于 2012年05月31日 16:32, Glauber Costa 写道: > On 05/31/2012 11:58 AM, Gao feng wrote: >>> > It's one of a option. But, I seriously doubt fuse can make simpler than kamezawa-san's >>> > idea. But yeah, I might NACK kamezawa-san's one if he will post ugly patch. >>> > >> It seams I should do some homework to make the implement beautifully. >> >> I think kamezawa-san's idea is more simpler. >> thanks for your advice. >> > > One think to keep in mind: A file in memcg does not need to follow the same format of /proc/meminfo so we can bind mount. We should be able to reconstruct that in userspace based on information > available from the kernel. You can even collect that from multiple locations, and *then* you bind mount. > > It helps to keep the churn out of the kernel, and in case of meminfo, you might need no extra kernel patches at all. And in the case of other files like /proc/stat, the relevant information comes from > more than one cgroup anyway, so there is not too much way around it. I got it,thank you very much,indeed we need no extra kernel patch at all. Maybe we should do this work in lxc or libvirt. thanks Glauber! -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href