[lldb-dev] Almost all lldb-mi tests failing from timeout

[Greg Clayton]

On MacOSX you should be building and testing by doing:

cd lldb
xcodebuild -configuration Debug -target desktop
cd test
./dotest.py

Please don't test using a make/cmake/ninja builds as it won't test the LLDB.framework as it appears on MacOSX. The make/cmake/ninja stuff builds a flat lldb.so (no framework).

Greg

> On Apr 2, 2015, at 12:52 PM, Chuck Ries <Chuck.Ries at microsoft.com> wrote:
> 
> I am still having difficult running tests on macos.  I have tried running the tests manually both with my built lldb-mi and lldb
>  
> When running with lldb, I get the following issue:
>  
> Chucks-Mac-mini:lldb-mi chuckr$ <build_dir>/bin/lldb
> (lldb) target create ./a.out
> Current executable set to './a.out' (x86_64).
> (lldb) process launch
> error: process launch failed: unable to locate debugserver
> (lldb)
>  
> Is there some step in testing I’m missing to get my built debuserver on the path?
>  
> From: Ilia K [mailto:ki.stfu at gmail.com] 
> Sent: Thursday, March 26, 2015 12:58 AM
> To: Chuck Ries
> Cc: lldb-dev at cs.uiuc.edu
> Subject: Re: [lldb-dev] Almost all lldb-mi tests failing from timeout
>  
> Hello,
>  
> >RE: Seems that it's lldb regression:
> I fixed it in r233258 (it was my error).
>  
> At the moment all lldb-mi tests pass for me on OS X:
> ./dotest.py -v --executable $BUILDDIR/bin/lldb tools/lldb-mi/
> [...]
> Ran 68 tests in 204.082s
>  
> OK (skipped=1, expected failures=3)
>  
>  As you see I use the same command.
>  
> Could you provide test outputs (using the -t option)?
>  
> Thanks,
> Ilia
>  
> On Thu, Mar 26, 2015 at 10:05 AM, Ilia K <ki.stfu at gmail.com> wrote:
> Seems that it's lldb regression:
>   1 -file-exec-and-symbols a.out
>   2 (gdb)       
>   3 ^done
>   4 (gdb)
>   5 =shlibs-added,shlib-info=[num="1",name="a.out",dyld-addr="-",reason="dyld",path="/Users/IliaK/p/llvm/tools/lldb/test/tools/lldb-mi/a.out",loa    ded_addr="-",dsym-objpath="/Users/IliaK/p/llvm/tools/lldb/test/tools/lldb-mi/a.out.dSYM/Contents/Resources/DWARF/a.out"]
>   6 -break-insert -f main
>   7 ^done,bkpt={number="1",type="breakpoint",disp="keep",enabled="y",addr="0xffffffffffffffff",func="main",file="main.cpp",fullname="/Users/IliaK    /p/llvm/tools/lldb/test/tools/lldb-mi/main.cpp",line="15",pending=["main"],times="0",original-location="main"}
>   8 (gdb)
>   9 =breakpoint-modified,bkpt={number="1",type="breakpoint",disp="keep",enabled="y",addr="0xffffffffffffffff",func="main",file="main.cpp",fullnam    e="/Users/IliaK/p/llvm/tools/lldb/test/tools/lldb-mi/main.cpp",line="15",pending=["main"],times="0",original-location="main"}
>  10 -exec-run
>  11 Assertion failed: (!"bad_weak_ptr"), function shared_ptr, file /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain    /usr/bin/../include/c++/v1/memory, line 5174.
>  
> Thanks,
> Ilia
>  
>  
> On Thu, Mar 26, 2015 at 7:58 AM, Ilia K <ki.stfu at gmail.com> wrote:
> Hello Chunk,
> 
> I checked them a few days ago on OSX and all lldb-mi tests passed. It's weird that you see so many errors. Today I'll check it again. Thanks for report.
> 
> Thanks, 
> Ilia 
> 
> On Mar 26, 2015 1:39 AM, "Chuck Ries" <Chuck.Ries at microsoft.com> wrote:
> I tried to run the lldb-mi test on MacOS 57 of 68 error’d out with timeout issues. Is this the standard state of the tests or am I doing something wrong?
>  
> python dotest.py --executable ~/llama/build/main_debug/bin/lldb tools/lldb-mi/
>  
>  
> 
> _______________________________________________
> lldb-dev mailing list
> lldb-dev at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev
> 
>  
>  
> _______________________________________________
> lldb-dev mailing list
> lldb-dev at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev