Re: [PATCH] Enhance AF_PACKET implementation to not require high order contiguous memory allocation

[Eric Dumazet <eric.dumazet@gmail.com>]

Le lundi 25 octobre 2010 à 15:06 -0400, nhorman@tuxdriver.com a écrit :
> From: Neil Horman <nhorman@tuxdriver.com>
> 
> It was shown to me recently that systems under high load were driven very deep
> into swap when tcpdump was run.  The reason this happened was because the
> AF_PACKET protocol has a SET_RINGBUFFER socket option that allows the user space
> application to specify how many entries an AF_PACKET socket will have and how
> large each entry will be.  It seems the default setting for tcpdump is to set
> the ring buffer to 32 entries of 64 Kb each, which implies 32 order 5
> allocation.  Thats difficult under good circumstances, and horrid under memory
> pressure.
> 
> I thought it would be good to make that a bit more usable.  I was going to do a
> simple conversion of the ring buffer from contigous pages to iovecs, but
> unfortunately, the metadata which AF_PACKET places in these buffers can easily
> span a page boundary, and given that these buffers get mapped into user space,
> and the data layout doesn't easily allow for a change to padding between frames
> to avoid that, a simple iovec change is just going to break user space ABI
> consistency.
> 
> So instead I've done this.  This patch does the aforementioned change,
> allocating an array of pages instead of one contiguous chunk, and then vmaps the
> array into a contiguous memory space, so that it can still be accessed in the
> same way it was before.  This allows for a consisten user and kernel space
> behavior for memory mapped AF_PACKET sockets, which at the same time relieving
> the memory pressure placed on a system when tcpdump defaults are used.
> 
> Tested successfully by me.
> 
> Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
> ---

Strange because last time I took a look at this stuff, libpcap was doing
several tries, reducing page orders until it got no allocation
failures...

(It tries to get high order pages, maybe to reduce TLB pressure...)

I remember adding __GFP_NOWARN to avoid a kernel message, while tcpdump
was actually working...