Age | Commit message (Collapse) | Author |
|
OTP-14122
* rickard/ds-fixes:
Fix VM global GC info for dirty schedulers
Leave dirty work in dirty run-queues on multi scheduling block
Fix premature removal of process struct
Fix crash due to GC of node entry on dirty scheduler
|
|
OTP-14121
* rickard/msb-fixes:
Multi scheduling block bug-fixes
|
|
|
|
|
|
|
|
Refc on process struct could reach zero while it was still
referred by dirty scheduler
|
|
The dirty scheduler failed to set a timer. Dispatch this job
to an ordinary scheduler.
|
|
* reith/erts/c_port_driver_tutorial_fix/PR-1281:
fix library open error in linked-in port driver tutorial
|
|
* ingela/inets/httpc-user-hangs/ERL-271/OTP-14091:
inets: httpc - Prevent hanging user process
inets: httpc - clean code
|
|
Input should be checked and httpc_handler process shall terminate
gracefully on errors so that user process will in hang in gen_server:call.
|
|
Remove dead code and redundant debug macros. Better to use tracing for
debugging.
|
|
* ingela/inets/httpc-stream/ERL-116/OTP-13571:
inets: httpc - Chunk size decoding could fail
|
|
Make depth of current_stacktrace configurable
|
|
* sverker/map-footprint-docs:
erts: Correct memory footprint for maps
|
|
Small map was wrong as we should include our own top Eterm and exclude
them for keys and values.
Large maps was wrong as it described the theoretical minimum of a
full tree, which does not happen in reality.
|
|
Correct chunk decoding by adding missing argument to match.
The symptom was that chunk decoding sometimes failed depending on
stream data arrival timing.
|
|
* siri/cdv/multi-line-slogan/ERL-318/OTP-14093:
[crashdump_viewer] Allow multiple lines in Slogan
|
|
* siri/etop/cpu-per-interval/OTP-14090:
[etop] Show CPU utilization for last interval only
|
|
* sverker/cuddle-port_SUITE:
erts: Fix faulty printout in port_SUITE
|
|
* legoscia/remove-watchdog-vestiges/PR-1255/OTP-14112:
Remove vestiges of watchdog support in heart
|
|
OTP-14055
* rickard/stacktrace-bugs:
Remove faulty release note for these fixes
New test cases testing stacktrace from apply on erlang:error()
Fix stactrace for apply on error/[1,2], exit/1, or throw/1
Fix stack-trace generated by a traced process
|
|
|
|
|
|
|
|
|
|
|
|
* ingela/ssl/terminate/OTP-14100:
ssl: Correct terminate behaviour
|
|
When the terminate function is called explicitly, to make guarantees
that for instance the reuseaddr option works as expected, we must
make sure that the clean up code is not run again when gen_statem
calls terminate. This check was broken in the rewrite from gen_fsm to
gen_statem.
Caused PEM cache errors, that in some cases would
cause unexpected connection failures.
|
|
* sverker/cuddle-code_SUITE-versions:
erts: Fix hanging race in code_SUITE:versions
|
|
hipe: one more test file and some cleanups
|
|
Add project-wide Emacs settings file
|
|
Allow escripts with only two lines
OTP-14098
|
|
|
|
into maint
* ingela/inets/httpc-redirect-host-header/ERL-316/OTP-14097:
inets: httpc - Correct redirection host header
|
|
* ingela/inets/httpc-PATCH-doc:
inets: httpc - patch method missing in doc
|
|
|
|
|
|
* egil/erts/handle-sigterm-nicely/OTP-14085:
erts: Document SIGTERM handler
erts: Add SIGTERM signal handler
|
|
|
|
* egil/doc-maps-type/ERL-197/OTP-14096:
doc: Enchance map pair optional/mandatory notes
doc: Change "stands for" to "denotes" in typespec
Remove whitespace errors
|
|
According to https://help.github.com/articles/setting-guidelines-for-repository-contributors, a link to a file named CONTRIBUTING.md will be shown when someone is about to create a pull request. That is more visible than our Wiki pages, which are easy to miss.
Create the CONTRIBUTING.md file, based on the existing Wiki page https://github.com/erlang/otp/wiki/Contribution-Guidelines.
|
|
Make sure that Travis fails when test cases fail
|
|
RFC2616 Sect 14.23: The Host request-header field specifies the
Internet host AND port number.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The current implementation assumes that escripts all have at least
three lines. But according to the escript documentation, comments
on the second line and emulator flags on the third line are all
optional. Thus escripts with only two lines -- shebang on the
first line and code on the second line -- will not run. This commit
fixes the problem by rearranging the escript header parsing logic,
and recognize the escript body on the second line if the header only
contains a line of shebang.
|