Provide all necessary fields in struct xunpcb. Adjust netstat(1) and
sockstat(1) accordingly. Drop kvm(3) support from netstat(1) UNIX
printing.
In unp_pcblist() put single struct xunpcb on stack, to avoid extra
allocation.
Differential D11111
Hide struct unpcb under #ifdef _KERNEL. glebius on Jun 9 2017, 7:44 AM. Authored by Tags None Referenced Files
Details
Diff Detail
Event TimelineComment Actions if you drop support from netstat what's the alternative to get this information out of crashdumps then? Comment Actions Well a python script won't help me with a base system crashdump on boot if it's not installed. The other problem I think I generally have is that your description does not give me a reason on why it has to go? Comment Actions I had this dialogue with jhb@ before. To not repeat myself, please read these two emails. https://lists.freebsd.org/pipermail/svn-src-head/2015-April/070483.html The base system is no longer helpful with crashdumps at all, since there is no longer debugger in the base system. What are you going to do with crashinfo output on a bare base system? However, now that kgdb has python support, I will create scripts, that will restore functionality of crashinfo. Yes, I am absolutely sure that they won't be used by anyone, but looks like this functionality is sacred. Comment Actions I am very much with @jhb in this case. Having written DDB and netstat support to easily, quickly, and without the need to think debug problems without having to switch the tool I normally use for system administration is a big plus. Reading your original email and the historic behaviour of netstat using kvm for certain functionality by default, I agree with you, is less preferred and a proper API is good, but for things that poke /dev/mem (such as kvm/libkvm) or an equivalent crashdump there is no API. It's best effort as a gdb, lldb, or other tool of choice support would be. I would very much like not to lose the ability to just type the normal netstat command lines w/ -M -N. |