Re: [PATCH 2/3] arm: Remove optpld macro

[Craig Howland <howland at LGSInnovations dot com>]

On 01/12/2017 09:56 AM, Corinna Vinschen wrote:
> On Jan 11 23:50, Pat Pannuto wrote:
> > LTO can re-order top-level assembly blocks, which can cause this
> > macro definition to appear after its use (or not at all), causing
> > compilation failures. As the macro has very few uses, simply removing
> > it by inlining is a simple fix.
> Am I the only one thinking this is a compiler bug, rather than correct
> behaviour?
>
> The macro definition is not an assembly block as such, but rather just
> that: a macro definition which is supposed to be evaluated at compile
> (or assemble) time.  To reorder it on the source level seems completely
> out of the scope of LTO.
>
> Am I missing something or should this be fixed in GCC?
>
>
> Corinna
I was asking the same question, so you are not the only one.  (I'm also confused 
because the patch is talking about link-time optimization, not compilation.  But 
Pat says it causes a compilation failure.  So then how is this related to LTO, 
even?)

In addition, I have another question/issue.  Sure, it is easy to modify the 
Newlib code, itself, to avoid this macro, but how can we know that there is not 
any user code which uses it that would now fail when it has been removed?  This 
is a higher-level include file that users might have used; it is not a 
lower-level, newlib-internal file, it does not seem.  That is, is it OK for 
compatibility reasons to just remove it (especially without a note as to why it 
is dangerous)?

Craig