Commit Graph

1318 Commits

Author SHA1 Message Date
Greg Clayton ab843393f2 Fixed an issue that would cause an assertion to fire when an inlined function was found during a regex function find call.
llvm-svn: 120814
2010-12-03 17:49:14 +00:00
Greg Clayton eb5596b4af Make sure timed_out is initialized to false just to be on the safe side.
llvm-svn: 120813
2010-12-03 17:47:00 +00:00
Greg Clayton e521966054 Fixed a race condition that could cause ProcessGDBRemote::DoResume() to return
an error saying the resume timed out. Previously the thread that was trying
to resume the process would eventually call ProcessGDBRemote::DoResume() which
would broadcast an event over to the async GDB remote thread which would sent the
continue packet to the remote gdb server. Right after this was sent, it would
set a predicate boolean value (protected by a mutex and condition) and then the
thread that issued the ProcessGDBRemote::DoResume() would then wait for that
condition variable to be set. If the async gdb thread was too quick though, the
predicate boolean value could have been set to true and back to false by the
time the thread that issued the ProcessGDBRemote::DoResume() checks the boolean
value. So we can't use the predicate value as a handshake. I have changed the code
over to using a Event by having the GDB remote communication object post an
event: 

	GDBRemoteCommunication::eBroadcastBitRunPacketSent

This allows reliable handshaking between the two threads and avoids the erroneous
ProcessGDBRemote::DoResume() errors.

Added a host backtrace service to allow in process backtraces when trying to track
down tricky issues. I need to see if LLVM has any backtracing abilities abstracted
in it already, and if so, use that, but I needed something ASAP for the current issue
I was working on. The static function is:

void
Host::Backtrace (Stream &strm, uint32_t max_frames);

And it will backtrace at most "max_frames" frames for the current thread and can be
used with any of the Stream subclasses for logging.

llvm-svn: 120793
2010-12-03 06:02:24 +00:00
Sean Callanan 4a5fcbb92b Removed a compiler warning.
llvm-svn: 120788
2010-12-03 03:02:31 +00:00
Sean Callanan 979f74d1dd Fixed object lifetimes in ClangExpressionDeclMap
so that it is not referring to potentially stale
state during IR execution.

This was done by introducing modular state (like
ClangExpressionVariable) where groups of state
variables have well-defined lifetimes:

- m_parser_vars are specific to parsing, and only
  exist between calls to WillParse() and DidParse().

- m_struct_vars survive for the entire execution
  of the ClangExpressionDeclMap because they
  provide the template for a materialized set of
  expression variables.

- m_material_vars are specific to a single
  instance of materialization, and only exist
  between calls to Materialize() and
  Dematerialize().

I also removed unnecessary references to long-
lived state that really didn't need to be referred
to at all, and also introduced several assert()s
that helped me diagnose a few bugs (fixed too).

llvm-svn: 120778
2010-12-03 01:38:59 +00:00
Greg Clayton 621801846e Fixed DoResume to watch for the correct return value from WaitForIsRunning to avoid spurious errors due to previous fix.
llvm-svn: 120762
2010-12-03 00:27:48 +00:00
Greg Clayton a975990111 Fixed bad logic that was trying to determine if the gdb remote resumed a process or not.
llvm-svn: 120761
2010-12-03 00:19:58 +00:00
Johnny Chen eb3411dc41 Log just the thread name, not process/thread id for easier human reading.
llvm-svn: 120751
2010-12-02 23:31:02 +00:00
Greg Clayton 38a614034a Updated to latest LLVM/Clang for external AST source changes that allow
TagDecl subclasses and Objective C interfaces to complete themselves through
the ExternalASTSource class.

llvm-svn: 120749
2010-12-02 23:20:03 +00:00
Johnny Chen 7528af2f86 When logging, enable process and thread ID recordings. Also add 'state' to the lldb
log-categories, so it now becomes "event process expr state".

llvm-svn: 120745
2010-12-02 22:47:52 +00:00
Johnny Chen c4221e47b0 Fixed a typo in comment.
llvm-svn: 120733
2010-12-02 20:53:05 +00:00
Sean Callanan d7a1ca2a12 Fixed IRForTarget so that it errors out when function
pointers are used.  Previously, they caused a crash
in the JIT because we didn't resolve them correctly.

llvm-svn: 120728
2010-12-02 19:47:57 +00:00
Johnny Chen 20d0b2198b For lldb loggings, prepend log lines with a timestamp. And change the default log options:
o lldb => event process expr
o gdb-remote => packets process

llvm-svn: 120725
2010-12-02 18:35:13 +00:00
Caroline Tice 82305fc59a Add proper EOF handling to Communication & Connection classes:
Add bool member to Communication class indicating whether the
Connection should be closed on receiving an EOF or not.  Update the
Connection read to return an EOF status when appropriate.  Modify the
Communication class to pass the EOF along or not, and to close the
Connection or not, as appropriate.

llvm-svn: 120723
2010-12-02 18:31:56 +00:00
Johnny Chen 5f46b89df2 On second thought, inserting a delay of 1.0 second slows down the test suite
for the types directory quite a bit and masking out potential problems.  Enable
the delay via an environment variable, instead.

llvm-svn: 120719
2010-12-02 18:03:37 +00:00
Johnny Chen 560befce38 GenericTester.generic_type_expr_tester():
Don't overwhelm the expression mechanism.  Previously, there was no delay
    between successive invocations of 'expression' command.  The test suite
    run has experienced 'Resume timed out' or similar error messages.  Now we
    currently add a sleep of 1.0 second before issuing the next 'expression'
    command.

llvm-svn: 120626
2010-12-01 23:46:33 +00:00
Johnny Chen b91cb31490 Add a mechanism of overwriting the default 1.0 second wait time between test
cases when specifying '-w' to the test driver, by using an environment variable
LLDB_TIME_WAIT_BETWEEN_TEST_CASES.

llvm-svn: 120625
2010-12-01 23:35:13 +00:00
Johnny Chen d6c05708eb Remove @expectedFailure decorator as the bug has been fixed.
llvm-svn: 120624
2010-12-01 23:15:58 +00:00
Johnny Chen 6bc1366366 TestBase.setUp() needs to make sure lldb.blacklist is not None
before iterating on it.

llvm-svn: 120623
2010-12-01 23:07:36 +00:00
Johnny Chen 75739045d3 Add a '-b blacklistFile' option to the test driver to take a file specifying the
test classes or test cases to be excludued from the test suite.

Check in an example blacklist file: blacklist.py:

"""
'blacklist' is a Python dictionary, it stores the mapping of a string describing
either a testclass or a testcase, i.e, testclass.testmethod, to the reason (a
string) it is blacklisted.

Following is an example which states that test class IntegerTypesExprTestCase
should be skipped because 'This test class crashed' and the test case
FoundationTestCase.test_data_type_and_expr_with_dsym should be skipped because
it is 'Temporarily disabled'.

blacklist = {'IntegerTypesExprTestCase': 'This test class crashed',
             'FoundationTestCase.test_data_type_and_expr_with_dsym': 'Temporarily disabled'
             }
"""

blacklist = {}

An example of invoking the test driver and specifying a blacklist file:

./dotest.py -b blacklist.py -v types

This runs the tests under 'types' directory but excludes the tests specified in
balcklist.py.

llvm-svn: 120620
2010-12-01 22:47:54 +00:00
Greg Clayton 3c14438fe8 Fixed 32 bit debugging after recent architecture changes to debugserver.
llvm-svn: 120618
2010-12-01 22:45:40 +00:00
Sean Callanan 3670ba5c87 Fixed ClangUserExpression's wrapping of expressions
in C++ methods.  There were two fixes involved:

 - For an object whose contents are not known, the
   expression should be treated as a non-member, and
   "this" should have no meaning.

 - For a const object, the method should be declared
   const as well.

llvm-svn: 120606
2010-12-01 21:35:54 +00:00
Johnny Chen 5237e90cbe Add test cases to test that:
1. 'help image du sym' is ambiguous.
2. 'help image du line' is not ambiguous.

llvm-svn: 120587
2010-12-01 19:10:59 +00:00
Sean Callanan 83b0918938 Allowed ClangExpressionDeclMap to dematerialize
persistent variables even after the parser has
finished running.

llvm-svn: 120521
2010-12-01 01:29:06 +00:00
Sean Callanan 1d47cafc1c Whitespace fix.
llvm-svn: 120520
2010-12-01 01:28:23 +00:00
Johnny Chen ce81496fab Make CommandObjectMultiword::GetSubcommandSP() more robust by appending the
exactly-matched sub_cmd to the passed in 'matches' string list.

llvm-svn: 120516
2010-12-01 01:04:22 +00:00
Johnny Chen 19e302bc70 Remove the @expectedFailure decorator as the bug has been fixed.
llvm-svn: 120515
2010-12-01 01:01:23 +00:00
Jim Ingham 97253e62e9 GetCommandObject returns no matches in the match array when there is only one match. That's odd, but I don't want to change that right now, just cope with it where I'm doing the command matching.
llvm-svn: 120511
2010-12-01 00:42:17 +00:00
Johnny Chen c7c03d919a Add @expectedFailure decorator for a known bug:
rdar://problem/8689814 test failure: test/macosx/universal (the i386 slice does not break?)

llvm-svn: 120505
2010-12-01 00:03:17 +00:00
Johnny Chen be6fb3a85f Add expectedFailure decorator for test_help_image_dump_symtab_should_not_crash().
Possibly due to http://llvm.org/viewvc/llvm-project?rev=120484&view=rev.

llvm-svn: 120503
2010-11-30 23:57:58 +00:00
Jim Ingham 271ad29a4e Fix completion for multi-word commands in the "help" command.
llvm-svn: 120484
2010-11-30 22:59:37 +00:00
Johnny Chen 1922a8dbea Remove @skip decorators for fixed crashes.
llvm-svn: 120479
2010-11-30 22:34:19 +00:00
Sean Callanan f5a998643c Fixed a problem where m_register_info was not being
copied by the copy constructor for ClangExpressionVariable.
This meant that a NULL m_register_info wouldn't be
copied, and instead the field was uninitialized, potentially
confusing the materializer.

llvm-svn: 120472
2010-11-30 22:01:58 +00:00
Johnny Chen 876e6b31f5 Skip two test classes due to:
rdar://problem/8710994 Assertion failed: (reg_info) GDBRemoteRegisterContext.cpp, line 199

llvm-svn: 120462
2010-11-30 21:09:06 +00:00
Johnny Chen 4bb6ceb764 Entire test class skipped due to:
rdar://problem/8710994 Assertion failed: (reg_info) GDBRemoteRegisterContext.cpp, line 199

llvm-svn: 120434
2010-11-30 17:59:40 +00:00
Johnny Chen ed6b6055b4 Entire test class skipped due to:
rdar://problem/8710994 Assertion failed: (reg_info) GDBRemoteRegisterContext.cpp, line 199

llvm-svn: 120432
2010-11-30 17:53:12 +00:00
Johnny Chen ad6e34609d Skipped two test cases due to seg fault while hadnling command:
Application Specific Information:
HandleCommand(command = "expression self->str")


radar:8711052

llvm-svn: 120431
2010-11-30 17:46:46 +00:00
Johnny Chen aa9bc2eca5 Entire test class skipped due to:
rdar://problem/8710994 Assertion failed: (reg_info) GDBRemoteRegisterContext.cpp, line 199

llvm-svn: 120430
2010-11-30 17:35:58 +00:00
Jim Ingham f48169bb4f Moved the code in ClangUserExpression that set up & ran the thread plan with timeouts, and restarting with all threads into a utility function in Process. This required a bunch of renaming.
Added a ThreadPlanCallUserExpression that differs from ThreadPlanCallFunction in that it holds onto a shared pointer to its ClangUserExpression so that can't go away before the thread plan is done using it.

Fixed the stop message when you hit a breakpoint while running a user expression so it is more obvious what has happened.

llvm-svn: 120386
2010-11-30 02:22:11 +00:00
Sean Callanan 348b5897f9 Added a feature where registers can be referred to
using special $-variables from expressions.

(lldb) expr $rip

These variables are available for reading and
writing.

llvm-svn: 120367
2010-11-30 00:27:43 +00:00
Johnny Chen a994070040 Add comment for a branch of test logic which applies to ['clang', 'llvm-gcc'] only.
llvm-svn: 120355
2010-11-29 23:58:04 +00:00
Johnny Chen 11d2bae187 Minor comment fix.
llvm-svn: 120352
2010-11-29 23:47:40 +00:00
Jason Molenda c1903406e5 Fix build error when lldb is being built i386.
llvm-svn: 120322
2010-11-29 21:38:58 +00:00
Johnny Chen c9ea0f8425 Add an expect() statement within breakpoint_conditions() method which tests that
the process status is indeed 'stopped' before invoking 'frame variable -t val'
command to inspect the frame variable.  When I run the test suite, it sometimes
emits an output like:

runCmd: frame variable -t val
runCmd failed!
error: you must be stopped in a valid stack frame to view frame variables.

The expect() statement inserted is to make sure that the process is stopped and
we have a valid frame.

llvm-svn: 120318
2010-11-29 20:49:09 +00:00
Johnny Chen 4921b11320 Change the variable TestBase.timeWait to the more descriptive timeWaitNextLaunch.
Increase the delta between successive test cases, enabled with the '-w' option to
the test driver, from 0.5 to 1.0 second, to arrive at a cleaner state before the
next test case.

llvm-svn: 120315
2010-11-29 20:20:34 +00:00
Johnny Chen bbda8ed39a Add comment.
llvm-svn: 120295
2010-11-29 17:52:43 +00:00
Johnny Chen 46389e9ddf When looping through the test suite more than once (repeat count > 1), we should
ignore and not enforce the signleton pattern for the LLDBTestResult class.

llvm-svn: 120294
2010-11-29 17:50:10 +00:00
Greg Clayton facfd06c4f When we resolve a forward declaration type, be sure to put it into the m_die_to_type map so we don't assert later when someone tries to resolve the type.
llvm-svn: 119910
2010-11-20 19:16:50 +00:00
Sean Callanan fbfd7ba547 Removed a stray dump() that made it into top-of-tree.
llvm-svn: 119890
2010-11-20 02:23:49 +00:00
Sean Callanan a3aa0cf6e6 Made GetVariableValue() more robust in the face
of failures in the AST importer.  Also ensured
that a variable will not be blindly added if
GetVariableValue() returns an error.

llvm-svn: 119889
2010-11-20 02:19:29 +00:00