At the moment the expression parser doesn't support evaluating expressions in
static member functions and just pretends the expression is evaluated within a
non-member function. This causes that all static members are inaccessible when
doing unqualified name lookup.
This patch adds support for evaluating in static member functions. It
essentially just does the same setup as what LLDB is already doing for
non-static member functions (i.e., wrapping the expression in a fake member
function) with the difference that we now mark the wrapping function as static
(to prevent access to non-static members).
Reviewed By: shafik, jarin
Differential Revision: https://reviews.llvm.org/D81550
Ever since Dave Zarzycki's patch to sort test start times based on prior
test timing data (https://reviews.llvm.org/D98179) the test suite aborts
with a SIGHUP. I don't believe his patch is to blame, but rather
uncovers an preexisting issue by making test runs more deterministic.
I was able to narrow down the issue to TestSimulatorPlatform.py. The
issue also manifests itself on the standalone bot on GreenDragon [1].
This patch disables the test until we can figure this out.
[1] http://green.lab.llvm.org/green/view/LLDB/job/lldb-cmake-standalone/
rdar://76995109
VSCode doesn't render multiple variables with the same name in the variables view. It only renders one of them. This is a situation that happens often when there are shadowed variables.
The nodejs debugger solves this by adding a number suffix to the variable, e.g. "x", "x2", "x3" are the different x variables in nested blocks.
In this patch I'm doing something similar, but the suffix is " @ <file_name:line>), e.g. "x @ main.cpp:17", "x @ main.cpp:21". The fallback would be an address if the source and line information is not present, which should be rare.
This fix is only needed for globals and locals. Children of variables don't suffer of this problem.
When there are shadowed variables
{F16182150}
Without shadowed variables
{F16182152}
Modifying these variables through the UI works
Reviewed By: clayborg
Differential Revision: https://reviews.llvm.org/D99989
In certain occasions times, like when LLDB is initializing and
evaluating the .lldbinit files, it tries to print to stderr and stdout
directly. This confuses the IDE with malformed data, as it talks to
lldb-vscode using stdin and stdout following the JSON RPC protocol. This
ends up terminating the debug session with the user unaware of what's
going on. There might be other situations in which this can happen, and
they will be harder to debug than the .lldbinit case.
After several discussions with @clayborg, @yinghuitan and @aadsm, we
realized that the best course of action is to simply redirect stdout and
stderr to the console, without modifying LLDB itself. This will prove to
be resilient to future bugs or features.
I made the simplest possible redirection logic I could come up with. It
only works for POSIX, and to make it work with Windows should be merely
changing pipe and dup2 for the windows equivalents like _pipe and _dup2.
Sadly I don't have a Windows machine, so I'll do it later once my office
reopens, or maybe someone else can do it.
I'm intentionally not adding a stop-redirecting logic, as I don't see it
useful for the lldb-vscode case (why would we want to do that, really?).
I added a test.
Note: this is a simpler version of D80659. I first tried to implement a
RIIA version of it, but it was problematic to manage the state of the
thread and reverting the redirection came with some non trivial
complexities, like what to do with unflushed data after the debug
session has finished on the IDE's side.
This diff ass postRunCommands, which are the counterpart of the preRunCommands. TThey will be executed right after the target is launched or attached correctly, which means that the targets can assume that the target is running.
Differential Revision: https://reviews.llvm.org/D100340
`RichManglingContext::FromCxxMethodName` allocates a m_cxx_method_parser, but never deletes it.
This fixes a `-DLLVM_USE_SANITIZER=Leaks` failure.
Reviewed By: teemperor
Differential Revision: https://reviews.llvm.org/D100795
Add initial tests for reading register sets from core dumps. This
includes a C++ program to write registers and dump core, resulting core
dumps for Linux, FreeBSD and NetBSD, and the tests to verify them.
The tests are split into generic part, verifying user-specified register
values, and coredump-specific tests that verify memory addresses that
differ for every dump.
At this moment, all platforms support GPRs and FPRs up to XMM for amd64
target. The i386 target does not work on NetBSD at all, and is missing
FPRs entirely on FreeBSD.
Differential Revision: https://reviews.llvm.org/D91963
The test had a race that could cause two threads to end up with the same
"thread local" value. I believe this would not cause the test to fail,
but it could cause it to succeed even when the functionality is broken.
The new implementation removes this uncertainty, and removes a lot of
cruft left over from the time this test was written using pthreads.
Just fixing a few things I noticed as I am working on another feature for format
strings in the prompt: forward decls, adding constexpr constructors, various
checks, and unit tests for FormatEntity::Parse and new Definition constructors,
etc.
Reviewed By: teemperor
Differential Revision: https://reviews.llvm.org/D98153
Support registering multiple callbacks for a single signal. This is
necessary to support multiple co-existing native process instances, with
separate SIGCHLD handlers.
The system signal handler is registered on first request, additional
callback are added on subsequent requests. The system signal handler
is removed when last callback is unregistered.
Differential Revision: https://reviews.llvm.org/D100418
The code used the total number of symbols to create a symbol ID for the
synthetic symbols. This is not correct because the IDs of real symbols
can be higher than their total number, as we do not add all symbols (and
in particular, we never add symbol zero, which is not a real symbol).
This meant we could have symbols with duplicate IDs, which caused
problems if some relocations were referring to the duplicated IDs. This
was the cause of the failure of the test D97786.
This patch fixes the code to use the ID of the highest (last) symbol
instead.
Landing this fix for Augusto Noronha. The code is getting the
Section from 'addr' passed in, but it may have been expressed as
a load address when it was created and Target::ReadMemory tries to
convert it to a Section+offset if that's now possible; use the
Section found from that cleanup if it exists.
Differential Revision: https://reviews.llvm.org/D100850
Use a variable of static storage duration to reference an intentionally
leaked variable. A static data area is in the GC-set of various leak
checkers.
This fixes 3 `check-lldb-shell` tests in a `-DLLVM_USE_SANITIZER={Leaks,Address}` build,
e.g. `test/Shell/Reproducer/TestHomeDir.test`
Differential Revision: https://reviews.llvm.org/D100806
The binary image list contains the following entry when a frame is not
found in any know binary image:
{
"size" : 0,
"source" : "A",
"base" : 0,
"uuid" : "00000000-0000-0000-0000-000000000000"
}
Note that this object is missing the name and path keys. This patch
makes the JSON parser resilient against their absence.
Commiting this patch for Augusto Noronha who is getting set
up still.
This patch changes Target::ReadMemory so the default behavior
when a read is in a Section that is read-only is to fetch the
data from the local binary image, instead of reading it from
memory. Update all callers to use their old preferences
(the old prefer_file_cache bool) using the new API; we should
revisit these calls and see if they really intend to read
live memory, or if reading from a read-only Section would be
equivalent and important for performance-sensitive cases.
rdar://30634422
Differential revision: https://reviews.llvm.org/D100338
Read the number of addressable bits from the qHostInfo packet and use it
to set the code and data address mask in the process. The data
(addressing_bits) is already present in the packet.
Differential revision: https://reviews.llvm.org/D100520
Implement FixCodeAddress and FixDataAddress for ABIMacOSX_arm64 and
ABISysV_arm64 and add missing calls to RegisterContextUnwind. We need
this to unwind on Apple Silicon where libraries like libSystem are
arm64e even when the program being debugged is arm64.
Differential revision: https://reviews.llvm.org/D100521
Add a code and data address mask to Process with respective getters and
setters and a setting that allows the user to specify the mast as a
number of addressable bits. The masks will be used by FixCodeAddress and
FixDataAddress respectively in the ABI classes.
Differential revision: https://reviews.llvm.org/D100515
DWARF allows .dwo file paths to be relative rather than absolute. When
they are relative, DWARF uses DW_AT_comp_dir to find the .dwo
file. DW_AT_comp_dir can also be relative, making the entire search
patch for the .dwo file relative. In this case, LLDB currently
searches relative to its current working directory, i.e. the directory
from which the debugger was launched. This is not right, as the
compiler, which generated the relative paths, can have no idea where
the debugger will be launched. The correct thing is to search relative
to the location of the executable binary. That is what this patch
does.
Differential Revision: https://reviews.llvm.org/D97786
debugserver's MachTask::DeallocateMemory when removing an
allocate entry from our map (in resposne to an '_m' packet),
copy the size from the entry before removing it from the
map and then using the iterator to fix an ASAN error on
the bots when running TestGdbRemoteMemoryAllocation.py
rdar://76595998
DWARF allows .dwo file paths to be relative rather than absolute. When
they are relative, DWARF uses DW_AT_comp_dir to find the .dwo
file. DW_AT_comp_dir can also be relative, making the entire search
patch for the .dwo file relative. In this case, LLDB currently
searches relative to its current working directory, i.e. the directory
from which the debugger was launched. This is not right, as the
compiler, which generated the relative paths, can have no idea where
the debugger will be launched. The correct thing is to search relative
to the location of the executable binary. That is what this patch
does.
Differential Revision: https://reviews.llvm.org/D97786
The armv6m entry in cores_match() got separated from its
friends armv7m and armv7em. Reuniting them to make it
easier to keep them updated in all at the same time.
Print a single line listing all the categories that are being skipped,
rather than relying on the check.*Support() functions specifying why a
particular category will be skipped. If we know why a category got
skipped, still print that in verbose mode.
The motivation for this change is that sometimes engineers misidentify
the output of these messages as the cause for a test failure (e.g. not
being able to build libc++ or libstdc++).
Differential revision: https://reviews.llvm.org/D100508
We have seen several crashes in LibCppStdFunctionCallableInfo(...) but we don't have a
reproducer. The last crash pointed to last call to line_entry_helper(...) and symbol
was a nullptr. So adding a check for this case.
Progress events internally have a completed count and a total count, which can mean that for a job with 20000 total counts, then there will be 20000 events fired. Sending all these events to the IDE can break it. For example, debugging a huge binary resulted in around 50 million messages, which rendered the IDE useless, as it was spending all of its resources simply parsing messages and updating the UI.
A way to fix this is to send unique percentage updates, which are at most 100 per job, which is not much. I was able to debug that big target and confirm that only unique percentage notifications are sent. I can't write a test for this because the current test is flaky. I'll figure out later how to make the test reliable, but fixing this will unblock us from deploy a new version of lldb-vscode.
Differential Revision: https://reviews.llvm.org/D100443
It looks like the goal of this code is to provide a more precise
architecture definition for the target when attaching to a process. When
attaching to a foreign debugserver, you might get into a situation where
the active (host) platform will give you bogus information on the target
process.
This change allows the platform to override the target arch only with a
compatible architecture. This fixes TestTargetXMLArch.py on Apple
Silicon. Another alternative would be to just fail in this scenario and
update the test(s).
This functionality is used exactly once, and it is trivial to implement
it differently (capture into two distinct variables, and compare for
equality afterwards).
These were in the shared llgs+platform code, but they only make sense
for llgs (as they deal with how the server reports information about
debugged processes).
By checking for cpu and toolchain features ahead
of time we don't need the custom return codes.
Reviewed By: omjavaid
Differential Revision: https://reviews.llvm.org/D97684
Also remove a superfluous semicolon after the braces for a switch
statement (that wasn't warned about).
Differential Revision: https://reviews.llvm.org/D100447
Introduce new m_current_process and m_continue_process variables that
keep the pointers to currently selected process. At this moment, this
is equivalent to m_debugged_process_up but it lays foundations for
the future multiprocess support.
Differential Revision: https://reviews.llvm.org/D100256
These two functions are doing the same thing, only one of them is
sending the packets immediately and the other "queues" them to be sent
later. The first one is better as in case of errors, the backtrace will
point straight to the place that caused them.
Modify the first method to avoid duplication, and ten standardize on it.
This test is flakey because it tries to read the proc/smaps
file of the first lldb-server process it finds. This process
can finish before we finish doing that.
http://lab.llvm.org:8011/#/builders/96/builds/6634/steps/6/logs/stdio
For now limit this to MTE targets which basically means
QEMU via lldb-dotest, which doesn't have this issue.
I'll fix the race condition shortly.
The annotation is now (since the introduction of @apple_simulator_test)
redundant, and the test could theoretically run on lldb-server too (if
it supported darwin hosts).
These tests fail if you build without the x86 llvm backend.
Either because they use an x86 triple or try to backtrace which
requires some x86 knowledge to see all frames.
Reviewed By: labath
Differential Revision: https://reviews.llvm.org/D100194
The original commit was reverted because of the problems it introduced
on Linux. However, FreeBSD should not be affected, so restore that part
and we will address Linux separately.
While at it, remove the dbreg hack as the underlying issue has been
fixed in the FreeBSD kernel and the problem is unlikely to happen
in real life use anyway.
Differential Revision: https://reviews.llvm.org/D98822
Refactor handling qSupported to use a virtual HandleFeatures() method.
The client-provided features are split into an array and passed
to the method. The method returns an array of server features that are
concatenated into the qSupported response to the server.
The base implementation of HandleFeatures()
in GDBRemoteCommunicationServerCommon now includes only flags common
to both platform server and llgs, while llgs-specific flags are inserted
in GDBRemoteCommunicationServerLLGS.
Differential Revision: https://reviews.llvm.org/D100140
Refactor the qSupported handler to split the reply into an array,
and identify features within the array rather than searching the string
for partial matches. While at it, use StringRef.split() to process
the compression list instead of reinventing the wheel.
Switch the arguments to MaybeEnableCompression() to use an ArrayRef
of StringRefs to simplify parameter passing from GetRemoteQSupported().
Differential Revision: https://reviews.llvm.org/D100146
This commit has caused the following tests to be flaky:
TestThreadSpecificBpPlusCondition.py
TestExitDuringExpression.py
The exact cause is not known yet, but since both tests deal with
threads, my guess is it has something to do with the tracking of
creation of new threads (which the commit touches upon).
This reverts the following commits:
d01bff8cbd,
ba62ebc48e,
e761b6b4c5,
a345419ee0.
The core file used is built for i386 so we
need the x86 backend to be able to load it.
Reviewed By: labath
Differential Revision: https://reviews.llvm.org/D100195
Previously the test would fail if you built on Arm/AArch64
but did not have the x86 llvm backend enabled.
Reviewed By: omjavaid
Differential Revision: https://reviews.llvm.org/D100192
By moving them into a folder with a local lit config
requiring x86. All these tests use x86 target triples.
There are two tests that require target-x86_64 because
they run program files (instead of just needing the backend).
Those are moved to the x86 folder also but their REQUIRES are
unchanged.
Reviewed By: JDevlieghere
Differential Revision: https://reviews.llvm.org/D100193
In all this time, we've never used more than one delegate. The logic to
support multiple delegates is therefore untested, and becomes
particularly unwieldy once we need to support multiple processes.
Just remove it.
This reverts commit 3842de49f6.
It fails to build, with errors such as:
GDBRemoteCommunicationClient.cpp:1005:20:
error: no viable overloaded '='
avail_name = compression;
Refactor the qSupported handler to split the reply into an array,
and identify features within the array rather than searching the string
for partial matches. While at it, use StringRef.split() to process
the compression list instead of reinventing the wheel.
Switch the arguments to MaybeEnableCompression() to use an ArrayRef
of StringRefs to simplify parameter passing from GetRemoteQSupported().
Differential Revision: https://reviews.llvm.org/D100146
That code is unused since it's check-in in 2010 (and I believe it would leak
memory when called as it releases the passed unique_ptr), so let's delete it.
Reviewed By: vsk
Differential Revision: https://reviews.llvm.org/D100212
When LLDB's DWARF parser is parsing the member DIEs of a struct/class it
currently fully resolves the types of static member variables in a class before
adding the respective `VarDecl` to the record.
For record types fully resolving the type will also parse the member DIEs of the
respective class. The other way of resolving is just 'forward' resolving the type
which will try to load only the minimum amount of information about the type
(for records that would only be the name/kind of the type). Usually we always
resolve types on-demand so it's rarely useful to speculatively fully resolve
them on the first use.
This patch changes makes that we only 'forward' resolve the types of static
members. This solves the fact that LLDB unnecessarily loads debug information
to parse the type if it's maybe not needed later and it also avoids a crash where
the parsed type might in turn reference the surrounding class that is currently
being parsed.
The new test case demonstrates the crash that might happen. The crash happens
with the following steps:
1. We parse class `ToLayout` and it's members.
2. We parse the static class member and fully resolve its type
(`DependsOnParam2<ToLayout>`).
3. That type has a non-static class member `DependsOnParam1<ToLayout>` for which
LLDB will try to calculate the size.
4. The layout (and size)`DependsOnParam1<ToLayout>` turns depends on the
`ToLayout` size/layout.
5. Clang will calculate the record layout/size for `ToLayout` even though we are
currently parsing it and it's missing it's non-static member.
The created is missing the offset for the yet unparsed non-static member. If we
later try to get the offset we end up hitting different asserts. Most common is
the one in `TypeSystemClang::DumpValue` where it checks that the record layout
has offsets for the current FieldDecl.
```
assert(field_idx < record_layout.getFieldCount());
```
Fixed rdar://67910011
Reviewed By: shafik
Differential Revision: https://reviews.llvm.org/D100180
When debugging LanguageRuntime unwindplans, it can be
helpful to disable their use and see the normal
stack walk. Add a setting for this.
Differential Revision: https://reviews.llvm.org/D99828
Watch for fork(2)/vfork(2) (also fork/vfork-style clone(2) on Linux)
notifications and explicitly detach the forked child process, and add
initial tests for these cases. The code covers FreeBSD, Linux
and NetBSD process plugins. There is no new user-visible functionality
provided -- this change lays foundations over subsequent work on fork
support.
Differential Revision: https://reviews.llvm.org/D98822
If the debug info is missing the terminating null die, we would crash
when trying to access the nonexisting children/siblings. This was
discovered because the test case for D98619 accidentaly produced such
input.
Add a minimal support for the multiprocess extension in gdb-remote
client. It accepts PIDs as part of thread-ids, and rejects PIDs that
do not match the current inferior.
Differential Revision: https://reviews.llvm.org/D99603
A little cleanup to how these firmware corefile tests are done; add
a test that loads a dSYM that loads an OS plugin, and confirm that
the OS plugin's threads are created.
When looking up user specified formatters qualifiers are removed from types before matching,
I have added a clarifying example to the document and added an example to a relevant test to demonstrate this behavior.
Differential Revision: https://reviews.llvm.org/D99827
An empty history entry can happen by entering the expression evaluator an immediately hitting enter:
```
$ lldb
(lldb) e
Enter expressions, then terminate with an empty line to evaluate:
1: <hit enter>
```
The next time the user enters the expression evaluator, if they hit the up arrow to load the previous expression, lldb crashes. This patch treats empty history sessions as a single expression of zero length, instead of an empty list of expressions.
Fixes http://llvm.org/PR49845.
Differential Revision: https://reviews.llvm.org/D100048
The memory read --outfile command should truncate the output when unless
--append-outfile. Fix the bug and add a test.
rdar://76062318
Differential revision: https://reviews.llvm.org/D99890
Problem:
On SystemZ we need to open text files in text mode. On Windows, files opened in text mode adds a CRLF '\r\n' which may not be desirable.
Solution:
This patch adds two new flags
- OF_CRLF which indicates that CRLF translation is used.
- OF_TextWithCRLF = OF_Text | OF_CRLF indicates that the file is text and uses CRLF translation.
Developers should now use either the OF_Text or OF_TextWithCRLF for text files and OF_None for binary files. If the developer doesn't want carriage returns on Windows, they should use OF_Text, if they do want carriage returns on Windows, they should use OF_TextWithCRLF.
So this is the behaviour per platform with my patch:
z/OS:
OF_None: open in binary mode
OF_Text : open in text mode
OF_TextWithCRLF: open in text mode
Windows:
OF_None: open file with no carriage return
OF_Text: open file with no carriage return
OF_TextWithCRLF: open file with carriage return
The Major change is in llvm/lib/Support/Windows/Path.inc to only set text mode if the OF_CRLF is set.
```
if (Flags & OF_CRLF)
CrtOpenFlags |= _O_TEXT;
```
These following files are the ones that still use OF_Text which I left unchanged. I modified all these except raw_ostream.cpp in recent patches so I know these were previously in Binary mode on Windows.
./llvm/lib/Support/raw_ostream.cpp
./llvm/lib/TableGen/Main.cpp
./llvm/tools/dsymutil/DwarfLinkerForBinary.cpp
./llvm/unittests/Support/Path.cpp
./clang/lib/StaticAnalyzer/Core/HTMLDiagnostics.cpp
./clang/lib/Frontend/CompilerInstance.cpp
./clang/lib/Driver/Driver.cpp
./clang/lib/Driver/ToolChains/Clang.cpp
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D99426
Use a with block for reading the cpuinfo file.
When loading the file fails (or we're not on Linux)
return an empty string. Since all the callers are
going to do "x in self.getCPUInfo()".
Reviewed By: omjavaid
Differential Revision: https://reviews.llvm.org/D99729
This reverts commit 602ab188a7.
The patch replicated an lldbassert for a certain type of NSNumber for tagged
pointers. This really shouldn't be an assert since we don't do anything wrong
with these numbers, we just don't print a summary. So this patch changed the
lldbassert to a log message in reverting the revert.
When referencing `NSObject`, it's enough to import `objc/NSObject.h`. Importing `Foundation` is unnecessary in these cases.
Differential Revision: https://reviews.llvm.org/D99867
Use `@import ObjectiveC` instead of `@import Foundation`, as the former is all
that's needed, and results in fewer clang modules being built.
This results in the following clang modules *not* being built for this test.
ApplicationServices
CFNetwork
ColorSync
CoreFoundation
CoreGraphics
CoreServices
CoreText
DiskArbitration
Dispatch
Foundation
IOKit
ImageIO
Security
XPC
_Builtin_intrinsics
launch
libkern
os_object
os_workgroup
Differential Revision: https://reviews.llvm.org/D99859
This reverts commit 4d9039c8dc.
This is causing the greendragon bots to fail most of the time when
running TestNSDictionarySynthetic.py. Reverting until Jim has a chance
to look at this on Monday. Running the commands from that test from
the command line, it fails 10-13% of the time on my desktop.
This is a revert of Jim's changes in https://reviews.llvm.org/D99694
On macOS Catalina, calling objc_debug_class_getNameRaw on some of the
ISA pointers returns NULL, causing us to crash and unwind before reading
all the Objective-C classes. This does not happen on macOS Big Sur.
Account for that possibility and skip the class when that happens.
Fill out ProcessMachCore::DoLoadCore to handle LC_NOTE hints with
a UUID or with a UUID+address, and load the binary at the specified
offset correctly. Add tests for all four combinations. Change
DynamicLoaderStatic to not re-set a Section's load address in the
Target if it's already been specified.
Differential Revision: https://reviews.llvm.org/D99571
rdar://51490545
Inline callstacks were being incorrectly displayed in the results of "image lookup --address". The deepest frame wasn't displaying the line table line entry, it was always showing the inline information's call file and line on the previous frame. This is now fixed and has tests to make sure it doesn't regress.
Differential Revision: https://reviews.llvm.org/D98761
Since quite a while Apple's LLDB fork (that contains the Swift debugging
support) is randomly crashing in `CommandLineParser::addOption` with an error
such as `CommandLine Error: Option 'h' registered more than once!`
The backtrace of the crashing thread is shown below. There are also usually many
other threads also performing similar clang::FrontendActions which are all
trying to generate (usually outdated) Clang modules which are used by Swift for
various reasons.
```
[ 6] LLDB`CommandLineParser::addOption(llvm:🆑:Option*, llvm:🆑:SubCommand*) + 856
[ 7] LLDB`CommandLineParser::addOption(llvm:🆑:Option*, llvm:🆑:SubCommand*) + 733
[ 8] LLDB`CommandLineParser::addOption(llvm:🆑:Option*, bool) + 184
[ 9] LLDB`llvm:🆑:ParseCommandLineOptions(...) [inlined] ::CommandLineParser::ParseCommandLineOptions(... + 1279
[ 9] LLDB`llvm:🆑:ParseCommandLineOptions(...) + 497
[ 10] LLDB`setCommandLineOpts(clang::CodeGenOptions const&) + 416
[ 11] LLDB`EmitAssemblyHelper::EmitAssemblyWithNewPassManager(...) + 98
[ 12] LLDB`clang::EmitBackendOutput(...) + 4580
[ 13] LLDB`PCHContainerGenerator::HandleTranslationUnit(clang::ASTContext&) + 871
[ 14] LLDB`clang::MultiplexConsumer::HandleTranslationUnit(clang::ASTContext&) + 43
[ 15] LLDB`clang::ParseAST(clang::Sema&, bool, bool) + 579
[ 16] LLDB`clang::FrontendAction::Execute() + 74
[ 17] LLDB`clang::CompilerInstance::ExecuteAction(clang::FrontendAction&) + 1808
```
The underlying reason for the crash is that the CommandLine code in LLVM isn't
thread-safe and will never be thread-safe with its current architecture. The way
LLVM's CommandLine logic works is that all parts of the LLVM can provide command
line arguments by defining `cl::opt` global variables and their constructors
(which are invoked during static initialisation) register the variable in LLVM's
CommandLineParser (which is also just a global variable). At some later point
after static initialization we actually try to parse command line arguments and
we ask the CommandLineParser to parse our `argv`. The CommandLineParser then
lazily constructs it's internal parsing state in a non-thread-safe way (this is
where the crash happens), parses the provided command line and then goes back to
the respective `cl::opt` global variables and sets their values according to the
parse result.
As all of this is based on global state, this whole mechanism isn't thread-safe
so the only time to ever use it is when we know we only have one active thread
dealing with LLVM logic. That's why nearly all callers of
`llvm:🆑:ParseCommandLineOptions` are at the top of the `main` function of the
some LLVM-based tool. One of the few exceptions to this rule is in the
`setCommandLineOpts` function in `BackendUtil.cpp` which is in our backtrace:
```
static void setCommandLineOpts(const CodeGenOptions &CodeGenOpts) {
SmallVector<const char *, 16> BackendArgs;
BackendArgs.push_back("clang"); // Fake program name.
if (!CodeGenOpts.DebugPass.empty()) {
BackendArgs.push_back("-debug-pass");
BackendArgs.push_back(CodeGenOpts.DebugPass.c_str());
}
if (!CodeGenOpts.LimitFloatPrecision.empty()) {
BackendArgs.push_back("-limit-float-precision");
BackendArgs.push_back(CodeGenOpts.LimitFloatPrecision.c_str());
}
BackendArgs.push_back(nullptr);
llvm:🆑:ParseCommandLineOptions(BackendArgs.size() - 1,
BackendArgs.data());
}
```
This is trying to set `cl::opt` variables in the LLVM backend to their right
value as the passed via CodeGenOptions by invoking the CommandLine parser. As
this is just in some generic Clang CodeGen code (where we allow having multiple
threads) this is code is clearly wrong. If we're unlucky it either overwrites
the value of the global variables or it causes the CommandLine parser to crash.
So the next question is why is this only crashing in LLDB? The main reason seems
to be that easiest way to crash this code is to concurrently enter the initial
CommandLineParser construction where it tries to collect all the registered
`cl::opt` options and checks for sanity:
```
// If it's a DefaultOption, check to make sure it isn't already there.
if (O->isDefaultOption() &&
SC->OptionsMap.find(O->ArgStr) != SC->OptionsMap.end())
return;
// Add argument to the argument map!
if (!SC->OptionsMap.insert(std::make_pair(O->ArgStr, O)).second) {
errs() << ProgramName << ": CommandLine Error: Option '" << O->ArgStr
<< "' registered more than once!\n";
HadErrors = true;
}
```
The `OptionsMap` here is global variable and if we end up in this code with two
threads at once then two threads at the same time can register an option (such
as 'h') when they pass the first `if` and then we fail with the sanity check in
the second `if`.
After this sanity check and initial setup code the only remaining work is just
parsing the provided CommandLine which isn't thread-safe but at least doesn't
crash in all my attempts at breaking it (as it's usually just reading from the
already generated parser state but not further modifying it). The exception to
this is probably that once people actually specify the options in the code
snippet above we might run into some new interesting ways to crash everything.
To go back to why it's only affecting LLDB: Nearly all LLVM tools I could find
(even if they are using threads) seem to call the CommandLine parser at the
start so they all execute the initial parser setup at a point where there is
only one thread. So once the code above is executed they are mostly safe from
the sanity check crashes. We even have some shady code for the gtest `main` in
`TestMain.cpp` which is why this also doesn't affect unit tests.
The only exception to this rule is ... *drum roll* ... LLDB! it's not using that
CommandLine library for parsing options so it also never ends up calling it in
`main`. So when we end up in the `FrontendAction` code from the backtrace we are
already very deep in some LLDB logic and usually already have several threads.
In a situation where Swift decides to compile a large amount of Clang modules in
parallel we then end up entering this code via several threads. If several
threads reach this code at the same time we end up in the situation where the
sanity-checking code of CommandLine crashes. I have a very reliable way of
demonstrating the whole thing in D99650 (just run the unit test several times,
it usually crashes after 3-4 attempts).
We have several ways to fix this:
1. Make the whole CommandLine mechanism in LLVM thread-safe.
2. Get rid of `setCommandLineOpts` in `BackendUtil.cpp` and other callers of the
command line parsing in generic Clang code.
3. Initialise the CommandLine library in a safe point in LLDB.
Option 1 is just a lot of work and I'm not even sure where to start. The whole
mechanism is based on global variables and global state and this seems like a
humongous task.
Option 2 is probably the best thing we can do in the near future. There are only
two callers of the command line parser in generic Clang code. The one in
`BackendUtils.cpp` looks like it can be replaced with some reasonable
refactoring (as it only deals with two specific options). There is another one
in `ExecuteCompilerInvocation` which deals with forwarding the generic `-mllvm`
options to the backend which seems like it will just end up requiring us to do
Option 1.
Option 3 is what this patch is doing. We just parse some dummy command line
invocation in a point of the LLDB execution where we only have one thread that
is dealing with LLVM/Clang stuff. This way we are at least prevent the frequent
crashes for users as parsing the dummy command line invocation will set up the
initial parser state safely.
Fixes rdar://70989856
Reviewed By: mib, JDevlieghere
Differential Revision: https://reviews.llvm.org/D99652
The ObjC runtime offers both signed & unsigned tagged pointer value
accessors to tagged pointer providers, but lldb's tagged pointer
code only implemented the unsigned one. This patch adds an
emulation of the signed one.
The motivation for doing this is that NSNumbers use the signed
accessor (they are always signed) and we need to follow that in our
summary provider or we will get incorrect values for negative
NSNumbers.
The data-formatter-objc test file had NSNumber examples (along with lots of other
goodies) but the NSNumber values weren't tested. So I also added
checks for those values to the test.
I also did a quick audit of the other types in that main.m file, and
it looks like pretty much all the other values are either intermediates
or are tested.
Differential Revision: https://reviews.llvm.org/D99694
Respect --apple-sdk <path> if it's specified. If the SDK is simply
mounted from some disk image, and not actually installed, this is the
only way to use it.
Differential Revision: https://reviews.llvm.org/D99746
Debugging tests sometimes involves debugging the Python source. This adds a paragraph to
the "Debugging Test Failures" section about using `pdb`, and also describes how to run
lldb commands from pdb.
Differential Revision: https://reviews.llvm.org/D99744
and probably other posix oses. Use extra_images to ensure
LD_LIBRARY_PATH is set correctly.
Also take the opportunity to remove hand-rolled library extension
management code in favor of the existing one.
The test uses debug info from one binary to debug a different one. This
does not work on macos, and its pure luck that it works elsewhere (the
variable that it inspects happens to have the same address in both).
The purpose of this test is to verify that lldb has not overwritten the
target executable. That can be more easily achieved by checking the exit
code of the binary, so change the test to do that.
Also remove the llgs_test decorator, as it's preventing the test from
running on macos. All the test needs is the platform functionality of
lldb-server, which is available everywhere.
This fixes flakiness in TestVSCode_launch.test_progress_events
vscode.progress_events some times failed to populate in time for
follow up iterations.
Adding a minor delay before the the for the loop fixes the issue.
Reviewed By: clayborg
Differential Revision: https://reviews.llvm.org/D99497
This reverts commit 71b648f715.
There was a typo in the last commit which was causing LLDB AArch64 Linux
buildbot testsuite failures. Now fixed in current version.
This fixes (works around) two errors with gcc-6.5.
- in the RegisterContext_x86 files, gcc is unable to synthesize a
default constructor -- it thinks it needs to initialize the virtual
base class, even though said classes are abstract. I fix that by
providing a dummy constructor.
- In ReproducerInstrumentationTest, it is not able to deduce that the
TestingRegistry class is movable (it contains a map of unique
pointers). I change the type from Optional<TestingRegistry> to
unique_ptr<TestingRegistry), so that moving is not required
(copying/moving a polymorphic type is not a very good idea in any
case).
Remove the remaining references to LLDB_CAPTURE_REPRODUCER. I removed
the functionality in an earlier commit but forgot that there was a
corresponding test and logic to unset it in our test suite.
Consistently use return with EXIT_SUCCESS or EXIT_FAILURE instead of
mix-and-matching return, exit 0, 1 etc.
Differential revision: https://reviews.llvm.org/D99701
Remove the LLDB_CAPTURE_REPRODUCER as it is inherently dangerous. The
reproducers require careful initialization which cannot be guaranteed by
overwriting the reproducer mode at this level.
If we want to provide this functionality, we should do it in the driver
instead. It was originally added to enable capture in CI, but we now
have a dedicated CI job that captures and replays the test suite.
This implements the interactive trace start and stop methods.
This diff ended up being much larger than I anticipated because, by doing it, I found that I had implemented in the beginning many things in a non optimal way. In any case, the code is much better now.
There's a lot of boilerplate code due to the gdb-remote protocol, but the main changes are:
- New tracing packets: jLLDBTraceStop, jLLDBTraceStart, jLLDBTraceGetBinaryData. The gdb-remote packet definitions are quite comprehensive.
- Implementation of the "process trace start|stop" and "thread trace start|stop" commands.
- Implementaiton of an API in Trace.h to interact with live traces.
- Created an IntelPTDecoder for live threads, that use the debugger's stop id as checkpoint for its internal cache.
- Added a functionality to stop the process in case "process tracing" is enabled and a new thread can't traced.
- Added tests
I have some ideas to unify the code paths for post mortem and live threads, but I'll do that in another diff.
Differential Revision: https://reviews.llvm.org/D91679
This patch adds a test case to test AArch64 dynamic register sets.
This tests for the availability of certain register sets and query
their registers accordingly.
Reviewed By: labath, DavidSpickett
Differential Revision: https://reviews.llvm.org/D96463
This patch adds two new dynamic register sets for AArch64 MTE and
Pointer Authentication features. These register sets are dynamic and
will only be available if underlying hardware support either of these
features. LLDB will pull in Aux vector information and create register
infos based on that information.
A follow up patch will add a test case to test these feature registers.
Reviewed By: labath, DavidSpickett
Differential Revision: https://reviews.llvm.org/D96460
This is patch adds support for adding dynamic register sets for
AArch64 dynamic features in LLDB. AArch64 has optional features like
SVE, Pointer Authentication and MTE which means LLDB needs to decide
at run time which registers it needs to pull in for the current
executable based on underlying support for a certain feature.
This patch makes necessary adjustments to make way for dynamic
register infos and dynamic register sets.
Reviewed By: labath
Differential Revision: https://reviews.llvm.org/D96458
This patch fixes an issue, where if the thread has a signal blocked when
we try to inject it into the process (via vCont), then instead of
executing straight away, the injected signal will trigger another stop
when the thread unblocks the signal.
As (linux) threads start their life with SIGUSR1 (among others)
disabled, and only enable it during initialization, injecting the signal
during this window did not behave as expected. The fix is to change the
test to ensure the signal gets injected with the signal unblocked.
The simplest way to do this was to write a dedicated inferior for this
test. I also created a new header to factor out the function retrieving
the (os-specific) thread id.
As discussed on lldb-dev
<https://lists.llvm.org/pipermail/lldb-dev/2021-March/016777.html> the
mips code is unmaintained and untested. It also carries a lot of
technical debt which is not limited to mips-specific code.
Generic mips support remains (and is going to be used by the upcoming
freebsd code). Resurrecting mips support should be a matter of re-adding
the relevant register context files (while avoiding reintroducing the
debt).
Add a minimal support for the multiprocess extension in lldb-server.
The server indicates support for it via qSupported, and accepts
thread-ids containing a PID. However, it still does not support
debugging more than one inferior, so any other PID value results
in an error.
Differential Revision: https://reviews.llvm.org/D98482
TestVSCode_disconnect.test_launch fails with clean up error because
disconnect gets called twice once from the test case and once from
the tear down hook.
This patch disables disconnect after its been called from test_launch
Reviewed By: clayborg
Differential Revision: https://reviews.llvm.org/D99491
This is apparently allowed in Objective-C so we should test this in LLDB.
Reviewed By: teemperor
Differential Revision: https://reviews.llvm.org/D99513
This test is not useful as the functions it's testing are just returning
a constant. It also fails in unoptimized builds as it's comparing
character strings by address.
Remove the "depth" variable, as the same information can be obtained
through die_index_stack.size().
Also add a test case for a one tricky case I noticed -- a unit
containing only a null unit die.
When LLVM error handling was introduced to the parsing of the .debug_aranges it would cause major issues if any DWARFDebugArangeSet::extract() calls returned any errors. The code in DWARFDebugInfo::GetCompileUnitAranges() would end up calling DWARFDebugAranges::extract() which would return an error if _any_ DWARFDebugArangeSet had any errors, but it default constructed a DWARFDebugAranges object into DWARFDebugInfo::m_cu_aranges_up and populated it partially, and returned an error prior to finishing much needed functionality in the DWARFDebugInfo::GetCompileUnitAranges() function. Subsequent callers to this function would see that the DWARFDebugInfo::m_cu_aranges_up was actually valid and return this partially populated DWARFDebugAranges reference _and_ it would not be sorted or minimized.
This above bugs would cause an incomplete .debug_aranges parsing, it would skip manually parsing any compile units for ranges, and would not sort the DWARFDebugAranges in m_cu_aranges_up.
This bug would also cause breakpoints set by file and line to fail to set correctly if a symbol context for an address could not be resolved properly, which the incomplete and unsorted DWARFDebugAranges object that DWARFDebugInfo::GetCompileUnitAranges() returned would cause symbol context lookups resolved by address (breakpoint address) to fail to find any DWARF debug info for a given address.
This patch fixes all of the issues that I found:
- DWARFDebugInfo::GetCompileUnitAranges() no longer returns a "llvm::Expected<DWARFDebugAranges &>", but just returns a "const DWARFDebugAranges &". Why? Because this code contained a fallback that would parse all of the valid DWARFDebugArangeSet objects, and would check which compile units had valid .debug_aranges set entries, and manually build an address ranges table using DWARFUnit::BuildAddressRangeTable(). If we return an error because any DWARFDebugArangeSet has any errors, then we don't do any of this code. Now we parse all DWARFDebugArangeSet objects that have no errors, if any calls to DWARFDebugArangeSet::extract() return errors, we skip that DWARFDebugArangeSet so that we can use the fallback call to DWARFUnit::BuildAddressRangeTable(). Since DWARFDebugInfo::GetCompileUnitAranges() needs to parse what it can from the .debug_aranges and build address ranges tables for any compile units that don't have any .debug_aranges sets, everything now works as expected.
- Fix an issue where a DWARFDebugArangeSet contains multiple terminator entries. The LLVM parser and llvm-dwarfdump properly warn about this because it happens with linux compilers and linkers and was the original cause of the bug I am fixing here. We now correctly warn about this issue if "log enable dwarf info" is enabled, but we continue to parse the DWARFDebugArangeSet correctly so we don't lose data that is contained in the .debug_aranges section.
- DWARFDebugAranges::extract() no longer returns a llvm::Error because we need to be able to parse all of the valid DWARFDebugArangeSet objects. It also will correctly skip a DWARFDebugArangeSet object that has errors in the middle of the stream by setting the start offsets of each DWARFDebugArangeSet to be calculated by the previous DWARFDebugArangeSet::extract() calculated offset that uses the header which contains the length of the DWARFDebugArangeSet. This means if do we run into real errors while parsing individual DWARFDebugArangeSet objects, we can continue to parse the rest of the validly encoded DWARFDebugArangeSet objects in the .debug_aranges section. This will allow LLDB to parse DWARF that contains a possibly newer .debug_aranges set format than LLDB currently supports because we will error out for the parsing of the DWARFDebugArangeSet, but be able to skip to the next DWARFDebugArangeSet object using the "DWARFDebugArangeSet.m_header.length" field to calculate the next starting offset.
Tests were added to cover all new functionality.
Differential Revision: https://reviews.llvm.org/D99401
Print LLVM's pretty stack trace when lldb-vscode crashes. Also removes
the unnecessary call to PrintStackTraceOnErrorSignal in lldb-server as
it's already part of InitLLVM.
Differential revision: https://reviews.llvm.org/D99535
Raphael Isemann