Re: gcc 4.1.1: char *p = "str" puts "str" into rodata

[Andrew Haley <aph at redhat dot com>]

Denis Vlasenko writes:
 > On Sunday 28 January 2007 17:15, Andrew Haley wrote:
 > > [ off-topic for gcc; redirected to gcc-help ]
 > > 
 > > Denis Vlasenko writes:
 > >  > char p;
 > >  > int main() {
 > >  >         p = "";
 > >  >         return 0;
 > >  > }
 > > 
 > > This is odd code: you're assigning a pointer to a string to a char
 > > variable.
 > 
 > Sorry, should be "char *p;"
 > > 
 > >  > Don't you think that "" should end up in rw data?
 > > 
 > > No: it's a literal string.  Many, many years ago, C compilers put
 > > literal strings into read/write memory and it was possible to alter
 > > them, but C89 outlawed that practice.
 > 
 > gcc doesn't warn me, it just produces buggy code.

Not at all.  This is normal C code, and the compiler is correct.  It's
a read-only string.

 > I remember that string literals are special - they decay to "const
 > char *" OR to "char*" depending on context. In this context, it
 > should decay to "char*", and it does - gcc doesn't complain
 > "assingment of const to non-const", the bug is that gcc placed
 > "str" in ro section.
 > 
 > I did get SEGV on this, in busybox project.

Well, yes.

Andrew.