I'm not sure about the need for this one. There's nothing ostensibly wrong with it, ...
13 years, 5 months ago
(2011-08-09 22:56:40 UTC)
#1
I'm not sure about the need for this one. There's nothing ostensibly wrong with
it, but you're refactoring out one line; I think it's preferable just to leave
the diagnostic expressions where they are unless there's some other reason you
want to factor them out.
On 2011/08/09 22:56:40, Sean Hunt wrote: > I'm not sure about the need for this ...
13 years, 5 months ago
(2011-08-11 03:29:46 UTC)
#2
On 2011/08/09 22:56:40, Sean Hunt wrote:
> I'm not sure about the need for this one. There's nothing ostensibly wrong
with
> it, but you're refactoring out one line; I think it's preferable just to leave
> the diagnostic expressions where they are unless there's some other reason you
> want to factor them out.
The main point is to reduce the number of places that a diagnostic is emitted
from, in case we need to change the diagnostic later.
On Wed, Aug 10, 2011 at 20:29, <rtrieu@google.com> wrote: > The main point is to ...
13 years, 5 months ago
(2011-08-11 03:32:12 UTC)
#3
On Wed, Aug 10, 2011 at 20:29, <rtrieu@google.com> wrote:
> The main point is to reduce the number of places that a diagnostic is
> emitted from, in case we need to change the diagnostic later.
I suppose that makes sense. I'd double-check with Doug on these though.
Sean
Issue 4856050: Refactor CheckAddressOfOperand()
(Closed)
Created 13 years, 5 months ago by rtrieu
Modified 13 years, 5 months ago
Reviewers: Sean Hunt
Base URL: https://llvm.org/svn/llvm-project/cfe/trunk/
Comments: 0