re: lwp resource limit

[christos%zoulas.com@localhost (Christos Zoulas)]

On Jun 5,  6:14am, mrg%eterna.com.au@localhost (matthew green) wrote:
-- Subject: re: lwp resource limit

| 
| > >1024 seems to small for me.  How about 2048 (if not 4096)?  Please make
| > >it a #define somewhere (perhaps lwp.h?).  When would the MD code override
| > >the value?
| > 
| > I just followed the pattern with maxproc. I will bump the limit, but I
| > think that putting in the header is a mistake because things might use
| > the constant thinking that it always represents reality.
| 
| make it a #define in kern_lwp.c itself, that can be pre-defined.

will do.

| > >For allowing process creation when maxlwp limit is exhausted, I would
| > >say this problem is better solved with maxprocperuid you mentioned.
| > 
| > Yes, perhaps. I think that the process/lwp limits need to be re-thought.
| 
| i would say that the best way to handle this would be to not count
| the first lwp in any process as part of the lwp rlimit.

That is a good idea! The only problem with it is that ps -sx and the sysctl
limit will not match (the limit will be lower than the sysctl number by the
number of processes). It does have the nice property though that the process
gets created anyway, as it is now (counting the limit).

It is a very simple change, and if people prefer it, I will do it.

christos