[lldb-dev] DNBConfig and debugserver

[Bruce Mitchener via lldb-dev]

Just in case the below came out with a resolution of "Let's remove it", I
submitted this for review:

    http://reviews.llvm.org/D11971

 - Bruce

On Wed, Aug 12, 2015 at 11:21 AM, Bruce Mitchener <bruce.mitchener at gmail.com
> wrote:

> Hello!
>
> This is probably for the Apple People since it relates to debugserver...
>
> I was going to fix an issue with the cmake build on Mac OS X related to
> the mig generation step and DNBConfig.h, but when I started digging into
> things, I noticed that DNBConfig.h isn't needed any longer.
>
> The only test that it performs is for whether or not 64 bit Mach
> exceptions can be used.
>
> The code that actually handled non-64 bit Mach exceptions was removed in
> 2011 by this commit:
>
> commit 0e8147bd867e4cdaae9400f56d02c7aacd40a9b3
> Author: Greg Clayton <gclayton at apple.com>
> Date:   Fri Oct 28 22:59:14 2011 +0000
>
>     Enabled the "printf" attribute on all debugserver logging functions
> and fixed
>     the ensuing mayhem.
>
>
>
>     git-svn-id: https://llvm.org/svn/llvm-project/llvdb/trunk@143244
> 91177308-0d34-0410-b5e6-96231b3b80d8
>
>
> So, I'm wondering what the right next step is:
>
>
>    - Just fix the issue that I was going to fix in the cmake build system.
>    - Remove the unused test for 64 bit Mach exception handling and that
>    usage of DNBConfig.h
>    - Remove DNBConfig.h entirely along with the machinery for generating
>    it since nothing actually needs or uses it.
>    - Something else.
>
> I'm happy to do any of the above options.
>
> Cheers,
>
>  - Bruce
>
> PS. HavAVX.s is missing an EOL at EOF, and so it generates a warning. Can
> I just commit an EOL at EOF without going through Phabricator? That won't
> break anything, right?
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/lldb-dev/attachments/20150812/0ceec605/attachment.html>