Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*Hardware\s+IP\s+checksums\s*$/: 12 ]

Total 12 documents matching your query.

1. Hardware IP checksums (score: 1)
Author: Andrew Morton <andrewm@xxxxxxxxxx>
Date: Mon, 24 Apr 2000 23:07:29 +1000
I spotted this discussion on the main kernel list: -- Original Message -- Subject: lockless poll() (was Re: namei() query) Date: Mon, 24 Apr 2000 21:36:00 +0900 From: kumon@xxxxxxxxxxxxxxxxxx Reply-T
/archives/netdev/2000-04/msg00064.html (7,815 bytes)

2. Re: Hardware IP checksums (score: 1)
Author: kuznet@xxxxxxxxxxxxx
Date: Mon, 24 Apr 2000 19:44:44 +0400 (MSK DST)
It appeared pretty useless. Jes noticed this first time, I had to check this experimentally, and yes... No visible improvements. HW checksum on transmit looks really not useful without zero copy. De
/archives/netdev/2000-04/msg00065.html (8,034 bytes)

3. Re: Hardware IP checksums (score: 1)
Author: Artur Skawina <skawina@xxxxxxxxxxxxx>
Date: Mon, 24 Apr 2000 19:12:34 +0200
for example on p2 the extra checksumming cost (vs a plain copy) is ~7%, and that's the worst case, ie everything cached. Add cache misses and the difference won't be visible...
/archives/netdev/2000-04/msg00066.html (7,536 bytes)

4. Re: Hardware IP checksums (score: 1)
Author: Andrew Morton <andrewm@xxxxxxxxxx>
Date: Tue, 25 Apr 2000 12:22:12 +1000
Ah. That's pretty convincing. But it would be a huge win for zero-copy Tx, as Alexey points out. I need to pay a bit more attention to what's happening on that front. Thanks. -- -akpm
/archives/netdev/2000-04/msg00074.html (8,016 bytes)

5. Re: Hardware IP checksums (score: 1)
Author: Artur Skawina <skawina@xxxxxxxxxxxxx>
Date: Tue, 25 Apr 2000 14:15:50 +0200
i didn't have the numbers in front of me and managed to remember the wrong figure :( Sorry. It really looks more like this: TIME-N+S TIME32 TIME33 TIME1480 TIMEXXXX FUNCTION 22109 9978 13303 18693 25
/archives/netdev/2000-04/msg00077.html (8,598 bytes)

6. Re: Hardware IP checksums (score: 1)
Author: kuznet@xxxxxxxxxxxxx
Date: Tue, 25 Apr 2000 20:24:51 +0400 (MSK DST)
I've said, I tried. On loopback improvement is invisible inside statistical bounds. On 100Mbit ethernet the difference does not affect throughput (it is saturated in any case), so that I looked at C
/archives/netdev/2000-04/msg00081.html (8,247 bytes)

7. Hardware IP checksums (score: 1)
Author: Andrew Morton <andrewm@xxxxxxxxxx>
Date: Mon, 24 Apr 2000 23:07:29 +1000
I spotted this discussion on the main kernel list: -- Original Message -- Subject: lockless poll() (was Re: namei() query) Date: Mon, 24 Apr 2000 21:36:00 +0900 From: kumon@xxxxxxxxxxxxxxxxxx Reply-T
/archives/netdev/2000-04/msg00184.html (7,815 bytes)

8. Re: Hardware IP checksums (score: 1)
Author: kuznet@xxxxxxxxxxxxx
Date: Mon, 24 Apr 2000 19:44:44 +0400 (MSK DST)
Hello! It appeared pretty useless. Jes noticed this first time, I had to check this experimentally, and yes... No visible improvements. HW checksum on transmit looks really not useful without zero co
/archives/netdev/2000-04/msg00185.html (8,059 bytes)

9. Re: Hardware IP checksums (score: 1)
Author: Artur Skawina <skawina@xxxxxxxxxxxxx>
Date: Mon, 24 Apr 2000 19:12:34 +0200
for example on p2 the extra checksumming cost (vs a plain copy) is ~7%, and that's the worst case, ie everything cached. Add cache misses and the difference won't be visible...
/archives/netdev/2000-04/msg00186.html (7,565 bytes)

10. Re: Hardware IP checksums (score: 1)
Author: Andrew Morton <andrewm@xxxxxxxxxx>
Date: Tue, 25 Apr 2000 12:22:12 +1000
Ah. That's pretty convincing. But it would be a huge win for zero-copy Tx, as Alexey points out. I need to pay a bit more attention to what's happening on that front. Thanks. -- -akpm
/archives/netdev/2000-04/msg00194.html (8,070 bytes)

11. Re: Hardware IP checksums (score: 1)
Author: Artur Skawina <skawina@xxxxxxxxxxxxx>
Date: Tue, 25 Apr 2000 14:15:50 +0200
i didn't have the numbers in front of me and managed to remember the wrong figure :( Sorry. It really looks more like this: TIME-N+S TIME32 TIME33 TIME1480 TIMEXXXX FUNCTION 22109 9978 13303 18693 25
/archives/netdev/2000-04/msg00197.html (8,677 bytes)

12. Re: Hardware IP checksums (score: 1)
Author: kuznet@xxxxxxxxxxxxx
Date: Tue, 25 Apr 2000 20:24:51 +0400 (MSK DST)
Hello! I've said, I tried. On loopback improvement is invisible inside statistical bounds. On 100Mbit ethernet the difference does not affect throughput (it is saturated in any case), so that I looke
/archives/netdev/2000-04/msg00201.html (8,272 bytes)


This search system is powered by Namazu