Hi Suchakra and Christian > Nice! I have done something similar here and I was able to build and insert > code > snippets. But then I am more inclined towards analysing performance of > dyninst.
According the document and my test, turn 'BPatch::setTrampRecursive(true)' on will reduce the 'mutatee' instrumentation overhead a lot. > You might also be interested in 'fast tracepoint' infrastructure and normal > tracepoints provided by gdb. > But getting gdb's infra for dynamic tracing with UST may be somewhat a > serious hack. Thanks for pointing out gdb's 'fast tracepoint' feature. I just compiled gdb and got 'ftrace' work. > Quote form gdb document [1]: "At present, the UST (LTTng Userspace Tracer, > http://lttng.org/ust) tracing engine is supported." So does it mean the dynamic instrumentation project is already done by the gdb guys? > The dynamic instrumentation project is very interesting indeed. I also > think it could be doable as a GSoC project. Do you have any plan on > how you would integrate dyninst with UST? > Also, a quick reminder that the GSoC application for student is open > until May 3rd. Don't forget to send us your proposal :)! I will finish my proposal as soon possible. Thanks. [1] http://sourceware.org/gdb/current/onlinedocs/gdb/Server.html#Server -- Best Regards, 仝子飞 (Zifei Tong) _______________________________________________ lttng-dev mailing list [email protected] http://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev
