Re: Proposal: immediate strings

[Ken Raeburn]

On May 24, 2012, at 01:17, Stefan Monnier wrote:
>> +#define SDATA(string)               (XSTRING (string)->u.imm.immbit ? \
>> +                             (XSTRING (string)->u.imm.data + 0) : \
>> +                             (XSTRING (string)->u.dat.data + 0))
> 
> IIUC the "+ 0" are unneeded here, because using a "..?..:.." already
> makes sure we have an rvalue.  Same for SCHARS.

FWIW, there used to be a GCC extension that caused the ternary operator to be 
an lvalue if both the second and third operands were lvalues.  I'm not sure 
when it was removed; possible in 4.0, which might be old enough that we can 
ignore it...

Ken