| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
|
|
| |
function `dbg:stop_clear/1` is not documented in the API but was kept
for compatibility reasons. in this commit we make a step forward to
deprecating its use such that it will be eventually removed.
closes GH-6903.
|
| |
|
| |
|
| |
|
| |
|
|\
| |
| |
| |
| | |
Move instrument.erl to runtime_tools
OTP-18487
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
instrument.erl can be a useful tool for debugging
memory allocation and memory leaks in production
systems. For example, eventually we could have a
button in Observer that invokes instrument and
prints a histogram on demand. For this purpose,
however, it must be part of runtime_tools.
|
|\ \
| |/
|/|
| |
| | |
* raimo/kernel/boot-on_load:
Use internal wildcard function from -on_load
|
| |
| |
| |
| |
| |
| | |
The file server may not be started in embedded mode, so we have
to use our own undocumented option to filelib:wildcard/2
to avoid hanging, waiting fo the file server, in embedded mode...
|
|/ |
|
| |
|
|
|
|
| |
This function can be used to trace to a file in clear text
|
|
|
|
|
|
|
| |
The CERL_DETACHED_PROG code seems to be a relic that is no
longer used. So we delete it doesn't work anyway.
Closes #4323
|
|\
| |
| |
| | |
OTP-18012
|
| |
| |
| |
| | |
OTP-18012
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Include "bad" options in the list, but not with a "value"
(since we did not get any). Instead with something that
will explain "something": A string; either "-",
"Not Supported" or "error: ..."
We can see what is missing and why (its better then nothing).
Also, this is approx like what is done for socket.
OTP-18012
|
| |
| |
| |
| | |
OTP-18012
|
| |
| |
| |
| |
| |
| |
| |
| | |
Reading port socket options is now done "individually".
This means that if one option is invalid its simply skipped,
and we continue with the rest.
OTP-18012
|
|/ |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
`dbg:stop/0` explicitly leaves global function tracing untouched,
which can lead to some very confusing situations:
1. As it clears the trace flags of all processes, it often
*looks* like all global function trace patterns have been
cleared, leaving a fun surprise for the next poor sod that
tries to trace anything.
2. It makes it really annoying to clear global function traces
afterwards, as we've discarded all information on what nodes
we've been tracing on. The user more or less has to call
`erlang:trace_pattern/3` themselves on all nodes they've
traced on.
Since clearing the trace flags of all processes means that `dbg`
can't coexist with other trace tools to begin with, there is
very little point in keeping global trace patterns around after
stop/0. This commit makes stop/0 behave like stop_clear/0,
ensuring that *ALL* tracing is turned off on *ALL* associated
nodes before shutting down.
|
| |
|
|\ |
|
| |\
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
OTP-17800
OTP-17830
* sverker/scheduler_wall_time-doc:
runtime_tools: Add scheduler:get_sample/0 and get_sample_all/0
runtime_tools: Fix bug in scheduler:utilization(Seconds)
erts: Clarify system_flag(scheduler_wall_time,_) docs
|
| | |
| | |
| | |
| | |
| | | |
And clarify documentation,
especially regarding erlang:system_flag(scheduler_wall_time,_).
|
| | |
| | |
| | |
| | |
| | |
| | | |
The returned OldFlag is the old node global state. That should
not prevent us from decreasing our process local logical ref counter
for scheduler_wall_time.
|
|/ / |
|
|/ |
|
| |
|
|
|
|
| |
OTP-17346
|
|
|
|
|
|
|
| |
Make it possible to "configure" the inclusion of non-supported
socket options.
OTP-17346
|
|
|
|
|
|
|
| |
Add even not supported sockets to the list, whith the
value set to 'Not Supported'.
OTP-17346
|
|
|
|
|
|
| |
Add a 'Options' section to the 'socket info' window.
OTP-17346
|
|
|
|
| |
OTP-17346
|
|
|
|
|
|
| |
A tab has been added for the 'new' socket(s).
OTP-17346
|
|\ |
|
| |\
| | |
| | |
| | |
| | | |
* kjell/runtime_tools/dbg/ERL-1371/OTP-16930/GH-4396:
GH-4396: exported function instead of fun to set up remote tracing
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The function `dbg:n/1` used a local fun to set up a tracer on a remote
node. This works fine as long as the remote node is running exactly
the same version of Erlang/OTP but does not work at all
otherwise. This is fixed by exporting the relevant function
(`dbg:do_relay/2`) and by calling this function on the remote node to
set up remote tracing.
|
| | | |
|
|\ \ \
| |/ /
|/| | |
Drop remnants of elib_malloc (removed in R15)
|
| | | |
|
| | |
| | |
| | | |
Co-authored-by: John Högberg <john@erlang.org>
|
| | | |
|
|/ / |
|
| |
| |
| |
| | |
This reverts commit eee1f3aba92b5a741eabfa1cb6eca0632562777a.
|
| | |
|
| |
| |
| |
| | |
This reverts commit 63638d668efbeb4539ba301328cabbc2636fb154.
|