[Lldb-commits] [PATCH] D26528: Fix uninitialized members.

[Zachary Turner via lldb-commits]

I think we can just add it to LLVM Coding Standards
<http://llvm.org/docs/CodingStandards.html> page.  I think part of why it's
not used more often is that people just don't know about it, so that would
help increase the visibility for everyone.

On Fri, Nov 11, 2016 at 10:51 AM Jim Ingham <jingham at apple.com> wrote:

> Is it worthwhile having some place where we can put this sort of policy so
> people don't have to follow our mailing list to know this sort of rule?
> The coding conventions were one possible site, but that fell victim to the
> code reformat...
>
> Jim
>
> > On Nov 11, 2016, at 9:11 AM, Zachary Turner <zturner at google.com> wrote:
> >
> > Yea, I agree doing it for a whole class at a time should be the standard
> > On Fri, Nov 11, 2016 at 9:10 AM Jim Ingham <jingham at apple.com> wrote:
> >
> > > On Nov 10, 2016, at 8:57 PM, Zachary Turner via lldb-commits <
> lldb-commits at lists.llvm.org> wrote:
> > >
> > > Maybe just inline the initializations so we don't have to repeat code
> across multiple constructors?  i.e.
> > >
> > > bool m_is_resolved = false;
> > >
> > > in the header file.
> >
> > I actually like the ability to do this in the header file, that seems
> clearer to me.  But if we're going to start doing this more widely I think
> it would be better to do it consistently  - maybe converting on a class by
> class basis when you touch one of the ivars in the class?  Having to look
> in two places for all these default initializations with only history as
> the pattern seems like it will make the code harder to read.
> >
> > Jim
> >
> >
> > >
> > > On Thu, Nov 10, 2016 at 8:54 PM Sam McCall via lldb-commits <
> lldb-commits at lists.llvm.org> wrote:
> > > sammccall created this revision.
> > > sammccall added a subscriber: lldb-commits.
> > >
> > > Fix uninitialized members.
> > >
> > >
> > > https://reviews.llvm.org/D26528
> > >
> > > Files:
> > >   source/Host/common/FileSpec.cpp
> > >   source/Target/Process.cpp
> > >
> > >
> > > Index: source/Target/Process.cpp
> > > ===================================================================
> > > --- source/Target/Process.cpp
> > > +++ source/Target/Process.cpp
> > > @@ -4581,7 +4581,7 @@
> > >        : IOHandler(process->GetTarget().GetDebugger(),
> > >                    IOHandler::Type::ProcessIO),
> > >          m_process(process), m_read_file(), m_write_file(write_fd,
> false),
> > > -        m_pipe() {
> > > +        m_pipe(), m_is_running(false) {
> > >      m_pipe.CreateNew(false);
> > >      m_read_file.SetDescriptor(GetInputFD(), false);
> > >    }
> > > Index: source/Host/common/FileSpec.cpp
> > > ===================================================================
> > > --- source/Host/common/FileSpec.cpp
> > > +++ source/Host/common/FileSpec.cpp
> > > @@ -278,8 +278,8 @@
> > >  }
> > >
> > >  FileSpec::FileSpec()
> > > -    : m_directory(), m_filename(),
> m_syntax(FileSystem::GetNativePathSyntax()) {
> > > -}
> > > +    : m_directory(), m_filename(), m_is_resolved(false),
> > > +      m_syntax(FileSystem::GetNativePathSyntax()) {}
> > >
> > >  //------------------------------------------------------------------
> > >  // Default constructor that can take an optional full path to a
> > >
> > >
> > > _______________________________________________
> > > lldb-commits mailing list
> > > lldb-commits at lists.llvm.org
> > > http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
> > > _______________________________________________
> > > lldb-commits mailing list
> > > lldb-commits at lists.llvm.org
> > > http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
> >
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/lldb-commits/attachments/20161111/20270898/attachment-0001.html>