| Age | Commit message (Collapse) | Author |
|---|
|
Usually, sub binaries are implemented as special sub binary terms
("Ys" lines). But if a sub binary is sent to another process, sub
binaries will be eliminated and incorporated into the ProcBin ("Yc"
line), and when that happens there can be several "Yc" lines that
refer to the same underlying binary object. Make sure that that the
crash dump viewer can handle multiple "Yc" lines referring to the same
binary object.
|
|
```erlang
Position = integer()
```
Parameter doesn't exist in `read/3,4`
|
|
Introudce erts_queue_release_literals() to queue a literal area to be
released.
|
|
|
|
* hans/ssh/convert_to_proper/OTP-15312:
ssh: Enable property_test/ssh_eqc_client_server for PropEr
ssh: Used fixed localhost address
ssh: Fix reporting functions in property_test/ssh_eqc_client_server.erl
ssh: Fix property_test/ssh_eqc_client_info_timing.erl
|
|
Eliminate an ets:select per process, which was used for reading
position data needed for progress reporting. Now this data is stored
in dictionary along with the last tag and id instead.
|
|
PropEr does not support the grouped statem-tests. This commits
calls the grouped functions from the functions supported by PropEr.
Optimized with memoization.
Previously only EQC was supported, but the changes to support PROPER is not
just a wrapper. Since I don't have access to eqc I can't test the changes
and therefore eqc is disabeled.
However, with access to eqc it ought to be quite easy to re-enable eqc by
studying the diff.
|
|
Problem on Darwin, Solaris and FreeBSD with local addresses in 127.1.0.0/8
|
|
|
|
Had problems on Darwin, Solaris and FreeBSD.
|
|
* maint-18:
Updated OTP version
Prepare release
|
|
Fix run_erl.c so it compiles on Solaris
|
|
Fix the default implementation of address_please
|
|
* jimdigriz/os_mon/fix_cpu_sup_android/OTP-15387:
Make Erlang's cpu_sup function better on Android
SELinux is another cause of MSG_CTRUNC
|
|
The removal set now uses a red-black tree instead of an array on
large inputs, decreasing runtime complexity from `n*n` to
`n*log(n)`. It will also exit early when there are no more items
left in the removal set, drastically improving performance and
memory use when the items to be removed are present near the head
of the list.
This got a lot more complicated than before as the overhead of
always using a red-black tree was unacceptable when either of the
inputs were small, but this compromise has okay-to-decent
performance regardless of input size.
Co-authored-by: Dmytro Lytovchenko <[email protected]>
|
|
Without this patch reltool would try to split strings like "C:\foo"
into ["C","\foo"].
|
|
Caught with unit test in ssl_pem_cache_SUITE.
When local timezone is PST (Pacific Standard Time), PEM cache was not evicting
expired entries due to file time converstion was done using
calendar:now_to_datetime, while file modification time is actually in local time.
Use os:system_time() to align with file_info modified time.
|
|
|
|
|
|
maint-18
* sverker/erts/18/memory-leak-terminating-port/OTP-14609:
erts: Fix memory leak when sending to terminating port
|
|
This greatly increases the performance of '--'/2 which does a lot
of term comparisons.
|
|
memory typo
|
|
|
|
|
|
The first stage wasn't bounded by reductions, and it bumped far
more reductions than it should have due to a logic bug.
|
|
It returns `ok`, instead of `noreply`.
|
|
In the implementation of the zero-copying term storage, we
want to preserve sharing, but not copy literals because the
modules holding the literals could be unloaded under our feet.
|
|
* bjorn/observer/fix-crashdump_viewer/ERL-722/OTP-15365:
Eliminate crash in crashdump_viewer reading some literal maps
|
|
Honor the max heap size when copying literals after purging
|
|
beam_lib: Clarify the beam() type in code and documentation
OTP-15378
|
|
* hans/crypto/cuddle_tests:
crypto: Fix tests failing if more than one test engine available This could happen in an Valgrind run:
crypto: Disable engine tests on Darwin and Windows Does not work.
|
|
This could happen in an Valgrind run:
=== Reason: no case clause matching ["otp_test_engine.so",
"otp_test_engine.valgrind.so"]
in function crypto:check_otp_test_engine/1 (crypto.erl, line 2026)
in call from engine_SUITE:engine_load_all_methods/1 (engine_SUITE.erl, line 154)
|
|
Does not work.
|
|
* hans/ssh/document_user/OTP-15314:
ssh: Re-phrase and adjust the documentation (ssh_file.xml)
ssh: Move some option's documentation to ssh_file user_dir user_dir_fun (missing previously) *_passphrase system_dir
ssh: Clearify a couple of options user_dir, system_dir and *_passphrase are only used in the default callback module ssh_file
ssh: Links updated in ssh.xml
ssh: Add reference manual page for the ssh_file module
ssh: Add new User's Guide chapter about SSH terminology The term "user" means different things in OpenSSH and in Erlang/SSH. This new chapter explains why.
|
|
|
|
user_dir
user_dir_fun (missing previously)
*_passphrase
system_dir
|
|
user_dir, system_dir and *_passphrase are only used in the default callback module ssh_file
|
|
|
|
This callback module has the knowledge about the different files
used by Erlang/OTP SSH.
It was unfortunatly not documented previously.
|
|
The term "user" means different things in OpenSSH and in Erlang/SSH.
This new chapter explains why.
|
|
* rickard/development_runtime_dependency/OTP-15341:
Fixup development runtime dependencies
|
|
The script 'make/fixup_development_runtime_dependencies' is run at
the end of a build of development branches in order to fixup future
not yet resolved versions (<app name>-@<ticket>(:<ticket>)+@) in
'runtime_dependencies'.
|
|
|
|
* ingela/public_key/DSA-SHA2/OTP-15367:
public_key: Add DSA with SHA2 support
|
|
* ingela/public_key/DSA-SHA2/OTP-15367:
public_key: Add DSA with SHA2 support
|
|
|
|
Before this change, it was always the aux thread that was woken
to handle thread progress events scheduled to happen when all
schedulers were going to sleep. This was ok in the pre-OTP-21
implementation when the aux thread just slept on a tse. Now that
it sleeps in the fallback pollset this uses too much cpu so
instead we wake the thread that is doing the request if it is
a managed thread, or else we wake scheduler 1.
|
|
The poll thread does a lot of waking up and then going
back to sleep. A large part of the waking up is managing
thread progress and a large part of that was using thread
specific data to get the thread progress data pointer.
With this refactor the tpd is passed to each of the functions
which greatly decreases the number of ethr_get_tsd calls
which in turn halves the CPU usage of the poller thread in
certain scenarios.
|
|
* ingela/ssl/use-correct-len-aead:
ssl: Correct length to be used in additional data in CHACHA20_POLY1305 cipher
|
|
|
