The problem is that in PlatformNetBSD::Initialize you call
Platform::SetHostPlatform unconditionally. You want to
call Platform::SetHostPlatform only if we are running on NetBSD so you have
to surround it with some conditional compilation (see example in
PlatformLinux::Initialize)
On Fri, Nov 20,
tberghammer added a comment.
I agree that it can be specified by CFALGS and CXXFLAGS but we will need some
way in the test suit to xfail a test based on the instruction set and sometime
we also need different test expectations based on it. Doing these based on
C(XX)FALGS is a bit problematic so
This revision was automatically updated to reflect the committed changes.
Closed by commit rL253657: Use thumb instruction set for ldb-server on android
arm (authored by tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D14816?vs=40635&id=40756#toc
Repository:
rL LLVM
http://re
Author: tberghammer
Date: Fri Nov 20 04:51:39 2015
New Revision: 253657
URL: http://llvm.org/viewvc/llvm-project?rev=253657&view=rev
Log:
Use thumb instruction set for ldb-server on android arm
Differential revision: http://reviews.llvm.org/D14816
Modified:
lldb/trunk/cmake/platforms/Android
tberghammer added a comment.
Can we create a setup where the test suit is capable of running in all 3 mode
(arm/thumb/mixed)?
I think we should introduce a new architecture called "arm-thumb" (open for
better name suggestions) what run the test suit in mixed mode while running the
test suit wi
tberghammer added a comment.
In http://reviews.llvm.org/D14816#292929, @omjavaid wrote:
> -mthumb will force T32 instruction set while -marm will force A32.
>
> Best is not to use any of these flags to let the compiler decide best
> possible instruction set combination.
I think because we spec
tberghammer created this revision.
tberghammer added reviewers: labath, ovyalov.
tberghammer added a subscriber: lldb-commits.
Herald added subscribers: srhines, danalbert, tberghammer, rengolin, aemerson.
Use thumb instruction set for ldb-server on android arm
This change reduce the size of the
tberghammer added a comment.
Hi,
I checked out the implementation of RegisterValue and this is my understanding
what is happening:
- When we call RegisterValue::SetUInt it will be creating a 32bit unsigned
llvm::APInt
- The llvm::APInt will store the data in a uint64_t
- Calling RegisterValue:
Author: tberghammer
Date: Wed Nov 18 06:11:34 2015
New Revision: 253455
URL: http://llvm.org/viewvc/llvm-project?rev=253455&view=rev
Log:
Revert 2 commits breaking the MSVC build
Revert "Remove a few vestigial typedefs from the old world"
This reverts commit 05872cda2a00fbd988c4fc761b1f87fe9edce2
tberghammer added a comment.
As far as I know the gdb remote protocol says that the registers in the 'p'
packet should be displayed in target byte order, but the protocol isn't too
well specified (and in my opinion target byte order is a silly decision).
If we accept that the 'p' packet is in t
tberghammer added a subscriber: tberghammer.
tberghammer added a comment.
Can't you address this issue with compiling LLDB on NetBSD with
-DLLDB_DISABLE_CURSES?
If you want to go with the current approach then you should add somebody who
can verify that committing it in to the LLVM source tree
tberghammer added a comment.
The feature Greg mentioned is already implemented in
DWARFDebugInfoEntry::GetAttributeHighPC and we shouldn't move it to
DWARFFormValue::Address because it is only needed for DW_AT_high_pc but
DWARFFormValue::Address used for other places as well (e.g. DW_AT_low_pc)
tberghammer added a comment.
Looks much better, but I think the root cause of your problem is that you are
using RegisterValue::SetBytes instead of RegisterValue::SetUInt. I would
suggest to use a code like this (I don't have a mips environment at the moment
to try it out):
Error
NativeReg
This revision was automatically updated to reflect the committed changes.
Closed by commit rL253027: Remove a broken hack from Scalar::ULongLong and fix
a test (authored by tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D14611?vs=40031&id=40128#toc
Repository:
rL LLVM
http:/
Author: tberghammer
Date: Fri Nov 13 04:51:35 2015
New Revision: 253027
URL: http://llvm.org/viewvc/llvm-project?rev=253027&view=rev
Log:
Remove a broken hack from Scalar::ULongLong and fix a test
Change Test-rdar-12481949.py to expect GetValueAsUnsigned() to return
0x if the variable is
We already have a test for it in TestConsecutiveBreakpoints.py what is
xfail-ed on all platform because of this bug (http://llvm.org/pr23478). As
far as I see from the build bots the test is also failing on OSX, but it
might fail from a different reason then on Linux/Windows.
On Fri, Nov 13, 2015
tberghammer requested changes to this revision.
tberghammer added a comment.
This revision now requires changes to proceed.
You are created a new ReadRegisterUnsigned method in the derived class what
will hide a method from the base class with the same name but with different
return type. I thin
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
LGTM
(Next time please upload the diff with full context because it is easier to
review that way)
Repository:
rL LLVM
http://reviews.llvm.org/D14631
_
tberghammer created this revision.
tberghammer added reviewers: clayborg, granata.enrico.
tberghammer added a subscriber: lldb-commits.
Remove a broken hack from Scalar::ULongLong and fix a test
Change Test-rdar-12481949.py to expect GetValueAsUnsigned() to return
0x if the variable is an
tberghammer abandoned this revision.
tberghammer added a comment.
I haven't noticed we already store the symbol index. Abandon it in favor of the
API you suggested
http://reviews.llvm.org/D14536
___
lldb-commits mailing list
lldb-commits@lists.llvm
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
Te test case looks good to me. Thanks for adding it.
For the actual code change I don' know enough about this area to say anything
so please wait for a feedback from Greg.
http://r
tberghammer added a subscriber: tberghammer.
tberghammer requested changes to this revision.
tberghammer added a reviewer: tberghammer.
tberghammer added a comment.
This revision now requires changes to proceed.
Please create a test case for this scenario so it won't break again
http://reviews.l
tberghammer added a comment.
If you want to autotools build support fro FenBSD then please go for it, but
wanted to mention that it will most likely go away within a few month so you
won't invest too much time into it.
Repository:
rL LLVM
http://reviews.llvm.org/D14531
__
tberghammer added a subscriber: tberghammer.
tberghammer accepted this revision.
tberghammer added a reviewer: tberghammer.
tberghammer added a comment.
This revision is now accepted and ready to land.
Looks good (I remember fixing some similar issues in SymbolFileDWARF where we
specify append=tr
tberghammer created this revision.
tberghammer added a reviewer: clayborg.
tberghammer added a subscriber: lldb-commits.
Herald added a subscriber: aemerson.
Add empty symbols to symtab for skipped symbols
LLDB skips several special purpose symbols during symtab parsing (e.g.:
arm/aarch64 mapping
tberghammer added a subscriber: tberghammer.
tberghammer added a comment.
I have no objection against this change but considering that LLVM plans to
remove the autotools/gmake support after branching the 3.8 release (most likely
in January 2016) it would be good if you can bring up CMake on NetB
tberghammer requested changes to this revision.
Comment at: source/Plugins/ObjectFile/ELF/ObjectFileELF.cpp:2191-2213
@@ -2170,25 +2190,25 @@
if (arch.GetMachine() == llvm::Triple::arm)
{
if (symbol_type == eSymbolTypeCode)
tberghammer added a comment.
If the purpose of the change to get closer to upstream then I am fine with
removing the bug number here. In general I don't feel it is a that high risk
change, but I might miss something.
http://reviews.llvm.org/D14406
___
tberghammer added inline comments.
Comment at: packages/Python/lldbsuite/test/lldbtest.py:605
@@ -611,2 +604,3 @@
if expected_fn(self):
-raise case._UnexpectedSuccess(sys.exc_info(), bugnumber)
+xfail_func = unittest2.expectedFailure(fu
Hi Enrico,
Skipping files this way isn't working because of the way the LLVM build
system works (see:
http://lab.llvm.org:8011/builders/lldb-x86_64-ubuntu-14.04-cmake/builds/8151/steps/cmake%20local/logs/stdio
)
Do you have any suggestion for a quick fix better then surrounding the
affected files
Author: tberghammer
Date: Wed Nov 4 13:39:32 2015
New Revision: 252067
URL: http://llvm.org/viewvc/llvm-project?rev=252067&view=rev
Log:
Revert r252001 as it brake TestCppIncompleteTypes on all platforms
Modified:
lldb/trunk/source/Plugins/SymbolFile/DWARF/DWARFASTParserClang.cpp
Modified:
Author: tberghammer
Date: Wed Nov 4 13:32:01 2015
New Revision: 252065
URL: http://llvm.org/viewvc/llvm-project?rev=252065&view=rev
Log:
Fix TestMemoryRead after r252035
Modified:
lldb/trunk/source/Commands/CommandObjectMemory.cpp
Modified: lldb/trunk/source/Commands/CommandObjectMemory.cpp
Author: tberghammer
Date: Tue Nov 3 16:31:50 2015
New Revision: 251997
URL: http://llvm.org/viewvc/llvm-project?rev=251997&view=rev
Log:
Revert change comitted in accidentally as r251992
Modified:
lldb/trunk/source/Symbol/Symtab.cpp
Modified: lldb/trunk/source/Symbol/Symtab.cpp
URL:
http:/
Author: tberghammer
Date: Tue Nov 3 16:29:20 2015
New Revision: 251993
URL: http://llvm.org/viewvc/llvm-project?rev=251993&view=rev
Log:
Use std::list::splice in TaskPool to avoid an allocation
Using std::list::splice to move an element from one list to an other
avoids the allocation of a new el
Author: tberghammer
Date: Tue Nov 3 16:29:16 2015
New Revision: 251992
URL: http://llvm.org/viewvc/llvm-project?rev=251992&view=rev
Log:
wip
Modified:
lldb/trunk/source/Symbol/Symtab.cpp
Modified: lldb/trunk/source/Symbol/Symtab.cpp
URL:
http://llvm.org/viewvc/llvm-project/lldb/trunk/sourc
Author: tberghammer
Date: Tue Nov 3 12:03:47 2015
New Revision: 251951
URL: http://llvm.org/viewvc/llvm-project?rev=251951&view=rev
Log:
Skip TestBacktraceAll on android-arm
The android compiler can't compile the inferior because of an issue
in the standard library.
Modified:
lldb/trunk/pa
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
LGTM
Comment at: source/Host/posix/ConnectionFileDescriptorPosix.cpp:182
@@ -156,2 +181,3 @@
{
-if (strstr(s, "listen://") == s)
+const char *ad
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
LGTM
http://reviews.llvm.org/D14264
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/lis
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
Looks good with a few minor suggestions inline
Comment at:
source/Plugins/DynamicLoader/POSIX-DYLD/DynamicLoaderPOSIXDYLD.cpp:578-579
@@ +577,4 @@
+}
+
+ret
Author: tberghammer
Date: Tue Oct 27 05:56:35 2015
New Revision: 251403
URL: http://llvm.org/viewvc/llvm-project?rev=251403&view=rev
Log:
Fix MSVC build after r251402
Modified:
lldb/trunk/source/Plugins/Language/CPlusPlus/CPlusPlusLanguage.cpp
Modified: lldb/trunk/source/Plugins/Language/CPl
This revision was automatically updated to reflect the committed changes.
Closed by commit rL251402: Some minor improvements on the symtab parsing code
(authored by tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D14088?vs=38445&id=38520#toc
Repository:
rL LLVM
http://reviews
Author: tberghammer
Date: Tue Oct 27 05:43:27 2015
New Revision: 251402
URL: http://llvm.org/viewvc/llvm-project?rev=251402&view=rev
Log:
Some minor improvements on the symtab parsing code
* Remove an unneccessary re-computaion on arch spec from the ELF file
* Use a local cache to optimize name b
tberghammer created this revision.
tberghammer added a reviewer: clayborg.
tberghammer added a subscriber: lldb-commits.
Some minor improvements on the symtab parsing code
* Remove an unnecessary re-computation on arch spec from the ELF file
* Use a local cache to optimize name based section look
Author: tberghammer
Date: Mon Oct 26 13:10:55 2015
New Revision: 251319
URL: http://llvm.org/viewvc/llvm-project?rev=251319&view=rev
Log:
Fix an API missues in ParseTypeFromDWARF
The arguments for ClangASTContext::CreateMemberPointerType was passed
in in the wrong order.
Modified:
lldb/trunk
tberghammer added a subscriber: tberghammer.
tberghammer added a comment.
Before commit please sync with Ying about updating all android build bot to
still skip the MI tests after the change.
http://reviews.llvm.org/D14060
___
lldb-commits mailing
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
LGTM
http://reviews.llvm.org/D14051
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/lis
Author: tberghammer
Date: Mon Oct 26 05:53:58 2015
New Revision: 251282
URL: http://llvm.org/viewvc/llvm-project?rev=251282&view=rev
Log:
Fix a fission expression evaluation issue
This fix should eliminate the duplicate definition errors when debug
info is available in multiple dwo symbol file fo
Author: tberghammer
Date: Fri Oct 23 08:36:31 2015
New Revision: 25
URL: http://llvm.org/viewvc/llvm-project?rev=25&view=rev
Log:
Fix arm lldb-server on aarch64 device
* Use PTRACE_GETVFPREGS/PTRACE_SETVFPREGS to access the floating point
registers instead of the old PTRACE_GETFPREGS/PT
4 PM Joerg Sonnenberger
wrote:
> On Fri, Oct 23, 2015 at 10:34:54AM -0000, Tamas Berghammer via
> lldb-commits wrote:
> > Author: tberghammer
> > Date: Fri Oct 23 05:34:53 2015
> > New Revision: 251107
> >
> > URL: http://llvm.org/viewvc/llvm-project?rev=251107&
Author: tberghammer
Date: Fri Oct 23 05:53:31 2015
New Revision: 251109
URL: http://llvm.org/viewvc/llvm-project?rev=251109&view=rev
Log:
Try to fix MSVC build after r251105
Modified:
lldb/trunk/include/lldb/Core/Timer.h
lldb/trunk/source/Core/Timer.cpp
Modified: lldb/trunk/include/lldb/
This revision was automatically updated to reflect the committed changes.
Closed by commit rL251107: Disable the strict-aliasing warnings produced by gcc
(authored by tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D13981?vs=38115&id=38226#toc
Repository:
rL LLVM
http://revie
Author: tberghammer
Date: Fri Oct 23 05:34:49 2015
New Revision: 251106
URL: http://llvm.org/viewvc/llvm-project?rev=251106&view=rev
Log:
Re-commit "Make dwarf parsing multi-threaded"
Re-commit the change after fixing a lot of race condition in LLDB
exposed by this change
Loading the debug info
Author: tberghammer
Date: Fri Oct 23 05:34:53 2015
New Revision: 251107
URL: http://llvm.org/viewvc/llvm-project?rev=251107&view=rev
Log:
Disable the strict-aliasing warnings produced by gcc
GCC produce a lot of strict-aliasing warning for the LLDB codebase
what makes reading the compile output v
This revision was automatically updated to reflect the committed changes.
Closed by commit rL251105: Fix race conditions in Core/Timer (authored by
tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D13940?vs=38118&id=38225#toc
Repository:
rL LLVM
http://reviews.llvm.org/D13940
Author: tberghammer
Date: Fri Oct 23 05:34:29 2015
New Revision: 251105
URL: http://llvm.org/viewvc/llvm-project?rev=251105&view=rev
Log:
Fix race conditions in Core/Timer
The Timer class already had some support for multi-threaded access
but it still contained several race conditions. This CL fi
tberghammer added a comment.
I decided to go with your approach primarily because I tried it out with lower
number of threads and it performed marginally better (~10%) in that case
http://reviews.llvm.org/D13662
___
lldb-commits mailing list
lldb-c
tberghammer added a comment.
I created a bug for it as https://llvm.org/bugs/show_bug.cgi?id=25296
I don't see too much value in creating a "janitor" mode because I think when
people want to do some cleanup, they will focus on one specific area where they
enable the necessary warnings manually.
tberghammer removed rL LLVM as the repository for this revision.
tberghammer updated this revision to Diff 38128.
tberghammer added a comment.
I tried out the implementation you suggests and made some measurements. The
difference between the 2 implementation when attaching to LLDB is negligible
tberghammer updated this revision to Diff 38118.
tberghammer added a comment.
We already used some thread local storage in the same file to store the current
Timer stack so I decided to move the depth to the same storage object (it is
using the Host::ThreadLocalStorage* methods).
Originally I d
tberghammer created this revision.
tberghammer added reviewers: clayborg, brucem.
tberghammer added a subscriber: lldb-commits.
Disable the strict-aliasing warnings produced by gcc
GCC produce a lot of strict-aliasing warning for the LLDB codebase
what makes reading the compile output very diffic
This revision was automatically updated to reflect the committed changes.
tberghammer marked 5 inline comments as done.
Closed by commit rL251006: Fix some race condition in ConstString around
Mangled name handling (authored by tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D139
Author: tberghammer
Date: Thu Oct 22 06:14:37 2015
New Revision: 251007
URL: http://llvm.org/viewvc/llvm-project?rev=251007&view=rev
Log:
Make SymbolFileDWARF::GetCachedSectionData thread safe
Differential revision: http://reviews.llvm.org/D13942
Modified:
lldb/trunk/source/Plugins/SymbolFil
This revision was automatically updated to reflect the committed changes.
Closed by commit rL251007: Make SymbolFileDWARF::GetCachedSectionData thread
safe (authored by tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D13942?vs=38019&id=38109#toc
Repository:
rL LLVM
http://rev
Author: tberghammer
Date: Thu Oct 22 06:14:31 2015
New Revision: 251006
URL: http://llvm.org/viewvc/llvm-project?rev=251006&view=rev
Log:
Fix some race condition in ConstString around Mangled name handling
Differential revision: http://reviews.llvm.org/D13941
Modified:
lldb/trunk/source/Core
tberghammer updated this revision to Diff 38019.
tberghammer added a comment.
Use std::call_once instead of a std::atomic+std::mutex combination based on
some offline discussion with Pavel
http://reviews.llvm.org/D13942
Files:
source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp
source/Plug
tberghammer created this revision.
tberghammer added reviewers: labath, clayborg.
tberghammer added a subscriber: lldb-commits.
Make SymbolFileDWARF::GetCachedSectionData thread safe
http://reviews.llvm.org/D13942
Files:
source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp
source/Plugins/Symb
tberghammer created this revision.
tberghammer added reviewers: labath, clayborg.
tberghammer added a subscriber: lldb-commits.
Fix race conditions in Core/Timer
The Timer class already had some support for multi-threaded access
but it still contained several race conditions. This CL fixes them
i
tberghammer created this revision.
tberghammer added reviewers: labath, clayborg.
tberghammer added a subscriber: lldb-commits.
Fix some race condition in ConstString around Mangled name handling
http://reviews.llvm.org/D13941
Files:
source/Core/ConstString.cpp
Index: source/Core/ConstString.
tberghammer accepted this revision.
Comment at: source/Plugins/Platform/gdb-server/PlatformRemoteGDBServer.cpp:39
@@ +38,3 @@
+#include
+#include
+
Please don't include here if it isn't absolutely necessary as there is
a bug on windows in the header.
If you n
tberghammer added a comment.
I reverted this change, as it caused some race condition, but see my comment
inline.
Comment at:
lldb/trunk/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp:2087-2088
@@ +2086,4 @@
+results.emplace_back(TaskPool::AddTask(parser_fn, c
t; chunks, or kicking off no more than hardware_concurrency compile units at a
> time, and having them all share a mutex around a queue of compilation
> units, so that when each one's work is finished, it can queue the next item
> (if one still exists).
>
> On Tue, Oct 20, 2015 at
Author: tberghammer
Date: Tue Oct 20 11:16:35 2015
New Revision: 250833
URL: http://llvm.org/viewvc/llvm-project?rev=250833&view=rev
Log:
Fix MSVC build after 250820
Work around a bug in MSVC 12 where _HAS_EXCEPTIONS=0 don't eliminate
all usage of __uncaught_exception with including eh.h what dec
Author: tberghammer
Date: Tue Oct 20 10:43:40 2015
New Revision: 250832
URL: http://llvm.org/viewvc/llvm-project?rev=250832&view=rev
Log:
Revert "Make dwarf parsing multi-threaded"
Revert it bacuse it introduces several race condition detected by
the build bots.
This reverts commit 5107a5ebdb7c4
Author: tberghammer
Date: Tue Oct 20 07:42:10 2015
New Revision: 250821
URL: http://llvm.org/viewvc/llvm-project?rev=250821&view=rev
Log:
Make dwarf parsing multi-threaded
Loading the debug info from a large application is the slowest task
LLDB do. This CL makes most of the dwarf parsing code mul
This revision was automatically updated to reflect the committed changes.
tberghammer marked 4 inline comments as done.
Closed by commit rL250820: Add a new task pool class to LLDB (authored by
tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D13727?vs=37566&id=37865#toc
Reposito
This revision was automatically updated to reflect the committed changes.
Closed by commit rL250821: Make dwarf parsing multi-threaded (authored by
tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D13662?vs=37353&id=37866#toc
Repository:
rL LLVM
http://reviews.llvm.org/D13662
Author: tberghammer
Date: Tue Oct 20 07:42:05 2015
New Revision: 250820
URL: http://llvm.org/viewvc/llvm-project?rev=250820&view=rev
Log:
Add a new task pool class to LLDB
The purpose of the class is to make it easy to execute tasks in parallel
Basic design goals:
* Have a very lightweight and e
tberghammer added a comment.
Looks good with a few minor comments inline
One additional design question:
What is your opinion about specifying the default value for the memory cache
line size (the 512 byte) in Platform::GetDefaultMemoryCacheLineSize() instead
of in the property definition? I th
Author: tberghammer
Date: Tue Oct 20 04:21:04 2015
New Revision: 250810
URL: http://llvm.org/viewvc/llvm-project?rev=250810&view=rev
Log:
Fix build and some warnings after r250798
Modified:
lldb/trunk/include/lldb/DataFormatters/CXXFunctionPointer.h
lldb/trunk/include/lldb/DataFormatters/
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
LGTM
http://reviews.llvm.org/D12522
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/lis
tberghammer added a comment.
I thought a bit about the points raised by Todd and in general I am happy with
overloading the system during testing with a very high number of tests as it
will most likely help us to detect race conditions and based on my testing (on
a 40 core Linux machine) it isn
tberghammer added a comment.
I agree with Greg that the Platform should hand out the value but I would
suggest to create a Platform::GetMemoryCacheLineSize() function what will be
overwritten in PlatfromAndroid.
> ADB packets have a fixed size of 4k.
It is not true. ADB packets have a maximum
tberghammer added a comment.
Based on Pavel's example and some additional experimenting we done I am not
sure if std::async will give us any benefit even on Windows as it don't limit
the number of threads to the number of cores (because it can't do if it want to
implement the standard). I have
tberghammer updated this revision to Diff 37566.
tberghammer added a comment.
Create optional std::async based implementation
http://reviews.llvm.org/D13727
Files:
include/lldb/Utility/TaskPool.h
source/Utility/CMakeLists.txt
source/Utility/TaskPool.cpp
unittests/Utility/CMakeLists.txt
tberghammer added a comment.
In http://reviews.llvm.org/D13727#268233, @zturner wrote:
> In http://reviews.llvm.org/D13727#268086, @clayborg wrote:
>
> > > Yes. But an implementation of std::async is either going to use a global
> > > thread pool or it's not. If it does there's no problem, and
tberghammer added a comment.
I agree that parking the threads in the kernel should be very cheap (I am not
sure about Windows) but on high end multi core machines having 40+ threads
around just to wait for work is a bit strange for me too and it definitely
makes debugging of LLDB more difficult
tberghammer updated this revision to Diff 37498.
tberghammer added a comment.
Addressing comments from the discussion with destroying the treads not in use
while keeping a global thread pool with at most hardware_concurrency threads.
IMO this update also simplifies the implementation of the Thre
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
Looks good
http://reviews.llvm.org/D13777
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailm
tberghammer added a comment.
Zach: I thought about your idea and I think it won't make the code any simpler
because your suggestion would only implement the functionality what is
currently implemented by TaskPool and TaskPoolImpl what is using only a single
condition variable and a single queue
tberghammer updated this revision to Diff 37492.
tberghammer marked 7 inline comments as done.
tberghammer added a comment.
- Address review comments
- Add some more documentation
- Add some unit tests
http://reviews.llvm.org/D13727
Files:
include/lldb/Utility/TaskPool.h
source/Utility/CMak
tberghammer added a comment.
Please see http://reviews.llvm.org/D13727
http://reviews.llvm.org/D13662
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
tberghammer added a comment.
We can change this class to use std::async on Windows and an std::thread based
implementation on Linux with the same interface (other OS-es should be decided)
but I would prefer to use just 1 logic as it is easier to maintain.
Limiting the number of threads with add
tberghammer updated this revision to Diff 37353.
tberghammer added a comment.
Herald added a subscriber: iancottrell.
Use the new ThreadPool class and make the Append+Finalize stage parallel.
http://reviews.llvm.org/D13662
Files:
source/Plugins/SymbolFile/DWARF/NameToDIE.cpp
source/Plugins/
tberghammer created this revision.
tberghammer added reviewers: labath, clayborg, vharron, zturner.
tberghammer added a subscriber: lldb-commits.
Herald added a subscriber: iancottrell.
Add a new task pool class to LLDB to make it easy to execute tasks in parallel
Basic design goals:
* Have a ver
tberghammer accepted this revision.
tberghammer added a comment.
This revision is now accepted and ready to land.
Looks good
http://reviews.llvm.org/D13695
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailm
Author: tberghammer
Date: Wed Oct 14 05:38:22 2015
New Revision: 250289
URL: http://llvm.org/viewvc/llvm-project?rev=250289&view=rev
Log:
Change ConstString to support massive multi-threaded access
Previously ConstString had a single mutex guarding the global string
pool for each access what beco
This revision was automatically updated to reflect the committed changes.
Closed by commit rL250289: Change ConstString to support massive multi-threaded
access (authored by tberghammer).
Changed prior to commit:
http://reviews.llvm.org/D13652?vs=37258&id=37328#toc
Repository:
rL LLVM
http:
Author: tberghammer
Date: Wed Oct 14 05:29:17 2015
New Revision: 250286
URL: http://llvm.org/viewvc/llvm-project?rev=250286&view=rev
Log:
Make use of lldv::Triple::isAndroid
It is a new function added to the llvm::Triple class to simplify the
checking if we are targeting android to clean up the c
Author: tberghammer
Date: Wed Oct 14 05:29:36 2015
New Revision: 250287
URL: http://llvm.org/viewvc/llvm-project?rev=250287&view=rev
Log:
Fix warning in CommandObjectMemory
Modified:
lldb/trunk/source/Commands/CommandObjectMemory.cpp
Modified: lldb/trunk/source/Commands/CommandObjectMemory.c
tberghammer added inline comments.
Comment at: source/Core/ArchSpec.cpp:859-870
@@ -853,1 +858,14 @@
+
+// If this and other are both arm ArchSpecs and this ArchSpec is a generic
"some kind of arm"
+// spec but the other ArchSpec is a specific arm core, adopt the specific
401 - 500 of 725 matches
Mail list logo
