https://github.com/dmpots approved this pull request.
LGTM
https://github.com/llvm/llvm-project/pull/136236
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots edited
https://github.com/llvm/llvm-project/pull/136236
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots commented:
After your change we can expect that no symbols are loaded by your simple
repro, right? Can we extend the test here (or add a new one) to show that no
symbols will be loaded in that case by checking the stats?
https://github.com/dmpots/llvm-project/blob/mai
@@ -761,3 +761,77 @@ TEST_F(SymtabTest, TestSymtabCreatedOnDemand) {
Symtab *cached_module_symtab = module_sp->GetSymtab(/*can_create=*/false);
ASSERT_EQ(module_symtab, cached_module_symtab);
}
+
+TEST_F(SymtabTest, TestSymbolTableCreatedOnDemand) {
+ const char *yamldata
@@ -749,10 +749,20 @@ TEST_F(SymtabTest, TestSymtabCreatedOnDemand) {
ASSERT_THAT_EXPECTED(ExpectedFile, llvm::Succeeded());
auto module_sp = std::make_shared(ExpectedFile->moduleSpec());
- // The symbol table should not be loaded by default.
+ // The symbol file should
@@ -749,10 +749,20 @@ TEST_F(SymtabTest, TestSymtabCreatedOnDemand) {
ASSERT_THAT_EXPECTED(ExpectedFile, llvm::Succeeded());
auto module_sp = std::make_shared(ExpectedFile->moduleSpec());
- // The symbol table should not be loaded by default.
+ // The symbol file should
@@ -168,23 +170,34 @@ def test_default_no_run(self):
"totalSymbolTableIndexTime",
"totalSymbolTablesLoadedFromCache",
"totalSymbolTablesSavedToCache",
+"totalSymbolsLoaded",
"totalDebugInfoByteSize",
"tot
https://github.com/dmpots approved this pull request.
LGTM
https://github.com/llvm/llvm-project/pull/136226
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
@@ -318,12 +321,15 @@ llvm::json::Value DebuggerStats::ReportStatistics(
module_stat.symtab_index_time = module->GetSymtabIndexTime().get().count();
Symtab *symtab = module->GetSymtab(/*can_create=*/false);
if (symtab) {
+ module_stat.num_symbols_loaded = symta
@@ -168,23 +170,34 @@ def test_default_no_run(self):
"totalSymbolTableIndexTime",
"totalSymbolTablesLoadedFromCache",
"totalSymbolTablesSavedToCache",
+"totalSymbolsLoaded",
dmpots wrote:
We should also check for
dmpots wrote:
@JDevlieghere @jimingham @clayborg
Thanks for the comments. I updated the PR based on the initial feedback. We
still have a few open issues that could use a bit more feedback:
1. Are we ok to start with a static table of plugin namespaces or do we want
dynamic registration from
@@ -46,12 +49,333 @@ class CommandObjectPluginLoad : public CommandObjectParsed
{
}
};
+namespace {
+#define LLDB_OPTIONS_plugin_list
+#include "CommandOptions.inc"
+
+// These option definitions are shared by the plugin list/enable/disable
+// commands.
+class PluginListCo
@@ -46,12 +49,333 @@ class CommandObjectPluginLoad : public CommandObjectParsed
{
}
};
+namespace {
+#define LLDB_OPTIONS_plugin_list
+#include "CommandOptions.inc"
+
+// These option definitions are shared by the plugin list/enable/disable
+// commands.
+class PluginListCo
@@ -46,12 +49,333 @@ class CommandObjectPluginLoad : public CommandObjectParsed
{
}
};
+namespace {
+#define LLDB_OPTIONS_plugin_list
+#include "CommandOptions.inc"
+
+// These option definitions are shared by the plugin list/enable/disable
+// commands.
+class PluginListCo
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/134418
>From e240bda8fcea9db4d9c456929ba811feb8d4152b Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Tue, 11 Mar 2025 13:02:14 -0700
Subject: [PATCH 1/7] Add commands to list/enable/disable plugins
This commit add
dmpots wrote:
Here are the relevant test result changes for this change.
Trace Hardware Unsupported
==
Before
---
-- Testing: 11 tests, 11 workers --
Unresolved Tests (7):
lldb-api :: commands/trace/TestTraceDumpFunctionCalls.py
lldb-api :: commands/trace/
https://github.com/dmpots closed
https://github.com/llvm/llvm-project/pull/134173
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/132884
>From 6edc108199044d30eed396145aab463db40d6351 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Thu, 13 Mar 2025 16:13:45 -0700
Subject: [PATCH 1/2] Remove raw access to PluginInstances vector
This commit mod
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/133826
The tests for the
[intel-pt](https://github.com/llvm/llvm-project/blob/348374028970c956f2e49ab7553b495d7408ccd9/lldb/docs/use/intel_pt.rst)
trace plugin were failing for multiple reasons.
On machines where traci
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/133794
This commit adds support for enabling and disabling plugins by name. The
changes are made generically in the `PluginInstances` class, but currently we
only expose the ability to SystemRuntime plugins. Other plug
@@ -65,6 +65,10 @@ class DebugNamesDWARFIndex : public DWARFIndex {
void GetFunctions(const RegularExpression ®ex,
llvm::function_ref callback) override;
+ StatsDuration::Duration GetIndexTime() override {
+return m_fallback.GetIndexTime();
-
https://github.com/dmpots closed
https://github.com/llvm/llvm-project/pull/131951
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
dmpots wrote:
> Another way to do this would be to have the list of disabled plugins show in
> the result of the `version` command, since that's a commonly provided bit of
> info for bug reports.
Thanks for the suggestion. Let me work on this and update the PR.
https://github.com/llvm/llvm-p
dmpots wrote:
> LLVM Buildbot has detected a new failure on builder `lldb-aarch64-ubuntu`
> running on `linaro-lldb-aarch64-ubuntu` while building `lldb` at step 6
> "test".
>
> Full details are available at:
> https://lab.llvm.org/buildbot/#/builders/59/builds/15362
>
> Here is the relevant
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/134418
>From e240bda8fcea9db4d9c456929ba811feb8d4152b Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Tue, 11 Mar 2025 13:02:14 -0700
Subject: [PATCH 1/4] Add commands to list/enable/disable plugins
This commit add
@@ -46,12 +49,344 @@ class CommandObjectPluginLoad : public CommandObjectParsed
{
}
};
+namespace {
+#define LLDB_OPTIONS_plugin_list
+#include "CommandOptions.inc"
+
+// These option definitions are shared by the plugin list/enable/disable
+// commands.
+class PluginListCo
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/134418
>From e240bda8fcea9db4d9c456929ba811feb8d4152b Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Tue, 11 Mar 2025 13:02:14 -0700
Subject: [PATCH 1/3] Add commands to list/enable/disable plugins
This commit add
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/132053
This commit modifes the `getSourceDir()` and `getBuildDir()` functions to use
os.path.realpath to resolve symlinks in the Base test class used for API tests.
A few tests were failing when the build and source di
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/132884
>From 6edc108199044d30eed396145aab463db40d6351 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Thu, 13 Mar 2025 16:13:45 -0700
Subject: [PATCH 1/2] Remove raw access to PluginInstances vector
This commit mod
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/134418
>From e240bda8fcea9db4d9c456929ba811feb8d4152b Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Tue, 11 Mar 2025 13:02:14 -0700
Subject: [PATCH 1/2] Add commands to list/enable/disable plugins
This commit add
@@ -46,12 +49,344 @@ class CommandObjectPluginLoad : public CommandObjectParsed
{
}
};
+namespace {
+#define LLDB_OPTIONS_plugin_list
+#include "CommandOptions.inc"
+
+// These option definitions are shared by the plugin list/enable/disable
+// commands.
+class PluginListCo
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/134418
This commit adds three new commands for managing plugins. The `list` command
will show which plugins are currently registered and their enabled state. The
`enable` and `disable` commands can be used to enable or
https://github.com/dmpots closed
https://github.com/llvm/llvm-project/pull/132884
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
dmpots wrote:
This is my guess on what is causing the test failures. Can revert this and the
original PR if it doesn't work.
https://github.com/llvm/llvm-project/pull/134173
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.
dmpots wrote:
The windows bots are now passing on this test so I think it was ICF that was
the problem: https://lab.llvm.org/buildbot/#/changes/32645
https://github.com/llvm/llvm-project/pull/134173
___
lldb-commits mailing list
lldb-commits@lists.llv
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/134173
This is an attempt to fix a test failure from #133794 when running on windows
builds. I suspect we are running into a case where the
[ICF](https://learn.microsoft.com/en-us/cpp/build/reference/opt-optimizations?
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/134183
Revert the below two commits while we investigate the test failures.
Revert "[lldb] Fix plugin manager test failure on windows (#134173)"
This reverts commit b55bab229228218341e2f24fc8529c7aaab51e2f.
Revert "Ad
dmpots wrote:
I have an attempted fix for these failures in #134173. I suspect ICF is
combining the create functions into a single address which make us unable to
remove them based on a unique address (which the plugin manager assumes).
https://github.com/llvm/llvm-project/pull/133794
dmpots wrote:
This PR modifies the PluginManager to support enable/disable. A following
[commit](https://github.com/dmpots/llvm-project/commit/aaa3ab3d063580747bb106f39e6792232cc0af00)
will expose the commands to the user.
This is currently only working for the SystemRuntime plugins. Once thes
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/132876
>From 21ac6ae78d53feba8322e3aca724ecb01ff7c847 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Thu, 13 Mar 2025 16:13:45 -0700
Subject: [PATCH 1/3] Remove raw access to PluginInstances vector
This commit mod
dmpots wrote:
> Alright, let's go with the copy. Unless you think the `enabled` property
> might change while we're iterating over the plugins, we should be able to
> store it in the PluginInstance. AFAIK, the instance itself is never stored
> away.
Sounds good. That seems to be the simplest
dmpots wrote:
> > We did consider this approach. There are some drawbacks beyond the
> > performance issue, but the tradeoff may be worth it.
> > If we define the method like: `std::vector
> > GetEnabledInstances()` then we are returning copies of the PluginInstance.
> > Anything that modifies
dmpots wrote:
> Did you consider having something like `GetEnabledInstances` that returns a
> vector with just the enabled instances? It seems like that would be a lot
> less intrusive and I'd be surprised if this code is hot enough that the
> overhead of creating a new vector instead of retur
https://github.com/dmpots edited
https://github.com/llvm/llvm-project/pull/132884
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots edited
https://github.com/llvm/llvm-project/pull/132053
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
dmpots wrote:
This PR is the first of a
[series](https://github.com/llvm/llvm-project/pull/132876) of commits to
implement a generic way to enable/disable plugins. The linked draft PR provides
more context about how these changes work to allow plugin enable/disable.
https://github.com/llvm/ll
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/132884
This commit modifies the PluginInstances class to remove direct access to the
m_instances vector. Instead, we expose a new `ForEachEnabledPlugin` method that
takes a callback to operate on each plugin. All exter
https://github.com/dmpots closed
https://github.com/llvm/llvm-project/pull/132053
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
dmpots wrote:
> Resolving the symlink is not the right thing to do in some cases (such as
> when you symlinks define the actual/expected layout of the directory tree and
> the real paths are meaningless (e.g. nodes in some CAS storage). I would very
> much like to see solution that applies the
https://github.com/dmpots edited
https://github.com/llvm/llvm-project/pull/132053
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/132053
>From efde68cf06da2f668f4d2ecd638b826c7f4c5b70 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Wed, 19 Mar 2025 08:44:59 -0700
Subject: [PATCH 1/3] [lldb] Change test paths to resolve symlinks
This commit mo
https://github.com/dmpots closed
https://github.com/llvm/llvm-project/pull/129614
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots edited
https://github.com/llvm/llvm-project/pull/129614
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots edited
https://github.com/llvm/llvm-project/pull/129614
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
dmpots wrote:
> The test really is big and splitting it up is a good idea, but instead of an
> arbitrary 1/2/3 split, I think it be better to do it by functionality.
> Looking at the part of the code you moved, I can see several "themes" that we
> could use to split this up. One of them is "me
dmpots wrote:
> You could make the test binary a cmake dependency, see
> `lldb/test/CMakeLists.txt`. Perhaps in:
>
> ```
> if(TARGET lldb-server)
> add_lldb_test_dependency(lldb-server)
> endif()
> ```
Just so it's clear, it is not lldb-server itself that is the slow dependency,
but the cpp
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/129593
>From bca07d83152df179f7784d0003262fa54834 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Wed, 26 Feb 2025 13:55:35 -0800
Subject: [PATCH 1/5] Avoid force loading symbol files in statistics collection
T
@@ -718,3 +720,40 @@ TEST_F(SymtabTest, TestDecodeCStringMaps) {
Symtab::eVisibilityAny);
ASSERT_NE(symbol, nullptr);
}
+
+TEST_F(SymtabTest, TestSymtabCreatedOnDemand) {
+ auto ExpectedFile = TestFile::fromYaml(R"(
+--- !ELF
dmpots wrote:
I looked into why this test (and TestGdbRemoteFork.py) is so slow. Turns out
its nothing related to lldb itself, but the build step that occurs as part of
the test.
These tests methods pretty much all call to `build()` (directly or indirectly)
https://github.com/llvm/llvm-proje
https://github.com/dmpots closed
https://github.com/llvm/llvm-project/pull/129817
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
dmpots wrote:
Thanks for all the suggestions. I'll close the PR and try to make some progress
internally on this.
https://github.com/llvm/llvm-project/pull/129817
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin
@@ -0,0 +1,31 @@
+# RUN: %clang_host -g %S/Inputs/main.c -o %t-main.exe
+
+# When we enable symbol preload and dump stats there should be a non-zero
+# time for parsing symbol tables for the main module.
dmpots wrote:
Added the comment, thanks!
https://github.co
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/129593
>From bca07d83152df179f7784d0003262fa54834 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Wed, 26 Feb 2025 13:55:35 -0800
Subject: [PATCH 1/4] Avoid force loading symbol files in statistics collection
T
dmpots wrote:
> That's not to say that things cannot be flakey sometimes: because of how we
> test the debugger, we depend on a lot of things, many of which are out of our
> control and can cause a test to fail. But that's different from a specific
> test being flakey, which is what this decor
dmpots wrote:
> How many others are close to the limit that you've found, just this one?
I was only seeing the one timeout in the runs so I had not thought to measure.
It's a good idea though. Based on the timings it looks like maybe
`TestGdbRemoteFork.py` is worth splitting as well? That one
@@ -5,9 +5,9 @@
lldb-server tests run where the lldb-server exe is
available.
-This class will be broken into smaller test case classes by
-gdb remote packet functional areas. For now it contains
-the initial set of tests implemented.
+The tests are split between the LldbGdbS
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/129614
>From bb034b3b8c43dae13d7f596c20863b47ce88e4d4 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Mon, 3 Mar 2025 15:39:20 -0800
Subject: [PATCH 1/2] Split test to avoid timeout
This test is running right up to
https://github.com/dmpots created
https://github.com/llvm/llvm-project/pull/129817
The dotest framework had an existing decorator to mark flakey tests. It was not
implemented and would simply run the underlying test. This commit modifies the
decorator to run the test multiple times in the case
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/129817
>From 1d1c3a6bf52ba47a4ff4c13e99209a6ae3d983b3 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Tue, 4 Mar 2025 11:39:03 -0800
Subject: [PATCH 1/2] [lldb] Support expectedFlakey decorator in dotest
The dotest
@@ -0,0 +1,31 @@
+# RUN: %clang_host -g %S/Inputs/main.c -o %t-main.exe
+
+# When we enable symbol preload and dump stats there should be a non-zero
+# time for parsing symbol tables for the main module.
+# RUN: %lldb %t-main.exe \
+# RUN: -O "settings set plugin.jit-loader.
https://github.com/dmpots updated
https://github.com/llvm/llvm-project/pull/129593
>From bca07d83152df179f7784d0003262fa54834 Mon Sep 17 00:00:00 2001
From: David Peixotto
Date: Wed, 26 Feb 2025 13:55:35 -0800
Subject: [PATCH 1/3] Avoid force loading symbol files in statistics collection
T
https://github.com/dmpots approved this pull request.
LGTM!
https://github.com/llvm/llvm-project/pull/128971
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/dmpots edited
https://github.com/llvm/llvm-project/pull/128971
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
@@ -40,7 +40,10 @@ SBProgress::~SBProgress() = default;
void SBProgress::Increment(uint64_t amount, const char *description) {
LLDB_INSTRUMENT_VA(amount, description);
- m_opaque_up->Increment(amount, description);
+ std::optional description_opt;
+ if (description && des
@@ -33,3 +33,15 @@ def test_without_external_bit_set(self):
expected_string = "Test progress first increment"
progress.Increment(1, expected_string)
self.assertFalse(listener.PeekAtNextEvent(event))
+
+def test_with_external_bit_set(self):
+"
@@ -40,7 +40,10 @@ SBProgress::~SBProgress() = default;
void SBProgress::Increment(uint64_t amount, const char *description) {
LLDB_INSTRUMENT_VA(amount, description);
- m_opaque_up->Increment(amount, description);
+ std::optional description_opt;
+ if (description && des
76 matches
Mail list logo
