It all works, now. I found that, for some reason, I ended up building debugserver on the BuildAndIntegration configuration, which wouldn't work, as expected.<div><br></div><div>Thanks,</div><div><br></div><div> Filipe</div>
<div class="gmail_extra"><br><br><div class="gmail_quote">On Fri, Nov 2, 2012 at 6:20 PM, Greg Clayton <span dir="ltr"><<a href="mailto:gclayton@apple.com" target="_blank">gclayton@apple.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im"><br>
On Nov 2, 2012, at 5:04 PM, Filipe Cabecinhas <<a href="mailto:filcab%2Blldb-dev@gmail.com">filcab+lldb-dev@gmail.com</a>> wrote:<br>
<br>
> Hmm, I have some problem here:<br>
><br>
> ± ./LLDB.framework/Resources/debugserver localhost:1234 /bin/ls<br>
> debugserver-193 for x86_64.<br>
> error: failed to launch process ./LLDB.framework/Resources/debugserver: failed to get the task for process 64342<br>
><br>
> The problem is: I just deleted the certificate, and followed all the instructions again. Can 10.8.2 have changed anything?<br>
<br>
</div>Probably not.<br>
<br>
- Delete all certs + public private key pairs.<br>
- Reboot.<br>
- remove ~/Desktop/lldb_codesign.cer, or you might get a ~/Desktop/lldb_codesign_1.cer when you drag your new lldb_codesign.cer to the desktop and cause the system to trust your old cert<br>
- nuke any and all LLDB build folders (the kernel caches code signing failures and needs to see the inode go away for the file)<br>
- follow instructions<br>
- reboot<br>
<br>
Working now??<br>
<div class="HOEnZb"><div class="h5"><br>
><br>
> Thanks,<br>
><br>
> Filipe<br>
><br>
><br>
><br>
> On Fri, Nov 2, 2012 at 4:58 PM, Greg Clayton <<a href="mailto:gclayton@apple.com">gclayton@apple.com</a>> wrote:<br>
> Just because it says it code signed ok, doesn't mean it will run.<br>
><br>
> debugserver and the XPC shlibs are the only things that need to be codesigned.<br>
><br>
> Try running debugserver directly:<br>
><br>
> ./LLDB.framework/Resources/debugserver localhost:1234 /bin/ls<br>
> debugserver-193 for x86_64.<br>
> Listening to port 1234...<br>
><br>
> If you see the "listening to port" message, it is working, else it isn't and it will complain.<br>
><br>
> On Nov 2, 2012, at 2:34 PM, Filipe Cabecinhas <<a href="mailto:filcab%2Blldb-dev@gmail.com">filcab+lldb-dev@gmail.com</a>> wrote:<br>
><br>
> > Okay, now it seems like something is not right.<br>
> > What should be signed? The lldb binary? debugserver?<br>
> ><br>
> > I have this on debugserver:<br>
> > ± codesign -vvv tools/debugserver/build/Debug/debugserver<br>
> > tools/debugserver/build/Debug/debugserver: valid on disk<br>
> > tools/debugserver/build/Debug/debugserver: satisfies its Designated Requirement<br>
> ><br>
> > The only other signed files are the XPC executables.<br>
> ><br>
> > Thanks,<br>
> ><br>
> > Filipe<br>
> ><br>
> ><br>
> > On Fri, Nov 2, 2012 at 2:12 PM, Filipe Cabecinhas <<a href="mailto:filcab%2Blldb-dev@gmail.com">filcab+lldb-dev@gmail.com</a>> wrote:<br>
> > But shouldn't Xcode say anything?<br>
> > I followed all the (weird) steps in the doc, but I will try again.<br>
> ><br>
> > Thanks,<br>
> ><br>
> > Filipe<br>
> ><br>
> ><br>
> > On Fri, Nov 2, 2012 at 2:07 PM, Greg Clayton <<a href="mailto:gclayton@apple.com">gclayton@apple.com</a>> wrote:<br>
> ><br>
> > On Nov 2, 2012, at 1:43 PM, "Malea, Daniel" <<a href="mailto:daniel.malea@intel.com">daniel.malea@intel.com</a>> wrote:<br>
> ><br>
> > > Hi Filipe,<br>
> > ><br>
> > > Yes, I am seeing the same behaviour on Mac OS X Mountain Lion (10.8.2) here, but not on Linux. It was not introduced recently, as we've been seeing the behaviour for the last month or two, and also with Lion.<br>
> > ><br>
> > > A workaround, though not a great one, is to invoke the tests with "sudo" then all the test inferior processes exit cleanly. I am not sure what the root-cause is.<br>
> > ><br>
> ><br>
> > It is usually your "lldb_codesign" code signing cert going bad, expired, or no longer valid due to OS upgrade/reinstall.<br>
> ><br>
> > > Dan<br>
> > ><br>
> > > On 2012-11-02, at 3:34 PM, Filipe Cabecinhas wrote:<br>
> > ><br>
> > >> Hi all,<br>
> > >><br>
> > >> I'm having some weird problems when testing lldb.<br>
> > >> Most tests don't stop the a.out file and leave it in an uninterruptible sleep.<br>
> > >> This will exhaust the machine's PID space and make it stop working.<br>
> > >><br>
> > >> Can you reproduce this?<br>
> > >><br>
> > >> I'm using lldb from trunk (updated a few minutes ago), and it happens the same way wether I compile it with the latest clang or clang from the latest Xcode.<br>
> > >><br>
> > >> I'm on Mountain Lion 10.8.2.<br>
> > >><br>
> > >> Regards,<br>
> > >><br>
> > >> Filipe<br>
> > >> _______________________________________________<br>
> > >> lldb-dev mailing list<br>
> > >> <a href="mailto:lldb-dev@cs.uiuc.edu">lldb-dev@cs.uiuc.edu</a><br>
> > >> <a href="http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev</a><br>
> > ><br>
> > ><br>
> > > _______________________________________________<br>
> > > lldb-dev mailing list<br>
> > > <a href="mailto:lldb-dev@cs.uiuc.edu">lldb-dev@cs.uiuc.edu</a><br>
> > > <a href="http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev</a><br>
> ><br>
> ><br>
> ><br>
><br>
><br>
<br>
</div></div></blockquote></div><br></div>