Age | Commit message (Collapse) | Author |
|
|
|
OTP-11529
|
|
A previous refactoring messed up the paths used in the start.boot
file used in intermediate releases generated for emulator upgrades.
|
|
* richcarl/fix-sasl-handler-docs:
Fix confusing documentation about error handlers in SASL
OTP-11507
|
|
* dgud/sasl/no_dot_erlang_boot/OTP-8479:
sasl: Add no_dot_erlang documentation and tests
sasl: Add no_dot_erlang start script
|
|
The SASL documentation confusingly talks about a handler 'error_logger_mf_h'
which doesn't actually exist. The handler is named 'log_mf_h', and it and
all other error handlers used by SASL are part of stdlib, not SASL itself.
This patch makes the documentation clearer.
|
|
The functions add_error_logger_mf/1 and delete_error_logger_mf/1 in sasl.erl
are designed to do nothing if the handler configuration as returned by
get_error_logger_mf/0 is 'undefined'. However, that function returns a
triple {undefined,undefined,undefined} when the error_logger_mf_*
configurations have not been set. This makes SASL always try to add the
log_mf_h handler on startup, passing 'undefined' for each of the Dir, MaxB
and MaxF parameters. The gen_event callback function log_mf_h:init/1 will
then crash with a function_clause error, causing the handler to be silently
removed again. This patch fixes the problem by making get_error_logger_mf/0
return the single atom 'undefined' as it is expected to do.
|
|
|
|
Sometimes it is wanted to start erlang without loading the user dependent
.erlang file, for example in scripts and configure tests.
|
|
|
|
|
|
rb was only accepting physical filenames as log file.
This patch allow rb to accept now any io_device, valid registered name,
except standard_error which is replaced by standard_io.
Creation of two new exported functions : rb:log_list/0-1
that print in log file if existing, otherwise on standard_io.
Add new functions documentation and tests in rb_SUITE.
|
|
This is to avoid lingering files after test runs on windows, since the
cleanup script often does not succeed in removing files with unicode
characters.
|
|
|
|
Using the low-level BEAM instructions, we can loop over each message in
the process queue and removes the first message that matches, without
receiving them all to later send them back to itself.
The function prim_eval:'receive'/2 is equivalent to the
following pseudo-code:
'receive'(F, T) ->
RESET MESSAGE QUEUE POINTER,
LOOP:
case PEEK CURRENT MESSAGE WITH TIMEOUT T of
{ok,Msg} ->
case F(Msg) of
nomatch ->
DECREMENT TIMEOUT T,
ADVANCE MESSAGE QUEUE POINTER,
GOTO LOOP;
Result ->
RESET MESSAGE QUEUE POINTER,
Result
end;
timeout ->
RESET MESSAGE QUEUE POINTER,
timeout
end.
To not break Dialyzer and other tools, we use a stub Erlang module which
abstract code is forcefully inserted into prim_inet.erl afterwards
compilation.
|
|
|
|
Tests that compare execution time for release_handler:install_release
sometimes fail. This has been corrected by multiplying the time with
the CPU utilization in order to disregard wait time for e.g. file
access and and non-erlang related load.
|
|
|
|
|
|
|
|
* Update systool_lib:read_term/1 to honour magic encoding comment in file.
* Add unicode option to re:compile and re:run when appropriate,
allowing e.g. unicode filenames.
* systools_make to allow unicode strings in .rel and .app files
|
|
|
|
|
|
|
|
|
|
This test often fails dues to too high expectations. Don't expect the
test to be more than 1000 times faster with no old code - it just
doesn't happen!
|
|
|
|
* nox/enable-silent-rules/OTP-10726:
Implement ./otp_build configure --enable-silent-rules
|
|
With silent rules, the output of make is less verbose and compilation
warnings are easier to spot. Silent rules are disabled by default and
can be disabled or enabled at will by make V=0 and make V=1.
|
|
|
|
|
|
rickard/r16/port-optimizations/OTP-10336
* rickard/port-optimizations/OTP-10336:
Change annotate level for emacs-22 in cerl
Update etp-commands
Add documentation on communication in Erlang
Add support for busy port message queue
Add driver callback epilogue
Implement true asynchronous signaling between processes and ports
Add erl_drv_[send|output]_term
Move busy port flag
Use rwlock for driver list
Optimize management of port tasks
Improve configuration of process and port tables
Remove R9 compatibility features
Use ptab functionality also for ports
Prepare for use of ptab functionality also for ports
Atomic port state
Generalize process table implementation
Implement functionality for delaying thread progress from unmanaged threads
Conflicts:
erts/doc/src/erl_driver.xml
erts/doc/src/erlang.xml
erts/emulator/beam/beam_bif_load.c
erts/emulator/beam/beam_bp.c
erts/emulator/beam/beam_emu.c
erts/emulator/beam/bif.c
erts/emulator/beam/copy.c
erts/emulator/beam/erl_alloc.c
erts/emulator/beam/erl_alloc.types
erts/emulator/beam/erl_bif_info.c
erts/emulator/beam/erl_bif_port.c
erts/emulator/beam/erl_bif_trace.c
erts/emulator/beam/erl_init.c
erts/emulator/beam/erl_message.c
erts/emulator/beam/erl_port_task.c
erts/emulator/beam/erl_process.c
erts/emulator/beam/erl_process.h
erts/emulator/beam/erl_process_lock.c
erts/emulator/beam/erl_trace.c
erts/emulator/beam/export.h
erts/emulator/beam/global.h
erts/emulator/beam/io.c
erts/emulator/sys/unix/sys.c
erts/emulator/sys/vxworks/sys.c
erts/emulator/test/port_SUITE.erl
erts/etc/unix/cerl.src
erts/preloaded/ebin/erlang.beam
erts/preloaded/ebin/prim_inet.beam
erts/preloaded/src/prim_inet.erl
lib/hipe/cerl/erl_bif_types.erl
lib/kernel/doc/src/inet.xml
lib/kernel/src/inet.erl
|
|
|
|
|
|
It was possible to insert {Mod::atom(),Vsn::term()} instead of
Mod::atom() in the modules list in a .app file. This was not visible
in the documentation of .app files, but it was visible in the
documentation of application:load/[1,2] (where the .app file term can
be used directly as first argument).
The Vsn part was never used, so this possibility has now been removed.
|
|
|
|
release_handler_SUITE:otp_9864 deleted parts of the
release_handler_SUITE_data directory so the test suite could not be
executed twice without re-installation. This has been corrected.
|
|
Some tests in rb_SUITE failed every now and then depending on timing:
1. sometimes SASL printouts from cth_log_redirect were captured in
addition to expected printouts
2. sometimes test supervisor and server were not completely stopped
before next test case was started
This has been corrected.
|
|
|
|
Release_handler_suite correctly checks that no calls to the file module
is done during a diskless upgrade, but file:native_name_encoding is
a BIF that does no real i/o and can therefore be called during command line
argunemt parsing. The testcase is updated to ignore such calls.
|
|
OTP-10106
OTP-10107
|
|
|
|
|
|
This is a temporarily fix for R15B01.
|
|
* jc/omit-undefined-start_phases-3:
Avoid undefined start_phases entry in .script
Avoid creating an undefined start_phases entry when generating a release
OTP-10003
|
|
* siri/sasl/correct-app-name-in-error/OTP-9888:
Fix error message if stdlib has faulty start type in .rel file
|
|
* siri/sasl/doc-regexp-in-appup/OTP-10001:
Add documentation of regular expressions for version numbers in appup
|
|
If stdlib had a different start type than permanent, systools_make
would fail but it would say that it was sasl that had faulty start
type. This has been corrected.
|
|
|
|
Inorder to align with reltool a {start_phases,undefined} entry in will
no longer occur in the application specification in .script files. The
start_phases entry will only exist if it contains real start phases.
Also, the default value for #application.start_phases is no longer set
to [] in systools.hrl. This value was actually never used, since it
was always explicitly set in systools_make:parse_applications - by
default to 'undefined', so the default value in the record definition
was only a confusing detail. Anyway - now corrected.
|