Age | Commit message (Collapse) | Author |
|
* hw/erl-clarify-detached-doc:
Mention that "-detached" implies "-noinput"
OTP-9086
|
|
* ta/driver-entry-typos:
Fix a couple typos in driver_entry(3)
OTP-9085
|
|
|
|
* fm/dict-man-page:
Fixes to the orddict module documentation
Fixes to the dict module documentation
OTP-9083
|
|
* pan/wx_macos_main_thread:
Remove race in main thread stealing due to reading and writing to pipe from same thread
OTP-9081
|
|
same thread
|
|
* [compiler] Added support for textual convention AGENT-CAPABILITIES
and full support for textual convention MODULE-COMPLIANCE:
OTP-8966
* [agent] Added a "complete" set of (snmp) table and variable
print functions:
OTP-8977
* [agent] For is_set_ok and set operations on the table
vacmAccessTable, all values of the vacmAccessSecurityModel
column was incorrectly translated into any:
OTP-8980
* [agent] Reconfigure of SNMP-VIEW-BASED-ACM-MIB could after
a change in the vacm config file (vacm.conf):
OTP-8981
* [compiler] Added a MIB compiler (frontend) escript, snmpc:
OTP-9004
|
|
Rewrote the code a bit to silence dialyzer. Seems dialyzer can't
handle a function that either returns or loops forever depending on
function argument.
|
|
* pan/wx_macos_main_thread:
Make MacOS (NS) wx use the main thread
OTP-9081
|
|
|
|
* pan/halfword-tmp-heap-fixes/OTP-8332:
Fix assymetric (Un)UseTmpHeap in erl_process.c
Conflicts:
erts/emulator/beam/beam_emu.c
|
|
* dgud/wx/unload-driver/OTP-9080:
Fix bad string copy during initialization
Fix start args
Fixed coredump when graphics could not be initiated
|
|
This may have caused the core dump on ubuntu-64 which have happend
for some people.
|
|
|
|
For instance when DISPLAY was not set
|
|
|
|
|
|
|
|
|
|
With help from Kostis
|
|
The second argument to the definitions_loop function should
be the record dldata and not the atome false...
|
|
* sverker/driver_SUITE_smp_select_nice_fail/OTP-9042:
Allow unexpected driver input event in driver_SUITE:smp_select
|
|
* sverker/doc-fixes/OTP-9002:
Fix minor typos in erl_nif documentation
|
|
* pan/valgrind-3.6.0:
Make crypto.c work with valgrind 3.6 and correct cerl.src
Teach cerl to handle newer valgrind
OTP-9079
|
|
* pan/inet-so-priority-ip-tos/OTP-9069:
Add patch from Per Hedeland
Ignore permission error when implicitly setting SO_PRIORITY
|
|
|
|
Joint effort by Kostis, pan, egil & sverker
|
|
* bjorn/erts/crash-dumps/OTP-9057:
Don't dump the contents of ordered_set ets tables
Write the value "Time left" for BIF timers as an unsigned integer
|
|
|
|
|
|
|
|
and end. Also some cosmetic stuff (indenting).
|
|
|
|
|
|
|
|
Merge branch 'bmk/snmp/minor_doc_improvements/OTP-9037' into bmk/snmp/snmp419_integration/OTP-9068
Conflicts:
lib/snmp/doc/src/notes.xml
lib/snmp/src/app/snmp.appup.src
|
|
Merge branch 'bmk/snmp/snmpc/OTP-9004' into bmk/snmp/snmp419_integration/OTP-9068
Conflicts:
lib/snmp/doc/src/notes.xml
lib/snmp/src/app/snmp.appup.src
lib/snmp/vsn.mk
|
|
Also, remove compiler warnings for crashdump_viewer_SUITE and fix
Makefile in test directory so Emakefile does not grow.
|
|
* lukas/common_test/test_order_wrong/OTP-9026:
Add test event info for the old tests
Update test spec parser to not group suites, groups and cases which are hierarchically related if merge_tests config i set to false
Add test cases for test spec ordering of testcases in spec
Conflicts:
lib/common_test/test/ct_testspec_1_SUITE.erl
|
|
* lukas/test_server/ts_migration/OTP-8768:
Update test_server tests to conform with common_test standard
Remove old test_server code
|
|
|
|
|
|
Since browsers no longer can provide the full path of a file selected
with a file-type input field (browse button), the input field for
loading a crashdump is now changed to a plain text input field. Since
this reduces the user-friendlyness, a shell script (and a .bat file)
has instead been added so the crashdump_viewer can be started directly
from the command line - and thus normal tab completion can be used for
selecting the crashdump file.
Usage: cdv file [ browser ]
|
|
This is a first attempt at fixing the problem described in seq11783 -
crashdump_viewer is very slow at parsing big crashdumps. To open the
first page for a dump of 17M takes about 2 minutes and a dump of 280M
takes 1.5-2 hours.
The main problmem is that the cdv_dump_index_table, which holds all
tags read from the dump, is a bag. Profiling shows that ~95% of the
time is spent in ets:insert. The table is now changed to an
ordered_set.
A second problem occured when a page with many table rows was
opened. These pages were sent to inet in one chunk, causing both
crashdump_viewer_server and the inets (mod_esi) process to grow very
much in memory usage. To overcome this, the pages are now sent to
inets in chunks of 1000 rows, and the data is coverted to binaries to
avoid data copying between the two processes.
Also, some new information in the crashdump was not recognized by the
crashdump_viewer. This has been fixed.
|
|
* lukas/stdlib/native_escript/OTP-9076:
Update escrips to allow the -n or -compile(native) flag, which compiles the code within the escript with the +native flag. You need an HiPE enabled emulator for this to work.
|
|
* lukas/tools/cover_mem_footprint/OTP-9043:
Update testcases which need crypto to be skipped on platforms which does not have crypto
Update internal pmap to have a process limit Add write concurrancy to cover masters ?COVER_TABLE
Update documentation to reflect performance enhancement changes of cover
Add aync_analyse_to_file function to cover
Split the cover ets tables into two tables, one with the clause info and one with the bump info. This will make it faster to search the tables when analyzing and exporting data.
Add process debug tags
Update remote collect to handle multiple requests at once
Remove io printout warnings when exporting an imported module
Make the call to cover parallel so that the test_server takes advantage of the new cool parallel cover features.
Update cover to allow multiple analyse and analyze_to_file calls at the same time. For each call a seperate process will be spawned to handle the request.
Refactor cover to prepare it for making analysis parallel
Update remote loading to only load a certain number of modules at a time to prevent memory usage explosion
Conflicts:
lib/tools/test/cover_SUITE.erl
|
|
|
|
|
|
|
|
Generation of the PLT is now performed without using OS commands.
We still try to copy in the default plt to make small scale testing
efficient. If generation/checking fails, suites are skipped except
plt_tests_SUITE which contains a bare PLT check that fails normally.
|