[lttv.git] / ltt / branches / poly / doc / developer / lttng-lttv-roadmap.html

<html>
<body>
<center><big><big>LTTV & LTTng roadmap<small><small></center>
<br>
<br>
Here are the roadmaps for the LTTV and LTTng development. I use a priority indice
for the TODO items :<br>
(1) : very high priority<br>
(10): lowest priority<br>
<br>
<br>
Dependencies are written between brackets [ ].<br>
The # symbol marks who is currently working on the item.<br>
The % symbol marks who is interested in the realisation of the item.<br>
<br>
<big>LTTV Roadmap<small><br>
<br>
* TODO<br>
(2) Finish pre-computing of states to a file.<br>
(3) Add cluster support.<br>
	# <A HREF="mailto:Eric Clement <eric.clement@polymtl.ca>">Eric Clement</A><br>
(3) Make LTTV aware of type formats (visual separators) defined in the XML
file.<br>
(3) Use a per architecture enumeration for traps.<br>
(3) Change the byte pair "facility, event" id for a short combining the
informatinon.<br>
(4) Statistics per time window.<br>
(4) Disable plugins when threshold reached (i.e. too much process in control
flow view). Draw, and, when the threshold is reached, stop drawing. The global
statistics view can inhibit showing the per process stats.<br>
(4) Add a visual artifact : PID 0 could be named swapper instead of UNNAMED for
cpus > 0.<br>
(4) Add event specific fields support to filter.<br>
(4) Add a periodic event interval view. (useful to verify event periodicity)<br>
(4) create a graphical per cpu activity view.<br>
(4) Filter by target process.<br>
(4) Compensate for time spent in probes in LTTV analysis.<br>
(4) Add CPU, network, disk, memory usage histogram. [Per interval statistics]<br>
(5) Add Python scripting hooks.<br>
(5) Flight recorder : start lttd automatically upon GUI trace control stop.<br>
(5) Automatically detect traces with too much processes and disable faulty operations.<br>
(5) Event sequence detector (inspired from regular expressions).<br>
(7) Create a hardware counter viewer (low cost rate counters : L1 cache miss,
page faults, interrupts...). This will be a generalisation of the event rate
view into a view of the evolution of a user definable event field.<br>
<br>
* TO FIX<br>
(10) Add cancel button to LTTV filter GUI window.<br>
(10) Sometimes, in the control flow view, a process with 0 creation time is
created in addition to the real process itself. Seems to be caused by end of
process life.<br>
(10) Statistics do not take in account the time spent in the mode present at
the beginning of the trace. Example : real time spent in system call on behalf
of process 0.<br>
<br>
<br>
<big>LTT Next Generation Roadmap<small><br>
<br>
* TODO<br>
(1) Announce LTTng patches on LKML.<br>
(3) Find a way to make logging know when it causes a minor page fault
recursively (without ignoring silently all nested events).<br>
(3) Have various subbuffer sizes for high and low rate channels.<br>
(3) Change the byte pair "facility, event" id for a short combining the
informatinon.<br>
(3) Add sysenter instrumentation on x86. Verify x86_64.<br>
(3) #IBM Fix ppc64 syscall time accounting. Maybe caused by a missing syscall
return instrumentation ?<br>
(3) Process and kernel stack dump for x86_64, for native and compat.<br>
(4) Add Xen support.<br>
	# <A HREF="mailto:Mathieu Desnoyers <compudj@krystal.dyndns.org>">Mathieu Desnoyers</A><br>
	# <A HREF="mailto:Parisa Heidari <parisa.heidari@polymtl.ca>">Parisa Heidari</A><br>
(4) efficient dynamic event filtering while recording trace.<br>
	% Sensis Corp. <A HREF="mailto:Bish, Tim <Tim.Bish@Sensis.com>">Tim Bish</A><br>
(4) instrument kernel bottom half irqsave, spinlocks, rwlocks, seqlocks, semaphores, mutexes, brlock.<br>
(4) In flight recorder mode, have the low traffic channels saved as normal
tracefiles and high traffic channels being flight recorder.<br>
(4) Try to use my own non LOCK prefixed version of atomic operations : we are
using per-CPU variables, so it should make it possible.<br>
(4) integrate NPTL instrumentation (see
<A HREF="http://nptltracetool.sourceforge.net/">PTT</A>).<br>
(4) Probe calibration kernel module.<br>
(5) Discuss with RelayFS upstream to integrate LTTng modifications.<br>
(5) Move to DebugFS for channel files.<br>
(5) Support CPUs with scalable frequency.<br>
(5) Integrate LTTng and lttd with LKCD.<br>
(7) Integrate periodical dump of perfctr hardware counters.<br>
(8) Integrate periodical dump of SystemTAP computed information.<br>
(8) Integrate SystemTAP logging with LTTng.<br>
(9) Add support for setjmp/longjmp and jump tables instrumentation to
ltt-instrument-functions.<br>
(9) Make ltt-usertrace-fast support internal heartbeat timer.<br>
(3) port LTTng to :<br>
x86_64<BR>
#Autodesk <A HREF="mailto:Francois L'Archeveque <fla@discreet.com>">Francois
L'Archeveque</A> for 2.6.9 RedHat Enterprise (patch available in contrib, still
has to be ported to 2.6.14, thanks!)<BR>
# Wind River for 2.6.14<BR>
PPC<BR>
PPC64<BR>
# Wind River for 2.6.14<BR>
alpha<BR>
user-mode Linux<BR>
Xen<BR>
# IBM<BR>
MIPS<BR>
MIPS64<BR>
% Wind River for 2.6.14<BR>
# Tim Bird <A HREF="mailto:tim.bird@as.sony.com">Tim Bird</A><br>
S/390<BR>
RTLinux<BR>
% Wind River for 2.6.14<BR>
<br>
<br>
<br>
* TODO (low priority)<br>
enhance RPM packages for lttng kernel<br>
Integrate header generation (genevent) in kernel build system.<br>
Export channels via network sockets instead of writing them to disk.<br>
Export buffers with time constraint for "live" visualisation. Use
ltt_force_switch periodically from a timer to insure slow channels do not
interfere with viewing.<br>
Have an optional round-robin mode to write information into multiple channels
from the same source.<br>
<br>
* Need to be discussed<br>
Use 2.6.14 RelayFS control files.<br>
Drop ltt-module-register and ltt-module-unregister, use exported variables.<br>
drop ltt_filter_control, use functions pointers instead.<br>
Merge facilities headers into one big header.<br>
Change the name of XML files from XML to something else.<br>
Remove ltt-base.c.<br>
<br>
<br>
* lttng patch division (0.5.10+) :<br>
<br>
instrumentation.diff<br>
instrumentation-arm.diff<br>
instrumentation-i386.diff<br>
instrumentation-mips.diff<br>
instrumentation-powerpc.diff<br>
instrumentation-ppc.diff<br>
instrumentation-s390.diff<br>
instrumentation-sh.diff<br>
ltt-facilities-headers.diff<br>
ltt-facilities-loader.diff<br>
ltt-facilities.diff<br>
ltt-relayfs.diff<br>
ltt-core.diff<br>
ltt-build.diff<br>
<br>
<br>
Mathieu Desnoyers<br>


</body>
</html>
Commit	Line	Data
	1	<html>
	2	<body>
	3	<center><big><big>LTTV & LTTng roadmap<small><small></center>
	4	<br>
	5	<br>
	6	Here are the roadmaps for the LTTV and LTTng development. I use a priority indice
	7	for the TODO items :<br>
	8	(1) : very high priority<br>
	9	(10): lowest priority<br>
	10	<br>
	11	<br>
	12	Dependencies are written between brackets [ ].<br>
	13	The # symbol marks who is currently working on the item.<br>
	14	The % symbol marks who is interested in the realisation of the item.<br>
	15	<br>
	16	<big>LTTV Roadmap<small><br>
	17	<br>
	18	* TODO<br>
	19	(2) Finish pre-computing of states to a file.<br>
	20	(3) Add cluster support.<br>
	21	# <A HREF="mailto:Eric Clement <eric.clement@polymtl.ca>">Eric Clement</A><br>
	22	(3) Make LTTV aware of type formats (visual separators) defined in the XML
	23	file.<br>
	24	(3) Use a per architecture enumeration for traps.<br>
	25	(3) Change the byte pair "facility, event" id for a short combining the
	26	informatinon.<br>
	27	(4) Statistics per time window.<br>
	28	(4) Disable plugins when threshold reached (i.e. too much process in control
	29	flow view). Draw, and, when the threshold is reached, stop drawing. The global
	30	statistics view can inhibit showing the per process stats.<br>
	31	(4) Add a visual artifact : PID 0 could be named swapper instead of UNNAMED for
	32	cpus > 0.<br>
	33	(4) Add event specific fields support to filter.<br>
	34	(4) Add a periodic event interval view. (useful to verify event periodicity)<br>
	35	(4) create a graphical per cpu activity view.<br>
	36	(4) Filter by target process.<br>
	37	(4) Compensate for time spent in probes in LTTV analysis.<br>
	38	(4) Add CPU, network, disk, memory usage histogram. [Per interval statistics]<br>
	39	(5) Add Python scripting hooks.<br>
	40	(5) Flight recorder : start lttd automatically upon GUI trace control stop.<br>
	41	(5) Automatically detect traces with too much processes and disable faulty operations.<br>
	42	(5) Event sequence detector (inspired from regular expressions).<br>
	43	(7) Create a hardware counter viewer (low cost rate counters : L1 cache miss,
	44	page faults, interrupts...). This will be a generalisation of the event rate
	45	view into a view of the evolution of a user definable event field.<br>
	46	<br>
	47	* TO FIX<br>
	48	(10) Add cancel button to LTTV filter GUI window.<br>
	49	(10) Sometimes, in the control flow view, a process with 0 creation time is
	50	created in addition to the real process itself. Seems to be caused by end of
	51	process life.<br>
	52	(10) Statistics do not take in account the time spent in the mode present at
	53	the beginning of the trace. Example : real time spent in system call on behalf
	54	of process 0.<br>
	55	<br>
	56	<br>
	57	<big>LTT Next Generation Roadmap<small><br>
	58	<br>
	59	* TODO<br>
	60	(1) Announce LTTng patches on LKML.<br>
	61	(3) Find a way to make logging know when it causes a minor page fault
	62	recursively (without ignoring silently all nested events).<br>
	63	(3) Have various subbuffer sizes for high and low rate channels.<br>
	64	(3) Change the byte pair "facility, event" id for a short combining the
	65	informatinon.<br>
	66	(3) Add sysenter instrumentation on x86. Verify x86_64.<br>
	67	(3) #IBM Fix ppc64 syscall time accounting. Maybe caused by a missing syscall
	68	return instrumentation ?<br>
	69	(3) Process and kernel stack dump for x86_64, for native and compat.<br>
	70	(4) Add Xen support.<br>
	71	# <A HREF="mailto:Mathieu Desnoyers <compudj@krystal.dyndns.org>">Mathieu Desnoyers</A><br>
	72	# <A HREF="mailto:Parisa Heidari <parisa.heidari@polymtl.ca>">Parisa Heidari</A><br>
	73	(4) efficient dynamic event filtering while recording trace.<br>
	74	% Sensis Corp. <A HREF="mailto:Bish, Tim <Tim.Bish@Sensis.com>">Tim Bish</A><br>
	75	(4) instrument kernel bottom half irqsave, spinlocks, rwlocks, seqlocks, semaphores, mutexes, brlock.<br>
	76	(4) In flight recorder mode, have the low traffic channels saved as normal
	77	tracefiles and high traffic channels being flight recorder.<br>
	78	(4) Try to use my own non LOCK prefixed version of atomic operations : we are
	79	using per-CPU variables, so it should make it possible.<br>
	80	(4) integrate NPTL instrumentation (see
	81	<A HREF="http://nptltracetool.sourceforge.net/">PTT</A>).<br>
	82	(4) Probe calibration kernel module.<br>
	83	(5) Discuss with RelayFS upstream to integrate LTTng modifications.<br>
	84	(5) Move to DebugFS for channel files.<br>
	85	(5) Support CPUs with scalable frequency.<br>
	86	(5) Integrate LTTng and lttd with LKCD.<br>
	87	(7) Integrate periodical dump of perfctr hardware counters.<br>
	88	(8) Integrate periodical dump of SystemTAP computed information.<br>
	89	(8) Integrate SystemTAP logging with LTTng.<br>
	90	(9) Add support for setjmp/longjmp and jump tables instrumentation to
	91	ltt-instrument-functions.<br>
	92	(9) Make ltt-usertrace-fast support internal heartbeat timer.<br>
	93	(3) port LTTng to :<br>
	94	x86_64<BR>
	95	#Autodesk <A HREF="mailto:Francois L'Archeveque <fla@discreet.com>">Francois
	96	L'Archeveque</A> for 2.6.9 RedHat Enterprise (patch available in contrib, still
	97	has to be ported to 2.6.14, thanks!)<BR>
	98	# Wind River for 2.6.14<BR>
	99	PPC<BR>
	100	PPC64<BR>
	101	# Wind River for 2.6.14<BR>
	102	alpha<BR>
	103	user-mode Linux<BR>
	104	Xen<BR>
	105	# IBM<BR>
	106	MIPS<BR>
	107	MIPS64<BR>
	108	% Wind River for 2.6.14<BR>
	109	# Tim Bird <A HREF="mailto:tim.bird@as.sony.com">Tim Bird</A><br>
	110	S/390<BR>
	111	RTLinux<BR>
	112	% Wind River for 2.6.14<BR>
	113	<br>
	114	<br>
	115	<br>
	116	* TODO (low priority)<br>
	117	enhance RPM packages for lttng kernel<br>
	118	Integrate header generation (genevent) in kernel build system.<br>
	119	Export channels via network sockets instead of writing them to disk.<br>
	120	Export buffers with time constraint for "live" visualisation. Use
	121	ltt_force_switch periodically from a timer to insure slow channels do not
	122	interfere with viewing.<br>
	123	Have an optional round-robin mode to write information into multiple channels
	124	from the same source.<br>
	125	<br>
	126	* Need to be discussed<br>
	127	Use 2.6.14 RelayFS control files.<br>
	128	Drop ltt-module-register and ltt-module-unregister, use exported variables.<br>
	129	drop ltt_filter_control, use functions pointers instead.<br>
	130	Merge facilities headers into one big header.<br>
	131	Change the name of XML files from XML to something else.<br>
	132	Remove ltt-base.c.<br>
	133	<br>
	134	<br>
	135	* lttng patch division (0.5.10+) :<br>
	136	<br>
	137	instrumentation.diff<br>
	138	instrumentation-arm.diff<br>
	139	instrumentation-i386.diff<br>
	140	instrumentation-mips.diff<br>
	141	instrumentation-powerpc.diff<br>
	142	instrumentation-ppc.diff<br>
	143	instrumentation-s390.diff<br>
	144	instrumentation-sh.diff<br>
	145	ltt-facilities-headers.diff<br>
	146	ltt-facilities-loader.diff<br>
	147	ltt-facilities.diff<br>
	148	ltt-relayfs.diff<br>
	149	ltt-core.diff<br>
	150	ltt-build.diff<br>
	151	<br>
	152	<br>
	153	Mathieu Desnoyers<br>
	154
	155
	156	</body>
	157	</html>