| Age | Commit message (Collapse) | Author |
|---|
|
|
|
* ingela/inets/http-benchmarks:
inets: Add comparable benchmark tests
inets: Create httpd benchmark framwork
|
|
|
|
|
|
|
|
|
|
maint-20
* hans/ssh/openssh_client_pubkey_sha2/ERL-531/OTP-14827:
ssh: Special treatment of OpenSSH clients >= 7.2 rsa-sha2-* public keys
|
|
Thoose clients signs with sha instead of sha2-*. Try first to verify with the correct one, and if that fails, retry with sha1.
|
|
|
|
* ingela/dtls/fallback-SCSV/OTP-14828:
ssl: Align (with DTLS) and correct TLS_FALLBACK_SCSV handling
|
|
|
|
Doing a proper transaction handling of user properties (in current
implmenentation) caused schema_transactions to be quadratic which
caused huge startup times for nodes with many tables since the merge schema
is a schema_transaction.
Do a direct read instead, cheap but implies that the plugin changes
will not be visible until the changes are commited and cannot be used
later in same transaction, which was possible before if undocumented
functions where used.
|
|
Add -MMD option to erlc
OTP-14830
|
|
digraph: Document a bad_edge error
|
|
There are four uncovered lines in combine_heap_needs/2 and
combine_alloc_lists/2. There is no way to reach starting from
Erlang source code using the standard compiler. However, they
can be reached starting from BEAM assembly code, so we don't
want to remove them.
We could add a test case that covers the lines using assembly
code, but an easier solution is to rewrite the code in a more
generic way using sofs so that the code can be covered with
existing test cases.
|
|
|
|
=== OTP-20.1.7.1 ===
Changed Applications:
- kernel-5.4.0.1
Unchanged Applications:
- asn1-5.0.3
- common_test-1.15.2
- compiler-7.1.3
- cosEvent-2.2.1
- cosEventDomain-1.2.1
- cosFileTransfer-1.2.1
- cosNotification-1.2.2
- cosProperty-1.2.2
- cosTime-1.2.2
- cosTransactions-1.3.2
- crypto-4.1
- debugger-4.2.3
- dialyzer-3.2.2
- diameter-2.1.2
- edoc-0.9.1
- eldap-1.2.2
- erl_docgen-0.7.1
- erl_interface-3.10
- erts-9.1.5
- et-1.6.1
- eunit-2.3.4
- hipe-3.16.1
- ic-4.4.2
- inets-6.4.4
- jinterface-1.8
- megaco-3.18.2
- mnesia-4.15.1
- observer-2.5
- odbc-2.12
- orber-3.8.3
- os_mon-2.4.3
- otp_mibs-1.1.1
- parsetools-2.1.5
- public_key-1.5.1
- reltool-0.7.5
- runtime_tools-1.12.2
- sasl-3.1
- snmp-5.2.8
- ssh-4.6.2
- ssl-8.2.2
- stdlib-3.4.2
- syntax_tools-2.1.3
- tools-2.11
- wx-1.8.2
- xmerl-1.3.15
Conflicts:
OTP_VERSION
lib/kernel/doc/src/notes.xml
lib/kernel/doc/src/os.xml
lib/kernel/src/os.erl
lib/kernel/vsn.mk
otp_versions.table
|
|
|
|
* bjorn/compiler/use-stacktrace-syntax:
Use the new syntax for retrieving stack traces
|
|
Slightly optimize reading of cooked files in list mode
|
|
01835845579e9 fixed some problems, but introduced a bug where
is_not_used/3 would report that a register was not used when it
in fact was.
|
|
|
|
|
|
* lukas/kernel/os_cmd_max_size/OTP-14823:
kernel: Add os:cmd/2 with max_size option
|
|
|
|
|
|
Symptom: "Closing pipe in state Waiting. Event lost?"
Snake oil:
Do erlang:yield() instead of busy spinning in "Waiting" state.
|
|
by limiting the memory usage.
Our valgrind test machine (pharazon) seems to get totally swamped
by this test case.
|
|
The Java 9 module system requires a module name to be defined. If none is provided, a default one is used derived from the JAR name, but that one is not very unique and can create problems. The recommended module name is the longest common package name for the source code included.
|
|
|
|
|
|
|
|
758712d6294 changed the need_heap/2 function so that it stopped
using its second argument.
Remove the second argument from need_heap(), and update all callers
to similarly remove unused arguments.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* lars/ssl/update-runtime-dependencies:
[ssl] Update runtime dependencies
|
|
* maint:
ssh: Update runtime dependencies of ssh
|
|
* hans/ssh/update_runtime_dep:
ssh: Update runtime dependencies of ssh
|
|
Add syntax in try/catch to retrieve the stacktrace directly
|
|
* bjorn/compiler/cover-v3_codegen:
v3_codegen: Simplify #k_guard_break{}
v3_codegen: Remove uncovered clause in bs_rename_ctx/4
Cover handling of #k_call{} in v3_codegen:bsm_rename_ctx/4
v3_codegen: Move guard_cg_list/6 to a more logical place
v3_codegen: Remove unnecessary clause for handling #cg_block{}
v3_codegen: Remove unnecessary adding of variable to set
|
|
|
|
|
|
In general, the new NIF-based file routines are faster than the old
efile driver.
However, on some computers, building the entire OTP system is somewhat
slower. It turns out that it is because 'erlc' cheated by turning off
the IO thread pool (using '+A0') to avoid context switches between
scheduler threads and threads in the IO thread pool. The new file
routines perform IO on dirty IO threads, and there is (by intent) no
way to force the operations to occur on scheduler threads to avoid
the context switches
What we can do to is to use a small (4Kb) read-ahead buffer for files
opened for reading (only) in list mode (which is how the compiler
opens its input files). The buffering reduces the number of context
switches between scheduler threads and dirty IO threads. On my
computer that seems to slightly speed up building of the entire OTP
system.
The buffer should do no harm. The only potential for harm I can
think of is random access to a file opened in read mode, where
the read-ahead buffer could slightly decrease performance. That
does not seems to be a likely use case in practice, though.
|
|
* hasse/stdlib/base64/OTP-14624:
stdlib: Add base64 benchmarks
stdlib: Do not check base64 input more than needed
stdlib: Minor optimization of base64
stdlib: Use binary_to_list in base64 when it is faster
stdlib: Optimize base64 functions
|
|
|
