On Fri, 4 Jun 2004 11:05:27 -0500, [EMAIL PROTECTED] wrote: >On Fri, Jun 04, 2004 at 08:30:22AM +1000, Keith Owens wrote: >> On Thu, 3 Jun 2004 11:51:45 -0500, >> [EMAIL PROTECTED] wrote: >> >Is anybody working on per_cpu variable printing for KDB? >> >I'm finding that looking up per-cpu values is tedious and >> >error prone, esp. on systems with more than a few cpu's. >> >> How about this? >> >> per_cpu variable_name length [cpu] >> >> If cpu is not specified, print the variable from all cpus. One line >> per cpu, in this format - "address cpu_number value". Zero values are >> suppressed, the last line is "zero suppressed: list of cpus". > >That works for me, are you planning on doing this? I was trying to set >up Ananth to volunteer to do this :) I'm wondering if there is some >way of making length implcit, although I admit I don't quite see how ..
I have already done it and mailed the patch to the kdb list, although oss.sgi.com is having problem right now. The length defaults to sizeof(machine register). The code will be included in the release of kdb that I plan for this weekend. --------------------------- Use http://oss.sgi.com/ecartis to modify your settings or to unsubscribe.
