Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Chaoren Lin via lldb-commits
Yeah, it fails to build if we don't have objc too. I'll upload an update. On Wed, Nov 4, 2015 at 6:12 PM, Jim Ingham wrote: > Yeah, sure. But I bet you could figure out something to put in > PlatformiOSSimulatorCoreSimulatorSupport.h that would cause this to fail to > link except on OS X. Not

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Jim Ingham via lldb-commits
Yeah, sure. But I bet you could figure out something to put in PlatformiOSSimulatorCoreSimulatorSupport.h that would cause this to fail to link except on OS X. Not that that's likely, but as a general strategy it seems wrong. Jim > On Nov 4, 2015, at 6:09 PM, Chaoren Lin wrote: > > There's

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Chaoren Lin via lldb-commits
There's no simulator support outside of Darwin, so none of it should get used anyway. On Wed, Nov 4, 2015 at 6:08 PM, Jim Ingham wrote: > In this case it's even a little weirder, 'cause we compile a source file > that contains a class declaration (since the #if defined is after the > #include),

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Jim Ingham via lldb-commits
In this case it's even a little weirder, 'cause we compile a source file that contains a class declaration (since the #if defined is after the #include), but not the methods that aren't inlined. Then presumably link in the .o file containing whatever was the result of that compilation. That se

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Chaoren Lin via lldb-commits
Yeah, I just looked at PlatformDarwinKernel.cpp as example for a quick fix. I agree that kind of logic probably belongs in the CMakeLists.txt. On Wed, Nov 4, 2015 at 6:02 PM, Jason Molenda wrote: > You're making a valid point. I was only noting that this is not the first > instance where've con

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Jason Molenda via lldb-commits
You're making a valid point. I was only noting that this is not the first instance where've conditionally compiled mac platform files. J > On Nov 4, 2015, at 6:00 PM, Jim Ingham wrote: > > Why? > > Jim > >> On Nov 4, 2015, at 5:43 PM, Jason Molenda wrote: >> >> We do the same thing in oth

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Jim Ingham via lldb-commits
Why? Jim > On Nov 4, 2015, at 5:43 PM, Jason Molenda wrote: > > We do the same thing in other MacOSX platform files where they depend on > Apple specific libraries. > > >> On Nov 4, 2015, at 5:39 PM, Jim Ingham via lldb-commits >> wrote: >> >> Why are we trying to build this at all on Lin

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Jason Molenda via lldb-commits
We do the same thing in other MacOSX platform files where they depend on Apple specific libraries. > On Nov 4, 2015, at 5:39 PM, Jim Ingham via lldb-commits > wrote: > > Why are we trying to build this at all on Linux? Seems odd to build an empty > file rather than just not build it. > > J

Re: [Lldb-commits] [lldb] r252124 - Fix build on Linux.

2015-11-04 Thread Jim Ingham via lldb-commits
Why are we trying to build this at all on Linux? Seems odd to build an empty file rather than just not build it. Jim > On Nov 4, 2015, at 5:23 PM, Chaoren Lin via lldb-commits > wrote: > > Author: chaoren > Date: Wed Nov 4 19:23:19 2015 > New Revision: 252124 > > URL: http://llvm.org/viewv