| | 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 | <br> |
| | 17 | <big>LTT Next Generation Roadmap<small><br> |
| | 18 | <br> |
| | 19 | * TODO<br> |
| | 20 | (1) LTTng event description: move from tracepoint/markers to Ftrace TRACE_EVENT |
| | 21 | declarations. Extend TRACE_EVENT as needed. (<A HREF="mailto: Douglas Santos |
| | 22 | <douglas.santos@polymtl.ca">Douglas Santos</A><BR>) |
| | 23 | (1) LTTng ring buffer adaptation for FTrace.<A HREF="mailto: |
| | 24 | mathieu.desnoyers@efficios.com">Mathieu Desnoyers</A><BR> |
| | 25 | (1) FTrace/LTTng trace format standardization.<BR> |
| | 26 | (1) Extend NO_HZ support vs trace streaming support to other architectures (need |
| | 27 | to add cpu idle notifiers and test).<br> |
| | 28 | (1) Make sure ltt-ascii kernel text dump fits well with streaming hooked into |
| | 29 | cpu idle.<br> |
| | 30 | (1) Support CPUs with scalable frequency. (done for ARM OMAP3 UP only)<br> |
| | 31 | <br> |
| | 32 | <br> |
| | 33 | * Nice to have<br> |
| | 34 | (3) Bring stack dump in sync with new lttng.<br> |
| | 35 | (4) Dump mounts. (to fix)<br> |
| | 36 | (4) Add Xen support. (Trace buffer desallocation needs to be fixed)<br> |
| | 37 | (4) integrate NPTL instrumentation (see |
| | 38 | <A HREF="http://nptltracetool.sourceforge.net/">PTT</A>).<br> |
| | 39 | (4) Probe calibration kernel module.<br> |
| | 40 | (4) Make page faults detect nested fault without nesting 4 times in the page |
| | 41 | fault handler. (or use vmalloc_sync_all at each tracing module load)<br> |
| | 42 | (5) Add boot time tracing support.<br> |
| | 43 | (5) Integrate LTTng and lttd with LKCD.<br> |
| | 44 | # <A HREF="mailto:Lai Jiangshan <laijs@cn.fujitsu.com>>">Lai Jiangshan</A><br> |
| | 45 | (7) Integrate periodical dump of perfctr hardware counters.<br> |
| | 46 | (8) Integrate SystemTAP logging with LTTng.<br> |
| | 47 | (8) Integrate periodical dump of SystemTAP computed information.<br> |
| | 48 | (9) Add support for setjmp/longjmp and jump tables instrumentation to |
| | 49 | ltt-instrument-functions.<br> |
| | 50 | <br> |
| | 51 | <br> |
| | 52 | <br> |
| | 53 | <big>LTTV Roadmap<small><br> |
| | 54 | <br> |
| | 55 | Note: new feature development is currently done in the Linux Tools Project: |
| | 56 | LTTng Integration. Mainwhile, LTTV is maintained as a known-stable viewer.<br> |
| | 57 | <br> |
| | 58 | <br> |
| | 59 | * Nice to have<br> |
| | 60 | (4) Statistics per time window.<br> |
| | 61 | (4) Add Xen per physical CPU view.<br> |
| | 62 | (4) Add Xen per vcpu view.<br> |
| | 63 | (4) Disable plugins when threshold reached (i.e. too much process in control |
| | 64 | flow view). Draw, and, when the threshold is reached, stop drawing. The global |
| | 65 | statistics view can inhibit showing the per process stats.<br> |
| | 66 | (4) Add a visual artifact : PID 0 could be named swapper instead of UNNAMED for |
| | 67 | cpus > 0.<br> |
| | 68 | (4) Add event specific fields support to filter.<br> |
| | 69 | (4) Add a periodic event interval view. (useful to verify event periodicity)<br> |
| | 70 | (4) create a graphical per cpu activity view.<br> |
| | 71 | (4) Filter by target process.<br> |
| | 72 | (4) Compensate for time spent in probes in LTTV analysis.<br> |
| | 73 | (4) Add CPU, network, disk, memory usage histogram. [Per interval statistics]<br> |
| | 74 | (4) Add sort by process priority in the control flow view (must also instrument |
| | 75 | priority information of the processes).<br> |
| | 76 | % Airbus<br> |
| | 77 | (5) Add Python scripting hooks.<br> |
| | 78 | (5) Add GUI interface to take an hybrid trace.<br> |
| | 79 | (5) Automatically detect traces with too much processes and disable faulty operations.<br> |
| | 80 | (5) Event sequence detector (inspired from regular expressions).<br> |
| | 81 | (7) Create a hardware counter viewer (low cost rate counters : L1 cache miss, |
| | 82 | page faults, interrupts...). This will be a generalisation of the event rate |
| | 83 | view into a view of the evolution of a user definable event field.<br> |
| | 84 | <br> |
| | 85 | * TO FIX<br> |
| | 86 | (10) Add cancel button to LTTV filter GUI window.<br> |
| | 87 | (10) Sometimes, in the control flow view, a process with 0 creation time is |
| | 88 | created in addition to the real process itself. Seems to be caused by end of |
| | 89 | process life.<br> |
| | 90 | (10) Statistics do not take in account the time spent in the mode present at |
| | 91 | the beginning of the trace. Example : real time spent in system call on behalf |
| | 92 | of process 0.<br> |
| | 93 | <br> |
| | 94 | <br> |
| | 95 | Mathieu Desnoyers<br> |
| | 96 | |
| | 97 | |
| | 98 | </body> |
| | 99 | </html> |
projects / lttv.git / blame_incremental