| Subject: | Re: Extend "struct malloc_type" for extended KMEMSTATS? |
|---|---|
| From: | christos@xxxxxxxxxx (Christos Zoulas) |
| Date: | Fri, 18 Jul 2008 13:47:48 UTC |
| Newsgroups: | fa.netbsd.tech.kern |
In article <20080718.110021.18308698.he@xxxxxxxxxx>, Havard Eidnes <he@xxxxxxxxxx> wrote: >-=-=-=-=-=- > >> while trying to get more info on the suspected kernel malloc leak >> documented in PR#39145, I found it useful to see how many >> allocations of each size each malloc type is currently holding. > >I got some private feedback, and a new diff is attached below. >Changes since the previous one: > > o Many of the fields in struct malloc_type ought to be size_t > instead of u_long. A few printf formats had to be adjusted to > correspond. > o The "spare" field is deleted. > o The counters per block size are u_int instead of u_long, > saving some bytes on lp64 platforms (and again printf formats > correspondingly adjusted). > >I've verified that this still builds correctly for a large number >of ports. > >Oh, yes, I also got a comment that a kernel version bump would be >required for this as well. > >Further comments? %zd's should be %zu's. christos |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: rfc: high-resolution timer framework, Alexander Shishkin |
|---|---|
| Next by Date: | __dead functions, Alexander Shishkin |
| Previous by Thread: | Re: Extend "struct malloc_type" for extended KMEMSTATS?, David Laight |
| Next by Thread: | wm @ 82571EB not working?, Jukka Marin |
| Indexes: | [Date] [Thread] [Top] [All Lists] |