[PATCH] part of implimenting a monotonic clock in rtems part of this is not the final patch.

Christian Mauderer oss at c-mauderer.de
Tue Jun 15 07:29:47 UTC 2021


If you add a new functionallity you _should_ add a test that tests the 
expected behaviour. You just shouldn't replace one but add a new test or 
a new test case to an existing test.

Best regards

Christian

On 14/06/2021 22:18, zack_on_the_speed_chanel wrote:
>   also i'll revert the posix test that I changed back to normal because it was only for testing.
> 
> Sent with ProtonMail Secure Email.
> 
> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
> 
> On Monday, June 14th, 2021 at 8:14 PM, zack_on_the_speed_chanel <zack_on_the_speed_chanel at protonmail.ch> wrote:
> 
>> I've made most of the corrections to the code. I fixed up the formatting but I still don't know if I have to add anything for the settime and delete_timer(). i assume as the monotonic clock only affects the value I think it's all I have to do. I also can try to modify a posix test to check my theory.
>>
>> Thanks
>>
>> Zack
>>
>> Sent with ProtonMail Secure Email.
>>
>> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
>>
>> On Saturday, June 12th, 2021 at 9:31 AM, Christian Mauderer oss at c-mauderer.de wrote:
>>
>>> Hello Zack,
>>>
>>> I don't really know a lot about the timer toppic. So this is more of a
>>>
>>> style and general suggestion review.
>>>
>>> On 09/06/2021 20:27, zack wrote:
>>>
>>>> From: zack zack_on_the_speed_chanel at protonmail.ch
>>>>
>>>> cpukit/include/rtems/posix/timer.h | 6 ++-
>>>>
>>>> cpukit/posix/src/psxtimercreate.c | 5 +-
>>>>
>>>> cpukit/posix/src/timergettime.c | 61 ++++++++++++++++++++---
>>>>
>>>> testsuites/psxtests/psxtimer02/psxtimer.c | 26 +++++++---
>>>>
>>>> 4 files changed, 81 insertions(+), 17 deletions(-)
>>>>
>>>> diff --git a/cpukit/include/rtems/posix/timer.h b/cpukit/include/rtems/posix/timer.h
>>>>
>>>> index bcbf07a65a..f8cf6115b2 100644
>>>>
>>>> --- a/cpukit/include/rtems/posix/timer.h
>>>>
>>>> +++ b/cpukit/include/rtems/posix/timer.h
>>>>
>>>> @@ -21,7 +21,7 @@
>>>>
>>>> #include <rtems/score/objectdata.h>
>>>>
>>>> #include <rtems/score/watchdog.h>
>>>> ---------------------------------
>>>
>>> I think the blank line separated rtems includes from others. So please
>>>
>>> don't remove it.
>>>
>>>> +#include <time.h>
>>>>
>>>> #include <pthread.h>
>>>>
>>>> #ifdef __cplusplus
>>>>
>>>> @@ -47,7 +47,9 @@ typedef struct {
>>>>
>>>> struct itimerspec timer_data; /* Timing data of the timer /
>>>>
>>>> uint32_t ticks; / Number of ticks of the initialization /
>>>>
>>>> uint32_t overrun; / Number of expirations of the timer */
>>>>
>>>> -   struct timespec time; /* Time at which the timer was started */
>>>>
>>>> -   struct timespec time; /* Time at which the timer was started */
>>>>
>>>
>>> Not sure what changed in that line. Looks like a whitespace change.
>>>
>>> Please avoid these.
>>>
>>>> -   clockid_t clock_type;
>>>
>>> Why a blank line?
>>>
>>>> } POSIX_Timer_Control;
>>>>
>>>> /**
>>>>
>>>> diff --git a/cpukit/posix/src/psxtimercreate.c b/cpukit/posix/src/psxtimercreate.c
>>>>
>>>> index a63cf1d100..e78c359bd5 100644
>>>>
>>>> --- a/cpukit/posix/src/psxtimercreate.c
>>>>
>>>> +++ b/cpukit/posix/src/psxtimercreate.c
>>>>
>>>> @@ -40,7 +40,7 @@ int timer_create(
>>>>
>>>> {
>>>>
>>>> POSIX_Timer_Control *ptimer;
>>>>
>>>> -   if ( clock_id != CLOCK_REALTIME )
>>>>
>>>> -   if ( clock_id != CLOCK_REALTIME || clock_id != CLOCK_MONOTONIC )
>>>>
>>>
>>> Not sure about that one. It's allways a bit difficult whith these
>>>
>>> constructions but I think this one is allways true and therefore will
>>>
>>> allways return an error.
>>>
>>> Let's assume clock_id is CLOCK_MONOTONIC. In that case (clock_id !=
>>>
>>> CLOCK_REALTIME) is true which means that the function will return with
>>>
>>> an error. The same works for CLOCK_RREALTIME and the second condition.
>>>
>>> If you apply De Morgan to the term you see that
>>>
>>> (clock_id != CLOCK_REALTIME || clock_id != CLOCK_MONOTONIC)
>>>
>>> is the same as
>>>
>>> !(clock_id == CLOCK_REALTIME && clock_id == CLOCK_MONOTONIC)
>>>
>>> The two inner comparisons are exclusive. So that won't work.
>>>
>>> What you most likely want is a
>>>
>>> !(clock_id == CLOCK_REALTIME || clock_id == CLOCK_MONOTONIC)
>>>
>>> or with De Morgan again:
>>>
>>> (clock_id != CLOCK_REALTIME && clock_id != CLOCK_MONOTONIC)
>>>
>>>>        rtems_set_errno_and_return_minus_one( EINVAL );
>>>>
>>>>      if ( !timerid )
>>>>
>>>>
>>>> @@ -91,7 +91,8 @@ int timer_create(
>>>>
>>>> ptimer->timer_data.it_value.tv_nsec = 0;
>>>>
>>>> ptimer->timer_data.it_interval.tv_sec = 0;
>>>>
>>>> ptimer->timer_data.it_interval.tv_nsec = 0;
>>>> -------------------------------------------
>>>>
>>>> -   ptimer->clock_type=clock_id;
>>>>
>>>> -   _Watchdog_Preinitialize( &ptimer->Timer, _Per_CPU_Get_snapshot() );
>>>>
>>>>      _Watchdog_Initialize( &ptimer->Timer, _POSIX_Timer_TSR );
>>>>
>>>>      _Objects_Open_u32(&_POSIX_Timer_Information, &ptimer->Object, 0);
>>>>
>>>>      diff --git a/cpukit/posix/src/timergettime.c b/cpukit/posix/src/timergettime.c
>>>>
>>>>      index ee2a566f0e..62011cde58 100644
>>>>
>>>>      --- a/cpukit/posix/src/timergettime.c
>>>>
>>>>      +++ b/cpukit/posix/src/timergettime.c
>>>>
>>>>      @@ -28,6 +28,7 @@
>>>>
>>>>      #include <rtems/score/todimpl.h>
>>>>
>>>>      #include <rtems/score/watchdogimpl.h>
>>>>
>>>>      #include <rtems/seterr.h>
>>>>
>>>>      +#include <rtems/timespec.h>
>>>>
>>>>      /*
>>>>
>>>>      -              - When a timer is initialized, the value of the time in
>>>>
>>>>
>>>>
>>>> @@ -43,21 +44,67 @@ int timer_gettime(
>>>>
>>>> {
>>>>
>>>> POSIX_Timer_Control *ptimer;
>>>>
>>>> ISR_lock_Context lock_context;
>>>>
>>>> -   uint64_t now;
>>>>
>>>>      uint32_t remaining;
>>>>
>>>> -         Per_CPU_Control *cpu;
>>>>
>>>>
>>>> -         struct  timespec * now; // get time now either with
>>>>
>>>>
>>>> -         struct  timespec * expire; // expire
>>>>
>>>>
>>>> -         struct  timespec * result;// get remaining time
>>>>
>>>>
>>>>      if ( !value )
>>>>
>>>> -   rtems_set_errno_and_return_minus_one( EINVAL );
>>>>
>>>> -   rtems_set_errno_and_return_minus_one( EINVAL );
>>>>
>>>>      ptimer = _POSIX_Timer_Get( timerid, &lock_context );
>>>>
>>>> -   if ( ptimer != NULL ) {
>>>>
>>>> -   Per_CPU_Control *cpu;
>>>>
>>>> -   if(ptimer== NULL ){
>>>>
>>>
>>> Style: There should be a space before the ==
>>>
>>>> -   rtems_set_errno_and_return_minus_one( EINVAL );
>>>> -   }
>>>>
>>>> +if ( ptimer->clock_type ==CLOCK_REALTIME) {
>>>
>>> Style again: Space after ==
>>>
>>> Why that many blank lines?
>>>
>>>> -   cpu = _POSIX_Timer_Acquire_critical( ptimer, &lock_context );
>>>>
>>>> -              _TOD_Get(now); // get current time
>>>>
>>>>
>>>> -            rtems_timespec_from_ticks (ptimer->Timer.expire,expire ); // get the time to expire
>>>>
>>>>
>>>
>>> That are C++ style comments. We normally have C-Style comments in RTEMS
>>>
>>> only. Beneath that the comments exceed the line length. And some of them
>>>
>>> are quite superflous. A _TOD_Get(now) doesn't need any comment. The line
>>>
>>> is already telling that.
>>>
>>> Again: Blank lines. I won't mention the further blank lines below!
>>>
>>>> -   if (now->tv_nsec+now->tv_sec > expire->tv_nsec+expire->tv_sec) { // check if the time expired
>>>>
>>>> -            rtems_timespec_subtract (now ,expire , result); //
>>>>
>>>>
>>>
>>> Empty comment at the end of the line?
>>>
>>>> -          remaining = (uint32_t) result->tv_nsec+result->tv_sec;
>>>>
>>>>
>>>> -   } else {
>>>>
>>>> -          remaining = 0;
>>>>
>>>>
>>>> -   }
>>>>
>>>> -   _Timespec_From_ticks( remaining, &value->it_value );
>>>>
>>>> -   value->it_interval = ptimer->timer_data.it_interval;
>>>>
>>>> -   _POSIX_Timer_Release( cpu, &lock_context );
>>>>
>>>> -   return 0;
>>>>
>>>> -   }
>>>>
>>>>
>>>> +if ( ptimer->clock_type ==CLOCK_MONOTONIC) {
>>>
>>> Odd indentation and missing blank before the closing bracket.
>>>
>>> Indentation and some blanks on brackets seems to be a problem on quite
>>>
>>> some more lines too.
>>>
>>>> -         cpu = _POSIX_Timer_Acquire_critical( ptimer, &lock_context );
>>>>
>>>>
>>>> -   now = cpu->Watchdog.ticks;
>>>>
>>>> -           _Timecounter_Nanouptime(now );
>>>>
>>>>
>>>> -            rtems_timespec_from_ticks (ptimer->Timer.expire,expire );
>>>>
>>>>
>>>> -   if (now->tv_nsec+now->tv_sec > expire->tv_nsec+expire->tv_sec) {
>>>>
>>>> -            rtems_timespec_subtract (now, expire, result);
>>>>
>>>>
>>>> -   if ( now < ptimer->Timer.expire ) {
>>>>
>>>> -          remaining = (uint32_t) ( ptimer->Timer.expire - now );
>>>>
>>>>
>>>> -          remaining = (uint32_t) result->tv_nsec+result->tv_sec;
>>>>            } else {
>>>>              remaining = 0;
>>>>            }
>>>>
>>>>
>>>>
>>>> diff --git a/testsuites/psxtests/psxtimer02/psxtimer.c b/testsuites/psxtests/psxtimer02/psxtimer.c
>>>>
>>>> index 9f79d33c42..029e638c76 100644
>>>>
>>>> --- a/testsuites/psxtests/psxtimer02/psxtimer.c
>>>>
>>>> +++ b/testsuites/psxtests/psxtimer02/psxtimer.c
>>>>
>>>> @@ -59,17 +59,31 @@ void *POSIX_Init (
>>>>
>>>> fatal_posix_service_status_errno( status, EINVAL, "bad clock id" );
>>>>
>>>>      puts( "timer_create - bad timer id pointer - EINVAL" );
>>>>
>>>>
>>>> -   status = timer_create( CLOCK_REALTIME, &event, NULL );
>>>>
>>>> -   status = timer_create( CLOCK_MONOTONIC, &event, NULL );
>>>>
>>>>      fatal_posix_service_status_errno( status, EINVAL, "bad timer id" );
>>>>
>>>
>>> It seems that you replace the test for REALTIME timers. Are they no
>>>
>>> longer supported with your changes? I'm not sure whether that is a good
>>>
>>> idea because it would break all existing applications that use REALTIME
>>>
>>> timer.
>>>
>>>>      puts( "timer_create - OK" );
>>>>
>>>>
>>>> -   status = timer_create( CLOCK_REALTIME, NULL, &timer );
>>>>
>>>> -   status = timer_create( CLOCK_MONOTONIC, NULL, &timer );
>>>>
>>>>      posix_service_failed( status, "timer_create OK" );
>>>>
>>>>      puts( "timer_create - too many - EAGAIN" );
>>>>
>>>> -   status = timer_create( CLOCK_REALTIME, NULL, &timer1 );
>>>>
>>>> -   status = timer_create( CLOCK_MONOTONIC, NULL, &timer1 );
>>>>
>>>>      fatal_posix_service_status_errno( status, EAGAIN, "too many" );
>>>>
>>>> -   puts( "timer_create (monotonic) - bad timer id pointer - EINVAL" );
>>>>
>>>> -   status = timer_create( CLOCK_MONOTONIC, &event, NULL );
>>>>
>>>> -   fatal_posix_service_status_errno( status, EINVAL, "bad timer id" );
>>>>
>>>> -   puts( "timer_create (Monotonic) - OK" );
>>>>
>>>> -   status = timer_create( CLOCK_MONOTONIC, NULL, &timer );
>>>>
>>>> -   posix_service_failed( status, "timer_create OK" );
>>>>
>>>> -   puts( "timer_create (monotonic)- too many - EAGAIN" );
>>>>
>>>> -   status = timer_create( CLOCK_MONOTONIC, NULL, &timer1 );
>>>>
>>>> -   fatal_posix_service_status_errno( status, EAGAIN, "too many" );
>>>>
>>>> -   puts( "timer_delete - bad id - EINVAL" );
>>>>
>>>>      status = timer_delete( timer1 + 1 );
>>>>
>>>>      fatal_posix_service_status_errno( status, EINVAL, "bad id" );
>>>>
>>>>      @@ -100,14 +114,14 @@ void *POSIX_Init (
>>>>
>>>>      status = timer_settime( timer, TIMER_ABSTIME, &itimer, NULL );
>>>>
>>>>      fatal_posix_service_status_errno( status, EINVAL, "bad itimer value #2" );
>>>>
>>>> -   clock_gettime( CLOCK_REALTIME, &now );
>>>>
>>>> -   clock_gettime( CLOCK_MONOTONIC, &now );
>>>>
>>>>      itimer.it_value = now;
>>>>
>>>>      itimer.it_value.tv_sec = itimer.it_value.tv_sec - 1;
>>>>
>>>>      puts( "timer_settime - bad itimer value - previous time - EINVAL" );
>>>>
>>>>      status = timer_settime( timer, TIMER_ABSTIME, &itimer, NULL );
>>>>
>>>>      fatal_posix_service_status_errno( status, EINVAL, "bad itimer value #3" );
>>>>
>>>> -   clock_gettime( CLOCK_REALTIME, &now );
>>>>
>>>> -   clock_gettime( CLOCK_MONOTONIC, &now );
>>>>
>>>>      itimer.it_value = now;
>>>>
>>>>      itimer.it_value.tv_sec = itimer.it_value.tv_sec + 1;
>>>>
>>>>      puts( "timer_settime - bad id - EINVAL" );
>>>>
>>>>      @@ -129,4 +143,4 @@ void *POSIX_Init (
>>>>
>>>>      TEST_END();
>>>>
>>>>      rtems_test_exit (0);
>>>>
>>>>      -}
>>>>
>>>>      +}
>>>>
>>>>      \ No newline at end of file
>>>>
>>>
>>> devel mailing list
>>>
>>> devel at rtems.org
>>>
>>> http://lists.rtems.org/mailman/listinfo/devel
> 


More information about the devel mailing list