sergiotarxz/nuttx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
32,412 Commits 1 Branch 0 Tags 291 MiB
341497484e
Commit Graph

1250 Commits

Author SHA1 Message Date
Gregory Nutt
7fbceacb46 Mark an untested driver as EXPERIMENTAL in its Kconfig; Update some comments. 2017-09-11 09:35:39 -06:00
Boris Astardzhiev
087a0cf2c7 waitpid(): Corrects two problems when CONFIG_SCHED_HAVE_PARENT+CONFIG_SCHED_CHILD_STATUS are enabled: (1) Was erroring out if the waited for task had already exited, and (2) was not freeing resources when a wait was completed. 2017-09-04 17:10:37 -06:00
Gregory Nutt
4993b0cb66 Work queue: In a recent change for a problem noted by Pascal Speck, it was noted (again by Pascal Speck) that the cancellation of existing work and replacement with new work must be atomic. Thanks, Pascal. 2017-08-31 07:43:47 -06:00
Gregory Nutt
92f44c5607 Networking: Move net/inet/net_monitor.c to net/tcp/tcp_monitor.c in preparation for design change to fix monitoring of duplicated sockets. 2017-08-29 08:40:13 -06:00
Gregory Nutt
bbf4d5048a work_queue() must cancel existing work prior to queuing new work, otherwise the work queue can become corrupted. Problem noted by Pascal Speck. 2017-08-28 07:46:48 -06:00
Arjun Hary
b274a97840 Miscellaneous fixes from astyle tool. 2017-08-14 17:19:27 -06:00
Gregory Nutt
6579078287 Eliminate another warning. 2017-08-12 11:24:11 -06:00
Gregory Nutt
baf4c0d011 Minor change to last PR: errno should not be modified if no error occurred. 2017-08-07 06:33:53 -06:00
Masayuki Ishikawa
4530a72379 sched/: Fix return value in sched_setaffinity() 2017-08-07 15:00:31 +09:00
Gregory Nutt
6b6d7ce9b9 sig_timedwait: Pending signal structure used after it has been releasd. From anonymous Bitbucket Issue 59. 2017-07-27 12:32:47 -06:00
Gregory Nutt
031fbfc6bc sched/: Remove dangling space at the end of lines. 2017-06-28 13:33:04 -06:00
Gregory Nutt
5e8037390c Fix problems in last commit noted by Jeongchan Kim: last change returned -EINVAL vs EINVAL; Treat the case where the mutex is already unlocked just like the case where the mutex is held by another thread -- return EPERM. 2017-06-22 06:26:00 -06:00
Gregory Nutt
0c3aef9591 Trivial typo fix 2017-06-21 06:48:13 -06:00
Gregory Nutt
6618336a39 pthread_mutex_unlock(): Missing check for the case where pthread_mutex_lock() is called when the mutex is not locked. In that case, it would increment the underlying semaphore avove 1. This is the fix for a problem noted by initialkjc@yahoo.com. 2017-06-21 06:40:09 -06:00
Gregory Nutt
2d4ba3ee3f IEEE 802.15.4 MAC: Add list management and prioritization logic that will permit the MAC layer to support bound multiple clients. 2017-06-18 11:00:47 -06:00
Gregory Nutt
2245dddaf9 Break up some long lines 2017-06-14 13:42:56 -06:00
Gregory Nutt
23a6ff5846 pthreads: Move pthread_barrier_init, pthread_barrier_destroy, and pthread_barrier_wait from sched/pthreads to libc/pthreads. This just coordinate other OS interface calls but are not a fundamental OS interfaces and, hence, do not belong within the OS. 2017-06-14 07:59:54 -06:00
Gregory Nutt
ae168bffa0 pthreads: Move pthread_cond_init and pthread_cond_destory from sched/pthreads to libc/pthreads. This just coordinate other OS interface calls but are not a fundamental OS interfaces and, hence, do not belong within the OS. 2017-06-14 07:48:45 -06:00
Gregory Nutt
557fd6504f pthreads: Move pthread_once from sched/pthreads to libc/pthreads. pthread_once just coordinates other OS interface calls but is not a fundamental OS interface and, hence, does not belong within the OS. 2017-06-14 07:31:10 -06:00
Gregory Nutt
8ae5450268 pthreads: Move pthread_yield from sched/pthreads to libc/pthreads. it is a simple wrapper for sched_yield and does not belong within the OS. 2017-06-14 07:31:10 -06:00
Gregory Nutt
0fe9c2f3f9 pthread mutex: Remove bogus DEBUGASSERT. Problem noted by Jussi Kivilinna 2017-06-01 06:28:23 -06:00
Gregory Nutt
90dda9357e pthread robust mutexes: Fix memmory trashing problem: the main task may also use mutexes; need to check thread type before accessing pthread-specific mutex data structures. Problem noted by Jussi Kivilinna. 2017-05-31 10:55:37 -06:00
Gregory Nutt
a8708424c2 pthread_trylock: Fixes a problem in pthread_trylock() noted by initialkjc@yahoo.com. When CONFIG_PTHREAD_MUTEX_UNSAFE=y, the special return value EAGAIN was not being detected due to differences in reporting of returned values. 2017-05-29 07:05:06 -06:00
Gregory Nutt
0de294a586 Fix lots of occurrences of 'the the', 'the there', 'the these', 'the then', 'the they. 2017-05-11 13:35:56 -06:00
Gregory Nutt
ba12817f9c Upate some comments 2017-05-11 13:15:31 -06:00
EunBong Song
03bd3688bc pthread: Fix compilation error on pthread_cond_wait when CONFIG_CANCELLATION_POINTS and CONFIG_PTHREAD_MUTEX_UNSAFE are enabled. 2017-05-02 07:14:04 -06:00
Jussi Kivilinna
e835803166 yslog: use monotonic clock for timestamp when available 2017-04-26 10:38:15 -06:00
Gregory Nutt
d5207efb5a Be consistent... Use Name: consistent in function headers vs Function: 2017-04-21 16:33:14 -06:00
Jussi Kivilinna
33ddaa0b10 clock: add testing for 32-bit overflow of 64-bit system timer 2017-04-21 08:55:07 -06:00
Jussi Kivilinna
c57d49f420 clock: Add new type ssystime_t for relative 64-bit ticks, change ticks<->time conversion functions to use ssystime_t 2017-04-21 08:51:31 -06:00
Jussi Kivilinna
325ba1a803 clock: add clock_resynchronize and use subseconds RTC 
Add clock_resynchronize for better synchronization of CLOCK_REALTIME and CLOCK_MONOTONIC to match RTC after resume from low-power state.

Add up_rtc_getdatetime_with_subseconds under CONFIG_ARCH_HAVE_RTC_SUBSECONDS to allow initializing (and resynchronizing) system clock with subseconds accuracy RTC.
 2017-04-21 08:45:57 -06:00
Nobutaka Toyoshima
150036eb8b sched: Fix tg_flags check with GROUP_FLAG_NOCLDWAIT 
Jira: PDFW15IS-208
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
 2017-04-14 14:50:44 +09:00
Juha Niskanen (Haltian)
b4747286b1 Add logic to disable cancellation points within the OS. This is useful when an internal OS function that is NOT a cancellation point calls an OS function which is a cancellation point. In that case, irrecoverable states may occur if the cancellation is within the OS. 2017-04-11 11:03:25 -06:00
Gregory Nutt
c08ba10d32 include/: Add some definitions needed by apps/wireless/wapi 2017-04-10 14:56:23 -06:00
Gregory Nutt
c36bf090f0 pthread: Minor logic fix in pthread_mutex_consistent. Updat some comments. 2017-04-10 10:10:41 -06:00
Gregory Nutt
948332ca34 pthreads: Backed most of last pthread changes. Found the 'real' root poblem. A one like error in pthread_mutex.c. 2017-04-10 09:51:03 -06:00
Gregory Nutt
b51b72b2db pthreads: Re-order some operations so that mutexes are placed in the inconsistent state BEFORE the clean-up callbacks are called. 2017-04-10 08:11:16 -06:00
Jussi Kivilinna
dffb8a67e3 Add entropy pool and strong random number generator 
Entropy pool gathers environmental noise from device drivers, user-space, etc., and returns good random numbers, suitable for cryptographic use. Based on entropy pool design from *BSDs and uses BLAKE2Xs algorithm for CSPRNG output.

Patch also adds /dev/urandom support for using entropy pool RNG and new 'getrandom' system call for getting randomness without file-descriptor usage (thus avoiding file-descriptor exhaustion attacks). The 'getrandom' interface is similar as 'getentropy' and 'getrandom' available on OpenBSD and Linux respectively.
 2017-03-30 07:38:37 -06:00
Gregory Nutt
eb344d7260 Fix an assertion noted by Jussi Kivilinna. 
This was a consequence of the recent robust mutex changes.  If robust mutexes are selected, then each mutex that a thread takes is retained in a list in threads TCB.  If the thread exits and that list is not empty, then we know that the thread exitted while holding mutexes.  And, in that case, each will be marked as inconsistent and the any waiter for the thread is awakened.

For the case of pthread_mutex_trywait(), the mutex was not being added to the list!  while not usually a fatal error, this was caught by an assertion when pthread_mutex_unlock() was called:  It tried to remove the mutex from the TCB list and it was not there when, of course, it shoule be.

The fix was to add pthread_mutex_trytake() which does sem_trywait() and if successful, does correctly add the mutext to the TCB list.  This should eliminated the assertion.
 2017-03-29 07:50:40 -06:00
Gregory Nutt
941360bf12 Correct some spacing 2017-03-27 10:49:41 -06:00
Gregory Nutt
b07d3fc305 Rename CONFIG_MUTEX_TYPES to CONFIG_PTHREAD_MUTEX_TYPES 2017-03-27 09:08:14 -06:00
Gregory Nutt
f2f798cb29 pthread mutexes: Finish logic to support configuration mutex robustness. 2017-03-27 08:50:45 -06:00
Gregory Nutt
666208cf23 pthread mutexes: Add option to support both unsafe and robust mutexes via pthread_mutexattr_get/setrobust(). 2017-03-26 18:37:24 -06:00
Gregory Nutt
8b3c554e45 pthreads: Add a configuration option to disable robust mutexes and revert to the traditional unsafe mutexes. 2017-03-26 17:37:28 -06:00
Gregory Nutt
86ab384d77 Forget to add some files in previous commits 2017-03-26 15:46:19 -06:00
Gregory Nutt
8b23c16b90 pthreads: pthread_mutex_consistent() needs to clear flags. 2017-03-26 14:44:57 -06:00
Gregory Nutt
34c5e1c18f Minor cleanup from recent changes. 2017-03-26 14:04:07 -06:00
Gregory Nutt
5a69453e16 pthreads: Add some assertions. 2017-03-26 13:54:43 -06:00
Gregory Nutt
fe03ef02c4 when pthread exits or is cancelled, mutexes held by thread are marked inconsistent and the highest priority thread waiting for the mutex is awakened. 2017-03-26 13:37:05 -06:00
Gregory Nutt
6e623ce06f pthreads: Partial implementation of final part of robust mutexes: Keep list of all mutexes held by a thread in a list in the TCB. 2017-03-26 12:46:57 -06:00
Gregory Nutt
2c37d369ab pthread: Fix return value of pthread_give/takesemaphore(). Add option to pthread_takesemaphore to ignore EINTR or not. 2017-03-26 11:22:17 -06:00
Gregory Nutt
363403fb1f pthreads: Add more robustness characteristics: pthread_mutex_lock() and trylock() will now return EOWNERDEAD if the mutex is locked by a thread that no longer exists. Add pthread_mutex_consistent() to recover from this situation. 2017-03-26 10:35:23 -06:00
Gregory Nutt
5fb85451cb Update some comments 2017-03-22 08:08:43 -06:00
Gregory Nutt
f5a957158d Review of last PR: Setting CONFIG_SMP_NCPUS=1 should only be permitted in a debug configuration. 2017-03-22 06:29:38 -06:00
Masayuki Ishikawa
a1f0802855 Kconfig: Change the minimum SMP_NCPUS to 1 2017-03-22 10:04:37 +09:00
Gregory Nutt
6893843cc5 sched/semaphore: Fix a warning aout an unused variable when priority inheritance is enabled. 2017-03-21 13:47:56 -06:00
Gregory Nutt
e8e3c2f362 sched/semaphore: Convert strange use of DEBUGASSERT to DEBUGPANIC. 2017-03-21 13:34:17 -06:00
David Sidrane
d76157db7b sem_holder:Clean up from Review 
Spelling and backward DEBUGASSERT along with one gem

  if (sem->holder[0].htcb != NULL || sem->holder[**1**].htcb != NULL)
 2017-03-21 08:03:06 -10:00
David Sidrane
7601a27cee sem_holder:The logic for the list version is unchanged 2017-03-16 14:16:18 -10:00
David Sidrane
3cc2a4f7c9 sem_holder: Fixes improper restoration of base_priority 
in the case of CONFIG_SEM_PREALLOCHOLDERS=0

   The call to sem_restorebaseprio_task context switches in the
   sem_foreachholder(sem, sem_restoreholderprioB, stcb); call
   prior to releasing the holder. So the running task is left
   as a holder as is the started task. Leaving both slots filled
   Thus failing to perforem the boost/or restoration on the
   correct tcb.

   This PR fixes this by releasing the running task slot prior
   to reprioritization that can lead to the context switch.
   To faclitate this, the interface to sem_restorebaseprio
   needed to take the tcb from the holder prior to the
   holder being freed. In the failure case where sched_verifytcb
   fails it added the overhead of looking up the holder.

   There is also the adfitinal thunking on the foreach to
   get from holer to holder->tcb.

   An alternate approach could be to leve the interface the same
   and allocate a holder on the stack of sem_restoreholderprioB
   copy the sem's holder to it, free it as is done in this pr
   and and then pass that address sem_restoreholderprio as the
   holder. It could then get the holder's tcb but we would
   keep the same sem_findholder in sched_verifytcb.
 2017-03-15 17:12:45 -10:00
David Sidrane
55fb9645a7 Guard from pend_reprios overlow 2017-03-15 07:42:55 -10:00
David Sidrane
caf8bac7fb missing semi 2017-03-13 22:54:13 +00:00
David Sidrane
d66fd9f965 semaphore:sem_boostholderprio prevent overrun of pend_reprios 
The second case rtcb->sched_priority <= htcb->sched_priority
  did not check if there is sufficient space in the pend_reprios
  array.
 2017-03-13 12:34:39 -10:00
David Sidrane
3c00651cfe semaphore:sem_holder sem_findholder missing inintalization of pholder 
sem_findholder would fail and code optimization coverd this up.
 2017-03-13 11:56:31 -10:00
David Sidrane
4d760c5ea4 semaphore:sem_holder add DEBUGASSERT s 2017-03-13 10:46:26 -10:00
Gregory Nutt
399f306744 A few cosmetic changes 2017-03-11 08:58:42 -06:00
David Sidrane
6cc8f9100b Priority Inversion fixes:typo 2017-03-10 06:37:46 -10:00
Gregory Nutt
360539afac Priority inheritance: When CONFIG_SEM_PREALLOCHOLDERS==0, there is only a single, hard-allocated holder structure. This is problem because in sem_wait() the holder is released, but needs to remain in the holder container until sem_restorebaseprio() is called. The call to sem_restorebaseprio() must be one of the last things the sem_wait() does because it can cause the task to be suspended. If in sem_wait(), a new task gets the semaphore count then it will fail to allocate the holder and will not participate in priority inheritance. This fix is to add two hard-allocated holders in the sem_t structure: One of the old holder and one for the new holder. 2017-03-10 09:30:15 -06:00
Gregory Nutt
a93e46d00c Cosmetic 2017-03-10 08:54:50 -06:00
Gregory Nutt
c976a66f8d net/drivers/skeleton.c: Back out serialization changes of the last commit. They are not necessary in the skeleton.c example because the calls to net_lock() at the beginning of each worker function will enforce serialization. 2017-03-04 11:33:36 -06:00
Gregory Nutt
b55a98e900 Olimex-STM32-P407: Enable task names 2017-03-04 08:49:24 -06:00
Gregory Nutt
47ebe1e320 Update some comments 2017-03-03 12:48:58 -06:00
Gregory Nutt
e1218c4b4b Smaller vector tables: Add irq_mapped_t. 2017-03-03 10:20:40 -06:00
Gregory Nutt
04c9ccdd2d Update TODO list and some Kconfig comments. 2017-03-03 09:20:25 -06:00
Gregory Nutt
c2b620b4f8 Implements support for smaller interrupt tables as described at http://www.nuttx.org/doku.php?id=wiki:howtos:smallvectors . This is largely the work of Mark Schulte. However, I have made several changes to match with the Wiki document. If you like the change, thanks go to Marc. For any errors you can blame me. 2017-03-03 09:20:25 -06:00
Masayuki Ishikawa
d78113e4ad Revert "CONFIG_START_YEAR/MONTH/DAY not required if an RTC is used" 
This reverts commit 4035ed8c6c.
 2017-03-03 10:42:10 +09:00
Masayuki Ishikawa
54b1f5ce2a Merged in masayuki2009/nuttx.nuttx/remove_unused_variable (pull request #233) 
Remove unused variable

Approved-by: Gregory Nutt
 2017-03-02 12:39:32 +00:00
Gregory Nutt
094795e0ed Review parameter usage in sigtimedwait(); update some comments. 2017-03-02 06:39:05 -06:00
Masayuki Ishikawa
237e041fba Remove an unused variable when calling sigwaitinfo() 2017-03-02 13:14:00 +09:00
Masayuki Ishikawa
4043239cfc Remove an unused variable when calling sigtimedwait() 2017-03-02 13:13:53 +09:00
Gregory Nutt
02b1e1ec1a Fixes for coding standard: '*' needs to 'snuggle' with following variable name 2017-02-28 18:22:57 -06:00
Gregory Nutt
840c5935a3 Correct a typo from one of the preceding commits. 2017-02-27 11:58:20 -06:00
Mark Schulte
b3222bbc8a irq_dispatch: Add argument pointer to irq_dispatch 
Provide a user defined callback context for irq's, such that when
registering a callback users can provide a pointer that will get
passed back when the isr is called.
 2017-02-27 06:27:56 -06:00
Gregory Nutt
4dbc0a27c6 pthread_create: g_pthreadname[] is not used if CONFIG_TASK_NAME_SIZE==0. 2017-02-26 18:15:16 -06:00
Gregory Nutt
12a4a58aa6 Update some wdog-related comments 2017-02-24 10:58:37 -06:00
Gregory Nutt
dca77fa06a sigtimedwait: When timer expires, up_unblock_task() is called. This is okay in the single CPU case because interrupts are disable in the timer interrupt handler. But it is insufficient in the SMP case. enter_ and leave_critical_section() must be called in order to manage spinlocks correctly. 2017-02-24 10:07:23 -06:00
Gregory Nutt
cb7c5f9921 Implement strings.h macros as inline functions when possible for better C++ compatibility. 2017-02-22 10:20:58 -06:00
Gregory Nutt
4f5879f998 C library: Add ffs(). Add strings.h. Move strcasecmp, strncasecmp, bzero, bcmp, and bcopy to where they belong in strings.h.h, not string.h. bzero, bcmp, and bcopy are legacy functions; the contemporary counterparts should be used instead. 2017-02-16 16:10:09 -06:00
Gregory Nutt
74189d84e2 Cancellation points: Fix some backward logic in conditional compilation. 2017-02-02 06:42:41 -06:00
Gregory Nutt
e0f199e42e Missing semicolon in previous commit 2017-01-31 14:18:20 -06:00
Gregory Nutt
154873a3d0 Timer logic: Add private function prototypes to eliminate a warning; Functions should not be inline because the may recurse. 2017-01-31 13:45:22 -06:00
Gregory Nutt
e30b2617e3 Shared Libs: Implement module based shared libraries for the PROTECTED mode build 2017-01-29 13:03:53 -06:00
Gregory Nutt
2c45f482b1 Module names are not needed in libc/modlib when the module library is used only for shared library support. 2017-01-29 12:23:24 -06:00
Gregory Nutt
86bdd0a4c7 Modules: Move last of many C files from sched/module to libc/modlib 2017-01-29 11:54:54 -06:00
Gregory Nutt
6da66fb5c8 Modules: Move next five of many C files from sched/module to libc/modlib 2017-01-29 11:17:29 -06:00
Gregory Nutt
5e94dd22bb Modules: Move first five of many C files from sched/module to libc/modlib 2017-01-29 10:05:15 -06:00
Gregory Nutt
00e46b5966 Move gnu-elf.ld from sched/module to libc/modlib. Update paths in Make.defs files to account for new location of file 2017-01-29 09:03:01 -06:00
Gregory Nutt
2717e781f0 libc/modlib: Add build a configuration logic for a shared module library. 2017-01-29 08:55:27 -06:00
Gregory Nutt
b79ca9dcdd Modules: Move sched/module/module.h to include/nuttx/lib/modlib.h 2017-01-29 08:24:42 -06:00
Gregory Nutt
68a7110195 sched/modules: Add support for dependencies between modules. 2017-01-27 11:43:27 -06:00
Gregory Nutt
6a4880f415 SMP: Fix timer related issues: Round robin and sporadic scheduling were only being performed for tasks running on the CPU that processes the system timer interrupt. Similary, CPU load measurements were only be processed for running on the CPU that receives the sampling interrupt. 2017-01-23 12:14:01 -06:00
Gregory Nutt
4574a1ca42 Modules: Extend the module interface so that we can access symbols exported by the module. 2017-01-22 14:26:22 -06:00
Gregory Nutt
d90e66a024 Kernel Modules: Module initializer may now return a symbol table (not yet used for anything) 2017-01-22 11:37:49 -06:00
Gregory Nutt
1823c78663 Update some comments 2017-01-22 10:17:51 -06:00
Gregory Nutt
686041f8e7 CPU load: Correct computation of the nominal period to use when the source is a oneshot timer. 2017-01-22 07:12:22 -06:00
rg
1e38884088 Fix a compile error: in sched_cpuload.c:Line136, the variables ts and secs are not defined if CONFIG_CPULOAD_ONESHOT_ENTROPY = 0. However, these variables are used regardless of CONFIG_CPULOAD_ONESHOT_ENTROPY at lines~180:onwards. 2017-01-22 06:48:11 -06:00
Gregory Nutt
edd9186540 ELF: Move ARMv7-A and ARMv7-R versions of ELF relocation logic to libc/machine 2017-01-21 14:40:26 -06:00
Gregory Nutt
fba247b119 Update some comments 2017-01-13 16:32:09 -06:00
Gregory Nutt
99bb2dda5d i.MX6: Corrects behavior of last SMP patch with i.MX6 2017-01-13 12:29:57 -06:00
Gregory Nutt
a51b5b7e17 Add REVISIT to comments. 2017-01-13 09:40:17 -06:00
Gregory Nutt
37e6e6a52e Eliminate a warning 2017-01-13 06:56:13 -06:00
Gregory Nutt
9ce4022096 SMP: Fix an error in critical section logic when performing a context switch from an interrupt handler. The g_cpu_irqset bit was not being set for the CPU so other CPUs did not know about the critical section. 2017-01-13 06:48:10 -06:00
Masayuki Ishikawa
acbfa47f70 sched_note: Fix spinlock instrumentation 2017-01-12 08:04:28 -06:00
Gregory Nutt
d5b1ca14e2 Update README and some comments. 2017-01-02 13:54:07 -06:00
Gregory Nutt
04557a1198 Eliminate a warning 2016-12-31 14:47:32 -06:00
Gregory Nutt
8f716a386b SMP: Use irq_cpu_locked() in sched_removereadytorun() and sched_setpriority() 2016-12-29 10:40:58 -06:00
Gregory Nutt
af92a67fc5 SMP: Use irq_cpu_locked() in sched_mergepending() 2016-12-29 08:53:31 -06:00
Gregory Nutt
e7d2b9f0e8 SMP: Move sharable function to common file as irq_cpu_locked(). Use irq_cpu_locked() in sched_unlock() 2016-12-29 08:17:10 -06:00
Gregory Nutt
96394f339e SMP: Fix a gap where we may try to make modifications to the task lists without being in a critical sections. That permits concurrent access to the tasks lists and many subtle problems. This fix just remains in the critical section throughout the operation (and possible until the task is restore in the event of a context switch). Makes a big difference in stability 2016-12-28 13:58:24 -06:00
Gregory Nutt
88da65ad8a sched_cpulocked: Avoid use of spinlock. That has been reported to cause a deadlock. 2016-12-28 12:10:17 -06:00
Gregory Nutt
72490f84e6 SMP: Make checks for CPU lock set more robust. There are certains conditions early in initialization on during interrupt handling where things need to be done a little differently. 2016-12-28 10:21:05 -06:00
Gregory Nutt
cfb876263a SMP: There were certain conditions that we must avoid by preventing releasing the pending tasks while withn a critical section. But this logic was incomplete; there was no logic to prevent other CPUs from adding new, running tasks while on CPU is in a critical section. 
This commit corrects this.  This is matching logic in sched_addreadytorun to avoid starting new tasks within the critical section (unless the CPU is the holder of the lock).  The holder of the IRQ lock must be permitted to do whatever it needs to do.
 2016-12-27 08:49:07 -06:00
Gregory Nutt
e0f3f4ae19 Update some coments 2016-12-26 16:12:27 -06:00
Gregory Nutt
849a5dc2a9 SMP: Enforce this rule: Tasks which are normally restored when sched_unlock() is called must remain pending (1) if we are in a critical section, i.e., g_cpu_irqlock is locked , or (2) other CPUs still have pre-emption disabled, i.e., g_cpu_schedlock is locked. In those cases, the release of the pending tasks must be deferred until those conditions are met. 2016-12-26 08:15:02 -06:00
Gregory Nutt
d45a81d643 Revert "SMP: Fix a error introduced in 36b676eea6 and fully decouple disabling of pre-emption from critical sections." 
This reverts commit 9aedf1bdec.
 2016-12-26 07:48:22 -06:00
Gregory Nutt
9aedf1bdec SMP: Fix a error introduced in 36b676eea6 and fully decouple disabling of pre-emption from critical sections. 2016-12-25 06:54:43 -06:00
Gregory Nutt
efb86382c3 SMP: Back out deferred IRQ locking. This was accidentally merged into master and it looks like it is going to be more work than I thought to get it working again. Changes will go to the irqlock branch. 2016-12-24 19:53:37 -06:00
Gregory Nutt
3af6b2a9a6 Update a comment 2016-12-24 19:01:07 -06:00
Gregory Nutt
f55bad863b SMP: Fix logic error in last change: Deferred restoration of IRQ lock only applies if the context switch was on this CPU. 2016-12-24 18:52:58 -06:00
Gregory Nutt
36b676eea6 SMP sched_unlock: Should merging depend on g_cpu_irqlock or not? Let's at least be consistent. And don't use irqcount; use g_cpu_irqlock. 2016-12-24 13:45:11 -06:00
Gregory Nutt
d75a7643d0 Fix a trivial typo 2016-12-23 13:01:20 -06:00
Gregory Nutt
e6fff09ef8 Implement deferred IRQ locking. So far only form ARMv7-M. 2016-12-23 07:55:41 -06:00
Gregory Nutt
d648f9c8b4 Add task_testcancel() 2016-12-10 16:34:14 -06:00
Gregory Nutt
f132960789 Add task_setcanceltype() 2016-12-10 16:06:14 -06:00
Gregory Nutt
5fb207eb36 Add task_setcancelstate() 2016-12-10 15:16:46 -06:00
Gregory Nutt
698597a838 task_delete() now obeys all cancellation point semantics. 2016-12-10 14:39:19 -06:00
Gregory Nutt
6997cda1b5 Grrr... cloned typos! 2016-12-10 09:45:55 -06:00
Gregory Nutt
842ec7e612 ifdef out some non-reachable code 2016-12-10 09:43:04 -06:00
Gregory Nutt
b52e4e5ecd Move cancellation point definitions to their own header file. 2016-12-10 09:08:26 -06:00
Gregory Nutt
bc3ca25cc7 Cancellation points: Close up some logic to eliminte some race conditions. 2016-12-10 08:36:58 -06:00
Gregory Nutt
e62b3bccd3 pthread_setcanceltype() and pthread_testcancel() do not have to be system calls if cancellation points are not enabled. 2016-12-10 07:40:48 -06:00
Gregory Nutt
a0f567f4a3 Update TODO, cosmetic changes, spelling -- US English spells it canceled and canceling vs cancelled and cancelling. No idea why. 2016-12-09 18:39:40 -06:00
Gregory Nutt
018eb7c1d2 Cancellation points no longer depend on EXPERIMENTAL 2016-12-09 17:44:37 -06:00
Gregory Nutt
a76e729b0d Trivial, cosmetic 2016-12-09 17:27:57 -06:00
Gregory Nutt
113d8bdcca Fix some SMP-related compilation errors 2016-12-09 17:10:59 -06:00
Gregory Nutt
78cdc9f113 Fix a typo in a debug assertion. 2016-12-09 16:59:10 -06:00
Gregory Nutt
7ee0862e67 Merge remote-tracking branch 'origin/master' into cancelpt 2016-12-09 16:54:15 -06:00
Gregory Nutt
b07964461e pthread_mutex_destroy(): Fix an error in destorynig a mutex which can occur after a pthread has been canceled while holding the mutex. 2016-12-09 16:53:29 -06:00
Gregory Nutt
7fce8022c6 Finishes all cancellation point logic 2016-12-09 16:50:34 -06:00
Gregory Nutt
d8783a7345 Update Documentation 2016-12-09 14:41:54 -06:00