Re: Kernel Event Notifications (was: [RFC] Parallelize IO for e2fsck)
Hi Jon > I looked at this a year or two back, then ran out of time. But the thing > I wanted to do was have libc's memory allocation routines extended to > handle these through reservations - the kernel should send a userspace > notification and then there should be some kind of concept of returning > memory that's been used for "opportunistic" userspace caching, e.g. in > firefox to cache the last 10 web pages. Let us know how you get on :) sorry for late response. (I didn't notice your mail ;-) You are right... stupid user space caching is very important problem. but I think this is no libc problem. glibc malloc hardly caches the memory. (its default behavior only caching 128K.) but some application use large memory for too opportunistic caching. I understood we need propagandize that using mem_notify to application guys after it merge mainline. I have no idea of solve it easily. - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Kernel Event Notifications (was: [RFC] Parallelize IO for e2fsck)
On Sat, 2008-01-26 at 16:55 +0300, Al Boldi wrote: > KOSAKI Motohiro wrote: > > > > And from a performance point of view letting applications voluntarily > > > > free some memory is better even than starting to swap. > > > > > > Absolutely. > > > > the mem_notify patch can realize "just before starting swapping" > > notification :) I looked at this a year or two back, then ran out of time. But the thing I wanted to do was have libc's memory allocation routines extended to handle these through reservations - the kernel should send a userspace notification and then there should be some kind of concept of returning memory that's been used for "opportunistic" userspace caching, e.g. in firefox to cache the last 10 web pages. Let us know how you get on :) Jon. - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: Kernel Event Notifications (was: [RFC] Parallelize IO for e2fsck)
Hi Al > > the mem_notify patch can realize "just before starting swapping" > > notification :) > > > > to be honest, I don't know fs guys requirement. > > if lacking feature of fs guys needed, I implement it with presure if > > you tell me it. > > These notifications are really useful, but it may be much wiser to pipe them > thru some kevent-notification sub-system, instead of introducing kernel > notifier-chain end-points left, right, and center. Aaahh Your feelings are understood well. but current design is decided through many poeple discussion. if anybody need kevent notification, I will add it to the current implementation instead replace. thanks. - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
Kernel Event Notifications (was: [RFC] Parallelize IO for e2fsck)
KOSAKI Motohiro wrote: > > > And from a performance point of view letting applications voluntarily > > > free some memory is better even than starting to swap. > > > > Absolutely. > > the mem_notify patch can realize "just before starting swapping" > notification :) > > to be honest, I don't know fs guys requirement. > if lacking feature of fs guys needed, I implement it with presure if > you tell me it. These notifications are really useful, but it may be much wiser to pipe them thru some kevent-notification sub-system, instead of introducing kernel notifier-chain end-points left, right, and center. Thanks! -- Al - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html