Files
clang-p2996/lldb/test/API/commands/trace/TestTraceStartStop.py
Walter Erquinigo ade59d5309 [trace] Dedup different source lines when dumping instructions + refactor
When dumping the traced instructions in a for loop, like this one

  4:  for (int a = 0; a < n; a++)
  5:    do something;

there might be multiple LineEntry objects for line 4, but with different address ranges. This was causing the dump command to dump something like this:

```
  a.out`main + 11 at main.cpp:4
    [1] 0x0000000000400518    movl   $0x0, -0x8(%rbp)
    [2] 0x000000000040051f    jmp    0x400529                  ; <+28> at main.cpp:4
  a.out`main + 28 at main.cpp:4
    [3] 0x0000000000400529    cmpl   $0x3, -0x8(%rbp)
    [4] 0x000000000040052d    jle    0x400521                  ; <+20> at main.cpp:5
```

which is confusing, as main.cpp:4 appears twice consecutively.

This diff fixes that issue by making the line entry comparison strictly about the line, column and file name. Before it was also comparing the address ranges, which we don't need because our output is strictly about what the user sees in the source.

Besides, I've noticed that the logic that traverses instructions and calculates symbols and disassemblies had too much coupling, and made my changes harder to implement, so I decided to decouple it. Now there are two methods for iterating over the instruction of a trace. The existing one does it on raw load addresses, but the one provides a SymbolContext and an InstructionSP, and does the calculations efficiently (not as efficient as possible for now though), so the caller doesn't need to care about these details. I think I'll be using that iterator to reconstruct the call stacks.

I was able to fix a test with this change.

Differential Revision: https://reviews.llvm.org/D100740
2021-05-04 19:40:52 -07:00

129 lines
5.1 KiB
Python

import lldb
from lldbsuite.test.lldbtest import *
from lldbsuite.test import lldbutil
from lldbsuite.test.decorators import *
ADDRESS_REGEX = '0x[0-9a-fA-F]*'
class TestTraceStartStop(TestBase):
mydir = TestBase.compute_mydir(__file__)
NO_DEBUG_INFO_TESTCASE = True
def setUp(self):
TestBase.setUp(self)
if 'intel-pt' not in configuration.enabled_plugins:
self.skipTest("The intel-pt test plugin is not enabled")
def expectGenericHelpMessageForStartCommand(self):
self.expect("help thread trace start",
substrs=["Syntax: thread trace start [<trace-options>]"])
def testStartStopSessionFileThreads(self):
# it should fail for processes from json session files
self.expect("trace load -v " + os.path.join(self.getSourceDir(), "intelpt-trace", "trace.json"))
self.expect("thread trace start", error=True,
substrs=["error: Process must be alive"])
# the help command should be the generic one, as it's not a live process
self.expectGenericHelpMessageForStartCommand()
self.expect("thread trace stop", error=True)
def testStartWithNoProcess(self):
self.expect("thread trace start", error=True,
substrs=["error: Process not available."])
def testStartSessionWithWrongSize(self):
self.expect("file " + os.path.join(self.getSourceDir(), "intelpt-trace", "a.out"))
self.expect("b main")
self.expect("r")
self.expect("thread trace start -s 2000", error=True,
substrs=["The trace buffer size must be a power of 2", "It was 2000"])
self.expect("thread trace start -s 5000", error=True,
substrs=["The trace buffer size must be a power of 2", "It was 5000"])
self.expect("thread trace start -s 0", error=True,
substrs=["The trace buffer size must be a power of 2", "It was 0"])
self.expect("thread trace start -s 1048576")
@skipIf(oslist=no_match(['linux']), archs=no_match(['i386', 'x86_64']))
def testStartStopLiveThreads(self):
# The help command should be the generic one if there's no process running
self.expectGenericHelpMessageForStartCommand()
self.expect("thread trace start", error=True,
substrs=["error: Process not available"])
self.expect("file " + os.path.join(self.getSourceDir(), "intelpt-trace", "a.out"))
self.expect("b main")
self.expect("thread trace start", error=True,
substrs=["error: Process not available"])
# The help command should be the generic one if there's still no process running
self.expectGenericHelpMessageForStartCommand()
self.expect("r")
# This fails because "trace start" hasn't been called yet
self.expect("thread trace stop", error=True,
substrs=["error: Process is not being traced"])
# the help command should be the intel-pt one now
self.expect("help thread trace start",
substrs=["Start tracing one or more threads with intel-pt.",
"Syntax: thread trace start [<thread-index> <thread-index> ...] [<intel-pt-options>]"])
# We start tracing with a small buffer size
self.expect("thread trace start 1 --size 4096")
# We fail if we try to trace again
self.expect("thread trace start", error=True,
substrs=["error: Thread ", "already traced"])
# We can reconstruct the single instruction executed in the first line
self.expect("n")
self.expect("thread trace dump instructions",
patterns=[f'''thread #1: tid = .*, total instructions = 1
a.out`main \+ 4 at main.cpp:2
\[0\] {ADDRESS_REGEX} movl'''])
# We can reconstruct the instructions up to the second line
self.expect("n")
self.expect("thread trace dump instructions",
patterns=[f'''thread #1: tid = .*, total instructions = 5
a.out`main \+ 4 at main.cpp:2
\[0\] {ADDRESS_REGEX} movl .*
a.out`main \+ 11 at main.cpp:4
\[1\] {ADDRESS_REGEX} movl .*
\[2\] {ADDRESS_REGEX} jmp .* ; <\+28> at main.cpp:4
\[3\] {ADDRESS_REGEX} cmpl .*
\[4\] {ADDRESS_REGEX} jle .* ; <\+20> at main.cpp:5'''])
# We stop tracing
self.expect("thread trace stop")
# We can't stop twice
self.expect("thread trace stop", error=True,
substrs=["error: Thread ", "not currently traced"])
# We trace again from scratch, this time letting LLDB to pick the current
# thread
self.expect("thread trace start")
self.expect("n")
self.expect("thread trace dump instructions",
patterns=[f'''thread #1: tid = .*, total instructions = 1
a.out`main \+ 20 at main.cpp:5
\[0\] {ADDRESS_REGEX} xorl'''])
self.expect("c")
# Now the process has finished, so the commands should fail
self.expect("thread trace start", error=True,
substrs=["error: Process must be launched"])
self.expect("thread trace stop", error=True,
substrs=["error: Process must be launched"])