netdev - Re: [PATCH 0/4] skb paged fragment destructors

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 21 Dec 2011 13:30:05 +0100
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Ian Campbell <Ian.Campbell@...rix.com>
Cc:	David Miller <davem@...emloft.net>,
	"jesse.brandeburg@...el.com" <jesse.brandeburg@...el.com>,
	"netdev@...r.kernel.org" <netdev@...r.kernel.org>
Subject: Re: [PATCH 0/4] skb paged fragment destructors

Le mercredi 21 décembre 2011 à 11:18 +0000, Ian Campbell a écrit :

> 
> An order 1 allocation is in multiples of PAGE_SIZE, isn't it, even
> though they happen to be contiguous?

Really an order-1 allocation allocates one page, a compound one.

> 
> Or are you considering the possibility of kmalloc returning a super page
> of some description? Isn't that to some extent transparent to the caller
> who (assuming PAGE_SIZE 4096) doesn't know if kmalloc(16384) returned
> 4*4096 contiguous pages or 1*16384 page?
> 

Its a part of a _single_ page, or the total of it.

Its mandatory, or else many drivers would break their DMA handling.

> Anyway I take your underlying point that 2*PAGE_SIZE is no kind of limit
> on the size of the linear region.
> 

Sure, since net/unix/af_unix.c can happily allocate huge linear skbs, if
memory is not too fragmented of course...

(The stream unix socket limit their skbs to SKB_MAX_ALLOC (up to 4
pages), but datagram ones have no limit)



--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html