**This patch adds a marker to make hidden frames more explicit.**
---
Hidden frames can be confusing for some users, who see that the indexes
of the frames in a backtrace are not contiguous. This patch aims to
lessen the confusion by adding a delimiter for the first and last non
hidden frame, i.e the boundaries.
IDE's like Xcode and VSCode represent those in the UI by having the
hidden frames either greyed out or collapsed.
It's not possible to do this in the CLI, therefore, this patch makes use
of 2 unicode characters to mark the beginning and end of the hidden
frames range.
This patch depends on:
- https://github.com/llvm/llvm-project/pull/168603
# Examples
In the example below, frame `#2` to `#7` are is hidden, and therefore,
frame `#1` is the first non hidden frame of the range while frame `#8`
is the last non hidden frame:
<img width="488" height="112" alt="Screenshot 2025-11-18 at 18 41 11"
src="https://github.com/user-attachments/assets/a21431da-9729-4cf0-a6bc-024aa306fc45"
/>
If the selected frame is one of the 2 boundary frames, we replace the
delimiter character with the select character (`*`).
<img width="487" height="111" alt="Screenshot 2025-11-18 at 18 41 03"
src="https://github.com/user-attachments/assets/5616fa81-6db6-457d-9d1e-bbe46e710c26"
/>
<img width="488" height="111" alt="Screenshot 2025-11-18 at 18 40 55"
src="https://github.com/user-attachments/assets/93dfa6cf-0956-4718-b31c-f965ec72b56d"
/>
This patch allows threads to have multiple SyntheticFrameProviderSP
instances that chain together sequentially. Each provider receives the
output of the previous provider as input, creating a transformation
pipeline.
It changes `Thread::m_frame_provider_sp` to a vector, adds provider
parameter to SyntheticStackFrameList to avoid calling back into
`Thread::GetFrameProvider()` during frame fetching, updated
`LoadScriptedFrameProvider()` to chain providers by wrapping each
previous provider's output in a `SyntheticStackFrameList` for the next
provider and finally, loads ALL matching providers in priority order
instead of just the first one.
The chaining works as follows:
```
Real Unwinder Frames
↓
Provider 1 (priority 10) → adds/transforms frames
↓
Provider 2 (priority 20) → transforms Provider 1's output
↓
Provider 3 (priority 30) → transforms Provider 2's output
↓
Final frame list shown to user
```
This patch also adds a test for this (test_chained_frame_providers) to verify that 3 providers chain correctly: `AddFooFrameProvider`, `AddBarFrameProvider`, `AddBazFrameProvider`.
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
Scripted frames that materialize Python functions or other non-native
code are PC-less by design, meaning they don't have valid program
counter values. Previously, these frames would display invalid addresses
(`0xffffffffffffffff`) in backtrace output.
This patch updates `FormatEntity` to detect and suppress invalid address
display for PC-less frames, adds fallback to frame methods when symbol
context is unavailable, and modifies `StackFrame::GetSymbolContext` to
skip PC-based symbol resolution for invalid addresses.
The changes enable PC-less frames to display cleanly with proper
function names, file paths, and line numbers, and allow for source
display of foreign sources (like Python). Includes comprehensive test
coverage demonstrating frames pointing to Python source files.
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
This change adds tracking of the StackFrameList that produced each frame
by storing a weak pointer (m_frame_list_wp) in both `StackFrame` and
`ExecutionContextRef`.
When resolving frames through `ExecutionContextRef::GetFrameSP`, the
code now first attempts to use the remembered frame list instead of
immediately calling `Thread::GetStackFrameList`. This breaks circular
dependencies that can occur during frame provider initialization, where
creating a frame provider might trigger `ExecutionContext` resolution,
which would then call back into `Thread::GetStackFrameList()`, creating
an infinite loop.
The `StackFrameList` now sets m_frame_list_wp on every frame it creates,
and a new virtual method `GetOriginatingStackFrameList` allows frames to
expose their originating list.
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
This patch extends ScriptedFrame to work with real (non-scripted)
threads,
enabling frame providers to synthesize frames for native processes.
Previously, ScriptedFrame only worked within
ScriptedProcess/ScriptedThread
contexts. This patch decouples ScriptedFrame from ScriptedThread,
allowing
users to augment or replace stack frames in real debugging sessions for
use
cases like custom calling conventions, reconstructing corrupted frames
from
core files, or adding diagnostic frames.
Key changes:
- ScriptedFrame::Create() now accepts ThreadSP instead of requiring
ScriptedThread, extracting architecture from the target triple rather
than ScriptedProcess.arch
- Added SBTarget::RegisterScriptedFrameProvider() and
ClearScriptedFrameProvider() APIs, with Target storing a
SyntheticFrameProviderDescriptor template for new threads
- Added "target frame-provider register/clear" commands for CLI access
- Thread class gains LoadScriptedFrameProvider(),
ClearScriptedFrameProvider(),
and GetFrameProvider() methods for per-thread frame provider management
- New SyntheticStackFrameList overrides FetchFramesUpTo() to lazily
provide
frames from either the frame provider or the real stack
This enables practical use of the SyntheticFrameProvider infrastructure
in
real debugging workflows.
rdar://161834688
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
[lldb] Track CFA pointer metadata in StackID
In this commit:
9c8e71644227 [lldb] Make StackID call Fix{Code,Data} pointers (#152796)
We made StackID keep track of the CFA without any pointer metadata in
it. This is necessary when comparing two StackIDs to determine which one
is "younger".
However, the CFA inside StackIDs is also used in other contexts through
the method StackID::GetCallFrameAddress. One notable case is
DWARFExpression: the computation of `DW_OP_call_frame_address` is done
using StackID. This feeds into many other places, e.g. expression
evaluation may require the address of a variable that is computed from
the CFA; to access the variable without faulting, we may need to
preserve the pointer metadata. As such, StackID must be able to provide
both versions of the CFA.
In the spirit of allowing consumers of pointers to decide what to do
with pointer metadata, this patch changes StackID to store both versions
of the cfa pointer. Two getter methods are provided, and all call sites
except DWARFExpression preserve their existing behavior (stripped
pointer). Other alternatives were considered:
* Just store the raw pointer. This would require changing the
comparisong operator `<` to also receive a Process, as the comparison
requires stripped pointers. It wasn't clear if all call-sites had a
non-null process, whereas we know we have a process when creating a
StackID.
* Store a weak pointer to the process inside the class, and then strip
metadata as needed. This would require a `weak_ptr::lock` in many
operations of LLDB, and it felt wasteful. It also prevents stripping
of the pointer if the process has gone away.
This patch also changes RegisterContextUnwind::ReadFrameAddress, which
is the method computing the CFA fed into StackID, to also preserve the
signature pointers.
The motivation for this patch is that
`StopInfo::GetSuggestedStackFrameIndex` would not take effect for
`InstrumentationRuntimeStopInfo` (which we plan to implement in
https://github.com/llvm/llvm-project/pull/133079). This was happening
because the instrumentation runtime plugins would run utility
expressions as part of the stop that would set the
`m_selected_frame_idx`. This means `SelectMostRelevantFrame` was never
called, and we would not be able to report the selected frame via the
`StopInfo` object.
This patch makes sure we clear the `m_selected_frame_idx` to allow
`GetSuggestedStackFrameIndex` to take effect, regardless of what the
frame recognizers choose to do.
This patch changes the way frames created from scripted affordances like
Scripted Threads are displayed. Currently, they're marked artificial
which is used usually for compiler generated frames.
This patch changes that behaviour by introducing a new synthetic
StackFrame kind and moves 'artificial' to be a distinct StackFrame
attribut.
On top of making these frames less confusing, this allows us to know
when a frame was created from a scripted affordance.
rdar://155949703
Signed-off-by: Med Ismail Bennani <ismail@bennani.ma>
Protected m_selected_frame_idx with a mutex. To avoid deadlocks, always acquire m_selected_frame_mutex after m_list_mutex. I'm using a recursive mutex because GetSelectedFrameIndex may indirectly call SetSelectedFrame.
This is an odd corner case of the use of scripts loaded from dSYM's - a
macOS only feature, which can load OS Plugins that re-present the thread
state of the program we attach to. If we find out about and load the
dSYM scripts when we discover a target in the course of attaching to it,
we can end up running the OS plugin before we've started up the private
state thread. However, the os_plugin in that case will be running before
we broadcast the stop event to the public event listener. So it should
formally use the private state and not the public state for the Python
code environment.
This patch says that if we have not yet started up the private state
thread, then any thread that is servicing events is doing so on behalf
of the private state machinery, and should see the private state, not
the public state.
Most of the patch is getting a test that will actually reproduce the
error. Only the test `test_python_os_plugin_remote` actually reproduced
the error. In `test_python_os_plugin` we actually do start up the
private state thread before handling the event. `test_python_os_plugin`
is there for completeness sake.
In fact, there's only one public API in StackFrameList that changes
the list explicitly. The rest only change the list if you happen to
ask for more frames than lldb has currently fetched and that
always adds frames "behind the user's back". So we were
much more prone to deadlocking than we needed to be.
This patch uses a shared_mutex instead, and when we have to add more
frames (in GetFramesUpTo) we switches to exclusive long enough to add
the frames, then goes back to shared.
Most of the work here was actually getting the stack frame list locking
to not
require a recursive mutex (shared mutexes aren't recursive).
I also added a test that has 5 threads progressively asking for more
frames simultaneously to make sure we get back valid frames and don't
deadlock.
This fixes the two test suite failures that I missed in the PR:
https://github.com/llvm/llvm-project/pull/112939
One was a poorly written test case - it assumed that on connect to a
gdb-remote with a running process, lldb MUST have fetched all the frame
0 registers. In fact, there's no need for it to do so (as the CallSite
patch showed...) and if we don't need to we shouldn't. So I fixed the
test to only expect a `g` packet AFTER calling read_registers.
The other was a place where some code had used 0 when it meant
LLDB_INVALID_LINE_NUMBER, which I had fixed but missed one place where
it was still compared to 0.
…ne stepping (#112939)"
This was breaking some gdb-remote packet counting tests on the bots. I
can't see how this patch could cause that breakage, but I'm reverting to
figure that out.
This reverts commit f14743794587db102c6d1b20f9c87a1ac20decfd.
Previously lldb didn't support setting breakpoints on call site
locations. This patch adds that ability.
It would be very slow if we did this by searching all the debug
information for every inlined subroutine record looking for a call-site
match, so I added one restriction to the call-site support. This change
will find all call sites for functions that also supply at least one
line to the regular line table. That way we can use the fact that the
line table search will move the location to that subsequent line (but
only within the same function). When we find an actually moved source
line match, we can search in the function that contained that line table
entry for the call-site, and set the breakpoint location back to that.
When I started writing tests for this new ability, it quickly became
obvious that our support for virtual inline stepping was pretty buggy.
We didn't print the right file & line number for the breakpoint, and we
didn't set the position in the "virtual inlined stack" correctly when we
hit the breakpoint. We also didn't step through the inlined frames
correctly. There was code to try to detect the right inlined stack
position, but it had been refactored a while back with the comment that
it was super confusing and the refactor was supposed to make it clearer,
but the refactor didn't work either.
That code was made much clearer by abstracting the job of "handling the
stack readjustment" to the various StopInfo's. Previously, there was a
big (and buggy) switch over stop info's. Moving the responsibility to
the stop info made this code much easier to reason about.
We also had no tests for virtual inlined stepping (our inlined stepping
test was actually written specifically to avoid the formation of a
virtual inlined stack... So I also added tests for that along with the
tests for setting the call-site breakpoints.
Compilers and language runtimes often use helper functions that are
fundamentally uninteresting when debugging anything but the
compiler/runtime itself. This patch introduces a user-extensible
mechanism that allows for these frames to be hidden from backtraces and
automatically skipped over when navigating the stack with `up` and
`down`.
This does not affect the numbering of frames, so `f <N>` will still
provide access to the hidden frames. The `bt` output will also print a
hint that frames have been hidden.
My primary motivation for this feature is to hide thunks in the Swift
programming language, but I'm including an example recognizer for
`std::function::operator()` that I wished for myself many times while
debugging LLDB.
rdar://126629381
Example output. (Yes, my proof-of-concept recognizer could hide even
more frames if we had a method that returned the function name without
the return type or I used something that isn't based off regex, but it's
really only meant as an example).
before:
```
(lldb) thread backtrace --filtered=false
* thread #1, queue = 'com.apple.main-thread', stop reason = breakpoint 1.1
* frame #0: 0x0000000100001f04 a.out`foo(x=1, y=1) at main.cpp:4:10
frame #1: 0x0000000100003a00 a.out`decltype(std::declval<int (*&)(int, int)>()(std::declval<int>(), std::declval<int>())) std::__1::__invoke[abi:se200000]<int (*&)(int, int), int, int>(__f=0x000000016fdff280, __args=0x000000016fdff224, __args=0x000000016fdff220) at invoke.h:149:25
frame #2: 0x000000010000399c a.out`int std::__1::__invoke_void_return_wrapper<int, false>::__call[abi:se200000]<int (*&)(int, int), int, int>(__args=0x000000016fdff280, __args=0x000000016fdff224, __args=0x000000016fdff220) at invoke.h:216:12
frame #3: 0x0000000100003968 a.out`std::__1::__function::__alloc_func<int (*)(int, int), std::__1::allocator<int (*)(int, int)>, int (int, int)>::operator()[abi:se200000](this=0x000000016fdff280, __arg=0x000000016fdff224, __arg=0x000000016fdff220) at function.h:171:12
frame #4: 0x00000001000026bc a.out`std::__1::__function::__func<int (*)(int, int), std::__1::allocator<int (*)(int, int)>, int (int, int)>::operator()(this=0x000000016fdff278, __arg=0x000000016fdff224, __arg=0x000000016fdff220) at function.h:313:10
frame #5: 0x0000000100003c38 a.out`std::__1::__function::__value_func<int (int, int)>::operator()[abi:se200000](this=0x000000016fdff278, __args=0x000000016fdff224, __args=0x000000016fdff220) const at function.h:430:12
frame #6: 0x0000000100002038 a.out`std::__1::function<int (int, int)>::operator()(this= Function = foo(int, int) , __arg=1, __arg=1) const at function.h:989:10
frame #7: 0x0000000100001f64 a.out`main(argc=1, argv=0x000000016fdff4f8) at main.cpp:9:10
frame #8: 0x0000000183cdf154 dyld`start + 2476
(lldb)
```
after
```
(lldb) bt
* thread #1, queue = 'com.apple.main-thread', stop reason = breakpoint 1.1
* frame #0: 0x0000000100001f04 a.out`foo(x=1, y=1) at main.cpp:4:10
frame #1: 0x0000000100003a00 a.out`decltype(std::declval<int (*&)(int, int)>()(std::declval<int>(), std::declval<int>())) std::__1::__invoke[abi:se200000]<int (*&)(int, int), int, int>(__f=0x000000016fdff280, __args=0x000000016fdff224, __args=0x000000016fdff220) at invoke.h:149:25
frame #2: 0x000000010000399c a.out`int std::__1::__invoke_void_return_wrapper<int, false>::__call[abi:se200000]<int (*&)(int, int), int, int>(__args=0x000000016fdff280, __args=0x000000016fdff224, __args=0x000000016fdff220) at invoke.h:216:12
frame #6: 0x0000000100002038 a.out`std::__1::function<int (int, int)>::operator()(this= Function = foo(int, int) , __arg=1, __arg=1) const at function.h:989:10
frame #7: 0x0000000100001f64 a.out`main(argc=1, argv=0x000000016fdff4f8) at main.cpp:9:10
frame #8: 0x0000000183cdf154 dyld`start + 2476
Note: Some frames were hidden by frame recognizers
```
This is another step towards supporting DWARF5 checksums and inline
source code in LLDB. This is a reland of #85468 but without the
functional change of storing the support file from the line table (yet).
This patch is rearranging code a bit to add WatchpointResources to
Process. A WatchpointResource is meant to represent a hardware
watchpoint register in the inferior process. It has an address, a size,
a type, and a list of Watchpoints that are using this
WatchpointResource.
This current patch doesn't add any of the features of
WatchpointResources that make them interesting -- a user asking to watch
a 24 byte object could watch this with three 8 byte WatchpointResources.
Or a Watchpoint on 1 byte at 0x1002 and a second watchpoint on 1 byte at
0x1003, these must both be served by a single WatchpointResource on that
doubleword at 0x1000 on a 64-bit target, if two hardware watchpoint
registers were used to track these separately, one of them may not be
hit. Or if you have one Watchpoint on a variable with a condition set,
and another Watchpoint on that same variable with a command defined or
different condition, or ignorecount, both of those Watchpoints need to
evaluate their criteria/commands when their WatchpointResource has been
hit.
There's a bit of code movement to rearrange things in the direction I'll
need for implementing this feature, so I want to start with reviewing &
landing this mostly NFC patch and we can focus on the algorithmic
choices about how WatchpointResources are shared and handled as they're
triggeed, separately.
This patch also stops printing "Watchpoint <n> hit: old value: <x>, new
vlaue: <y>" for Read watchpoints. I could make an argument for print
"Watchpoint <n> hit: current value <x>" but the current output doesn't
make any sense, and the user can print the value if they are
particularly interested. Read watchpoints are used primarily to
understand what code is reading a variable.
This patch adds more fallbacks for how to print the objects being
watched if we have types, instead of assuming they are all integral
values, so a struct will print its elements. As large watchpoints are
added, we'll be doing a lot more of those.
To track the WatchpointSP in the WatchpointResources, I changed the
internal API which took a WatchpointSP and devolved it to a Watchpoint*,
which meant touching several different Process files. I removed the
watchpoint code in ProcessKDP which only reported that watchpoints
aren't supported, the base class does that already.
I haven't yet changed how we receive a watchpoint to identify the
WatchpointResource responsible for the trigger, and identify all
Watchpoints that are using this Resource to evaluate their conditions
etc. This is the same work that a BreakpointSite needs to do when it has
been tiggered, where multiple Breakpoints may be at the same address.
There is not yet any printing of the Resources that a Watchpoint is
implemented in terms of ("watchpoint list", or
SBWatchpoint::GetDescription).
"watchpoint set var" and "watchpoint set expression" take a size
argument which was previously 1, 2, 4, or 8 (an enum). I've changed this
to an unsigned int. Most hardware implementations can only watch 1, 2,
4, 8 byte ranges, but with Resources we'll allow a user to ask for
different sized watchpoints and set them in hardware-expressble terms
soon.
I've annotated areas where I know there is work still needed with
LWP_TODO that I'll be working on once this is landed.
I've tested this on aarch64 macOS, aarch64 Linux, and Intel macOS.
https://discourse.llvm.org/t/rfc-large-watchpoint-support-in-lldb/72116
(cherry picked from commit fc6b72523f3d73b921690a713e97a433c96066c6)
...and follow ups.
As it has caused test failures on Linux Arm and AArch64:
https://lab.llvm.org/buildbot/#/builders/96/builds/49126https://lab.llvm.org/buildbot/#/builders/17/builds/45824
```
lldb-shell :: Subprocess/clone-follow-child-wp.test
lldb-shell :: Subprocess/fork-follow-child-wp.test
lldb-shell :: Subprocess/vfork-follow-child-wp.test
```
This reverts commit a6c62bf1a4717accc852463b664cd1012237d334,
commit a0a1ff3ab40e347589b4e27d8fd350c600526735 and commit
fc6b72523f3d73b921690a713e97a433c96066c6.
This patch is rearranging code a bit to add WatchpointResources to
Process. A WatchpointResource is meant to represent a hardware
watchpoint register in the inferior process. It has an address, a size,
a type, and a list of Watchpoints that are using this
WatchpointResource.
This current patch doesn't add any of the features of
WatchpointResources that make them interesting -- a user asking to watch
a 24 byte object could watch this with three 8 byte WatchpointResources.
Or a Watchpoint on 1 byte at 0x1002 and a second watchpoint on 1 byte at
0x1003, these must both be served by a single WatchpointResource on that
doubleword at 0x1000 on a 64-bit target, if two hardware watchpoint
registers were used to track these separately, one of them may not be
hit. Or if you have one Watchpoint on a variable with a condition set,
and another Watchpoint on that same variable with a command defined or
different condition, or ignorecount, both of those Watchpoints need to
evaluate their criteria/commands when their WatchpointResource has been
hit.
There's a bit of code movement to rearrange things in the direction I'll
need for implementing this feature, so I want to start with reviewing &
landing this mostly NFC patch and we can focus on the algorithmic
choices about how WatchpointResources are shared and handled as they're
triggeed, separately.
This patch also stops printing "Watchpoint <n> hit: old value: <x>, new
vlaue: <y>" for Read watchpoints. I could make an argument for print
"Watchpoint <n> hit: current value <x>" but the current output doesn't
make any sense, and the user can print the value if they are
particularly interested. Read watchpoints are used primarily to
understand what code is reading a variable.
This patch adds more fallbacks for how to print the objects being
watched if we have types, instead of assuming they are all integral
values, so a struct will print its elements. As large watchpoints are
added, we'll be doing a lot more of those.
To track the WatchpointSP in the WatchpointResources, I changed the
internal API which took a WatchpointSP and devolved it to a Watchpoint*,
which meant touching several different Process files. I removed the
watchpoint code in ProcessKDP which only reported that watchpoints
aren't supported, the base class does that already.
I haven't yet changed how we receive a watchpoint to identify the
WatchpointResource responsible for the trigger, and identify all
Watchpoints that are using this Resource to evaluate their conditions
etc. This is the same work that a BreakpointSite needs to do when it has
been tiggered, where multiple Breakpoints may be at the same address.
There is not yet any printing of the Resources that a Watchpoint is
implemented in terms of ("watchpoint list", or
SBWatchpoint::GetDescription).
"watchpoint set var" and "watchpoint set expression" take a size
argument which was previously 1, 2, 4, or 8 (an enum). I've changed this
to an unsigned int. Most hardware implementations can only watch 1, 2,
4, 8 byte ranges, but with Resources we'll allow a user to ask for
different sized watchpoints and set them in hardware-expressble terms
soon.
I've annotated areas where I know there is work still needed with
LWP_TODO that I'll be working on once this is landed.
I've tested this on aarch64 macOS, aarch64 Linux, and Intel macOS.
https://discourse.llvm.org/t/rfc-large-watchpoint-support-in-lldb/72116
StreamFile subclasses Stream (from lldbUtility) and is backed by a File
(from lldbHost). It does not depend on anything from lldbCore or any of its
sibling libraries, so I think it makes sense for this to live in
lldbHost instead.
Differential Revision: https://reviews.llvm.org/D157460
Also, make it possible for new Targets which haven't been added to
the TargetList yet to check for interruption, and add a few more
places in building modules where we can check for interruption.
Differential Revision: https://reviews.llvm.org/D154542
wrong answer. Plus, it's useful in some places to have a way to force
the full stack to be created even in the face of
interruption. Moreover, most of the time when you're just getting
frames, you don't need to know the number of frames in the stack to
start with. You just keep calling
Thread::GetStackFrameAtIndex(index++) and when you get a null
StackFrameSP back, you're done. That's also more amenable to
interruption if you are doing some work frame by frame.
So this patch makes GetStackFrameCount always return the full count,
suspending interruption. I also went through all the places that use
GetStackFrameCount to make sure that they really needed the full stack
walk. In many cases, they did not. For instance frame select -r 10 was
getting the number of frames just to check whether cur_frame_idx + 10
was within the stack. It's better in that case to see if that frame
exists first, since that doesn't force a full stack walk, and only
deal with walking off the end of the stack if it doesn't...
I also added a test for some of these behaviors.
Differential Revision: https://reviews.llvm.org/D150236
This is a user facing action, it is meant to focus the user's attention on
something other than the 0th frame when you stop somewhere where that's
helpful. For instance, stopping in pthread_kill after an assert will select
the assert frame.
This is not something you want to have happen internally in lldb, both
because internally you really don't want the selected frame changing out
from under you, and because the recognizers can do arbitrary work, and that
can cause deadlocks or other unexpected behavior.
However, it's not something that the current code does
explicitly after a stop has been delivered, it's expected to happen implicitly
as part of stopping. I changing this to call SMRF explicitly after a user
stop, but that got pretty ugly quickly.
So I added a bool to control whether to run this and audited all the current
uses to determine whether we're returning to the user or not.
Differential Revision: https://reviews.llvm.org/D148863
SelectMostRelevantFrame triggers the StackFrameRecognizer construction,
which can run arbitrary Python code, call expressions etc. WillStop gets
called on every private stop while the recognizers are a user-facing
feature, so first off doing this work on every stop is inefficient. But
more importantly, you can get in to situations where the recognizer
causes an expression to get run, then when we fetch the stop event at
the end of the expression evaluation, we call WillStop again on the
expression handling thread, which will do the same StackFrameRecognizer
work again. If anyone is locking along that path, you will end up with a
deadlock between the two threads.
The example that brought this to my attention was the
objc_exception_throw recognizer which can cause the objc runtime
introspection functions to get run, and those take a lock in
AppleObjCRuntimeV2::DynamicClassInfoExtractor::UpdateISAToDescriptorMap
along this path, so the second thread servicing the expression deadlocks
against the first thread waiting for the expression to complete.
It makes more sense to have the frame recognizers run on demand, either
when someone asks for the variables for the frame, or when someone does
GetSelectedFrame. The former already worked that way, the only reason
this was being done in WillStop was because the StackFrameRecognizers
can change the SelectedFrame, so you needed to run them before the
anyone requested the SelectedFrame.
This patch moves SelectMostRelevantFrame to StackFrameList, and runs it
when GetSelectedFrame is called for the first time on a given stop. If
you call SetSelectedFrame before GetSelectedFrame, then you should NOT
run the recognizer & change the frame out from under you. This patch
also makes that work. There were already tests for this behavior, and
for the feature that caused the hang, but the hang is racy, and it
doesn't trigger all the time, so I don't have a way to test that
explicitly.
One more detail: it's actually pretty easy to end up calling
GetSelectedFrame, for instance if you ask for the best ExecutionContext
from an ExecutionContextRef it will fill the StackFrame with the result
of GetSelectedFrame and that would still have the same problems if this
happens on the Private State Thread. So this patch also short-circuits
SelectMostRelevantFrame if run on the that thread. I can't think of any
reason the computations that go on on the Private State Thread would
actually want the SelectedFrame - that's a user-facing concept, so
avoiding that complication is the best way to go.
rdar://107643231
Differential revision: https://reviews.llvm.org/D147753
Most of our code was including Log.h even though that is not where the
"lldb" log channel is defined (Log.h defines the generic logging
infrastructure). This worked because Log.h included Logging.h, even
though it should.
After the recent refactor, it became impossible the two files include
each other in this direction (the opposite inclusion is needed), so this
patch removes the workaround that was put in place and cleans up all
files to include the right thing. It also renames the file to LLDBLog to
better reflect its purpose.
Introduce three new stop reasons for fork, vfork and vforkdone events.
This includes server support for serializing fork/vfork events into
gdb-remote protocol. The stop infos for the two base events take a pair
of PID and TID for the newly forked process.
Differential Revision: https://reviews.llvm.org/D100196
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
Add calls into LanguageRuntime when finding the unwind method to
use out of the 0th (currently executing) stack frame.
Allow for the LanguageRuntimes to indicate if this stack frames
should be treated like a zeroth-frame -- symbolication should be
done based on the saved pc address, not decremented like normal ABI
function calls.
Add methods to RegisterContext and StackFrame to get a pc value
suitable for symbolication, to reduce the number of places in lldb
where we decrement the saved pc values before symbolication.
<rdar://problem/70398009>
Differential Revision: https://reviews.llvm.org/D97644
Summary:
The way that the support for the GNU dialect of tail call frames was
implemented in D80519 meant that the were reporting very bogus PC values
which pointed into the middle of an instruction: the -1 trick is
necessary for the address to resolve to the right function, but we
should still be reporting a more realistic PC value -- I say "realistic"
and not "real", because it's very debatable what should be the correct
PC value for frames like this.
This patch achieves that my moving the -1 from SymbolFileDWARF into the
stack frame computation code. The idea is that SymbolFileDWARF will
merely report whether it has provided an address of the instruction
after the tail call, or the address of the call instruction itself. The
StackFrameList machinery uses this information to set the "behaves like
frame zero" property of the artificial frames (the main thing this flag
does is it controls the -1 subtraction when looking up the function
address).
This required a moderate refactor of the CallEdge class, because it was
implicitly assuming that edges pointing after the call were real calls
and those pointing the the call insn were tail calls. The class now
carries this information explicitly -- it carries three mostly
independent pieces of information:
- an address of interest in the caller
- a bit saying whether this address points to the call insn or after it
- whether this is a tail call
Reviewers: vsk, dblaikie
Subscribers: aprantl, mgrang, lldb-commits
Tags: #lldb
Differential Revision: https://reviews.llvm.org/D81010
Reland with changes: the test modified in this change originally failed
on a Debian/x86_64 builder, and I suspect the cause was that lldb looked
up the line location for an artificial frame by subtracting 1 from the
frame's address. For artificial frames, the subtraction must not happen
because the address is already exact.
---
lldb currently guesses the address to use when creating an artificial
frame (i.e., a frame constructed by determining the sequence of (tail)
calls which must have happened).
Guessing the address creates problems -- use the actual address provided
by the DW_AT_call_pc attribute instead.
Depends on D76336.
rdar://60307600
Differential Revision: https://reviews.llvm.org/D76337
This reverts commit 6905394d153960ded3a7b884a9747ed2d4a6e8d8. The
changed test is failing on Debian/x86_64, possibly because lldb is
subtracting an offset from the DW_AT_call_pc address used for the
artificial frame:
http://lab.llvm.org:8011/builders/lldb-x86_64-debian/builds/7171/steps/test/logs/stdio
/home/worker/lldb-x86_64-debian/lldb-x86_64-debian/llvm-project/lldb/test/API/functionalities/tail_call_frames/unambiguous_sequence/main.cpp:6:17: error: CHECK-NEXT: expected string not found in input
// CHECK-NEXT: frame #1: 0x{{[0-9a-f]+}} a.out`func3() at main.cpp:14:3 [opt] [artificial]
^
<stdin>:3:2: note: scanning from here
frame #1: 0x0000000000401127 a.out`func3() at main.cpp:13:4 [opt] [artificial]
lldb currently guesses the address to use when creating an artificial
frame (i.e., a frame constructed by determining the sequence of (tail)
calls which must have happened).
Guessing the address creates problems -- use the actual address provided
by the DW_AT_call_pc attribute instead.
Depends on D76336.
rdar://60307600
Differential Revision: https://reviews.llvm.org/D76337
In order to synthesize tail call frames, the stack frame list must not
be empty (otherwise, there is no "previous" frame to infer a tail call
from).
This case is hard to hit. To trigger it, we must first fail to push
`unwind_frame_sp` because we either fail to get its SymbolContext, or
given its SymbolContext the GetParentOfInlineScope call fails. This
causes m_concrete_frames_fetched to be incremented while m_frames
remains empty. Then, the next frame in the stack may fail within
SynthesizeTailCallFrames. This crash arose during a kernel debugging
session.
rdar://59147051
Summary:
A *.cpp file header in LLDB (and in LLDB) should like this:
```
//===-- TestUtilities.cpp -------------------------------------------------===//
```
However in LLDB most of our source files have arbitrary changes to this format and
these changes are spreading through LLDB as folks usually just use the existing
source files as templates for their new files (most notably the unnecessary
editor language indicator `-*- C++ -*-` is spreading and in every review
someone is pointing out that this is wrong, resulting in people pointing out that this
is done in the same way in other files).
This patch removes most of these inconsistencies including the editor language indicators,
all the different missing/additional '-' characters, files that center the file name, missing
trailing `===//` (mostly caused by clang-format breaking the line).
Reviewers: aprantl, espindola, jfb, shafik, JDevlieghere
Reviewed By: JDevlieghere
Subscribers: dexonsmith, wuzish, emaste, sdardis, nemanjai, kbarton, MaskRay, atanasyan, arphaman, jfb, abidh, jsji, JDevlieghere, usaxena95, lldb-commits
Tags: #lldb
Differential Revision: https://reviews.llvm.org/D73258
Split CallEdge into DirectCallEdge and IndirectCallEdge. Teach
DWARFExpression how to evaluate entry values in cases where the current
activation was created by an indirect call.
rdar://57094085
Differential Revision: https://reviews.llvm.org/D70100
