I thinkg it might be useful if I could get some representation
(graphical or otherwise) about the number of RUNNable tasks at any
given time to get an idea of how much contention is ongoing on the CPU.
So I guess something like the following information:
- peak measurement of the running
TRACE_DEBUG_SYSTEM = 7,
TRACE_DEBUG_PROGRAM = 8,
TRACE_DEBUG_PROCESS = 9,
TRACE_DEBUG_MODULE = 10,
TRACE_DEBUG_UNIT= 11,
TRACE_DEBUG_FUNCTION= 12,
TRACE_DEBUG_LINE= 13,
I still don't
Hi, we are looking for a research associate and a postdoctoral fellow to work
on developing further LTTng for Integrated tracing, profiling and debugging
for tuning large heterogeneous clusters. These positions require interacting
with the team and are thus based in Montreal. If you could be
I am interested in some of the problems that the LTTNG tool can be
used to address. I have some time available, many years experience in
vaguely related areas and am willing to take direction.
If my help as a developer, tester, or documenter would be appreciated,
Your help is indeed most
-C limits the maximum file size and -W the amount of such files to
keep on disk.
Do you round up to the next packet / subbuffer / event?
In the upcoming releases, there should be small indexes telling where each
packet / subbuffer starts in each file. How are these index files affected by
It still doesn't explain why I was able to add procname to the
context but adding also vpid and vtid failed when the number of cores was 24
rather than 12.
No idea. These are unrelated.
Perhaps you were not tracing that much and not filling all buffers with the
smaller events with less
I would suggest something relatively simple and basic as far as
performance characteristics measurements goes - x apps, y traces /
second - measure CPU, memory, i/o, file size across the various
mechanisms.
Yes, something relatively simple should do. For a process writing to syslog,
you
Don't forget, it's a beer that must be made in the Quebec province and
from a microbrewery. For 2.4, the beer must begin with the letter E.
Some help:
http://www.bieresduquebec.ca/bieres
There is quite a choice, Ete indien, Elixir celeste, Ephemere... Any indication
on the character of
I would expect that the ctf writer API recently added to babeltrace
(currently in master branch), along with the Python bindings that cover
trace read and write APIs, should allow you to implement things like:
- A plugin to read a CTF trace, and output it in an intermediate format
to
That's what I thought, but benchmarking showed that there's
practically no difference.
The filter is a simple ID comparison of the form 'id % 1,000 == 0',
so 999 out of 1K tracepoints are filtered out.
Could you please point me to some references on this topic?
What is the running time and
Studying the API of URCU it seems that it's mostly intended for
link-based data structures, such a lists. Suppose I have a large
struct that is often accessed by many readers, and some writers that
occasionally update few fields each time. If I understand correctly,
using URCU, the writers
Yes, if you look at the options available, you can split the trace in smaller
files and even limit the total size accumulated.
- Mail original -
Hi,
It's well known that tracing can produce huge data easily.So if we
keep tracing on a production system for a while(hour/day/week) ,we
Normally, the two traces are stored separately and they are merged at display
time in TraceCompass. If they were taken on the same host, they use the same
clock and should be naturally aligned. If they come from different hosts and
you want very fine symchronization, you need some communication
- is it possible to use LTTng from Xenomai RT threads? (old mails on the
Xenomai list suggest so, but it is unclear if special
precautions/incantations/patches are needed)
I have not looked into this. There should not be big problem since LTTng is
fairly self standing to minimize the impact
and could help you by getting your
feedback and proposing adjustments and improvements to the LTTng toolchain as a
result.
- Mail original -
Michel,
Am 29.03.2015 um 16:31 schrieb Michel Dagenais
michel.dagen...@polymtl.ca:
- is there a complete example out-of-tree kernel
The problem at hand has these characteristics:
- it happens only on low-end ARM platforms
- the code is part of the trajectory planner, which involves fp and
transcendental functions
- the code is executed from a period Xenomai thread and once in a while
execution time exceeds the cycle
That said, the folks which have timing problems have not caught on yet.. it
looks some more guidance, examples and a bit of a machinekit-specific
writeup is needed, competing with the other 247 priority projects on my desk
;)
Many people interested in MachineKit wonder about the control
Here is a short answer to start. Others will have more precise suggestions.
Something that can seriously affect the performance is what is used to obtain
information such as the timestamp and possibly the CPU id. More details on your
platform would help in that respect. The second thing is
?
>
> Adding Paul E. McKenney in CC, who may have some thoughts on this
> topic.
>
> Thanks,
>
> Mathieu
>
> >
> > Thanks again!
> >
> > On Thu, Mar 10, 2016 at 1:52 PM, Michel Dagenais
> > <michel.dagen...@polymtl.ca> wrote:
> >> Real-ti
We have been using LTTng-UST for hard real-time systems for a while. Once
tracing is setup, there are few interactions between the traced application and
the rest. The tracing code simply fills the sub-buffers and switches to another
sub-buffer when the current one is full. In one mode, the
On Linux, with pthreads, thread scheduling takes place in the kernel. You can
then obtain this information through lttng tracing in the kernel domain
(scheduling events) and view the trace in Trace Compass.
- Le 15 Mai 19, à 12:42, Zvi Vered a écrit :
> Hello,
> Can I use lttng to
>From the start of these tools, Polytechnique was involved in Research projects
>to design and prototype new advanced features for tracing, profiling and
>debugging parallel distributed heterogeneous systems with LTTng and Trace
>Compass. This work is conducted in collaboration with Ericsson,
> Without recompiling, how would that be implemented?
As you mentioned, this is possible when "jump patching" 5 bytes instructions.
Fast tracepoints in GDB and in kprobe do it. Kprobe goes further and patches
sequences of instructions (because the target instruction is less than 5 bytes)
if
You are probably referring to the work of Mohammad Gebai:
Survey and Analysis of Kernel and Userspace Tracers on Linux: Design,
Implementation, and Overhead
https://dl.acm.org/doi/abs/10.1145/3158644
- Le 11 Sep 23, à 11:52, Mathieu Desnoyers mathieu.desnoy...@efficios.com a
écrit :
>
24 matches
Mail list logo
