Age | Commit message (Collapse) | Author |
|
Microstate accounting is a way to track which state the
different threads within ERTS are in. The main usage area
is to pin point performance bottlenecks by checking which
states the threads are in and then from there figuring out
why and where to optimize.
Since checking whether microstate accounting is on or off is
relatively expensive if done in a short loop only a few of the
states are enabled by default and more states can be enabled
through configure.
I've done some benchmarking and the overhead with it turned off
is not noticible and with it on it is a fraction of a percent.
If you enable the extra states, depending on the benchmark,
the ovehead when turned off is about 1% and when turned on
somewhere inbetween 5-15%.
OTP-12345
|
|
Conflicts:
OTP_VERSION
erts/doc/src/notes.xml
erts/vsn.mk
otp_versions.table
|
|
* sverk/poll-lost-wakeup/OTP-12859:
erts: Fix race in poller thread wake up
|
|
|
|
OTP-12845
* bruce/change-license:
fix errors caused by changed line numbers
Change license text to APLv2
|
|
|
|
|
|
to have one common implementation for both _kp and _nkp.
|
|
* egil/erts/high_accuracy_sleep/OTP-12236:
erts: Add high accuracy poll timeouts
|
|
|
|
Different poll/select implementations have different ways
to handle timeouts of < ms accuracy. Most have extended
API like pselect or such, while others rely on using
timerfds (epoll_wait). If no high accuracy timeout is
available, we simply round up to nearest ms. If we do not
roundup we will spin the last ms when waiting for a timeout
which is not desirable.
|
|
The old time API is based on erlang:now/0. The major issue with
erlang:now/0 is that it was intended to be used for so many
unrelated things. This tied these unrelated operations together
and unnecessarily caused performance, scalability as well as
accuracy, and precision issues for operations that do not need
to have such issues. The new API spreads different functionality
over multiple functions in order to improve on this.
The new API consists of a number of new BIFs:
- erlang:convert_time_unit/3
- erlang:monotonic_time/0
- erlang:monotonic_time/1
- erlang:system_time/0
- erlang:system_time/1
- erlang:time_offset/0
- erlang:time_offset/1
- erlang:timestamp/0
- erlang:unique_integer/0
- erlang:unique_integer/1
- os:system_time/0
- os:system_time/1
and a number of extensions of existing BIFs:
- erlang:monitor(time_offset, clock_service)
- erlang:system_flag(time_offset, finalize)
- erlang:system_info(os_monotonic_time_source)
- erlang:system_info(time_offset)
- erlang:system_info(time_warp_mode)
- erlang:system_info(time_correction)
- erlang:system_info(start_time)
See the "Time and Time Correction in Erlang" chapter of the
ERTS User's Guide for more information.
|
|
|
|
The warning is:
sys/common/erl_poll.c:2513:72: warning: for loop has empty body [-Wempty-body]
for (prev_ps = pollsets; ps != prev_ps->next; prev_ps = prev_ps->next);
|
|
|
|
|
|
Make sure the fd-sets (read and write) are of equal size.
|
|
An attempt to beautify the code by reducing the amount of #ifdef's
in the hot code path.
|
|
|
|
This allows erl_poll to select() on file descriptors greater than
FD_SETSIZE (1024) on Darwin. The fd sets are manually (re)allocated and
freed.
|
|
Conflicts:
lib/diameter/autoconf/vxworks/sed.general
xcomp/README.md
|
|
|
|
|
|
|
|
|
|
I/O events could potentially be delayed for ever when enabling
kernel-poll on a non-SMP runtime system executing on Solaris. When
also combined with async-threads the runtime system hung before
completing the boot phase. This bug was introduced in
erts-5.9/OTP-R15B.
|
|
|
|
A number of memory allocation optimizations have been implemented. Most
optimizations reduce contention caused by synchronization between
threads during allocation and deallocation of memory. Most notably:
* Synchronization of memory management in scheduler specific allocator
instances has been rewritten to use lock-free synchronization.
* Synchronization of memory management in scheduler specific
pre-allocators has been rewritten to use lock-free synchronization.
* The 'mseg_alloc' memory segment allocator now use scheduler specific
instances instead of one instance. Apart from reducing contention
this also ensures that memory allocators always create memory
segments on the local NUMA node on a NUMA system.
|
|
Conflicts:
erts/aclocal.m4
erts/emulator/beam/erl_db.c
erts/emulator/sys/win32/sys.c
erts/include/internal/ethread_header_config.h.in
|
|
|
|
All uses of the old deprecated atomic API in the runtime system
have been replaced with the use of the new atomic API. In a lot of
places this change imply a relaxation of memory barriers used.
|
|
|
|
|
|
|
|
|
|
A race condition in erts_poll() could cause
delay of poll for I/O.
|
|
Missing memory barriers in erts_poll() could cause the runtime system to
hang indefinitely.
|
|
erts_poll_info_kp() [defined in erts/emulator/sys/common/erl_poll.c
via some name-mangling trickery] contains a code path that can end
up in an infinite loop, causing a livelock. There is a block of code
inside #if ERTS_POLL_USE_UPDATE_REQUESTS_QUEUE that is supposed to
iterate over a linked list of ErtsPollSetUpdateRequestsBlocks and
update two variables based on the sizes of these blocks. The bug is
that the loop forgets to advance the list pointer to the next element,
so if the loop is entered at all (the initial list pointer is non-NULL),
the thread falls into an infinite loop.
This patch, against R13B03 but applies fine to today's git, fixes the
bug by adding a statement to advance the list pointer in the loop.
All other loops over this list appear to be correct.
Thanks to Chetan Ahuja for the original report of a livelock problem
in erts_poll_info_kp().
|
|
tile-cc 2.0.1.78377 when compiling the runtime system.
|
|
|