[perf-discuss] Re: Proposal for in-kernel buffering of CPC overflow

2006-05-04 Thread WILLIAM CHEN
The way I had prototyped it only requires minor changes to the current libcpc structure support. All the underlying functionality and code are already in the kernel. 1. All the cpc data structures are already available. Only addition to support buffering is adding the buffer array into the st

[perf-discuss] Re: Re: Proposal for in-kernel buffering of CPC overflow events

2006-03-13 Thread WILLIAM CHEN
Since this case can be categorized as a synchronous event, that one can read the 'unfilled' buffer at anytime, it is simple enough to add an interface to read and cleanup the buffer as though the buffer was full and interrupt given. Thanks, William This message posted from opensolaris.org __

[perf-discuss] Re: Proposal for in-kernel buffering of CPC overflow events

2006-03-08 Thread WILLIAM CHEN
Thread id is very useful information to include also. William This message posted from opensolaris.org ___ perf-discuss mailing list perf-discuss@opensolaris.org

[perf-discuss] Re: Proposal for in-kernel buffering of CPC overflow events

2006-03-08 Thread WILLIAM CHEN
In response to the question of what happens if the buffer never overflows. One will not get an interrupt if the buffer does not overflow. In this case, the event being monitored is very rare or the application being monitored has a very short execution time. In either case, the overhead of th