You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@nuttx.apache.org by GitBox <gi...@apache.org> on 2022/04/11 12:14:41 UTC
[GitHub] [incubator-nuttx] anjiahao1 opened a new issue, #6037: thread and task behavior should be consistent?
anjiahao1 opened a new issue, #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037
```c
#include <pthread.h>
#include <stdio.h>
#include <errno.h>
static void *a_thread_func()
{
printf("123456\n");
pthread_exit(0);
return NULL;
}
int main(void)
{
pthread_t new_th;
pthread_attr_t new_attr;
int ret_val;
/* Initialize attribute */
if (pthread_attr_init(&new_attr) != 0) {
printf("Cannot initialize attribute object\n");
return 1;
}
/* Set the attribute object to PTHREAD_CREATE_DETACHED. */
if (pthread_attr_setdetachstate(&new_attr, PTHREAD_CREATE_DETACHED) !=
0) {
printf("Error in pthread_attr_setdetachstate()\n");
return 1;
}
/* Create a new thread passing it the new attribute object */
if (pthread_create(&new_th, &new_attr, a_thread_func, NULL) != 0) {
printf("Error creating thread\n");
return 1;
}
/* If pthread_join() or pthread_detach fail, that means that the
* test fails as well. */
ret_val = pthread_join(new_th, NULL);
if (ret_val != EINVAL) {
printf("Test FAILED\n");
return 1;
}
printf("123\n");
if (ret_val != EINVAL) {
printf("Test FAILED\n");
return 1;
}
printf("Test PASSED\n");
pthread_exit(0);
}
```
it is a testcase,i try it use
`gcc hello_main.c -o test -lpthread` and then `./test` output is
```
123
Test PASSED
123456
```
Then i use it in nuttx `hello` , output is
```
123
Test PASSED
```
`123456` is miss,i debug `ptread_exit`code,`pthread_completejoin` filed,then call `exit` kill all threads.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] anjiahao1 commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
anjiahao1 commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1094992729
> > And in NuttX case both `main` and `pthread` have the same priorities? From the first look it look like `a_thread_func` didn't have a chance to be scheduled.
>
> This case cannot be used directly on nuttx, it will cause crash, this should be the problem of nuttx, I will patch the last time later
[pr6038](https://github.com/apache/incubator-nuttx/pull/6038) will fix it
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] anjiahao1 commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
anjiahao1 commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1094986574
> And in NuttX case both `main` and `pthread` have the same priorities? From the first look it look like `a_thread_func` didn't have a chance to be scheduled.
This case cannot be used directly on nuttx, it will cause crash, this should be the problem of nuttx, I will patch the last time later
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] anjiahao1 commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
anjiahao1 commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1094980789
> `gcc hello_main.c -o test -lpthread` -- is this tested on RT-Linux? What are the priorities of `main` vs `pthread`?
i use ubuntu 20.04
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] patacongo commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
patacongo commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1095584488
This might be correct behavior for NuttX. In NuttX, the scheduler and priority are determined by the parent task. So the pthread inherits SCHED_FIFO (probably) and same priority as the parent task. Since the parent never gave the child an opportunity to run, the output from the pthread never occurred.
Linux works differently. The scheduler will default to SCHED_OTHER and priority 5 (I think). That will be inherited by the pthread (unless PTHREAD_EXPLICIT is used). SCHED_OTHER is a fair scheduler, SCHED_FIFO is not and from my understanding of the FIFO scheduler, the behavior you show is correct.
What you are suggesting is that pthread_create() should yield. I don't think would be correct.
In order to do a correct comparison with Linux, you would need to set SCHED_FIFO and an appropriate priority. I would be interested in the result. I don't know what would happen!!!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] pkarashchenko commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
pkarashchenko commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1094979570
`gcc hello_main.c -o test -lpthread` -- is this tested on RT-Linux? What are the priorities of `main` vs `pthread`?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] pkarashchenko commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
pkarashchenko commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1094983635
And in NuttX case both `main` and `pthread` have the same priorities? From the first look it look like `a_thread_func` didn't have a chance to be scheduled.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] xiaoxiang781216 commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
xiaoxiang781216 commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1152128990
Close, already fixed by #6038
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] xiaoxiang781216 closed issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
xiaoxiang781216 closed issue #6037: thread and task behavior should be consistent?
URL: https://github.com/apache/incubator-nuttx/issues/6037
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [incubator-nuttx] patacongo commented on issue #6037: thread and task behavior should be consistent?
Posted by GitBox <gi...@apache.org>.
patacongo commented on issue #6037:
URL: https://github.com/apache/incubator-nuttx/issues/6037#issuecomment-1095599932
There are quite a few Stack Overflow discussions of this. Here are a couple:
1. https://stackoverflow.com/questions/12536649/pthread-run-a-thread-right-after-its-creation
2. https://stackoverflow.com/questions/15454019/main-thread-does-not-continue-execution-after-pthread-create
3. https://stackoverflow.com/questions/13700143/thread-does-nothing-after-successful-pthread-create
4. https://stackoverflow.com/questions/34859627/c-force-the-thread-to-start-immediately-after-pthread-create-pthread-yield-afte
And others. Some of those are more relevant than others.
The results are not consistent. I think that this is because people are not carefully managing the scheduler parameters. Certainly you need to investigate this further before committing any code changes.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org