supression of 'matching constraint does not allow a register' warning

[Andrew Haley]

Peter Barada writes:
 > 
 > In the Linux-2.4.x kernel for m68k, the atomic operations use the
 > addql/subql instructions for incrementing/decrementing atomic
 > variables, and force the variables to memory to make the operation
 > truly atomic.  The definition from linux-2.4.x/include/asm-m68k/atomic.h:
 > 
 > static __inline__ void atomic_inc(volatile atomic_t *v)
 > {
 > 	__asm__ __volatile__("addql #1,%0" : "=m" (*v): "0" (*v));
 > }
 > 
 > static __inline__ void atomic_dec(volatile atomic_t *v)
 > {
 > 	__asm__ __volatile__("subql #1,%0" : "=m" (*v): "0" (*v));
 > }
 > 
 > Unfortunately when used they generate the following warning from
 > parse_input_contstraint in gcc/stmt.c:
 > 
 >   if (saw_match && !*allows_reg)
 >     warning ("matching constraint does not allow a register");
 > 
 > 
 > Since atomic_inc/atomic_dec are defined that way to enforce an
 > atomic operation to memory, when building the kernel, *lots* of
 > warnings are generated making it hard to wade through the log looking
 > for other warnings/errors.
 > 
 > I'd like to add a flag to suppress this warning, and was wondering if
 > a patch for this would be accepted, and what best to call the flag.

I don't think gcc supports what you want to do.  That's what the
warning means.

In glibc, these matching constraints were removed some time ago.

glibc changed to use

#define atomic_increment_and_test(mem) \

...
       __asm __volatile ("addq%.l %#1,%1; seq %0"                             \
                         : "=dm" (__result), "=m" (*(mem))                    \
                         : "m" (*(mem)));                                     \

Andrew.