The function was extremely messy in that it, depending on the set of
arguments, it could either modify the Connection object in `this` or
not. It had a lot of arguments, with each call site passing a different
combination of null values. This PR:
- packs "url" and "comm_fd" arguments into a variant as they are
mutually exclusive
- removes the (surprising) "null url *and* null comm_fd" code path which
is not used as of https://github.com/llvm/llvm-project/pull/145017
- marks the function as `static` to make it clear it (now) does not
operate on the `this` object.
Depends on #145017
This lets get rid of platform-specific code in ProcessGDBRemote and use
the
same code path (module differences in socket types) everywhere. It also
unlocks
further cleanups in the debugserver launching code.
The main effect of this change is that lldb on windows will now use the
`--fd` lldb-server argument for "local remote" debug sessions instead of
having lldb-server connect back to lldb. This is the same method used by
lldb on non-windows platforms (for many years) and "lldb-server
platform" on windows for truly remote debug sessions (for ~one year).
Depends on #145015.
This patch should be mostly obvious, but in one place, this patch
changes:
const auto &it = std::find(...)
to:
auto it = llvm::find(...)
We do not need to bind to a temporary with const ref.
The new test added in https://github.com/llvm/llvm-project/pull/132783
was failing on Windows because it created a new error to say it did not
support the feature, but then returned the existing, default constructed
error. Which was a success value.
This also changes the GDBRemote error message to the same phrasing used
in all the other places so we don't have to special case any platform.
This reapplies commit
232525f069.
The original commit triggered a sanitizer failure when `Target` was
destroyed. In `Target::Destroy`, `DeleteCurrentProcess` was called, but
it did not destroy the thread creation breakpoints for the underlying
`ProcessGDBRemote` because `ProcessGDBRemote::Clear` was not called in
that path.
`Target `then proceeded to destroy its breakpoints, which resulted in a
call to the destructor of a `std::vector` containing the breakpoints.
Through a sequence of complicated events, destroying breakpoints caused
the reference count of the underlying `ProcessGDBRemote` to finally
reach zero. This, in turn, called `ProcessGDBRemote::Clear`, which
attempted to destroy the breakpoints. To do that, it would go back into
the Target's vector of breakpoints, which we are in the middle of
destroying.
We solve this by moving the breakpoint deletion into
`Process:DoDestroy`, which is a virtual Process method that will be
called much earlier.
This reverts commit 232525f06942adb3b9977632e38dcd5f08c0642d.
This change is causing test crashes while running
TestCompletion.py on Darwin systems, most of the CI runs
have failed since it has been merged in.
Currently, these breakpoints are being accumulated every time a new
process if created (e.g. through a `run`). Depending on the
circumstances, the old breakpoints are even left enabled, interfering
with subsequent processes. This is addressed by removing the breakpoints
in ProcessGDBRemote::Clear
Note that these breakpoints are more of a PlatformDarwin thing, so in
the future we should look into moving them there.
This reverts commit
87b7f63a11,
reapplying
7e66cf74fb
with a small (and probably temporary)
change to generate more debug info to help with diagnosing buildbot
issues.
Make StreamAsynchronousIO an unique_ptr instead of a shared_ptr. I tried
passing the class by value, but the llvm::raw_ostream forwarder stored
in the Stream parent class isn't movable and I don't think it's worth
changing that. Additionally, there's a few places that expect a
StreamSP, which are easily created from a StreamUP.
lldb today has two rules: When a thread stops at a BreakpointSite, we
set the thread's StopReason to be "breakpoint hit" (regardless if we've
actually hit the breakpoint, or if we've merely stopped *at* the
breakpoint instruction/point and haven't tripped it yet). And second,
when resuming a process, any thread sitting at a BreakpointSite is
silently stepped over the BreakpointSite -- because we've already
flagged the breakpoint hit when we stopped there originally.
In this patch, I change lldb to only set a thread's stop reason to
breakpoint-hit when we've actually executed the instruction/triggered
the breakpoint. When we resume, we only silently step past a
BreakpointSite that we've registered as hit. We preserve this state
across inferior function calls that the user may do while stopped, etc.
Also, when a user adds a new breakpoint at $pc while stopped, or changes
$pc to be the address of a BreakpointSite, we will silently step past
that breakpoint when the process resumes. This is purely a UX call, I
don't think there's any person who wants to set a breakpoint at $pc and
then hit it immediately on resuming.
One non-intuitive UX from this change, butt is necessary: If you're
stopped at a BreakpointSite that has not yet executed, you `stepi`, you
will hit the breakpoint and the pc will not yet advance. This thread has
not completed its stepi, and the ThreadPlanStepInstruction is still on
the stack. If you then `continue` the thread, lldb will now stop and
say, "instruction step completed", one instruction past the
BreakpointSite. You can continue a second time to resume execution.
The bugs driving this change are all from lldb dropping the real stop
reason for a thread and setting it to breakpoint-hit when that was not
the case. Jim hit one where we have an aarch64 watchpoint that triggers
one instruction before a BreakpointSite. On this arch we are notified of
the watchpoint hit after the instruction has been unrolled -- we disable
the watchpoint, instruction step, re-enable the watchpoint and collect
the new value. But now we're on a BreakpointSite so the watchpoint-hit
stop reason is lost.
Another was reported by ZequanWu in
https://discourse.llvm.org/t/lldb-unable-to-break-at-start/78282 we
attach to/launch a process with the pc at a BreakpointSite and
misbehave. Caroline Tice mentioned it is also a problem they've had with
putting a breakpoint on _dl_debug_state.
The change to each Process plugin that does execution control is that
1. If we've stopped at a BreakpointSite that has not been executed yet,
we will call Thread::SetThreadStoppedAtUnexecutedBP(pc) to record that.
When the thread resumes, if the pc is still at the same site, we will
continue, hit the breakpoint, and stop again.
2. When we've actually hit a breakpoint (enabled for this thread or
not), the Process plugin should call
Thread::SetThreadHitBreakpointSite(). When we go to resume the thread,
we will push a step-over-breakpoint ThreadPlan before resuming.
The biggest set of changes is to StopInfoMachException where we
translate a Mach Exception into a stop reason. The Mach exception codes
differ in a few places depending on the target (unambiguously), and I
didn't want to duplicate the new code for each target so I've tested
what mach exceptions we get for each action on each target, and
reorganized StopInfoMachException::CreateStopReasonWithMachException to
document these possible values, and handle them without specializing
based on the target arch.
I first landed this patch in July 2024 via
https://github.com/llvm/llvm-project/pull/96260
but the CI bots and wider testing found a number of test case failures
that needed to be updated, I reverted it. I've fixed all of those issues
in separate PRs and this change should run cleanly on all the CI bots
now.
rdar://123942164
Generally speaking, process plugins (e.g. ProcessGDBRemote) should not
be aware of OS plugin threads. However, ProcessGDBRemote attempts to
check for the existence of OS threads when calculating stop info. When
OS threads are present, it sets the stop info directly on the OS plugin
thread and leaves the ThreadGDBRemote without a StopInfo.
This is problematic for a few reasons:
1. No other process plugins do this, as they shouldn't. They should set
the stop info for their own process threads, and let the abstractions
built on top propagate StopInfos.
2. This conflicts with the expectations of ThreadMemory, which checks
for the backing threads's info, and then attempts to propagate it (in
the future, it should probably ask the plugin itself too...). We see
this happening in the code below. The `if` condition will not trigger,
because `backing_stop_info_sp` will be null (remember, ProcessGDB remote
is ignoring its own threads), and then this method returns false.
```
bool ThreadMemory::CalculateStopInfo() {
...
lldb::StopInfoSP backing_stop_info_sp(
m_backing_thread_sp->GetPrivateStopInfo());
if (backing_stop_info_sp &&
backing_stop_info_sp->IsValidForOperatingSystemThread(*this)) {
backing_stop_info_sp->SetThread(shared_from_this());
```
```
Thread::GetPrivateStopInfo
...
if (!CalculateStopInfo())
SetStopInfo(StopInfoSP());
```
To solve this, we change ProcessGDB remote so that it does the
principled thing: it now only sets the stop info of its own threads.
This change by itself breaks the tests TestPythonOSPlugin.py and
TestOSPluginStepping.py and probably explains why ProcessGDB had
originally "violated" this isolation of layers.
To make this work, BreakpointSites must be aware of BackingThreads when
answering the question: "Is this breakpoint valid for this thread?".
Why? Breakpoints are created on top of the OS threads (that's what the
user sees), but breakpoints are hit by process threads. In the presence
of OS threads, a TID-specific breakpoint is valid for a process thread
if it is backing an OS thread with that TID.
This reverts commit a774de807e56c1147d4630bfec3110c11d41776e.
This is the same changes as last time, plus:
* We load the binary into the target object so that on Windows, we can
resolve the locations of the functions.
* We now assert that each required breakpoint has at least 1 location,
to prevent an issue like that in the future.
* We are less strict about the unsupported error message, because it
prints "error: windows" on Windows instead of "error: gdb-remote".
Reverts llvm/llvm-project#123945
Has failed on the Windows on Arm buildbot:
https://lab.llvm.org/buildbot/#/builders/141/builds/5865
```
********************
Unresolved Tests (2):
lldb-api :: functionalities/reverse-execution/TestReverseContinueBreakpoints.py
lldb-api :: functionalities/reverse-execution/TestReverseContinueWatchpoints.py
********************
Failed Tests (1):
lldb-api :: functionalities/reverse-execution/TestReverseContinueNotSupported.py
```
Reverting while I reproduce locally.
This reverts commit 22561cfb443267905d4190f0e2a738e6b412457f and fixes
b7b9ccf44988edf49886743ae5c3cf4184db211f (#112079).
The problem is that x86_64 and Arm 32-bit have memory regions above the
stack that are readable but not writeable. First Arm:
```
(lldb) memory region --all
<...>
[0x00000000fffcf000-0x00000000ffff0000) rw- [stack]
[0x00000000ffff0000-0x00000000ffff1000) r-x [vectors]
[0x00000000ffff1000-0xffffffffffffffff) ---
```
Then x86_64:
```
$ cat /proc/self/maps
<...>
7ffdcd148000-7ffdcd16a000 rw-p 00000000 00:00 0 [stack]
7ffdcd193000-7ffdcd196000 r--p 00000000 00:00 0 [vvar]
7ffdcd196000-7ffdcd197000 r-xp 00000000 00:00 0 [vdso]
ffffffffff600000-ffffffffff601000 --xp 00000000 00:00 0 [vsyscall]
```
Compare this to AArch64 where the test did pass:
```
$ cat /proc/self/maps
<...>
ffffb87dc000-ffffb87dd000 r--p 00000000 00:00 0 [vvar]
ffffb87dd000-ffffb87de000 r-xp 00000000 00:00 0 [vdso]
ffffb87de000-ffffb87e0000 r--p 0002a000 00:3c 76927217 /usr/lib/aarch64-linux-gnu/ld-linux-aarch64.so.1
ffffb87e0000-ffffb87e2000 rw-p 0002c000 00:3c 76927217 /usr/lib/aarch64-linux-gnu/ld-linux-aarch64.so.1
fffff4216000-fffff4237000 rw-p 00000000 00:00 0 [stack]
```
To solve this, look up the memory region of the stack pointer (using
https://lldb.llvm.org/resources/lldbgdbremote.html#qmemoryregioninfo-addr)
and constrain the read to within that region. Since we know the stack is
all readable and writeable.
I have also added skipIfRemote to the tests, since getting them working
in that context is too complex to be worth it.
Memory write failures now display the range they tried to write, and
register write errors will show the name of the register where possible.
The patch also includes a workaround for a an issue where the test code
could mistake an `x` response that happens to begin with an `O` for an
output packet (stdout). This workaround will not be necessary one we
start using the [new
implementation](https://discourse.llvm.org/t/rfc-fixing-incompatibilties-of-the-x-packet-w-r-t-gdb/84288)
of the `x` packet.
---------
Co-authored-by: Pavel Labath <pavel@labath.sk>
This commit adds support for a
`SBProcess::ContinueInDirection()` API. A user-accessible command for
this will follow in a later commit.
This feature depends on a gdbserver implementation (e.g. `rr`) providing
support for the `bc` and `bs` packets. `lldb-server` does not support
those packets, and there is no plan to change that. For testing
purposes, this commit adds a Python implementation of *very limited*
record-and-reverse-execute functionality, implemented as a proxy between
lldb and lldb-server in `lldbreverse.py`. This should not (and in
practice cannot) be used for anything except testing.
The tests here are quite minimal but we test that simple breakpoints and
watchpoints work as expected during reverse execution, and that
conditional breakpoints and watchpoints work when the condition calls a
function that must be executed in the forward direction.
Reverting this again; I added a commit which added @skipIfDarwin
markers to the TestReverseContinueBreakpoints.py and
TestReverseContinueNotSupported.py API tests, which use lldb-server
in gdbserver mode which does not work on Darwin. But the aarch64 ubuntu
bot reported a failure on TestReverseContinueBreakpoints.py,
https://lab.llvm.org/buildbot/#/builders/59/builds/6397
File "/home/tcwg-buildbot/worker/lldb-aarch64-ubuntu/llvm-project/lldb/test/API/functionalities/reverse-execution/TestReverseContinueBreakpoints.py", line 63, in test_reverse_continue_skip_breakpoint
self.reverse_continue_skip_breakpoint_internal(async_mode=False)
File "/home/tcwg-buildbot/worker/lldb-aarch64-ubuntu/llvm-project/lldb/test/API/functionalities/reverse-execution/TestReverseContinueBreakpoints.py", line 81, in reverse_continue_skip_breakpoint_internal
self.expect(
File "/home/tcwg-buildbot/worker/lldb-aarch64-ubuntu/llvm-project/lldb/packages/Python/lldbsuite/test/lldbtest.py", line 2372, in expect
self.runCmd(
File "/home/tcwg-buildbot/worker/lldb-aarch64-ubuntu/llvm-project/lldb/packages/Python/lldbsuite/test/lldbtest.py", line 1002, in runCmd
self.assertTrue(self.res.Succeeded(), msg + output)
AssertionError: False is not true : Process should be stopped due to history boundary
Error output:
error: Process must be launched.
This reverts commit 4f297566b3150097de26c6a23a987d2bd5fc19c5.
This commit only adds support for the
`SBProcess::ReverseContinue()` API. A user-accessible command for this
will follow in a later commit.
This feature depends on a gdbserver implementation (e.g. `rr`) providing
support for the `bc` and `bs` packets. `lldb-server` does not support
those packets, and there is no plan to change that. So, for testing
purposes, `lldbreverse.py` wraps `lldb-server` with a Python
implementation of *very limited* record-and-replay functionality for use
by *tests only*.
The majority of this PR is test infrastructure (about 700 of the 950
lines added).
This commit only adds support for the
`SBProcess::ReverseContinue()` API. A user-accessible command for this
will follow in a later commit.
This feature depends on a gdbserver implementation (e.g. `rr`) providing
support for the `bc` and `bs` packets. `lldb-server` does not support
those packets, and there is no plan to change that. So, for testing
purposes, `lldbreverse.py` wraps `lldb-server` with a Python
implementation of *very limited* record-and-replay functionality for use
by *tests only*.
The majority of this PR is test infrastructure (about 700 of the 950
lines added).
Fixes#107864
QEMU decided that when SVE is enabled it will only tell us about SVE
registers in the XML, and not include Neon registers. On the grounds
that the Neon V registers can be read from the bottom 128 bits of a SVE
Z register (SVE's vector length is always >= 128 bits).
To support this we create sub-registers just as we do for S and D
registers of the V registers. Except this time we use part of the Z
registers.
This change also updates our fallback for registers with unknown types
that are > 128 bit. This is detailed in
https://github.com/llvm/llvm-project/issues/87471, though that covers
more than this change fixes.
We'll now treat any register of unknown type that is >= 128 bit as a
vector of bytes. So that the user gets to see something
even if the order might be wrong.
And until lldb supports vector and union types for registers, this is
also the only way we can get a value to apply the sub-reg to, to make
the V registers.
Some gdb remote serial protocol stubs will send the thread IDs and PCs
for all threads in a process in the stop-reply packet. lldb often needs
to know the pc values for all threads while at a private stop, and that
results in <n-1> read-register packets for <n> threads, and can be a big
performance problem when this is a hot code path.
GDBRemoteRegisterContext tracks the StopID of when its values were set,
and when the thread's StopID has incremented, it marks all values it has
as Invalid, and knows to refetch them.
We have a code path that resulted in setting the PCs for all the
threads, and then `ProcessGDBRemote::CalculateThreadStopInfo` *forcing*
an invalidation of all the register contexts, forcing us to re-read the
pc values for all threads except the one that stopped.
There are times when it is valid to force an invalidation of the
regsiter cache - for instance, if the layout of the registers has
changed because the processor state is different, or we've sent a
write-all-registers packet to the inferior and we want to make sure we
stay in sync with the inferior. But there was no reason for this method
to be forcing the register context to be invalid.
I added a test when running on Darwin systems, where debugserver always
sends the thread IDs and PCs, which turns on packet logging. The test
runs against an inferior which has 4 threads; it steps over a dlopen()
call, steps in to a user function with debug info, steps-over and
steps-in across source lines with multiple function calls, and then
examines the packet log and flags it as an error if lldb asked for the
pc value of any thread at any point in the debug session.
For this program and the operations we're doing, with debugserver that
provides thread IDs and PCs, we should never ask for the value of a pc
register.
rdar://136247381
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.
xusheng added support for swbreak/hwbreak a month ago, and no special
support was needed in ProcessGDBRemote when they're received because
lldb already marks a thread as having hit a breakpoint when it stops at
a breakpoint site. However, with changes I am working on, we need to
know the real stop reason a thread stopped or the breakpoint hit will
not be recognized.
This is similar to how lldb processes the "watch/rwatch/awatch" keys in
a thread stop packet -- we set the `reason` to `watchpoint`, and these
set it to `breakpoint` so we set the stop reason correctly later in
these methods.
This patch removes all of the Set.* methods from Status.
This cleanup is part of a series of patches that make it harder use the
anti-pattern of keeping a long-lives Status object around and updating
it while dropping any errors it contains on the floor.
This patch is largely NFC, the more interesting next steps this enables
is to:
1. remove Status.Clear()
2. assert that Status::operator=() never overwrites an error
3. remove Status::operator=()
Note that step (2) will bring 90% of the benefits for users, and step
(3) will dramatically clean up the error handling code in various
places. In the end my goal is to convert all APIs that are of the form
` ResultTy DoFoo(Status& error)
`
to
` llvm::Expected<ResultTy> DoFoo()
`
How to read this patch?
The interesting changes are in Status.h and Status.cpp, all other
changes are mostly
` perl -pi -e 's/\.SetErrorString/ = Status::FromErrorString/g' $(git
grep -l SetErrorString lldb/source)
`
plus the occasional manual cleanup.
This PR is in reference to porting LLDB on AIX.
Link to discussions on llvm discourse and github:
1. https://discourse.llvm.org/t/port-lldb-to-ibm-aix/80640
2. #101657
The complete changes for porting are present in this draft PR:
https://github.com/llvm/llvm-project/pull/102601
The changes on this PR are intended to avoid namespace collision for
certain typedefs between lldb and other platforms:
1. tid_t --> lldb::tid_t
2. offset_t --> lldb::offset_t
This fixes https://github.com/llvm/llvm-project/issues/56125 and
https://github.com/vadimcn/codelldb/issues/666, as well as the
downstream issue in our binary ninja debugger:
https://github.com/Vector35/debugger/issues/535
Basically, lldb does not claim to support the `swbreak` packet so the
gdbserver would not use it. As a result, the gdbserver always sends the
unmodified program counter value which, on systems like x86, causes the
program counter to be off-by-one (or otherwise wrong). For reference,
the lldb-server always sends the modified program counter value so it
works perfectly with lldb.
https://sourceware.org/gdb/current/onlinedocs/gdb.html/Stop-Reply-Packets.html#swbreak-stop-reason
No new code is added to add support `swbreak`, since the way lldb works
already expects the remote to have adjusted the program counter. The
change just lets the gdbserver know that lldb supports it, so that it
will send the adjusted program counter.
To test this PR, you can use lldb to connect to a gdbserver running on
e.g., Ubuntu 22.04, and see the program counter is off-by-one without
the patch. With the patch, things work as expected
This PR introduces a new `ThreadPlanSingleThreadTimeout` that will be
used to address potential deadlock during single-thread stepping.
While debugging a target with a non-trivial number of threads (around
5000 threads in one example target), we noticed that a simple step over
can take as long as 10 seconds. Enabling single-thread stepping mode
significantly reduces the stepping time to around 3 seconds. However,
this can introduce deadlock if we try to step over a method that depends
on other threads to release a lock.
To address this issue, we introduce a new
`ThreadPlanSingleThreadTimeout` that can be controlled by the
`target.process.thread.single-thread-plan-timeout` setting during
single-thread stepping mode. The concept involves counting the elapsed
time since the last internal stop to detect overall stepping progress.
Once a timeout occurs, we assume the target is not making progress due
to a potential deadlock, as mentioned above. We then send a new async
interrupt, resume all threads, and `ThreadPlanSingleThreadTimeout`
completes its task.
To support this design, the major changes made in this PR are:
1. `ThreadPlanSingleThreadTimeout` is popped during every internal stop
and reset (re-pushed) to the top of the stack (as a leaf node) during
resume. This is achieved by always returning `true` from
`ThreadPlanSingleThreadTimeout::DoPlanExplainsStop()` and
`ThreadPlanSingleThreadTimeout::MischiefManaged()`.
2. A new thread-specific async interrupt stop is introduced, which can
be detected/consumed by `ThreadPlanSingleThreadTimeout`.
3. The clearing of branch breakpoints in the range thread plan has been
moved from `DoPlanExplainsStop()` to `ShouldStop()`, as it is not
guaranteed that it will be called.
The detailed design is discussed in the RFC below:
[https://discourse.llvm.org/t/improve-single-thread-stepping/74599](https://discourse.llvm.org/t/improve-single-thread-stepping/74599)
---------
Co-authored-by: jeffreytan81 <jeffreytan@fb.com>
This reverts commit 05f0e86cc895181b3d2210458c78938f83353002.
The debuginfo dexter tests are failing, probably because the way
stepping over breakpoints has changed with my patches. And there
are two API tests fails on the ubuntu-arm (32-bit) bot. I'll need
to investigate both of these, neither has an obvious failure reason.
lldb today has two rules: When a thread stops at a BreakpointSite, we
set the thread's StopReason to be "breakpoint hit" (regardless if we've
actually hit the breakpoint, or if we've merely stopped *at* the
breakpoint instruction/point and haven't tripped it yet). And second,
when resuming a process, any thread sitting at a BreakpointSite is
silently stepped over the BreakpointSite -- because we've already
flagged the breakpoint hit when we stopped there originally.
In this patch, I change lldb to only set a thread's stop reason to
breakpoint-hit when we've actually executed the instruction/triggered
the breakpoint. When we resume, we only silently step past a
BreakpointSite that we've registered as hit. We preserve this state
across inferior function calls that the user may do while stopped, etc.
Also, when a user adds a new breakpoint at $pc while stopped, or changes
$pc to be the address of a BreakpointSite, we will silently step past
that breakpoint when the process resumes. This is purely a UX call, I
don't think there's any person who wants to set a breakpoint at $pc and
then hit it immediately on resuming.
One non-intuitive UX from this change, but I'm convinced it is
necessary: If you're stopped at a BreakpointSite that has not yet
executed, you `stepi`, you will hit the breakpoint and the pc will not
yet advance. This thread has not completed its stepi, and the thread
plan is still on the stack. If you then `continue` the thread, lldb will
now stop and say, "instruction step completed", one instruction past the
BreakpointSite. You can continue a second time to resume execution. I
discussed this with Jim, and trying to paper over this behavior will
lead to more complicated scenarios behaving non-intuitively. And mostly
it's the testsuite that was trying to instruction step past a breakpoint
and getting thrown off -- and I changed those tests to expect the new
behavior.
The bugs driving this change are all from lldb dropping the real stop
reason for a thread and setting it to breakpoint-hit when that was not
the case. Jim hit one where we have an aarch64 watchpoint that triggers
one instruction before a BreakpointSite. On this arch we are notified of
the watchpoint hit after the instruction has been unrolled -- we disable
the watchpoint, instruction step, re-enable the watchpoint and collect
the new value. But now we're on a BreakpointSite so the watchpoint-hit
stop reason is lost.
Another was reported by ZequanWu in
https://discourse.llvm.org/t/lldb-unable-to-break-at-start/78282 we
attach to/launch a process with the pc at a BreakpointSite and
misbehave. Caroline Tice mentioned it is also a problem they've had with
putting a breakpoint on _dl_debug_state.
The change to each Process plugin that does execution control is that
1. If we've stopped at a BreakpointSite that has not been executed yet,
we will call Thread::SetThreadStoppedAtUnexecutedBP(pc) to record
that. When the thread resumes, if the pc is still at the same site, we
will continue, hit the breakpoint, and stop again.
2. When we've actually hit a breakpoint (enabled for this thread or not),
the Process plugin should call Thread::SetThreadHitBreakpointSite().
When we go to resume the thread, we will push a step-over-breakpoint
ThreadPlan before resuming.
The biggest set of changes is to StopInfoMachException where we
translate a Mach Exception into a stop reason. The Mach exception codes
differ in a few places depending on the target (unambiguously), and I
didn't want to duplicate the new code for each target so I've tested
what mach exceptions we get for each action on each target, and
reorganized StopInfoMachException::CreateStopReasonWithMachException to
document these possible values, and handle them without specializing
based on the target arch.
rdar://123942164
This reverts commit d9e659c538516036e40330b6a98160cbda4ff100.
I could not reproduce the Mac OS ASAN failure locally but I narrowed it
down to the test `test_many_fields_same_enum`. This test shares an enum
between x0, which is 64 bit, and cpsr, which is 32 bit.
My theory is that when it does `register read x0`, an enum type is
created where the undlerying enumerators are 64 bit, matching the
register size.
Then it does `register read cpsr` which used the cached enum type, but
this register is 32 bit. This caused lldb to try to read an 8 byte value
out of a 4 byte allocation:
READ of size 8 at 0x60200014b874 thread T0
<...>
=>0x60200014b800: fa fa fd fa fa fa fd fa fa fa fd fa fa fa[04]fa
To fix this I've added the register's size in bytes to the constructed
enum type's name. This means that x0 uses:
__lldb_register_fields_enum_some_enum_8
And cpsr uses:
__lldb_register_fields_enum_some_enum_4
If any other registers use this enum and are read, they will use the
cached type as long as their size matches, otherwise we make a new type.
Fixes#92541
When e69a3d18f48bc0d81b5dd12e735a2ec898ce64d added fallback register
layouts, it assumed that the choices were target XML with registers, or
no target XML at all.
In the linked issue, a user has a debug stub that does have target XML,
but it's missing register information.
This caused us to finalize the register information using an empty set
of registers got from target XML, then fail an assert when we attempted
to add the fallback set. Since we think we've already completed the
register information.
This change adds a check to prevent that first call and expands the
existing tests to check each architecture without target XML and with
target XML missing register information.
This teaches lldb to parse the enum XML elements sent by lldb-server,
and make use of the information in `register read` and `register info`.
The format is described in
https://sourceware.org/gdb/current/onlinedocs/gdb.html/Enum-Target-Types.html.
The target XML parser will drop any invalid enum or evalue. If we find
multiple evalue for the same value, we will use the last one we find.
The order of evalues from the XML is preserved as there may be good
reason they are not in numerical order.
AddressableBits is in the Utility module of LLDB. It currently directly
refers to Process, which is from the Target LLDB module. This is a
layering violation which concretely means that it is impossible to link
anything that uses Utility without it also using Target as well. This is
generally not an issue for LLDB (since everything is built together) but
it may make it difficult to write unit tests for AddressableBits later
on.
We got user reporting lldb crash while the debuggee is calling vfork()
concurrently from multiple threads.
The crash happens because the current implementation can only handle
single vfork, vforkdone protocol transaction.
This diff fixes the crash by lldb-server storing forked debuggee's <pid,
tid> pair in jstopinfo which will be decoded by lldb client to create
StopInfoVFork for follow parent/child policy. Each StopInfoVFork will
later have a corresponding vforkdone packet. So the patch also changes
the `m_vfork_in_progress` to be reference counting based.
Two new test cases are added which crash/assert without the changes in
this patch.
---------
Co-authored-by: jeffreytan81 <jeffreytan@fb.com>
Partly, there's just a lot of unnecessary boiler plate. It's also
possible to define combinations of arguments that make no sense (e.g.
eArgRepeatPlus followed by eArgRepeatPlain...) but these are never
checked since we just push_back directly into the argument definitions.
This commit is step 1 of this cleanup - do the obvious stuff. In it, all
the simple homogenous argument lists and the breakpoint/watchpoint
ID/Range types, are set with common functions. This is an NFC change, it
just centralizes boiler plate. There's no checking yet because you can't
get a single argument wrong.
The end goal is that all argument definition goes through functions and
m_arguments is hidden so that you can't define inconsistent argument
sets.
This is next in my series of "fix the racey tests that fail on
greendragon" addressing the failure of TestConcurrentManyBreakpoints.py
where we set a breakpoint in a function that 100 threads execute, and we
check that we hit the breakpoint 100 times. But sometimes it is only hit
99 times, and the test fails.
When we hit a software breakpoint, the pc value for the thread is the
address of the breakpoint instruction - as if it had not been hit yet.
And because a user might ADD a breakpoint for the current pc from the
commandline, when we go to resume execution, any thread that is sitting
at a breakpoint site will be silently advanced past the breakpoint
instruction (disable bp, instruction step that thread, re-enable bp)
before resuming -- whether that thread has hit its breakpoint or not.
What this test is exposing is that there is another corner case, a
thread that is sitting at a breakpoint site but has not yet executed the
breakpoint instruction. The thread will have no stop reason, no mach
exception, so it will not be recorded as having hit the breakpoint
(because it hasn't yet). But when we resume execution, because it is
sitting at a breakpoint site, we advance past it and miss the breakpoint
hit.
In 2016 Abhishek Aggarwal handled a similar issue with a patch in
`ProcessGDBRemote::SetThreadStopInfo()`, adding a breakpoint StopInfo
for a thread sitting at a breakpoint site that has no stop reason.
debugserver's `jThreadsInfo` would not correctly execute Abhishek's code
though because it would respond with `"reason":"none"` for a thread with
no stop reason, and `SetThreadStopInfo()` expected an empty reason here.
The first part of my patch is to clear the `reason` if it is `"none"` so
we flow through the code correctly.
On Darwin, though, our stop reply packet (Txx...) includes the
`threads`, `thread-pcs`, and `jstopinfo` keys, which give us the tids
for all current threads, the pc values for those threads, and
`jstopinfo` has a JSON dictionary with the mach exceptions for all
threads that have a mach exception. In
`ProcessGDBRemote::CalculateThreadStopInfo()` we set the StopInfo for
each thread for a private stop and if we have `jstopinfo` it is the
source of all the StopInfos. I have to add the same logic here, to give
the thread a breakpoint StopInfo even though it hasn't executed the
breakpoint yet. In this case we are very early in thread construction
and I only have the information in the Txx stop reply packet -- tids,
pcs, and jstopinfo, so I can't use the normal general mechanisms of
going through the RegisterContext to get the pc, it's a bit different.
If I hack debugserver to not issue `jstopinfo`,
`CalculateThreadStopInfo` will fall back to sending `qThreadStopInfo`
for each thread and going through
`ProcessGDBRemote::SetThreadStopInfo()` to set the stop infos (and with
the `reason:none` fix, use Abhishek's code).
rdar://110549165
debugserver on arm64 devices can manage both Byte Address Select
watchpoints (1-8 bytes) and MASK watchpoints (8 bytes-2 gigabytes). This
adds a SupportedWatchpointTypes key to the QSupported response from
debugserver with a list of these, so lldb can take full advantage of
them when creating larger regions with a single hardware watchpoint.
Also add documentation for this, and two other lldb extensions, to the
lldb-gdb-remote.txt documentation.
Re-enable TestLargeWatchpoint.py on Darwin systems when testing with the
in-tree built debugserver. I can remove the "in-tree built debugserver"
in the future when this new key is handled by an Xcode debugserver.
This patch is the next piece of work in my Large Watchpoint proposal,
https://discourse.llvm.org/t/rfc-large-watchpoint-support-in-lldb/72116
This patch breaks a user's watchpoint into one or more
WatchpointResources which reflect what the hardware registers can cover.
This means we can watch objects larger than 8 bytes, and we can watched
unaligned address ranges. On a typical 64-bit target with 4 watchpoint
registers you can watch 32 bytes of memory if the start address is
doubleword aligned.
Additionally, if the remote stub implements AArch64 MASK style
watchpoints (e.g. debugserver on Darwin), we can watch any power-of-2
size region of memory up to 2GB, aligned to that same size.
I updated the Watchpoint constructor and CommandObjectWatchpoint to
create a CompilerType of Array<UInt8> when the size of the watched
region is greater than pointer-size and we don't have a variable type to
use. For pointer-size and smaller, we can display the watched granule as
an integer value; for larger-than-pointer-size we will display as an
array of bytes.
I have `watchpoint list` now print the WatchpointResources used to
implement the watchpoint.
I added a WatchpointAlgorithm class which has a top-level static method
that takes an enum flag mask WatchpointHardwareFeature and a user
address and size, and returns a vector of WatchpointResources covering
the request. It does not take into account the number of watchpoint
registers the target has, or the number still available for use. Right
now there is only one algorithm, which monitors power-of-2 regions of
memory. For up to pointer-size, this is what Intel hardware supports.
AArch64 Byte Address Select watchpoints can watch any number of
contiguous bytes in a pointer-size memory granule, that is not currently
supported so if you ask to watch bytes 3-5, the algorithm will watch the
entire doubleword (8 bytes). The newly default "modify" style means we
will silently ignore modifications to bytes outside the watched range.
I've temporarily skipped TestLargeWatchpoint.py for all targets. It was
only run on Darwin when using the in-tree debugserver, which was a proxy
for "debugserver supports MASK watchpoints". I'll be adding the
aforementioned feature flag from the stub and enabling full mask
watchpoints when a debugserver with that feature is enabled, and
re-enable this test.
I added a new TestUnalignedLargeWatchpoint.py which only has one test
but it's a great one, watching a 22-byte range that is unaligned and
requires four 8-byte watchpoints to cover.
I also added a unit test, WatchpointAlgorithmsTests, which has a number
of simple tests against WatchpointAlgorithms::PowerOf2Watchpoints. I
think there's interesting possible different approaches to how we cover
these; I note in the unit test that a user requesting a watch on address
0x12e0 of 120 bytes will be covered by two watchpoints today, a
128-bytes at 0x1280 and at 0x1300. But it could be done with a 16-byte
watchpoint at 0x12e0 and a 128-byte at 0x1300, which would have fewer
false positives/private stops. As we try refining this one, it's helpful
to have a collection of tests to make sure things don't regress.
I tested this on arm64 macOS, (genuine) x86_64 macOS, and AArch64
Ubuntu. I have not modifed the Windows process plugins yet, I might try
that as a standalone patch, I'd be making the change blind, but the
necessary changes (see ProcessGDBRemote::EnableWatchpoint) are pretty
small so it might be obvious enough that I can change it and see what
the Windows CI thinks.
There isn't yet a packet (or a qSupported feature query) for the gdb
remote serial protocol stub to communicate its watchpoint capabilities
to lldb. I'll be doing that in a patch right after this is landed,
having debugserver advertise its capability of AArch64 MASK watchpoints,
and have ProcessGDBRemote add eWatchpointHardwareArmMASK to
WatchpointAlgorithms so we can watch larger than 32-byte requests on
Darwin.
I haven't yet tackled WatchpointResource *sharing* by multiple
Watchpoints. This is all part of the goal, especially when we may be
watching a larger memory range than the user requested, if they then add
another watchpoint next to their first request, it may be covered by the
same WatchpointResource (hardware watchpoint register). Also one "read"
watchpoint and one "write" watchpoint on the same memory granule need to
be handled, making the WatchpointResource cover all requests.
As WatchpointResources aren't shared among multiple Watchpoints yet,
there's no handling of running the conditions/commands/etc on multiple
Watchpoints when their shared WatchpointResource is hit. The goal beyond
"large watchpoint" is to unify (much more) the Watchpoint and Breakpoint
behavior and commands. I have a feeling I may be slowly chipping away at
this for a while.
Re-landing this patch after fixing two undefined behaviors in
WatchpointAlgorithms found by UBSan and by failures on different
CI bots.
rdar://108234227
