Re: [PATCH 3/8] riscv: init: merge split string literals in preprocessor directive

[Luc Van Oostenryck]

On Thu, Oct 17, 2019 at 09:38:18PM -0700, Paul Walmsley wrote:
> On Fri, 18 Oct 2019, Luc Van Oostenryck wrote:
> 
> > On Thu, Oct 17, 2019 at 05:49:24PM -0700, Paul Walmsley wrote:
> > > sparse complains loudly when string literals associated with
> > > preprocessor directives are split into multiple, separately quoted
> > > strings across different lines:
> > 
> > ...
> >  
> > >  #ifndef __riscv_cmodel_medany
> > > -#error "setup_vm() is called from head.S before relocate so it should "
> > > -	"not use absolute addressing."
> > > +#error "setup_vm() is called from head.S before relocate so it should not use absolute addressing."
> > >  #endif
> > 
> > Using a blacslash should do the trick :
> > 	#error "blablablablablablablablablablablabla" \
> > 			"and blablabla again"
> > Or if need I cn fix Sparse if needed and desiable.
> 
> Thanks for the kind offer!
> 
> The backslashless syntax is pretty horrible to my eyes.  As far as I can 
> tell from a brief glance, the instance fixed by this patch was the only 
> instance of its kind in the kernel.  The existing kernel precedents appear 
> to be to simply use a single long line.  Example:
> 
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/include/linux/compiler-gcc.h#n3
> 
> So, from a kernel point of view, we should just fix this specific 
> instance.  It doesn't seem worth changing sparse for such a rare case.
> 
> On the other hand, gcc seems to support the non-backslashed syntax.  So if 
> the intention is for sparse to follow the gcc practice, and to be used 
> beyond the kernel, maybe it's worth aligning sparse to gcc?  Only if 
> you're bored, I suppose...

I'll first see what is acceptable for the point of view of the standard
(and maybe some justifications from GCC).

-- Luc