[lldb-dev] check-lldb will start using in-tree clang by default

2017-10-26 Thread Pavel Labath via lldb-dev
I am going to check in a change (D39215) which causes the check-lldb
target to use the just-built clang for compiling the test inferiors
(instead of the system compiler, which was the old default). The main
reason for this is to provide better reproducibility of test results
between different machines/developers, by removing one of the main
sources of nondeterminism. This behavior can be overridden by setting
the LLDB_TEST_C_COMPILER and LLDB_TEST_CXX_COMPILER cmake variables.

For the change to take effect you will need to clean your build folder
(or at least, remove the affected variables from your CMakeCache.txt).
After this you may observe a change in the test results from the
check-lldb run.

Note that this change only affect cmake code -- if you run your tests
by running dotest.py directly, nothing will change for you.

regards,
pavel
___
lldb-dev mailing list
lldb-dev@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev


Re: [lldb-dev] did anyone konw LLDB support lldb + openocd to run dotest.py in bare board like ARM or other non-x86 architecture?

2017-10-26 Thread cui bixiong via lldb-dev
Hi :

I build a armhf qemu enviroment to test ARM lldb + lldb-server
testsuite.  source code base on LLVM + LLDB + LLD + CLANG formal release
5.0.0.

Target: lldb-server platform --listen *: --server

Host :

(lldb) platform select remote-linux
Platform: remote-linux
Connected: no
(lldb) platform connect connect://0.0.0.0:
Platform: remote-linux
Triple: arm-*-linux-gnueabihf
OS Version: 3.2.0 (3.2.0-4-vexpress)
Kernel: #1 SMP Debian 3.2.51-1
Hostname: debian-armhf.""
Connected: yes
WorkingDir: /root
(lldb) file ./hello.exe
Current executable set to './hello.exe' (arm).
(lldb) log enable gdb-remote packets
(lldb) process launch
history[1] tid=0x6192 <   1> send packet: +
history[2] tid=0x6192 <  19> send packet: $QStartNoAckMode#b0
history[3] tid=0x6192 <   1> read packet: +
history[4] tid=0x6192 <   6> read packet: $OK#9a
history[5] tid=0x6192 <   1> send packet: +
history[6] tid=0x6192 <  13> send packet: $qHostInfo#9b
history[7] tid=0x6192 < 297> read packet:
$triple:61726d2d2d6c696e75782d676e75656162696866;ptrsize:4;distribution_id:64656269616e;watchpoint_exceptions_received:before;endian:little;os_version:3.2.0;os_build:332e322e302d342d7665787072657373;os_kernel:233120534d502044656269616e20332e322e35312d31;hostname:64656269616e2d61726d68662e;#1a
history[8] tid=0x6192 <  18> send packet: $qGetWorkingDir#91
history[9] tid=0x6192 <  14> read packet: $2f726f6f74#a4
history[10] tid=0x6192 <  19> send packet: $qQueryGDBServer#cb
history[11] tid=0x6192 <   7> read packet: $E04#a9
history[12] tid=0x6192 <  73> send packet:
$qModuleInfo:2e2f68656c6c6f2e657865;61726d2d2d6c696e75782d656162696866#b7
history[13] tid=0x6192 <   7> read packet: $E03#a8
history[14] tid=0x6192 <  69> send packet:
$qModuleInfo:6c6962632e736f2e36;61726d2d2d6c696e75782d656162696866#7b
history[15] tid=0x6192 <   7> read packet: $E03#a8
<  36> send packet: $qLaunchGDBServer;host:mtkslt202;#b1
error: unable to launch a GDB server on 'debian-armhf.""'

 i show host and port is uncorrect, is it a bug?


Best Regards
--cuibixiong

On Tue, Oct 24, 2017 at 11:04 PM, Greg Clayton  wrote:

>
> On Oct 24, 2017, at 12:02 AM, cui bixiong  wrote:
>
> Hi
>
> sorry, i'm confuse, in my mind, lldb-server maybe like gdb-server,
> running on Linux-like platform  listening RSP command which send form gdb
> and use ptrace syscall to debug which you want to debug program
>
> is it support remote download to batch mode run all testsuite?
> certainly i think reset baseboard is very important feature too,  but in
> lldb-server maybe not support currently i guess.
>
>
> Yes lldb-server does support a full connection to a remote OS. On
> baseboards you are typically debugging the entire OS so lldb-server won't
> work because you have to OS to run it in the background. Supporting the
> test suite on baseboards will take some effort.
>
>
> Best Regards
> --cuibixiong
>
> On Mon, Oct 23, 2017 at 11:06 PM, Greg Clayton  wrote:
>
>>
>> > On Oct 22, 2017, at 6:21 AM, cuibixiong via lldb-dev <
>> lldb-dev@lists.llvm.org> wrote:
>> >
>> > Hi
>> >
>> >   did anyone konw LLDB support lldb + openocd to run dotest.py in
>> bare board like ARM or other non-x86 architecture?
>>
>> We run the test suite on iOS devices using the platform stuff. This
>> requires a running lldb-server on the remote system, so we only have the
>> test suite running when we have the lldb-server running in platform mode on
>> the other side. For this to work with baseboards, we would need the JTAG
>> box to respond to many GDB remote protocol packets that "lldb-server
>> platform" implements. I am sure for baseboards the test suite would need to
>> be modified. A few ideas there:
>> - Have the test suite watch for a triple with no os (like
>> "arm64-none-none") and have it go into a baseboard mode
>> - Many tests that might rely on writing to files, reading from files for
>> stdin, and others, would need to be skipped in this mode
>> - Any tests that build and debug shared libraries would either need to be
>> modified to build multiple static binaries or skipped
>> - We might need to make a Bareboard platform and would load the ELF files
>> into memory instead of copying them over like the current platforms do
>>
>> So there would be quite a lot of modifications required to get the test
>> suite running. We will be happy to help you if you choose to try this.
>>
>> Greg Clayton
>>
>> >
>> > Best Regards
>> > —cuibixiong
>> > ___
>> > lldb-dev mailing list
>> > lldb-dev@lists.llvm.org
>> > http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev
>>
>>
>
>
___
lldb-dev mailing list
lldb-dev@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev