Re: [PATCH] Improve scalability of epoll_ctl

On Jan 9, 2008 12:56 AM, Andi Kleen <[EMAIL PROTECTED]> wrote: > > Are you sure it's faster? vmalloc/vfree can be quite slow because > they have to change the MMU state of the kernel and take global > locks. I don't think so. In fact, the memory of fdtable is allocated by vmalloc when its size is b

Re: [PATCH] Improve scalability of epoll_ctl

On Tue, 8 Jan 2008, Eric Dumazet wrote: > Changli Gao a écrit : > > Replace the epitem rbtree with a dynamic array to get the constant > > insertion/deletion/modification time of the file descriptors. Reuse the > > size argument of epoll_create, however the size must be smaller than the > > max

Re: [PATCH] Improve scalability of epoll_ctl

Changli Gao <[EMAIL PROTECTED]> writes: > Replace the epitem rbtree with a dynamic array to get the constant > insertion/deletion/modification time of the file descriptors. Reuse the size > argument of epoll_create, however the size must be smaller than the max file > descriptor number: ether t

Re: [PATCH] Improve scalability of epoll_ctl

Changli Gao a écrit : > Replace the epitem rbtree with a dynamic array to get the constant > insertion/deletion/modification time of the file descriptors. Reuse the size > argument of epoll_create, however the size must be smaller than the max file > descriptor number: ether the resource limitat

[PATCH] Improve scalability of epoll_ctl

Replace the epitem rbtree with a dynamic array to get the constant insertion/deletion/modification time of the file descriptors. Reuse the size argument of epoll_create, however the size must be smaller than the max file descriptor number: ether the resource limitation or the compiling time lim