| Age | Commit message (Collapse) | Author |
|---|
|
Communication between Erlang processes has conceptually always been
performed through asynchronous signaling. The runtime system
implementation has however previously preformed most operation
synchronously. In a system with only one true thread of execution, this
is not problematic (often the opposite). In a system with multiple threads
of execution (as current runtime system implementation with SMP support)
it becomes problematic. This since it often involves locking of structures
when updating them which in turn cause resource contention. Utilizing
true asynchronous communication often avoids these resource contention
issues.
The case that triggered this change was contention on the link lock due
to frequent updates of the monitor trees during communication with a
frequently used server. The signal order delivery guarantees of the
language makes it hard to change the implementation of only some signals
to use true asynchronous signaling. Therefore the implementations
of (almost) all signals have been changed.
Currently the following signals have been implemented as true
asynchronous signals:
- Message signals
- Exit signals
- Monitor signals
- Demonitor signals
- Monitor triggered signals (DOWN, CHANGE, etc)
- Link signals
- Unlink signals
- Group leader signals
All of the above already defined as asynchronous signals in the
language. The implementation of messages signals was quite
asynchronous to begin with, but had quite strict delivery constraints
due to the ordering guarantees of signals between a pair of processes.
The previously used message queue partitioned into two halves has been
replaced by a more general signal queue partitioned into three parts
that service all kinds of signals. More details regarding the signal
queue can be found in comments in the erl_proc_sig_queue.h file.
The monitor and link implementations have also been completely replaced
in order to fit the new asynchronous signaling implementation as good
as possible. More details regarding the new monitor and link
implementations can be found in the erl_monitor_link.h file.
|
|
|
|
|
|
|
|
The retry loop wasn't working anyway as it tested fd<0
instead of res.
So, there is no real semantic change here.
|
|
|
|
to run beam with "rr record".
|
|
|
|
|
|
|
|
|
|
|
|
* lukas/erts/etp-processes-ports-optimization:
erts: Optimize etp-processes and etp-ports
|
|
Not calculating the max, table and invalid addresses
each loop speeds up iterating through the processes list
by about 5x.
|
|
OTP-14152
* rickard/ds-runqs:
Make statistics/1 aware of dirty run-queues and tasks
|
|
|
|
"./configure --enable-lock-counter" would modify the behavior
of the default emulator (usually SMP).
To make lock counting more accessible, change --enable-lock-counter
to build an additional emulator in the same way as:
(cd erts/emulator && make lcnt)
(The next commit will make it easier to start the built lock-counter
emulator.)
|
|
|
|
OTP-14356
* rickard/timer-improvements:
Fix of later timer wheel
Minimum timeout position in each timer wheel
Manage timers to trigger at once in a slot similar to other timers
Introduce timer slot range counters
Timer wheel divided into a "soon wheel" and a "later wheel"
Remove unnecessary cancel callback from timer-wheel timers
Rearrange timer struct fields in order to simplify
Use timer wheel for short BIF timers
Use magic refs for BIF timers
Remove accessor BIF timer implementation
Fix aux-work timer implementation
|
|
* bjorn/dialyzer/add-typer/OTP-14336:
Add smoke test for TypER
Add back TypEr to the main OTP repository
|
|
It turned out that the dependencies between Dialyzer
and TypEr makes it impractical to have TypEr in a
separate repository.
Add it back to the OTP repository, but put the Erlang
module 'typer' in the dialyzer application.
|
|
Minimum known timeout position is saved in bot far and near
wheel. This information is used to avoid scanning from current
position in the cases were we know the minimum timeout position.
|
|
|
|
The application now has an own repo, https://github.com/erlang/typer
|
|
|
|
Magic references are *intentionally* indistinguishable from ordinary
references for the Erlang software. Magic references do not change
the language, and are intended as a pure runtime internal optimization.
An ordinary reference is typically used as a key in some table. A
magic reference has a direct pointer to a reference counted magic
binary. This makes it possible to implement various things without
having to do lookups in a table, but instead access the data directly.
Besides very fast lookups this can also improve scalability by
removing a potentially contended table. A couple of examples of
planned future usage of magic references are ETS table identifiers,
and BIF timer identifiers.
Besides future optimizations using magic references it should also
be possible to replace the exposed magic binary cludge with magic
references. That is, magic binaries that are exposed as empty
binaries to the Erlang software.
|
|
|
|
|
|
* maint:
Update copyright-year
Conflicts:
lib/dialyzer/src/dialyzer.hrl
lib/dialyzer/src/dialyzer_options.erl
lib/dialyzer/test/opaque_SUITE_data/src/recrec/dialyzer.hrl
lib/dialyzer/test/opaque_SUITE_data/src/recrec/dialyzer_races.erl
lib/hipe/icode/hipe_icode.erl
lib/hipe/main/hipe.erl
lib/hipe/main/hipe.hrl.src
lib/hipe/main/hipe_main.erl
|
|
|
|
* maint:
Update etp-commands for dirty schedulers
Fix scheduling of system tasks on processes executing dirty
Fix call time tracing with dirty schedulers
Fix send of exit signal to process executing dirty
Fix dirty scheduler process priority
Fix alloc-util hard-debug
Silence debug warning when no beam jump table is used with dirty schedulers
Fix check_process_code() when NifExport is in use
Fix GC when NifExport is in use
Fix saving of original arguments when rescheduling via NifExport
Conflicts:
erts/emulator/beam/beam_bif_load.c
erts/emulator/beam/erl_nif.c
|
|
OTP-14051
* rickard/dirty-scheduling-fixes:
Update etp-commands for dirty schedulers
Fix scheduling of system tasks on processes executing dirty
Fix call time tracing with dirty schedulers
Fix send of exit signal to process executing dirty
Fix dirty scheduler process priority
Fix alloc-util hard-debug
Silence debug warning when no beam jump table is used with dirty schedulers
Conflicts:
erts/etc/unix/etp-commands.in
|
|
|
|
|
|
|
|
|
|
Specifically etp-stacktrace/stackdump/process-info have been changed
|
|
This commit adds two new structs to be used to represent
erlang code in erts.
ErtsCodeInfo is used to describe the i_func_info header
that is part of all Export entries and the prelude of
each function. This replaces all the BeamInstr * that
were previously used to point to these locations.
After this change the code should never use BeamInstr *
with offsets to figure out different parts of the
func_info header.
ErtsCodeMFA is a struct that is used to descripe a
MFA in code. It is used within ErtsCodeInfo and also
in Process->current.
All function that previously took Eterm * or BeamInstr *
to identify a MFA now use the ErtsCodeMFA or ErtsCodeInfo
where appropriate.
The code has been tested to work when adding a new field to the
ErtsCodeInfo struct, but some updates are needed in ops.tab to
make it work.
|
|
|
|
|
|
|
|
with run_erl option -sleepy-child
to provoke race when slave pty is late
|
|
Symptom: run_erl does exit(0) and child program (erl) does
not seem to start. Running run_erl again however brings the previous
child program to life.
Problem: A race causing run_erl to read 0 from master pty and exit
because the child has not yet opened its corresponding slave pty.
Solution: Use the non standard openpty() function instead that does
not expose this race as the slave fd is opened in the parent.
Question:
Is there a race free way to do this with posix_openpt on OpenBSD?
|
|
|
|
|
|
|
|
|
|
|
|
* henrik/update-copyrightyear:
update copyright-year
|
|
|
