Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*Q\:\s+\(ab\)using\s+zerocopy\s+for\s+drivers\s+with\s+alignment\s+contraints\s*$/: 18 ]

Total 18 documents matching your query.

1. Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Date: Sun, 10 Jun 2001 16:58:57 +0200
Several cheap busmaster nics only accept tx buffers that are 32-bit Currently they memcpy into transfer buffers. What about replacing that memcpy with csum_copy_partial_nocheck and enabling NETIF_F_{
/archives/netdev/2001-06/msg00066.html (11,573 bytes)

2. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Jeff Garzik <jgarzik@xxxxxxxxxxxxxxxx>
Date: Sun, 10 Jun 2001 12:56:15 -0400
I'm definitely for reducing copies, so this patch looks nice. I would prefer to call it "single-copy" or something other than zero-copy, though... -- Jeff Garzik | Andre the Giant has a posse. Buildi
/archives/netdev/2001-06/msg00068.html (9,030 bytes)

3. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Sun, 10 Jun 2001 17:48:34 -0700 (PDT)
I'm going to assume that it is safe to bet that such cards cannot take multiple buffers for a TX packet too. Because if they could, then we could do something like copy the header forward a few bytes
/archives/netdev/2001-06/msg00071.html (9,117 bytes)

4. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Sun, 10 Jun 2001 17:50:15 -0700 (PDT)
Call it skb_copy_and_csum_iovec_kernel() which is what it is :-) Later, David S. Miller davem@xxxxxxxxxx
/archives/netdev/2001-06/msg00072.html (8,848 bytes)

5. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Donald Becker <becker@xxxxxxxxx>
Date: Sun, 10 Jun 2001 21:13:30 -0400 (EDT)
Incorrect assumption. If it were that easy, don't you think the device driver writer would have used the same technique to avoid the bulk copy? There wouldn't be an alignment requirement if it were n
/archives/netdev/2001-06/msg00074.html (9,660 bytes)

6. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Date: Mon, 11 Jun 2001 18:46:57 +0200
skb_copy_datagram & friends follow the fragment list. My function doesn't/mustn't follow skb_shinfo(skb)->frag_list. Should I still call it skb_copy_datagram{,_iovec}_kernel? I don't like functions w
/archives/netdev/2001-06/msg00086.html (9,792 bytes)

7. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 11 Jun 2001 20:14:41 -0700 (PDT)
Why "mustn't it" follow the frag list? I think it would be "absolutely fantastic" if it did follow the frag list! Then we could optimize the forwarding of fragmented packets. There is no subtle diffe
/archives/netdev/2001-06/msg00087.html (9,465 bytes)

8. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Andi Kleen <ak@xxxxxx>
Date: Wed, 13 Jun 2001 03:19:29 +0200
[...] I'll probably not give you much gain in 2.4 anymore. Both TCP and UDP do csum and copy to user in most fast path cases. -Andi
/archives/netdev/2001-06/msg00094.html (8,814 bytes)

9. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Date: Wed, 13 Jun 2001 08:28:45 +0200
^^^^^^^^^^^^ It's an improvement for the tx codepath: If an application uses sendfile with an 8139too [or via-rhine,...] nic then currently 2 copies are made: 1) copy_and_csum into skb->data 2) memco
/archives/netdev/2001-06/msg00099.html (9,274 bytes)

10. Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Date: Sun, 10 Jun 2001 16:58:57 +0200
Several cheap busmaster nics only accept tx buffers that are 32-bit aligned. Currently they memcpy into transfer buffers. What about replacing that memcpy with csum_copy_partial_nocheck and enabling
/archives/netdev/2001-06/msg00256.html (11,573 bytes)

11. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Jeff Garzik <jgarzik@xxxxxxxxxxxxxxxx>
Date: Sun, 10 Jun 2001 12:56:15 -0400
I'm definitely for reducing copies, so this patch looks nice. I would prefer to call it "single-copy" or something other than zero-copy, though... -- Jeff Garzik | Andre the Giant has a posse. Buildi
/archives/netdev/2001-06/msg00258.html (9,055 bytes)

12. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Sun, 10 Jun 2001 17:48:34 -0700 (PDT)
I'm going to assume that it is safe to bet that such cards cannot take multiple buffers for a TX packet too. Because if they could, then we could do something like copy the header forward a few bytes
/archives/netdev/2001-06/msg00261.html (9,167 bytes)

13. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Sun, 10 Jun 2001 17:50:15 -0700 (PDT)
Call it skb_copy_and_csum_iovec_kernel() which is what it is :-) Later, David S. Miller davem@xxxxxxxxxx
/archives/netdev/2001-06/msg00262.html (8,923 bytes)

14. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Donald Becker <becker@xxxxxxxxx>
Date: Sun, 10 Jun 2001 21:13:30 -0400 (EDT)
Incorrect assumption. If it were that easy, don't you think the device driver writer would have used the same technique to avoid the bulk copy? There wouldn't be an alignment requirement if it were n
/archives/netdev/2001-06/msg00264.html (9,707 bytes)

15. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Date: Mon, 11 Jun 2001 18:46:57 +0200
skb_copy_datagram & friends follow the fragment list. My function doesn't/mustn't follow skb_shinfo(skb)->frag_list. Should I still call it skb_copy_datagram{,_iovec}_kernel? I don't like functions w
/archives/netdev/2001-06/msg00276.html (9,849 bytes)

16. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxx>
Date: Mon, 11 Jun 2001 20:14:41 -0700 (PDT)
Why "mustn't it" follow the frag list? I think it would be "absolutely fantastic" if it did follow the frag list! Then we could optimize the forwarding of fragmented packets. There is no subtle diffe
/archives/netdev/2001-06/msg00277.html (9,572 bytes)

17. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Andi Kleen <ak@xxxxxx>
Date: Wed, 13 Jun 2001 03:19:29 +0200
[...] I'll probably not give you much gain in 2.4 anymore. Both TCP and UDP do csum and copy to user in most fast path cases. -Andi
/archives/netdev/2001-06/msg00284.html (8,907 bytes)

18. Re: Q: (ab)using zerocopy for drivers with alignment contraints (score: 1)
Author: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Date: Wed, 13 Jun 2001 08:28:45 +0200
^^^^^^^^^^^^ It's an improvement for the tx codepath: If an application uses sendfile with an 8139too [or via-rhine,...] nic then currently 2 copies are made: 1) copy_and_csum into skb->data 2) memco
/archives/netdev/2001-06/msg00289.html (9,327 bytes)


This search system is powered by Namazu