| To: | Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx> |
|---|---|
| Subject: | Re: [PATCH 3/12 2.4] e1000 - use vmalloc for data structures not shared with h/w |
| From: | Christoph Hellwig <hch@xxxxxxxxxxxxx> |
| Date: | Wed, 4 Aug 2004 17:48:56 +0100 |
| Cc: | Christoph Hellwig <hch@xxxxxxxxxxxxx>, ganesh.venkatesan@xxxxxxxxx, jgarzik@xxxxxxxxx, netdev@xxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx |
| In-reply-to: | <E1BqoRX-0004DH-00@gondolin.me.apana.org.au>; from herbert@gondor.apana.org.au on Sat, Jul 31, 2004 at 05:38:11PM +1000 |
| Mail-followup-to: | Christoph Hellwig <hch@xxxxxxxxxxxxx>, Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>, ganesh.venkatesan@xxxxxxxxx, jgarzik@xxxxxxxxx, netdev@xxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx |
| References: | <20040729192519.A6235@infradead.org> <E1BqoRX-0004DH-00@gondolin.me.apana.org.au> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.2.5.1i |
On Sat, Jul 31, 2004 at 05:38:11PM +1000, Herbert Xu wrote: > > No, it's not. vmalloc needs virtual space that's rather limited (e.g. 64MB > > on PAE x86) in addition to physical memory. Unless you do really big > > allocations stay away from vmalloc. > > How big is really big? 64K? 256K? 1M? Well, the VM deals with big-order (aka bigger than page size) allocations rather bad, so for allocation during any I/O I'd stick to allocation smaller than that (and certainly no vmalloc!), for init-time allocations order 1 is fine, maybe even order 2 or three. |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH 2.6 5/5]: remove noop_qdisc assignments in destroy functions, David S. Miller |
|---|---|
| Next by Date: | Re: [PATCH 2.6] cache align qdisc data, David S. Miller |
| Previous by Thread: | Inaccuracies with SCH_CLK_CPU, Stephen Hemminger |
| Next by Thread: | [PATCH] ibmtr: fix init (devinit) section usage, Randy.Dunlap |
| Indexes: | [Date] [Thread] [Top] [All Lists] |