Age | Commit message (Collapse) | Author |
|
erts_sys_aligned_alloc() is currently implemented using posix_memalign if
it exist, or using _aligned_malloc on Windows.
If erts_sys_aligned_alloc() exist allocators will create sys_alloc
carriers similar to how this was done pre-R16.
|
|
'bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r16' into maint-r16
* bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r16:
[megaco] Appup cleanup
[megaco] Retroactive version bump
[megaco] Used wrong downgrade tag for (flex) handler
[megaco] Fixrd appup sub-test suite
[megaco] Debug printouts in megaco test server
[megaco] Internal test server not handling failing cases
[megaco] Fix merge
[megaco] Misc codec tickets test case fixes
[megaco] Misc fixes of the (internal) megaco test server
Changed name of test case functions (from seq12263 to otp10998)
Assign correct version and updated release notes accordingly.
Removed comments
Ensure of incorrect number of terms when scanning property groups
Uncommented test cases
Add new messages
Calculation of term_spec size (initial and next) changed
Minor test case rework
Test case cleanup
Add rule for test beam'es
Preliminary test case updates
|
|
|
|
|
|
* pan/monitor_long_schedule:
Skip testcase on platforms w/o pipes/usleep
Add documentation for new system_monitor
Add long_schedule detection for ports and testcases
Add erlang:system_monitor(Pid,[{long_schedule,Millis}])
|
|
|
|
|
|
|
|
Removed instruction for 3.16.0.2 since this version has
never existed in R16 and since we would otherwise have to
include a *lot* of module delete instructions.
|
|
'bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r15' into bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r16
Conflicts:
lib/megaco/doc/src/notes.xml
lib/megaco/src/flex/megaco_flex_scanner_drv.flex.src
lib/megaco/test/megaco_codec_prev3a_test.erl
lib/megaco/test/megaco_codec_prev3b_test.erl
lib/megaco/test/megaco_codec_prev3c_test.erl
lib/megaco/test/megaco_codec_v1_test.erl
lib/megaco/test/megaco_codec_v2_test.erl
lib/megaco/test/megaco_codec_v3_test.erl
lib/megaco/vsn.mk
|
|
|
|
|
|
'bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r14' into bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r15
[megaco] Fixed felx driver handler code-change tags.
|
|
* maint-r15:
Prepare release
Dummy merge.
|
|
'bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r13' into bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r14
Fixed flex handler code-change version tag.
|
|
|
|
* rickard/sched/OTP-11022:
Do not count offline run-queues as empty
|
|
|
|
|
|
|
|
|
|
|
|
|
|
'bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r14' into bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r15
Conflicts:
lib/megaco/.gitignore
lib/megaco/doc/src/notes.xml
lib/megaco/src/app/megaco.appup.src
lib/megaco/src/flex/megaco_flex_scanner_drv.flex.src
lib/megaco/test/megaco_codec_v2_test.erl
lib/megaco/vsn.mk
|
|
|
|
|
|
* bjorn/asn1/fix-lost-extension-mark/OTP-10995:
Prevent loss of objects after the extension marker
Don't lose the extension mark for object set parameters
|
|
|
|
'bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r13' into bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r14
Conflicts:
lib/megaco/doc/src/notes.xml
lib/megaco/src/app/megaco.appup.src
lib/megaco/test/megaco_codec_flex_lib.erl
lib/megaco/test/megaco_codec_v2_test.erl
lib/megaco/test/megaco_test_lib.erl
lib/megaco/vsn.mk
|
|
|
|
* bjorn/asn1/fix-lost-extension-mark/OTP-10995:
Prevent loss of objects after the extension marker
Don't lose the extension mark for object set parameters
|
|
|
|
Also added (and cleaned up) appup.
|
|
bmk/megaco/flex_buffer_overrun_while_scanning_prop_parms/OTP-10998/r13
Conflicts:
lib/megaco/src/flex/megaco_flex_scanner_drv.flex.src
|
|
|
|
* egil/r16/fix-delete_element/OTP-10932:
erts: Refactor erlang:insert_element/3 for clarity
erts: Fix copy error in erlang:delete_element/2
|
|
* rickard/sys-flags/OTP-11000:
Fix `+sws' and `+swt' system flags
|
|
* rickard/thr_prgr_later_op/OTP-10994:
Make port close not be delayed
Be less eager requesting wakeup for cleanup jobs
|
|
* rickard/is_process_alive/OTP-10926:
Fix erlang:is_process_alive/1
|
|
Drivers are dereferenced when the port is completely gone, in the operation
scheduled with erts_schedule_thr_prgr_later_op. If we use
erts_schedule_thr_prgr_later_cleanup_op, it may not happen
for a long time and the driver can not be unloaded.
We can not dereference the driver before the thread_progress_later
thing, because references to the driver may exist in other threads, so
the unloading of the driver may free locks held by other schedulers and
whatnot.
|
|
|
|
|
|
In an object set with a single root object, objects after
the extension marker would be lost.
|
|
When an object set is an actual parameter, the extension marker
for the object set could get lost.
|
|
|
|
At the end of the mfs_load_property_groups function, the
final property group list and property groups list is
"terminated". A call to mfs_ensure_term_spec with (incorrect)
size 4 was made prior to this to ensure enough data was available.
The correct size was 6!
|
|
|
|
|
|
There seem to be some problem with realloc (core dump)
so the calculation of the term_spec size needs to be
"corrected".
|
|
|