diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2017-07-15 22:58:58 +0300 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2017-07-15 22:58:58 +0300 |
commit | 486088bc4689f826b80aa317b45ac9e42e8b25ee (patch) | |
tree | adf5847a6119d24da990d9e336f005c4a316e6be /Documentation/futex-requeue-pi.txt | |
parent | 52f6c588c77b76d548201470c2a28263a41b462b (diff) | |
parent | 43e5f7e1fa66531777c49791014c3124ea9208d8 (diff) | |
download | linux-486088bc4689f826b80aa317b45ac9e42e8b25ee.tar.xz |
Merge tag 'standardize-docs' of git://git.lwn.net/linux
Pull documentation format standardization from Jonathan Corbet:
"This series converts a number of top-level documents to the RST format
without incorporating them into the Sphinx tree. The hope is to bring
some uniformity to kernel documentation and, perhaps more importantly,
have our existing docs serve as an example of the desired formatting
for those that will be added later.
Mauro has gone through and fixed up a lot of top-level documentation
files to make them conform to the RST format, but without moving or
renaming them in any way. This will help when we incorporate the ones
we want to keep into the Sphinx doctree, but the real purpose is to
bring a bit of uniformity to our documentation and let the top-level
docs serve as examples for those writing new ones"
* tag 'standardize-docs' of git://git.lwn.net/linux: (84 commits)
docs: kprobes.txt: Fix whitespacing
tee.txt: standardize document format
cgroup-v2.txt: standardize document format
dell_rbu.txt: standardize document format
zorro.txt: standardize document format
xz.txt: standardize document format
xillybus.txt: standardize document format
vfio.txt: standardize document format
vfio-mediated-device.txt: standardize document format
unaligned-memory-access.txt: standardize document format
this_cpu_ops.txt: standardize document format
svga.txt: standardize document format
static-keys.txt: standardize document format
smsc_ece1099.txt: standardize document format
SM501.txt: standardize document format
siphash.txt: standardize document format
sgi-ioc4.txt: standardize document format
SAK.txt: standardize document format
rpmsg.txt: standardize document format
robust-futexes.txt: standardize document format
...
Diffstat (limited to 'Documentation/futex-requeue-pi.txt')
-rw-r--r-- | Documentation/futex-requeue-pi.txt | 93 |
1 files changed, 47 insertions, 46 deletions
diff --git a/Documentation/futex-requeue-pi.txt b/Documentation/futex-requeue-pi.txt index 77b36f59d16b..14ab5787b9a7 100644 --- a/Documentation/futex-requeue-pi.txt +++ b/Documentation/futex-requeue-pi.txt @@ -1,5 +1,6 @@ +================ Futex Requeue PI ----------------- +================ Requeueing of tasks from a non-PI futex to a PI futex requires special handling in order to ensure the underlying rt_mutex is never @@ -20,28 +21,28 @@ implementation would wake the highest-priority waiter, and leave the rest to the natural wakeup inherent in unlocking the mutex associated with the condvar. -Consider the simplified glibc calls: - -/* caller must lock mutex */ -pthread_cond_wait(cond, mutex) -{ - lock(cond->__data.__lock); - unlock(mutex); - do { - unlock(cond->__data.__lock); - futex_wait(cond->__data.__futex); - lock(cond->__data.__lock); - } while(...) - unlock(cond->__data.__lock); - lock(mutex); -} - -pthread_cond_broadcast(cond) -{ - lock(cond->__data.__lock); - unlock(cond->__data.__lock); - futex_requeue(cond->data.__futex, cond->mutex); -} +Consider the simplified glibc calls:: + + /* caller must lock mutex */ + pthread_cond_wait(cond, mutex) + { + lock(cond->__data.__lock); + unlock(mutex); + do { + unlock(cond->__data.__lock); + futex_wait(cond->__data.__futex); + lock(cond->__data.__lock); + } while(...) + unlock(cond->__data.__lock); + lock(mutex); + } + + pthread_cond_broadcast(cond) + { + lock(cond->__data.__lock); + unlock(cond->__data.__lock); + futex_requeue(cond->data.__futex, cond->mutex); + } Once pthread_cond_broadcast() requeues the tasks, the cond->mutex has waiters. Note that pthread_cond_wait() attempts to lock the @@ -53,29 +54,29 @@ In order to support PI-aware pthread_condvar's, the kernel needs to be able to requeue tasks to PI futexes. This support implies that upon a successful futex_wait system call, the caller would return to user space already holding the PI futex. The glibc implementation -would be modified as follows: - - -/* caller must lock mutex */ -pthread_cond_wait_pi(cond, mutex) -{ - lock(cond->__data.__lock); - unlock(mutex); - do { - unlock(cond->__data.__lock); - futex_wait_requeue_pi(cond->__data.__futex); - lock(cond->__data.__lock); - } while(...) - unlock(cond->__data.__lock); - /* the kernel acquired the mutex for us */ -} - -pthread_cond_broadcast_pi(cond) -{ - lock(cond->__data.__lock); - unlock(cond->__data.__lock); - futex_requeue_pi(cond->data.__futex, cond->mutex); -} +would be modified as follows:: + + + /* caller must lock mutex */ + pthread_cond_wait_pi(cond, mutex) + { + lock(cond->__data.__lock); + unlock(mutex); + do { + unlock(cond->__data.__lock); + futex_wait_requeue_pi(cond->__data.__futex); + lock(cond->__data.__lock); + } while(...) + unlock(cond->__data.__lock); + /* the kernel acquired the mutex for us */ + } + + pthread_cond_broadcast_pi(cond) + { + lock(cond->__data.__lock); + unlock(cond->__data.__lock); + futex_requeue_pi(cond->data.__futex, cond->mutex); + } The actual glibc implementation will likely test for PI and make the necessary changes inside the existing calls rather than creating new |