Author: jmolenda
Date: Thu Jan 8 00:41:12 2015
New Revision: 225431
URL: http://llvm.org/viewvc/llvm-project?rev=225431view=rev
Log:
Have AssemblyParse_x86::get_non_call_site_unwind_plan track
which registers have been spilled (saved to the stack) - and
if we see that same register being saved
Author: jmolenda
Date: Wed Jan 7 21:57:48 2015
New Revision: 225427
URL: http://llvm.org/viewvc/llvm-project?rev=225427view=rev
Log:
Rearrange RegisterContextLLDB::SavedLocationForRegister a tiny bit
so that we will use the UnwindPlan's rule for providing the stack
pointer BEFORE we use the
Author: jmolenda
Date: Mon Dec 22 05:02:02 2014
New Revision: 224698
URL: http://llvm.org/viewvc/llvm-project?rev=224698view=rev
Log:
Add support for frameless function compact unwind encodings on x86_64/i386.
This completes the compact unwind support for x86 targets.
I'm still skipping the
Author: jmolenda
Date: Sun Dec 21 04:44:54 2014
New Revision: 224689
URL: http://llvm.org/viewvc/llvm-project?rev=224689view=rev
Log:
Various unwinder work.
Most of the changes are to the FuncUnwinders class -- as we've added
more types of unwind information, the way this class was written was
Author: jmolenda
Date: Fri Dec 19 21:12:51 2014
New Revision: 224670
URL: http://llvm.org/viewvc/llvm-project?rev=224670view=rev
Log:
Re-enable compact unwind use on Mac platforms.
When lldb has a binary with protected section contents,
don't use the on-disk representation of that compact
uwnind
Author: jmolenda
Date: Tue Dec 16 00:20:30 2014
New Revision: 224321
URL: http://llvm.org/viewvc/llvm-project?rev=224321view=rev
Log:
Temporarily disable CompactUnwindInfo::GetCompactUnwindInfoForFunction.
The compact unwind importer is getting the wrong unwind info for one
case that I found. I
Author: jmolenda
Date: Tue Dec 9 14:52:26 2014
New Revision: 223826
URL: http://llvm.org/viewvc/llvm-project?rev=223826view=rev
Log:
Change AddressSanitzierRuntime to print its info message via
the Debugger's output stream instead of logging to the module.
http://reviews.llvm.org/D6577
Landed in r223826.
http://reviews.llvm.org/D6577
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Author: jmolenda
Date: Tue Dec 9 16:28:10 2014
New Revision: 223843
URL: http://llvm.org/viewvc/llvm-project?rev=223843view=rev
Log:
Add a new 'eRegisterInLiveRegisterContext' RegisterLocation to track
a register value that is live in the stack frame 0 register context.
Fixes a problem where
Author: jmolenda
Date: Tue Dec 2 17:21:05 2014
New Revision: 223174
URL: http://llvm.org/viewvc/llvm-project?rev=223174view=rev
Log:
Mark the armv7 q0-q3 and q8-q15 registers as volatile (not callee
preserved) in the ABI.
Realistically lldb isn't able to track register saves of any of
the neon
Author: jmolenda
Date: Tue Dec 2 22:02:03 2014
New Revision: 223196
URL: http://llvm.org/viewvc/llvm-project?rev=223196view=rev
Log:
Update setMCJITMemoryManager call to keep in line with llvm r223183.
Patch from Ryan Goodfellow.
Modified:
lldb/trunk/source/Expression/IRExecutionUnit.cpp
Author: jmolenda
Date: Sat Nov 29 01:05:07 2014
New Revision: 222951
URL: http://llvm.org/viewvc/llvm-project?rev=222951view=rev
Log:
The finishing touches on getting the compact unwind dumping completed
for x86_64. i386, arm, arm64 aren't handled yet but those are minor
variations on this
Author: jmolenda
Date: Thu Nov 27 07:21:38 2014
New Revision: 222884
URL: http://llvm.org/viewvc/llvm-project?rev=222884view=rev
Log:
A little more work on the compact unwind dumper.
UNWIND_X86_64_MODE_STACK_IND mode is almost correct; extra stack
space allocated before the reg saves isn't
Author: jmolenda
Date: Thu Nov 27 21:54:13 2014
New Revision: 222910
URL: http://llvm.org/viewvc/llvm-project?rev=222910view=rev
Log:
Add support for UNWIND_X86_64_MODE_STACK_IND entries.
Correct the function offset computations in UNWIND_SECOND_LEVEL_REGULAR
tables. A few other small fixes.
Author: jmolenda
Date: Fri Nov 21 18:00:17 2014
New Revision: 222592
URL: http://llvm.org/viewvc/llvm-project?rev=222592view=rev
Log:
Fix mispelled 'ling' Python property to be 'line' in
SBLineEntry and SBDeclaration. Patch from Chris Willmore.
rdar://problem/19054323
Modified:
Author: jmolenda
Date: Fri Nov 21 19:52:03 2014
New Revision: 222601
URL: http://llvm.org/viewvc/llvm-project?rev=222601view=rev
Log:
When a RegisterContext produces an invalid CFA address, change
UnwindLLDB::AddOneMoreFrame to try the fallback unwind plan on
that same stack frame before it tries
Author: jmolenda
Date: Tue Nov 18 20:29:52 2014
New Revision: 222301
URL: http://llvm.org/viewvc/llvm-project?rev=222301view=rev
Log:
Add additional checks to the SavedLocationForRegister method
where it is retrieving the Return Address register contents
on a target where that's a thing. If we
Author: jmolenda
Date: Mon Nov 17 05:43:37 2014
New Revision: 222127
URL: http://llvm.org/viewvc/llvm-project?rev=222127view=rev
Log:
Add a little sketch of a program that can extract unwind
information from the compact unwind section used on darwin
for exception handling, and dump that
Author: jmolenda
Date: Mon Nov 17 20:27:42 2014
New Revision: 14
URL: http://llvm.org/viewvc/llvm-project?rev=14view=rev
Log:
Read the LSDA and Personality Routine function address out of the
eh_frame data. These two pieces of information are used in the
process of exception handler
Author: jmolenda
Date: Mon Nov 17 22:57:28 2014
New Revision: 21
URL: http://llvm.org/viewvc/llvm-project?rev=21view=rev
Log:
Make the mutex ivar in Unwind recursive so we don't have a thread
deadlocking when we have the base Unwind class and the HistoryUnwind
subclass both trying to
Author: jmolenda
Date: Mon Nov 17 23:43:11 2014
New Revision: 25
URL: http://llvm.org/viewvc/llvm-project?rev=25view=rev
Log:
Add documentation for the SBTarget::ReadInstructions and
SBTarget::GetInstructions APIs so it's a little clearer to
understand which should be used.
Author: jmolenda
Date: Mon Nov 17 23:57:42 2014
New Revision: 26
URL: http://llvm.org/viewvc/llvm-project?rev=26view=rev
Log:
Fix up the code in the FuncUnwinders class that
retrieves the personality routine addr and the
LSDA addr. Don't bother checking with the
non-call site unwind plan
One small comment but otherwise looks good to me, fwiw.
Comment at: tools/debugserver/source/RNBRemote.cpp:2219
@@ -2219,3 +2218,3 @@
{
-for (i = buf_size-1; i = 0; i--)
+for (long i = buf_size-1; i = 0; i--)
ostrm RAWHEX8(buf[i]);
Good point, thanks.
http://reviews.llvm.org/D6271
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Author: jmolenda
Date: Wed Nov 12 13:51:43 2014
New Revision: 221810
URL: http://llvm.org/viewvc/llvm-project?rev=221810view=rev
Log:
Use PRIx64 when printing addr_t's. Don't need to force full-width 0 padding
with addresses that aren't designed to be column-aligned across multiple lines.
Author: jmolenda
Date: Thu Nov 13 01:31:45 2014
New Revision: 221866
URL: http://llvm.org/viewvc/llvm-project?rev=221866view=rev
Log:
A pretty big overhaul of the TryFallbackUnwindPlan method in
RegisterContextLLDB. I have core files of half a dozen tricky
unwind situations on x86/arm and
Author: jmolenda
Date: Tue Nov 11 02:26:44 2014
New Revision: 221677
URL: http://llvm.org/viewvc/llvm-project?rev=221677view=rev
Log:
Add an operator== to the RegisterNumber class; it simplifies
RegisterContextLLDB a bit more in a few places.
Modified:
Author: jmolenda
Date: Tue Nov 11 04:32:04 2014
New Revision: 221682
URL: http://llvm.org/viewvc/llvm-project?rev=221682view=rev
Log:
Put the current pc arrow back into the default disassembly format.
I went back and forth on removing this - and tried dropping it for
a few weeks. But when you're
Author: jmolenda
Date: Tue Nov 11 04:59:15 2014
New Revision: 221683
URL: http://llvm.org/viewvc/llvm-project?rev=221683view=rev
Log:
Add support for 32-bit core file dumping. Add support for i386 process core
file dumping.
Modified:
Author: jmolenda
Date: Tue Nov 11 20:39:14 2014
New Revision: 221760
URL: http://llvm.org/viewvc/llvm-project?rev=221760view=rev
Log:
Enable armv7 core file writing for Mach-O binaries.
The problems with the dyld all image infos struct
seems to be specific to arm64.
Modified:
Nice catch on the UnwindAssembly bug, thanks.. LazyBools only have three
possible states (and aren't going to gain extra states in the future) so the
code in CommandInterpreter was safe - but silencing warnings is always good to
do so we can spot real issues. Thanks.
Looks good, some very minor comments inlined, fine to commit after those are
tweaked. Have you tried exercising the code? Either with an architecture
default unwindplan in your ABI or with your eh_frame parser? I'd like to know
that it actually works for you. :)
Comment at:
Thanks, please commit.
http://reviews.llvm.org/D6182
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Comment at: source/Plugins/ABI/SysV-ppc/ABISysV_ppc.cpp:1015
@@ -1014,4 +1014,3 @@
-const int32_t ptr_size = 8;
-row-SetCFARegister (LLDB_REGNUM_GENERIC_SP);
-row-SetCFAOffset (8);
+const int32_t ptr_size = 4;
+row-SetCFARegister (sp_reg_num);
On Nov 9, 2014, at 1:46 PM, Justin Hibbits jr...@alumni.cwru.edu wrote:
On Nov 9, 2014 1:25 PM, Jason Molenda jmole...@apple.com wrote:
You should be calling SetCFARegisterNumberToDereference() here, shouldn't
you?
I completely missed that API when making the last patch
Author: jmolenda
Date: Sat Nov 8 02:09:22 2014
New Revision: 221577
URL: http://llvm.org/viewvc/llvm-project?rev=221577view=rev
Log:
Add a RegisterNumber class to RegisterContextLLDB.h and start using
it in RegisterContext.cpp.
There's a lot of bookkeeping code in RegisterContextLLDB where it
Author: jmolenda
Date: Fri Nov 7 23:38:17 2014
New Revision: 221575
URL: http://llvm.org/viewvc/llvm-project?rev=221575view=rev
Log:
Fix a corner case with the handling of noreturn functions.
If a noreturn function was the last function in a section,
we wouldn't correctly back up the saved-pc
FWIW the use of an Address object to represent addresses was motivated by the
hassle of address handling we had with gdb. If you start the debugger and give
it an executable and a bunch of solibs, set an address breakpoint, and then run
the process, how does that address breakpoint get re-set
Author: jmolenda
Date: Mon Nov 3 23:48:11 2014
New Revision: 221241
URL: http://llvm.org/viewvc/llvm-project?rev=221241view=rev
Log:
Add recognition for another x86 epilogue sequence (ret followed by
a nop). Fixes an instruction stepping problem when trying to step
over the final instructions
Author: jmolenda
Date: Fri Oct 31 16:30:59 2014
New Revision: 221002
URL: http://llvm.org/viewvc/llvm-project?rev=221002view=rev
Log:
SBAddress currently *may* have an Address object or it may not.
If it has an Address object, it is assumed to be Valid.
Change SBAddress to always have an Address
Author: jmolenda
Date: Thu Oct 30 22:39:11 2014
New Revision: 220947
URL: http://llvm.org/viewvc/llvm-project?rev=220947view=rev
Log:
Update xcode project file to build new ppc files.
Modified:
lldb/trunk/lldb.xcodeproj/project.pbxproj
Modified: lldb/trunk/lldb.xcodeproj/project.pbxproj
Author: jmolenda
Date: Thu Oct 30 22:40:06 2014
New Revision: 220948
URL: http://llvm.org/viewvc/llvm-project?rev=220948view=rev
Log:
Update default disassembly format string so we get
better output when we don't have any symbol name.
It looked like this:
0x1097fd029 ud2
0x1097fd02b addb
I think Todd is a little busy with non-lldb stuff right now. I've reviewed
this.
http://reviews.llvm.org/D5923
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Landed in r220762.
http://reviews.llvm.org/D5923
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Author: jmolenda
Date: Mon Oct 27 22:43:54 2014
New Revision: 220762
URL: http://llvm.org/viewvc/llvm-project?rev=220762view=rev
Log:
Add breakpoint instruction byte sequences for arm to
PlatformLinux::GetSoftwareBreakpointTrapOpcode.
Patch by Stephane Sezer.
http://reviews.llvm.org/D5923
OK I think this patch is OK to commit. Do you have llvm commit access? I'll
commit it for you if not.
J
On Oct 22, 2014, at 4:33 PM, Stephane Sezer s...@fb.com wrote:
On Oct 21, 2014, at 6:18 PM, Jason Molenda jmole...@apple.com wrote:
Stephane, this looks good to me but I'm
This looks good to me but yeah, we should prob wait for Todd to sign off on it
before committing.
http://reviews.llvm.org/D5923
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Author: jmolenda
Date: Thu Oct 16 02:47:37 2014
New Revision: 219905
URL: http://llvm.org/viewvc/llvm-project?rev=219905view=rev
Log:
Remove dead store.
clang static analyzer fixit.
Modified:
lldb/trunk/source/Symbol/LineTable.cpp
Modified: lldb/trunk/source/Symbol/LineTable.cpp
URL:
Author: jmolenda
Date: Thu Oct 16 02:49:27 2014
New Revision: 219907
URL: http://llvm.org/viewvc/llvm-project?rev=219907view=rev
Log:
Remove dead store.
clang static analyzer fixit.
Modified:
lldb/trunk/source/Target/Process.cpp
Modified: lldb/trunk/source/Target/Process.cpp
URL:
Author: jmolenda
Date: Thu Oct 16 03:05:22 2014
New Revision: 219911
URL: http://llvm.org/viewvc/llvm-project?rev=219911view=rev
Log:
Remove unreachable code.
Modified:
lldb/trunk/source/DataFormatters/CXXFormatterFunctions.cpp
Modified:
Author: jmolenda
Date: Thu Oct 16 03:08:13 2014
New Revision: 219915
URL: http://llvm.org/viewvc/llvm-project?rev=219915view=rev
Log:
Remove unreachable code.
Modified:
lldb/trunk/source/Interpreter/OptionValuePathMappings.cpp
Modified:
Author: jmolenda
Date: Thu Oct 16 03:43:27 2014
New Revision: 219920
URL: http://llvm.org/viewvc/llvm-project?rev=219920view=rev
Log:
Most of this function checks to see if m_process is non-null before
dereferencing it, except for this one section of code. Add a null
check around it.
clang
Author: jmolenda
Date: Thu Oct 16 11:59:23 2014
New Revision: 219936
URL: http://llvm.org/viewvc/llvm-project?rev=219936view=rev
Log:
Fix MemoryHistory plugin to check whether the plugin
was able to create itself before returning the shared
pointer to it.
clang warning.
Modified:
Author: jmolenda
Date: Thu Oct 16 18:10:03 2014
New Revision: 219983
URL: http://llvm.org/viewvc/llvm-project?rev=219983view=rev
Log:
Change a use of mktemp() to mkstemp() for better security.
We have two more uses of mktemp still in the source base
but they'll take a little more consideration.
Author: jmolenda
Date: Wed Oct 15 18:27:12 2014
New Revision: 219865
URL: http://llvm.org/viewvc/llvm-project?rev=219865view=rev
Log:
Give user_id a default invalid value so we don't
possibly use it uninitialized in a log message later.
clang static analyzer fixit.
Modified:
Author: jmolenda
Date: Wed Oct 15 20:23:06 2014
New Revision: 219886
URL: http://llvm.org/viewvc/llvm-project?rev=219886view=rev
Log:
Check that process is non-null before calling a method in it.
clang static analyzer fixit.
Modified:
lldb/trunk/source/Expression/ClangUserExpression.cpp
Author: jmolenda
Date: Wed Oct 15 20:27:31 2014
New Revision: 219888
URL: http://llvm.org/viewvc/llvm-project?rev=219888view=rev
Log:
Ah, accidentally committed a patch I didn't mean to.
Modified:
lldb/trunk/source/Expression/DWARFExpression.cpp
Modified:
Author: jmolenda
Date: Wed Oct 15 20:40:16 2014
New Revision: 219889
URL: http://llvm.org/viewvc/llvm-project?rev=219889view=rev
Log:
Guard against NULL derefs.
clang static analyzer fixits.
Modified:
lldb/trunk/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp
Modified:
Author: jmolenda
Date: Wed Oct 15 21:53:57 2014
New Revision: 219895
URL: http://llvm.org/viewvc/llvm-project?rev=219895view=rev
Log:
Fix accidental over-checking of args in launcherXPCService.
Modified:
lldb/trunk/source/Host/macosx/launcherXPCService/main.mm
Modified:
LGTM.
http://reviews.llvm.org/D5779
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Author: jmolenda
Date: Tue Oct 14 22:07:39 2014
New Revision: 219771
URL: http://llvm.org/viewvc/llvm-project?rev=219771view=rev
Log:
Make sure local var cu is non-NULL before dereferencing.
(it was checked for NULL-ness in some places, not in others)
clang static analyzer fixit.
Modified:
Author: jmolenda
Date: Tue Oct 7 17:55:13 2014
New Revision: 219247
URL: http://llvm.org/viewvc/llvm-project?rev=219247view=rev
Log:
When we detect a stack unwind loop, before we abort
the backtrace, try falling back to the architecture default
unwind plan and see if we can backtrace a little
Author: jmolenda
Date: Tue Oct 7 18:37:28 2014
New Revision: 219251
URL: http://llvm.org/viewvc/llvm-project?rev=219251view=rev
Log:
Update TestInlinedBreakpoints to reflect the fact that
the default search method is always as of r218405.
For the purposes of this test, set it back to headers
to
Author: jmolenda
Date: Mon Oct 6 17:15:29 2014
New Revision: 219151
URL: http://llvm.org/viewvc/llvm-project?rev=219151view=rev
Log:
Update xcode project file to match zach's move of connetionfiledescriptor in
r219143.
Modified:
lldb/trunk/lldb.xcodeproj/project.pbxproj
Modified:
Author: jmolenda
Date: Mon Oct 6 17:23:30 2014
New Revision: 219152
URL: http://llvm.org/viewvc/llvm-project?rev=219152view=rev
Log:
Add another address to check for the kernel's load addr in debug configs.
rdar://problem/18560328
Modified:
Author: jmolenda
Date: Thu Oct 2 01:17:15 2014
New Revision: 218853
URL: http://llvm.org/viewvc/llvm-project?rev=218853view=rev
Log:
Update to match strings output by lldb.
Modified:
lldb/trunk/test/functionalities/platform/TestPlatformCommand.py
Modified:
Author: jmolenda
Date: Wed Oct 1 20:08:16 2014
New Revision: 218845
URL: http://llvm.org/viewvc/llvm-project?rev=218845view=rev
Log:
Add a mutex lock to StackFrame to protect race conditions for
updating its ivars. We've had a lot of crash reports and careful
analysis shows that we've got
Author: jmolenda
Date: Wed Oct 1 23:34:44 2014
New Revision: 218850
URL: http://llvm.org/viewvc/llvm-project?rev=218850view=rev
Log:
Add SBExecutionContext.h to the list of headers to copy into the LLDB framework.
Modified:
lldb/trunk/lldb.xcodeproj/project.pbxproj
Modified:
Author: jmolenda
Date: Thu Oct 2 00:15:07 2014
New Revision: 218851
URL: http://llvm.org/viewvc/llvm-project?rev=218851view=rev
Log:
Add a couple of extra exepct strings so this test file keeps in sync with lldb
a little better.
Modified:
Looks good; one small suggestion, otherwise please commit.
Comment at: source/Plugins/SymbolFile/DWARF/DWARFDebugLine.cpp:427
@@ -426,1 +426,3 @@
prologue-min_inst_length = debug_line_data.GetU8(offset_ptr);
+if (prologue-version = 4)
+
Author: jmolenda
Date: Mon Sep 22 21:43:35 2014
New Revision: 218291
URL: http://llvm.org/viewvc/llvm-project?rev=218291view=rev
Log:
ConnectionFileDescriptor::Connect()'s handling of the fd:// method
was broken in r214984 by the addition of an unconditional error
return at the start of the code
Looks good, a couple changes to match the lldb coding style better; please
commit.
Comment at: source/Plugins/MemoryHistory/asan/MemoryHistoryASan.cpp:119
@@ +118,3 @@
+
+if (count = 0) return;
+
Please split into two lines,
```
if (count = 0)
return;
```
Committed. (I don't think Paul has commit access)
Sendingsource/Interpreter/CommandObject.cpp
Transmitting file data .
Committed revision 218181.
I also updated eFlagRequiresThread and eFlagRequiresProcess in a similar fashion
to Paul's changes to eFlagRequiresFrame.
On Sep 19, 2014,
Looks good to me! Anyone else?
On Sep 19, 2014, at 1:40 PM, Paul Osmialowski paw...@king.net.pl wrote:
Maybe this looks better?
http://reviews.llvm.org/D5374
Files:
source/Interpreter/CommandObject.cpp
D5374.13888.patch___
lldb-commits
Looks good to me! Anyone else?
http://reviews.llvm.org/D5374
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
I applied the patch and was going to commit it, then I realized I don't clearly
understand what behavior we're fixing. e.g. these commands all report a
reasonable error msg when a thread/process/frame is missing before the patch is
applied.
% lldb
(lldb) f
error: invalid thread
(lldb) bt
create' command
instead of listing all of the things that the command needs (frame, thread,
process, target)?
On Sep 18, 2014, at 7:49 PM, Jason Molenda jmole...@apple.com wrote:
I applied the patch and was going to commit it, then I realized I don't
clearly understand what behavior we're
This looks like a good change, thank you for coming up with it. Could you put
the statements on separate lines from the conditional expressions? Instead of
if (!m_exe_ctx.HasTargetScope()) result.AppendError
(GetInvalidTargetDescription());
more like this -
if (!m_exe_ctx.HasTargetScope())
This looks like a good change, thank you for coming up with it. Could you put
the statements on separate lines from the conditional expressions? Instead of
if (!m_exe_ctx.HasTargetScope()) result.AppendError
(GetInvalidTargetDescription());
more like this -
if (!m_exe_ctx.HasTargetScope())
Author: jmolenda
Date: Thu Sep 11 19:09:04 2014
New Revision: 217649
URL: http://llvm.org/viewvc/llvm-project?rev=217649view=rev
Log:
Change SystemRuntimeMacOSX::ReadLibdispatchTSDIndexes to use the
ProcessStructReader instead of reading the structure out of libdispatch
by hand. I will convert
Author: jmolenda
Date: Thu Sep 11 20:50:46 2014
New Revision: 217653
URL: http://llvm.org/viewvc/llvm-project?rev=217653view=rev
Log:
Fix the ctor ivar initialization formatting for Debugger,
TypeValidatorImpl, FileAction, and ProcessLaunchInfo to match the
lldb coding convention.
Modified:
Greg has done most of the DWARF parser work, he should comment.
In DWARFDebugInfoEntry.cpp DW_FORM_ref_addr you use cu-GetAddressByteSize()
for the size (or cu_addr_size, used in a different part of the file) -- but if
this is a 64-bit binary using DWARF32, that would be wrong - this is a
Looks good Kuba, please commit.
J
http://reviews.llvm.org/D4596
___
lldb-commits mailing list
lldb-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits
Looks good Kuba, please commit.
J
On Sep 2, 2014, at 7:33 PM, Kuba Brecka kuba.bre...@gmail.com wrote:
Replaced the MemoryHistory* to use MemoryHistorySP instead. Also changed
the behavior of the lldbtest.py method for finding the just-built clang to
return the path, instead of setting
This looks fine to me - but Jim/Enrico/Greg may want to comment.
One thing I'd change is that you document the
SBValue::GetNumMemoryHistoryThreads and SBValue::GetMemoryHistoryThreadAtIndex
in include/lldb/API/SBValue.h but not in scripts/Python/interface/SBValue.i.
The generated
This looks fine to me - but Jim/Enrico/Greg may want to comment.
One thing I'd change is that you document the
SBValue::GetNumMemoryHistoryThreads and SBValue::GetMemoryHistoryThreadAtIndex
in include/lldb/API/SBValue.h but not in scripts/Python/interface/SBValue.i.
The generated
For what it's worth, with the temporal/logical backtrace feature we added
SBThread::GetExtendedBacktraceThread() which hands out a fake SBThread similar
to the ones Kuba is doing here. (in that case, we don't want to fetch all of
the extended backtraces, e.g. to support a
Thanks Paul.
Does the ReadRegister method really do the correct thing for the x0/w0 or d0/s0
registers? If we ask for w0, it's going to realize this is a subreg of x0, and
call it self again asking for x0:
if (is_subreg)
{
// Read the full aligned 64-bit register.
sneaky
problems to find when we're trying to run lldb on a BE PPC system or something.
:)
On Sep 2, 2014, at 12:50 PM, Jason Molenda jmole...@apple.com wrote:
Thanks Paul.
Does the ReadRegister method really do the correct thing for the x0/w0 or
d0/s0 registers? If we ask for w0, it's
Yeah, sorry for not looking more closely at the patch earlier.
As to my so we're going to call RegisterValue::SetUInt64 on that -- I don't
see where we'd recognize we only want the lower 32 bits and return those
comment, we may just be lucking out when the target / host are both little
endian.
Looks good to me, please commit. I'll be the one who has to fix it if our
arm64 back end is broken :) but I don't anticipate any problems.
Is this necessary in RegisterInfos_arm64.h?
+#include ../../../Utility/ARM64_GCC_Registers.h
+#include ../../../Utility/ARM64_DWARF_Registers.h
The
In RegisterContextPOSIXProcessMonitor_arm64::ReadRegister this code seems to be
copied from an x86 version of this impl:
if (success)
{
// If our read was not aligned (for ah,bh,ch,dh), shift our
returned value one byte to the right.
if (is_subreg
Hey Paul,
http://reviews.llvm.org/D4579 - accepted, not commited
I assume this on Todd.
http://reviews.llvm.org/D4580 - waiting on Apple review
I replied separately but this looks good to me.
http://reviews.llvm.org/D4430 - waiting for review
Greg said on July 10 that this looks fine.
Author: jmolenda
Date: Tue Aug 26 17:39:03 2014
New Revision: 216498
URL: http://llvm.org/viewvc/llvm-project?rev=216498view=rev
Log:
Don't have ProcessMachCore::CanDebug use the Target's arch when
creating the ModuleSpec to load the core file - we won't have a fat
core file and we can end up
Author: jmolenda
Date: Tue Aug 26 18:31:00 2014
New Revision: 216504
URL: http://llvm.org/viewvc/llvm-project?rev=216504view=rev
Log:
Clean up after the multithreaded test case finishes.
Patch by Sean Callanan.
rdar://problem/18140875
Modified:
lldb/trunk/test/api/multithreaded/Makefile
trying to debug the
crash like I did. ;)
On Aug 21, 2014, at 7:46 PM, Jason Molenda jmole...@apple.com wrote:
Author: jmolenda
Date: Thu Aug 21 21:46:46 2014
New Revision: 216258
URL: http://llvm.org/viewvc/llvm-project?rev=216258view=rev
Log:
When adding a dSYM to an existing ObjectFile
Author: jmolenda
Date: Thu Aug 21 18:22:33 2014
New Revision: 216243
URL: http://llvm.org/viewvc/llvm-project?rev=216243view=rev
Log:
Two small fixes to get Mac native + debugserver working after the
HostInfo et al changes from Zachary. Changes suggested by Zachary
- fixes the problems I was
Author: jmolenda
Date: Fri Aug 15 16:38:13 2014
New Revision: 215771
URL: http://llvm.org/viewvc/llvm-project?rev=215771view=rev
Log:
Fix dependency ordering for the xpc file installs.
Modified:
lldb/trunk/lldb.xcodeproj/project.pbxproj
Modified: lldb/trunk/lldb.xcodeproj/project.pbxproj
Hi Tong, sorry for not replying earlier.
I don't understand why the current code is wrong.
eRegisterKindGCC = 0,// the register numbers seen in eh_frame
eRegisterKindDWARF, // the register numbers seen DWARF
and
SectionSP sect = sl-FindSectionByType
, Aug 12, 2014 at 3:51 PM, Jason Molenda jmole...@apple.com wrote:
Hm, on my Mac I'm getting the correct registers from the eh_frame.
* thread #1: tid = 0xce36, 0x1f30 a.out`main, queue =
'com.apple.main-thread', stop reason = breakpoint 1.1
#0: 0x1f30 a.out`main
a.out`main
101 - 200 of 304 matches
Mail list logo