive/code/__torch__/caffe2/torch
8.0K    archive/code/__torch__/caffe2
20M     archive/code/__torch__/torch/fx/graph_module
20M     archive/code/__torch__/torch/fx
8.0K    archive/code/__torch__/torch/classes
20M     archive/code/__torch__/torch
20M     archive/code/__torch__
20M     archive/code
2.7M    archive/constants
35M     archive
[qihan@devvm5585.vll0 ~]$ du resaved -h
4.0K    resaved/extra
8.0K    resaved/code/__torch__/caffe2/torch/fb/model_transform/splitting
8.0K    resaved/code/__torch__/caffe2/torch/fb/model_transform
8.0K    resaved/code/__torch__/caffe2/torch/fb
8.0K    resaved/code/__torch__/caffe2/torch
8.0K    resaved/code/__torch__/caffe2
1.3M    resaved/code/__torch__/torch/fx/graph_module
1.3M    resaved/code/__torch__/torch/fx
8.0K    resaved/code/__torch__/torch/classes
1.4M    resaved/code/__torch__/torch
1.4M    resaved/code/__torch__
1.4M    resaved/code
2.7M    resaved/constants
13M     resaved
[qihan@devvm5585.vll0 ~]$
```
## Additional test:
`buck test mode/dev-tsan //caffe2/benchmarks/static_runtime:static_runtime_cpptest -- --exact 'caffe2/benchmarks/static_runtime:static_runtime_cpptest - StaticRuntime.to'` passes

 test jest.fbios.startup_cold_start.local.simulator f333356873 -

Differential Revision: D35196883

Pull Request resolved: https://github.com/pytorch/pytorch/pull/74869
Approved by: https://github.com/gmagogsfm
2022-04-18 22:34:21 +00:00

586 lines
17 KiB
C++

//==-- llvm/Support/FileCheck.h ---------------------------*- C++ -*-==//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
// API modified from llvm::FileCheck
#include <c10/util/Exception.h>
#include <c10/util/Optional.h>
#include <c10/util/StringUtil.h>
#include <c10/util/irange.h>
#include <torch/csrc/Export.h>
#include <torch/csrc/jit/frontend/source_range.h>
#include <torch/csrc/jit/ir/ir.h>
#include <torch/csrc/jit/testing/file_check.h>
#include <algorithm>
#include <iostream>
#include <sstream>
#include <string>
namespace torch {
namespace jit {
namespace testing {
enum CheckType {
CHECK,
CHECK_NEXT,
CHECK_SAME,
CHECK_NOT,
CHECK_COUNT,
CHECK_DAG,
CHECK_SOURCE_HIGHLIGHTED,
};
struct Check {
Check(
CheckType type,
std::string str,
c10::optional<size_t> count = c10::nullopt)
: type_(type), search_str_(std::move(str)) {
count_ = count;
};
Check(
CheckType type,
c10::string_view str,
c10::optional<size_t> count = c10::nullopt)
: Check(type, std::string(str.begin(), str.end()), count) {}
CheckType type_;
c10::optional<size_t> count_;
const std::string search_str_;
friend std::ostream& operator<<(std::ostream& out, const Check& c);
};
std::ostream& operator<<(std::ostream& out, const Check& c) {
switch (c.type_) {
case CHECK:
out << "CHECK";
break;
case CHECK_NEXT:
out << "CHECK-NEXT";
break;
case CHECK_SAME:
out << "CHECK-SAME";
break;
case CHECK_NOT:
out << "CHECK-NOT";
break;
case CHECK_DAG:
out << "CHECK-DAG";
break;
case CHECK_COUNT:
out << "CHECK-COUNT-" << *c.count_;
break;
case CHECK_SOURCE_HIGHLIGHTED:
out << "CHECK-SOURCE-HIGHLIGHTED";
break;
}
out << ": " << c.search_str_;
return out;
};
namespace {
size_t assertFind(
const SourceRange& search_range,
const std::string& sub,
const std::function<void(std::ostream& out)>& extra_msg = nullptr) {
auto pos = search_range.source()->text_str().find(sub, search_range.start());
if (pos == std::string::npos || (pos + sub.size()) > search_range.end()) {
auto found_range =
SourceRange(search_range.source(), search_range.start(), sub.size());
std::stringstream ss;
ss << "Expected to find ";
c10::printQuotedString(ss, sub);
ss << " but did not find it" << std::endl;
ss << "Searched string:" << std::endl;
found_range.highlight(ss);
if (extra_msg) {
extra_msg(ss);
}
throw std::runtime_error(ss.str());
}
return pos;
}
size_t assertFind(
const SourceRange& search_range,
const std::string& sub,
const Check& check) {
return assertFind(search_range, sub, [&](std::ostream& out) {
out << "From " << check << "\n";
});
}
size_t assertFind(
const std::shared_ptr<Source>& source,
const std::string& sub,
size_t start,
const Check& check) {
return assertFind(SourceRange(source, start, source->size()), sub, check);
}
void assertNotFind(
const SourceRange& search_range,
const std::string& sub,
const Check& check) {
auto pos = search_range.source()->text_str().find(sub, search_range.start());
if (pos != std::string::npos && (pos + sub.size()) <= search_range.end()) {
auto found_range =
SourceRange(search_range.source(), pos, sub.size() + pos);
std::stringstream ss;
ss << "Expected to not find ";
c10::printQuotedString(ss, sub);
ss << " but found it\n";
found_range.highlight(ss);
ss << "From " << check << "\n";
throw std::runtime_error(ss.str());
}
}
} // namespace
struct FileCheckImpl {
TORCH_API explicit FileCheckImpl() = default;
TORCH_API void run(const std::string& test_file) {
has_run = true;
if (groups.size() == 0 || groups[0].size() == 0) {
throw std::runtime_error(
"No checks have been added to this instance of"
"Filecheck! Check for bad input.");
}
doChecks(std::make_shared<Source>(test_file));
}
TORCH_API void run(
const std::string& checks_file,
const std::string& test_file) {
auto source = std::make_shared<Source>(checks_file);
parseStrings(source);
run(test_file);
}
TORCH_API void addCheck(const Check& check) {
// consecutive CHECK_DAGs & CHECK_NOTs need to be evaluated as a group
if (groups.size() == 0 ||
(check.type_ != CHECK_NOT && check.type_ != CHECK_DAG)) {
groups.push_back({check});
} else {
auto& last_group = groups.back();
if (last_group.at(0).type_ == check.type_) {
last_group.push_back(check);
} else {
groups.push_back({check});
}
}
has_run = false;
}
TORCH_API void addCheck(
CheckType type,
const std::string& s,
c10::optional<size_t> count = c10::nullopt) {
addCheck(Check(type, s, count));
}
// NOLINTNEXTLINE(cppcoreguidelines-non-private-member-variables-in-classes)
bool has_run = false;
friend std::ostream& operator<<(std::ostream& out, const FileCheckImpl& fc);
private:
bool parseSingleCheck(const std::shared_ptr<Source>& source, size_t* start) {
const static std::vector<std::pair<CheckType, std::string>> check_pairs = {
{CHECK, ": "},
{CHECK_NEXT, "-NEXT: "},
{CHECK_SAME, "-SAME: "},
{CHECK_NOT, "-NOT: "},
{CHECK_DAG, "-DAG: "},
{CHECK_COUNT, "-COUNT-"}, // needs special parsing
{CHECK_SOURCE_HIGHLIGHTED, "-SOURCE-HIGHLIGHTED: "},
};
for (const auto& check_pair : check_pairs) {
const std::string& check_suffix = check_pair.second;
auto suffix_pos = source->text_str().find(check_suffix, *start);
if (suffix_pos != *start) {
continue;
}
size_t end_check_string = suffix_pos + check_suffix.size();
CheckType type = check_pair.first;
c10::optional<size_t> count = c10::nullopt;
auto end_line = source->text_str().find("\n", end_check_string);
bool exactly = false;
if (type == CHECK_COUNT) {
const std::string exact = "EXACTLY-";
if (source->text_str().find(exact, end_check_string) ==
end_check_string) {
exactly = true;
end_check_string += exact.size();
}
size_t end =
assertFind(SourceRange(source, end_check_string, end_line), ":");
auto count_view = source->text_str()
.substr(end_check_string, end - end_check_string)
.str();
count = c10::stoll(std::string(count_view.begin(), count_view.end()));
end_check_string = end + 2; // add ':' and the space
}
auto check = Check(
type,
source->text_str()
.substr(end_check_string, end_line - end_check_string)
.str(),
count);
addCheck(check);
if (exactly) {
addCheck(CHECK_NOT, check.search_str_);
}
*start = end_line;
return true;
}
return false;
}
size_t findNextStart(const std::shared_ptr<Source>& source, size_t prev_end) {
size_t start = source->text_str().find("#", prev_end);
if (start == std::string::npos) {
return start;
}
start += 1;
static constexpr size_t max_whitespace = 6;
size_t i = 0;
while (start + i < source->size() && i < max_whitespace) {
auto c = source->char_at(start + i);
if (c != ' ' && c != '\t') {
break;
}
i++;
}
static const std::string check = "CHECK";
if (source->text_str().substr(start + i, check.size()) == check) {
return start + i + check.size();
} else {
return findNextStart(source, start + i + 1);
}
}
void parseStrings(const std::shared_ptr<Source>& source) {
size_t start = 0;
start = findNextStart(source, 0);
while (start != std::string::npos) {
bool found_match = parseSingleCheck(source, &start);
if (!found_match) {
std::ostringstream ss;
ss << "Could not parse check at:\n";
SourceRange(source, start, start + 1).highlight(ss);
ss << "Check for bad input.";
has_run = true;
throw std::runtime_error(ss.str());
}
start = findNextStart(source, start);
}
}
void doCheckNot(
const std::vector<Check>& nots,
const std::shared_ptr<Source>& source,
const SourceRange& prev,
const SourceRange& next) {
auto start = prev.end(); // inclusive
auto end = next.start(); // exclusive
if (end < start) {
return;
}
for (const auto& check : nots) {
AT_ASSERT(check.type_ == CHECK_NOT);
assertNotFind(SourceRange(source, start, end), check.search_str_, check);
}
}
// Checks that source token is highlighted, does not advance search range.
void doCheckSourceHighlighted(
const Check& check,
const std::shared_ptr<Source>& source,
size_t start_offset) {
auto construct_error_and_throw = [&](size_t error_start_pos) {
SourceRange error_range(
source, error_start_pos, check.search_str_.size());
std::stringstream ss;
ss << "Expected to find ";
c10::printQuotedString(ss, check.search_str_);
ss << "highlighted but it is not." << std::endl;
error_range.highlight(ss);
throw std::runtime_error(ss.str());
};
size_t search_start_offset = start_offset;
bool found_token_at_least_once = false;
size_t pos = search_start_offset;
while (pos < source->size()) {
pos = source->text_str().find(check.search_str_, search_start_offset);
if (pos == std::string::npos) {
break;
}
found_token_at_least_once = true;
auto lineno = source->lineno_for_offset(pos);
auto col = pos - source->offset_for_line(lineno);
auto highlight_lineno = lineno + 1;
if (highlight_lineno >= source->num_lines()) {
construct_error_and_throw(pos);
}
auto highlight_start_offset =
source->offset_for_line(highlight_lineno) + col;
auto highlight_end_offset = std::min(
highlight_start_offset + check.search_str_.size(), source->size());
if (highlight_end_offset >= source->size()) {
construct_error_and_throw(pos);
}
bool found_highlight = true;
for (const auto posi :
c10::irange(highlight_start_offset, highlight_end_offset)) {
if (source->char_at(posi) != '~') {
found_highlight = false;
}
}
if (found_highlight) {
assertNotFind(
SourceRange(
source, highlight_start_offset - 1, highlight_start_offset),
"~",
check);
assertNotFind(
SourceRange(source, highlight_end_offset, highlight_end_offset + 1),
"~",
check);
return;
}
search_start_offset = pos + 1;
}
if (!found_token_at_least_once) {
// Guaranteed to fail to generate error message.
assertFind(source, check.search_str_, start_offset, check);
}
construct_error_and_throw(start_offset);
}
SourceRange matchDagGroup(
const std::vector<Check>& group,
const std::shared_ptr<Source>& source,
const SourceRange& prev) {
size_t group_beg = std::string::npos;
size_t group_end = 0;
AT_ASSERT(groups.size() != 0);
for (const auto& check : group) {
AT_ASSERT(check.type_ == group[0].type_);
auto pos = assertFind(source, check.search_str_, prev.end(), check);
group_beg = std::min(pos, group_beg);
group_end = std::max(pos + check.search_str_.size(), group_end);
}
return SourceRange(source, group_beg, group_end);
}
SourceRange matchGroup(
const std::vector<Check>& group,
const std::shared_ptr<Source>& source,
const SourceRange& prev) {
AT_ASSERT(group.size() != 0);
CheckType type = group[0].type_;
if (type == CHECK_DAG) {
return matchDagGroup(group, source, prev);
}
AT_ASSERT(type != CHECK_NOT);
AT_ASSERT(group.size() == 1);
const auto& check = group[0];
size_t start_range = prev.end();
size_t end_range = start_range;
switch (check.type_) {
case CHECK: {
start_range = assertFind(source, check.search_str_, start_range, check);
end_range = start_range + check.search_str_.size();
} break;
case CHECK_SAME: {
auto pos = assertFind(source, check.search_str_, start_range, check);
assertNotFind(SourceRange(source, prev.end(), pos), "\n", check);
start_range = pos;
end_range = pos + check.search_str_.size();
} break;
case CHECK_NEXT: {
auto line_end = assertFind(source, "\n", start_range, check);
auto pos = assertFind(source, check.search_str_, line_end + 1, check);
assertNotFind(SourceRange(source, line_end + 1, pos), "\n", check);
start_range = pos;
end_range = pos + check.search_str_.size();
} break;
case CHECK_COUNT: {
auto group_start_range = std::string::npos;
AT_ASSERT(check.count_ && *check.count_ != 0);
for (size_t i = 0; i < *check.count_; ++i) {
start_range =
assertFind(source, check.search_str_, start_range, check);
group_start_range = std::min(start_range, group_start_range);
end_range = start_range + check.search_str_.size();
start_range = end_range;
}
start_range = group_start_range;
} break;
case CHECK_SOURCE_HIGHLIGHTED: {
doCheckSourceHighlighted(check, source, start_range);
break;
}
case CHECK_DAG: {
AT_ERROR();
} break;
case CHECK_NOT: {
AT_ERROR();
} break;
}
return SourceRange(source, start_range, end_range);
}
void doChecks(const std::shared_ptr<Source>& source) {
SourceRange prev(source, 0, 0);
for (size_t i = 0; i < groups.size(); i++) {
const auto& curr_group = groups[i];
CheckType type = curr_group.at(0).type_;
if (type != CHECK_NOT) {
prev = matchGroup(curr_group, source, prev);
} else {
if (i + 1 < groups.size()) {
const auto& next_group = groups[i + 1];
AT_ASSERT(next_group.at(0).type_ != CHECK_NOT);
SourceRange after_not = matchGroup(next_group, source, prev);
doCheckNot(curr_group, source, prev, after_not);
prev = after_not;
++i; // already checked the group after
} else {
SourceRange end_of_file(
source, source->size() + 1, source->size() + 1);
doCheckNot(curr_group, source, prev, end_of_file);
}
}
}
}
std::vector<Check> checks;
std::vector<std::vector<Check>> groups;
};
FileCheck::FileCheck() : fcImpl(new FileCheckImpl()){};
std::ostream& operator<<(std::ostream& out, const FileCheckImpl& fc) {
out << "FileCheck checks:\n";
for (const Check& c : fc.checks) {
out << "\t" << c << "\n";
}
return out;
};
FileCheck::~FileCheck() {
if (!fcImpl->has_run) {
std::cout << "You have not run this instance of FileCheck!\n";
std::cout << *fcImpl;
}
fcImpl.reset();
};
void FileCheck::run(const std::string& test_file) {
fcImpl->run(test_file);
};
void FileCheck::run(const Graph& graph) {
std::stringstream graph_str;
graph_str << graph;
fcImpl->run(graph_str.str());
};
void FileCheck::run(
const std::string& input_checks_string,
const std::string& test_string) {
fcImpl->run(input_checks_string, test_string);
}
void FileCheck::run(
const std::string& input_checks_string,
const Graph& graph) {
std::stringstream graph_str;
graph_str << graph;
fcImpl->run(input_checks_string, graph_str.str());
}
FileCheck* FileCheck::check(const std::string& str) {
fcImpl->addCheck(CHECK, str);
return this;
}
FileCheck* FileCheck::check_not(const std::string& str) {
fcImpl->addCheck(CHECK_NOT, str);
return this;
}
FileCheck* FileCheck::check_same(const std::string& str) {
fcImpl->addCheck(CHECK_SAME, str);
return this;
}
FileCheck* FileCheck::check_next(const std::string& str) {
fcImpl->addCheck(CHECK_NEXT, str);
return this;
}
FileCheck* FileCheck::check_count(
const std::string& str,
size_t count,
bool exactly) {
TORCH_INTERNAL_ASSERT(
count != 0 || exactly, "Count == 0 && !exactly doesn't do anything");
if (count) {
fcImpl->addCheck(CHECK_COUNT, str, count);
}
if (exactly) {
fcImpl->addCheck(CHECK_NOT, str);
}
return this;
}
FileCheck* FileCheck::check_dag(const std::string& str) {
fcImpl->addCheck(CHECK_DAG, str);
return this;
}
FileCheck* FileCheck::check_source_highlighted(const std::string& str) {
fcImpl->addCheck(CHECK_SOURCE_HIGHLIGHTED, str);
return this;
}
} // namespace testing
} // namespace jit
} // namespace torch