Comment some parts of portmacro.h

This commit is contained in:
Angus Gratton 2015-05-07 14:14:37 +10:00
parent c636adf42c
commit 1e15cc266f

View file

@ -93,10 +93,8 @@ extern "C" {
#define portSTACK_TYPE unsigned portLONG
#define portBASE_TYPE long
typedef unsigned portLONG portTickType;
#define portMAX_DELAY (( portTickType ) UINT32_MAX)
/*-----------------------------------------------------------*/
typedef uint32_t portTickType;
#define portMAX_DELAY ( portTickType ) 0xffffffff
/* Architecture specifics. */
#define portSTACK_GROWTH ( -1 )
@ -126,13 +124,27 @@ void PendSV(enum SVC_ReqType);
/*-----------------------------------------------------------*/
/* NMIIrqIsOn flag is defined in libpp.a, and appears to be set when an NMI
(int level 3) is currently runnning (during which time libpp.a might
call back into parts of the OS?)
The esp_iot_rtos_sdk disables all interrupt manipulations while this
flag is set.
It's also referenced from some other blob libraries (not known if
read or written there).
ESPTODO: It may be possible to just read the 'ps' register instead
of accessing thisvariable.
*/
extern char NMIIrqIsOn;
extern char level1_int_disabled;
extern unsigned cpu_sr;
/* ESPTODO: Currently we store the old interrupt level (ps) in a global variable
cpu_sr. It may not be necessary to do this, but it depends on how the blob libraries
call into these functions.
/* ESPTODO: Currently we store the old interrupt level (ps) in a
global variable cpu_sr. It may not be necessary to do this,
especially as lx106 has only one real interrupt level + NMI, but it
all depends on how the blob libraries call into these functions.
*/
inline static __attribute__((always_inline)) void _esp_disable_interrupts(void)
{