As specified in the docs,
1) raw_string_ostream is always unbuffered and
2) the underlying buffer may be used directly
( 65b13610a5226b84889b923bae884ba395ad084d for further reference )
* Don't call raw_string_ostream::flush(), which is essentially a no-op.
* Avoid unneeded calls to raw_string_ostream::str(), to avoid excess
indirection.
This PR adds a statistics provider cache, which allows an individual
target to keep a rolling tally of it's total time and number of
invocations for a given summary provider. This information is then
available in statistics dump to help slow summary providers, and gleam
more into insight into LLDB's time use.
# Added/changed options
The following options are **added** to the `statistics dump` command:
* `--targets=bool`: Boolean. Dumps the `targets` section.
* `--modules=bool`: Boolean. Dumps the `modules` section.
When both options are given, the field `moduleIdentifiers` will be
dumped for each target in the `targets` section.
The following options are **changed**:
* `--transcript=bool`: Changed to a boolean. Dumps the `transcript`
section.
# Behavior of `statistics dump` with various options
The behavior is **backward compatible**:
- When no options are provided, `statistics dump` dumps all sections.
- When `--summary` is provided, only dumps the summary info.
**New** behavior:
- `--targets=bool`, `--modules=bool`, `--transcript=bool` overrides the
above "default".
For **example**:
- `statistics dump --modules=false` dumps summary + targets +
transcript. No modules.
- `statistics dump --summary --targets=true --transcript=true` dumps
summary + targets (in summary mode) + transcript.
# Added options into public API
In `SBStatisticsOptions`, add:
* `Set/GetIncludeTargets`
* `Set/GetIncludeModules`
* `Set/GetIncludeTranscript`
**Alternative considered**: Thought about adding
`Set/GetIncludeSections(string sections_spec)`, which receives a
comma-separated list of section names to be included ("targets",
"modules", "transcript"). The **benefit** of this approach is that the
API is more future-proof when it comes to possible adding/changing of
section names. **However**, I feel the section names are likely to
remain unchanged for a while - it's not like we plan to make big changes
to the output of `statistics dump` any time soon. The **downsides** of
this approach are: 1\ the readability of the API is worse (requires
reading doc to understand what string can be accepted), 2\ string input
are more prone to human error (e.g. typo "target" instead of expected
"targets").
# Tests
```
bin/llvm-lit -sv ../external/llvm-project/lldb/test/API/commands/statistics/basic/TestStats.py
```
```
./tools/lldb/unittests/Interpreter/InterpreterTests
```
New test cases have been added to verify:
* Different sections are dumped/not dumped when different
`StatisticsOptions` are given through command line (CLI or
`HandleCommand`; see `test_sections_existence_through_command`) or API
(see `test_sections_existence_through_api`).
* The order in which the options are given in command line does not
matter (see `test_order_of_options_do_not_matter`).
---------
Co-authored-by: Roy Shi <royshi@meta.com>
The summary already includes other size information, e.g. total debug
info size in bytes. The only other way I can get this information is by
dumping all statistics which can be quite large. Adding it to the
summary seems fair.
# Changes
1. Changes to the structured transcript.
1. Add fields `commandName` and `commandArguments`. They will hold the
name and the arguments string of the expanded/executed command (e.g.
`breakpoint set` and `-f main.cpp -l 4`). This is not to be confused
with the `command` field, which holds the user input (e.g. `br s -f
main.cpp -l 4`).
2. Add field `timestampInEpochSeconds`. It will hold the timestamp when
the command is executed.
3. Rename field `seconds` to `durationInSeconds`, to improve
readability, especially since `timestampInEpochSeconds` is added.
2. When transcript is available and the newly added option
`--transcript` is present, add the transcript to the output of
`statistics dump`, as a JSON array under a new field `transcript`.
3. A few test name and comment changes.
Updates:
- The previous patch changed the default behavior to not load dwos in
`DWARFUnit`
~~`SymbolFileDWARFDwo *GetDwoSymbolFile(bool load_all_debug_info =
false);`~~
`SymbolFileDWARFDwo *GetDwoSymbolFile(bool load_all_debug_info = true);`
- This broke some lldb-shell tests (see
https://green.lab.llvm.org/green/view/LLDB/job/as-lldb-cmake/16273/)
- TestDebugInfoSize.py
- with symbol on-demand, by default statistics dump only reports
skeleton debug info size
- `statistics dump -f` will load all dwos. debug info = skeleton debug
info + all dwo debug info
Currently running `statistics dump` will trigger lldb to load debug info
that's not yet loaded (eg. dwo files). Resulted in a delay in the
command return, which, can be interrupting.
This patch also added a new option `--load-all-debug-info` asking
statistics to dump all possible debug info, which will force loading all
debug info available if not yet loaded.
Currently running `statistics dump` will trigger lldb to load debug info
that's not yet loaded (eg. dwo files). Resulted in a delay in the
command return, which, can be interrupting.
This patch also added a new option `--load-all-debug-info` asking
statistics to dump all possible debug info, which will force loading all
debug info available if not yet loaded.
Adding command interpreter statistics into "statistics dump" command so
that we can track the command usage frequency for telemetry purpose.
This is useful to answer questions like what is the most frequently used
lldb commands across all our users.
---------
Co-authored-by: jeffreytan81 <jeffreytan@fb.com>
This patch mechanically replaces None with std::nullopt where the
compiler would warn if None were deprecated. The intent is to reduce
the amount of manual work required in migrating from Optional to
std::optional.
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
A previous patch added the ability for us to tell if types were forcefully completed. This patch adds the ability to see which modules have forcefully completed types and aggregates the number of modules with forcefully completed types at the root level.
We add a module specific setting named "debugInfoHadIncompleteTypes" that is a boolean value. We also aggregate the number of modules at the root level that had incomplete debug info with a key named "totalModuleCountWithIncompleteTypes" that is a count of number of modules that had incomplete types.
Differential Revision: https://reviews.llvm.org/D138638
We have a statistic on each module named "debugInfoHadVariableErrors" which tracks when we have debug info, but an error prevented the variables from being displayed. This patch adds a new top level statistic named "totalModuleCountWithVariableErrors" which is a count of the modules that have "debugInfoHadVariableErrors" set to true.
Differential Revision: https://reviews.llvm.org/D138383
When a process gets restarted TypeSystem objects associated with it
may get deleted, and any CompilerType objects holding on to a
reference to that type system are a use-after-free in waiting. Because
of the SBAPI, we don't have tight control over where CompilerTypes go
and when they are used. This is particularly a problem in the Swift
plugin, where the scratch TypeSystem can be restarted while the
process is still running. The Swift plugin has a lock to prevent
abuse, but where there's a lock there can be bugs.
This patch changes CompilerType to store a std::weak_ptr<TypeSystem>.
Most of the std::weak_ptr<TypeSystem>* uglyness is hidden by
introducing a wrapper class CompilerType::WrappedTypeSystem that has a
dyn_cast_or_null() method. The only sites that need to know about the
weak pointer implementation detail are the ones that deal with
creating TypeSystems.
rdar://101505232
Differential Revision: https://reviews.llvm.org/D136650
Context: I plan on using this change primarily downstream in the apple
fork of llvm to track swift module loading time.
Reviewed By: clayborg, tschuett
Differential Revision: https://reviews.llvm.org/D137191
Now that we display an error when users try to get variables, but something in the debug info is preventing variables from showing up, track this with a new bool in each module's statistic information named "debugInfoHadVariableErrors".
This patch modifies the code to track when we have variable errors in a module and adds accessors to get/set this value. This value is used in the module statistics and we added a test to verify this value gets set correctly.
Differential Revision: https://reviews.llvm.org/D134508
This patch adds auto source map deduce count as a target level statistics.
This will help telemetry to track how many debug sessions benefit from this feature.
Differential Revision: https://reviews.llvm.org/D134483
This diff introduces a new symbol on-demand which skips
loading a module's debug info unless explicitly asked on
demand. This provides significant performance improvement
for application with dynamic linking mode which has large
number of modules.
The feature can be turned on with:
"settings set symbols.load-on-demand true"
The feature works by creating a new SymbolFileOnDemand class for
each module which wraps the actual SymbolFIle subclass as member
variable. By default, most virtual methods on SymbolFileOnDemand are
skipped so that it looks like there is no debug info for that module.
But once the module's debug info is explicitly requested to
be enabled (in the conditions mentioned below) SymbolFileOnDemand
will allow all methods to pass through and forward to the actual SymbolFile
which would hydrate module's debug info on-demand.
In an internal benchmark, we are seeing more than 95% improvement
for a 3000 modules application.
Currently we are providing several ways to on demand hydrate
a module's debug info:
* Source line breakpoint: matching in supported files
* Stack trace: resolving symbol context for an address
* Symbolic breakpoint: symbol table match guided promotion
* Global variable: symbol table match guided promotion
In all above situations the module's debug info will be on-demand
parsed and indexed.
Some follow-ups for this feature:
* Add a command that allows users to load debug info explicitly while using a
new or existing command when this feature is enabled
* Add settings for "never load any of these executables in Symbols On Demand"
that takes a list of globs
* Add settings for "always load the the debug info for executables in Symbols
On Demand" that takes a list of globs
* Add a new column in "image list" that shows up by default when Symbols On
Demand is enable to show the status for each shlib like "not enabled for
this", "debug info off" and "debug info on" (with a single character to
short string, not the ones I just typed)
Differential Revision: https://reviews.llvm.org/D121631
This mainly affects Darwin targets (macOS, iOS, tvOS and watchOS) when these targets don't use dSYM files and the debug info was in the .o files. All modules, including the .o files that are loaded by the debug maps, were in the global module list. This was great because it allows us to see each .o file and how much it contributes. There were virtual functions on the SymbolFile class to fetch the symtab/debug info parse and index times, and also the total debug info size. So the main executable would add all of the .o file's stats together and report them as its own data. Then the "totalDebugInfoSize" and many other "totalXXX" top level totals were all being added together. This stems from the fact that my original patch only emitted the modules for a target at the start of the patch, but as comments from the reviews came in, we switched to emitting all of the modules from the global module list.
So this patch fixes it so when we have a SymbolFileDWARFDebugMap that loads .o files, the main executable will have no debug info size or symtab/debug info parse/index times, but each .o file will have its own data as a separate module. Also, to be able to tell when/if we have a dSYM file I have added a "symbolFilePath" if the SymbolFile for the main modules path doesn't match that of the main executable. We also include a "symbolFileModuleIdentifiers" key in each module if the module does have multiple lldb_private::Module objects that contain debug info so that you can track down the information for a module and add up the contributions of all of the .o files.
Tests were added that are labeled with @skipUnlessDarwin and @no_debug_info_test that test all of this functionality so it doesn't regress.
For a module with a dSYM file, we can see the "symbolFilePath" is included:
```
"modules": [
{
"debugInfoByteSize": 1070,
"debugInfoIndexLoadedFromCache": false,
"debugInfoIndexSavedToCache": false,
"debugInfoIndexTime": 0,
"debugInfoParseTime": 0,
"identifier": 4873280600,
"path": "/Users/gclayton/Documents/src/lldb/main/Debug/lldb-test-build.noindex/commands/statistics/basic/TestStats.test_dsym_binary_has_symfile_in_stats/a.out",
"symbolFilePath": "/Users/gclayton/Documents/src/lldb/main/Debug/lldb-test-build.noindex/commands/statistics/basic/TestStats.test_dsym_binary_has_symfile_in_stats/a.out.dSYM/Contents/Resources/DWARF/a.out",
"symbolTableIndexTime": 7.9999999999999996e-06,
"symbolTableLoadedFromCache": false,
"symbolTableParseTime": 7.8999999999999996e-05,
"symbolTableSavedToCache": false,
"triple": "arm64-apple-macosx12.0.0",
"uuid": "E1F7D85B-3A42-321E-BF0D-29B103F5F2E3"
},
```
And for the DWARF in .o file case we can see the "symbolFileModuleIdentifiers" in the executable's module stats:
```
"modules": [
{
"debugInfoByteSize": 0,
"debugInfoIndexLoadedFromCache": false,
"debugInfoIndexSavedToCache": false,
"debugInfoIndexTime": 0,
"debugInfoParseTime": 0,
"identifier": 4603526968,
"path": "/Users/gclayton/Documents/src/lldb/main/Debug/lldb-test-build.noindex/commands/statistics/basic/TestStats.test_no_dsym_binary_has_symfile_identifiers_in_stats/a.out",
"symbolFileModuleIdentifiers": [
4604429832
],
"symbolTableIndexTime": 7.9999999999999996e-06,
"symbolTableLoadedFromCache": false,
"symbolTableParseTime": 0.000112,
"symbolTableSavedToCache": false,
"triple": "arm64-apple-macosx12.0.0",
"uuid": "57008BF5-A726-3DE9-B1BF-3A9AD3EE8569"
},
```
And the .o file for 4604429832 looks like:
```
{
"debugInfoByteSize": 1028,
"debugInfoIndexLoadedFromCache": false,
"debugInfoIndexSavedToCache": false,
"debugInfoIndexTime": 0,
"debugInfoParseTime": 6.0999999999999999e-05,
"identifier": 4604429832,
"path": "/Users/gclayton/Documents/src/lldb/main/Debug/lldb-test-build.noindex/commands/statistics/basic/TestStats.test_no_dsym_binary_has_symfile_identifiers_in_stats/main.o",
"symbolTableIndexTime": 0,
"symbolTableLoadedFromCache": false,
"symbolTableParseTime": 0,
"symbolTableSavedToCache": false,
"triple": "arm64-apple-macosx"
}
```
Differential Revision: https://reviews.llvm.org/D119400
Add statistics about the memory usage of the string pool. I'm
particularly interested in the memory used by the allocator, i.e. the
number of bytes actually used by the allocator it self as well as the
number of bytes allocated through the allocator.
Differential revision: https://reviews.llvm.org/D117914
std::chrono::duration types are not thread-safe, and they cannot be
concurrently updated from multiple threads. Currently, we were doing
such a thing (only) in the DWARF indexing code
(DWARFUnit::ExtractDIEsRWLocked), but I think it can easily happen that
someone else tries to update another statistic like this without
bothering to check for thread safety.
This patch changes the StatsDuration type from a simple typedef into a
class in its own right. The class stores the duration internally as
std::atomic<uint64_t> (so it can be updated atomically), but presents it
to its users as the usual chrono type (duration<float>).
Differential Revision: https://reviews.llvm.org/D117474
This patch add the ability to cache the manual DWARF indexing results to disk for faster subsequent debug sessions. Manual DWARF indexing is time consuming and causes all DWARF to be fully parsed and indexed each time you debug a binary that doesn't have an acceptable accelerator table. Acceptable accelerator tables include .debug_names in DWARF5 or Apple accelerator tables.
This patch breaks up testing by testing all of the encoding and decoding of required C++ objects in a gtest unit test, and then has a test to verify the debug info cache is generated correctly.
This patch also adds the ability to track when a symbol table or DWARF index is loaded or saved to the cache in the "statistics dump" command. This is essential to know in statistics as it can help explain why a debug session was slower or faster than expected.
Reviewed By: labath, wallace
Differential Revision: https://reviews.llvm.org/D115951
It is great to know how many times the target has stopped over its lifetime as each time the target stops, and possibly resumes without the user seeing it for things like shared library loading and signals that are not notified and auto continued, to help explain why a debug session might be slow. This is now included as "stopCount" inside each target JSON.
Differential Revision: https://reviews.llvm.org/D113810
Android and other platforms make wide use of signals when running applications and this can slow down debug sessions. Tracking this statistic can help us to determine why a debug session is slow.
The new data appears inside each target object and reports the signal hit counts:
"signals": [
{
"SIGSTOP": 1
},
{
"SIGUSR1": 1
}
],
Differential Revision: https://reviews.llvm.org/D112683
This patch adds breakpoints to each target's statistics so we can track how long it takes to resolve each breakpoint. It also includes the structured data for each breakpoint so the exact breakpoint details are logged to allow for reproduction of slow resolving breakpoints. Each target gets a new "breakpoints" array that contains breakpoint details. Each breakpoint has "details" which is the JSON representation of a serialized breakpoint resolver and filter, "id" which is the breakpoint ID, and "resolveTime" which is the time in seconds it took to resolve the breakpoint. A snippet of the new data is shown here:
"targets": [
{
"breakpoints": [
{
"details": {...},
"id": 1,
"resolveTime": 0.00039291599999999999
},
{
"details": {...},
"id": 2,
"resolveTime": 0.00022679199999999999
}
],
"totalBreakpointResolveTime": 0.00061970799999999996
}
]
This provides full details on exactly how breakpoints were set and how long it took to resolve them.
Differential Revision: https://reviews.llvm.org/D112587
The new key/value pairs that are added to each module's stats are:
"debugInfoByteSize": The size in bytes of debug info for each module.
"debugInfoIndexTime": The time in seconds that it took to index the debug info.
"debugInfoParseTime": The time in seconds that debug info had to be parsed.
At the top level we add up all of the debug info size, parse time and index time with the following keys:
"totalDebugInfoByteSize": The size in bytes of all debug info in all modules.
"totalDebugInfoIndexTime": The time in seconds that it took to index all debug info if it was indexed for all modules.
"totalDebugInfoParseTime": The time in seconds that debug info was parsed for all modules.
Differential Revision: https://reviews.llvm.org/D112501
The new module stats adds the ability to measure the time it takes to parse and index the symbol tables for each module, and reports modules statistics in the output of "statistics dump" along with the path, UUID and triple of the module. The time it takes to parse and index the symbol tables are also aggregated into new top level key/value pairs at the target level.
Differential Revision: https://reviews.llvm.org/D112279
This patch is a smaller version of a previous patch https://reviews.llvm.org/D110804.
This patch modifies the output of "statistics dump" to be able to get stats from the current target. It adds 3 new stats as well. The output of "statistics dump" is now emitted as JSON so that it can be used to track performance and statistics and the output could be used to populate a database that tracks performance. Sample output looks like:
(lldb) statistics dump
{
"expressionEvaluation": {
"failures": 0,
"successes": 0
},
"firstStopTime": 0.34164492800000001,
"frameVariable": {
"failures": 0,
"successes": 0
},
"launchOrAttachTime": 0.31969605400000001,
"targetCreateTime": 0.0040863039999999998
}
The top level keys are:
"expressionEvaluation" which replaces the previous stats that were emitted as plain text. This dictionary contains the success and fail counts.
"frameVariable" which replaces the previous stats for "frame variable" that were emitted as plain text. This dictionary contains the success and fail counts.
"targetCreateTime" contains the number of seconds it took to create the target and load dependent libraries (if they were enabled) and also will contain symbol preloading times if that setting is enabled.
"launchOrAttachTime" is the time it takes from when the launch/attach is initiated to when the first private stop occurs.
"firstStopTime" is the time in seconds that it takes to stop at the first stop that is presented to the user via the LLDB interface. This value will only have meaning if you set a known breakpoint or stop location in your code that you want to measure as a performance test.
This diff is also meant as a place to discuess what we want out of the "statistics dump" command before adding more funcionality. It is also meant to clean up the previous code that was storting statistics in a vector of numbers within the lldb_private::Target class.
Differential Revision: https://reviews.llvm.org/D111686
