I'm not a programmer. I'm a hardware designer that hacks at programming. So my questions are often naive and without enough detail. So bare with me, please
I would like to look at the code that defines cli(); and soi(); in the attached example for the this code is in the Inferred Library IRremote.
Where is the underlying code?
in the library
For an Uno anyway, it's in interrupt.h.
It does depend on the compiler, which differs for different processors.
sei(); cli(); are macros that compile to the single instructions SEI CLI in AVR code.
Thanks for your help,
Like gfvalvo said, it's in avr/interrupt.h (for all AVRs, anyway.)
If you use interrupts() and noInterrupts(), your code should be more portable to other architectures, and produce the same code...
/** \name Global manipulation of the interrupt flag
The global interrupt flag is maintained in the I bit of the status
register (SREG).
Handling interrupts frequently requires attention regarding atomic
access to objects that could be altered by code running within an
interrupt context, see <util/atomic.h>.
Frequently, interrupts are being disabled for periods of time in
order to perform certain operations without being disturbed; see
\ref optim_code_reorder for things to be taken into account with
respect to compiler optimizations.
*/
#if defined(__DOXYGEN__)
/** \def sei()
\ingroup avr_interrupts
Enables interrupts by setting the global interrupt mask. This function
actually compiles into a single line of assembly, so there is no function
call overhead. However, the macro also implies a <i>memory barrier</i>
which can cause additional loss of optimization.
In order to implement atomic access to multi-byte objects,
consider using the macros from <util/atomic.h>, rather than
implementing them manually with cli() and sei().
*/
#define sei()
#else /* !DOXYGEN */
# define sei() __asm__ __volatile__ ("sei" ::: "memory")
#endif /* DOXYGEN */
#if defined(__DOXYGEN__)
/** \def cli()
\ingroup avr_interrupts
Disables all interrupts by clearing the global interrupt mask. This function
actually compiles into a single line of assembly, so there is no function
call overhead. However, the macro also implies a <i>memory barrier</i>
which can cause additional loss of optimization.
In order to implement atomic access to multi-byte objects,
consider using the macros from <util/atomic.h>, rather than
implementing them manually with cli() and sei().
*/
#define cli()
#else /* !DOXYGEN */
# define cli() __asm__ __volatile__ ("cli" ::: "memory")
#endif /* DOXYGEN */