Mail Archive Search: <+subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*\[patch\]\s+TCP\s+throughput\s+after\s+2\.2\.17\-pre1\s*$/>

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*\[patch\]\s+TCP\s+throughput\s+after\s+2\.2\.17\-pre1\s*$/: 16 ]

Total 16 documents matching your query.

1. [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxx>; Date: Sun, 18 Jun 2000 23:35:10 +1000; With 2.2.17-pre1 on a 400MHz uniprocessor it is possible to saturate a 100baseT with netperf (just a TCP transmitter) with 40% CPU left over. With -pre3 and -pre4 it maxes out at 38 Mbits/sec. A five; /archives/netdev/2000-06/msg00201.html (13,192 bytes)
2. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxx>; Date: Sun, 18 Jun 2000 07:26:13 -0700; The problem is caused by this patch: The change aparently came from Alexey to cure some deadlock issues. I just noticed the change today, and I also am skeptical about it's correctness. It I understa; /archives/netdev/2000-06/msg00202.html (8,601 bytes)
3. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxx>; Date: Sun, 18 Jun 2000 10:47:01 -0400 (EDT); Its from me. Its critical to fix the problem Its fixed a huge range of hangs especially on SMP boxes. Its staying until the proper fix is done. That probably needs a sock_wmalloc_err() and also the; /archives/netdev/2000-06/msg00203.html (8,621 bytes)
4. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxx>; Date: Sun, 18 Jun 2000 20:09:52 +0400 (MSK DST); The fix, which is in 2.3 now, should be OK. Seems, it is easy to copy it to 2.2. I can do it. Alexey; /archives/netdev/2000-06/msg00204.html (7,991 bytes)
5. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxxx; Date: Sun, 18 Jun 2000 18:16:37 +0200; Your patch is unfortunately not the right solution. It goes back to the old state and doesn't fix the original bug (tcp spinning in case of real oom). Also you cannot just use tcp_* in generic code a; /archives/netdev/2000-06/msg00205.html (15,604 bytes)
6. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxx>; Date: Sun, 18 Jun 2000 22:43:53 -0400 (EDT); isn't it.. interruptable (or is my english weakening)? -Statux; /archives/netdev/2000-06/msg00206.html (7,964 bytes)
7. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxx>; Date: Mon, 19 Jun 2000 20:58:29 +1000; Andi's patch works for me. I've attached the 2.2.17-pre4 version here. He didn't appear to do that bit. It just polls. The sleep_on() would be good to have; wait_for_tcp_memory() will be called quite; /archives/netdev/2000-06/msg00208.html (15,508 bytes)
8. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: xxxx>; Date: Mon, 19 Jun 2000 13:33:25 +0200; It does actually. The schedule_timeout(1) will be woken earlier with the kfree_skb wakeup. -Andi; /archives/netdev/2000-06/msg00209.html (8,559 bytes)
9. [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: Andrew Morton <andrewm@xxxxxxxxxx>; Date: Sun, 18 Jun 2000 23:35:10 +1000; With 2.2.17-pre1 on a 400MHz uniprocessor it is possible to saturate a 100baseT with netperf (just a TCP transmitter) with 40% CPU left over. With -pre3 and -pre4 it maxes out at 38 Mbits/sec. A five; /archives/netdev/2000-06/msg00500.html (13,192 bytes)
10. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: "David S. Miller" <davem@xxxxxxxxxx>; Date: Sun, 18 Jun 2000 07:26:13 -0700; The problem is caused by this patch: The change aparently came from Alexey to cure some deadlock issues. I just noticed the change today, and I also am skeptical about it's correctness. It I understa; /archives/netdev/2000-06/msg00501.html (8,651 bytes)
11. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: Alan Cox <alan@xxxxxxxxxx>; Date: Sun, 18 Jun 2000 10:47:01 -0400 (EDT); Its from me. Its critical to fix the problem Its fixed a huge range of hangs especially on SMP boxes. Its staying until the proper fix is done. That probably needs a sock_wmalloc_err() and also the; /archives/netdev/2000-06/msg00502.html (8,650 bytes)
12. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: kuznet@xxxxxxxxxxxxx; Date: Sun, 18 Jun 2000 20:09:52 +0400 (MSK DST); Hello! The fix, which is in 2.3 now, should be OK. Seems, it is easy to copy it to 2.2. I can do it. Alexey; /archives/netdev/2000-06/msg00503.html (8,020 bytes)
13. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: Andi Kleen <ak@xxxxxx>; Date: Sun, 18 Jun 2000 18:16:37 +0200; Your patch is unfortunately not the right solution. It goes back to the old state and doesn't fix the original bug (tcp spinning in case of real oom). Also you cannot just use tcp_* in generic code a; /archives/netdev/2000-06/msg00504.html (15,654 bytes)
14. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: Statux <statux@xxxxxxxxxxx>; Date: Sun, 18 Jun 2000 22:43:53 -0400 (EDT); isn't it.. interruptable (or is my english weakening)? -Statux; /archives/netdev/2000-06/msg00505.html (7,989 bytes)
15. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: Andrew Morton <andrewm@xxxxxxxxxx>; Date: Mon, 19 Jun 2000 20:58:29 +1000; Andi's patch works for me. I've attached the 2.2.17-pre4 version here. He didn't appear to do that bit. It just polls. The sleep_on() would be good to have; wait_for_tcp_memory() will be called quite; /archives/netdev/2000-06/msg00507.html (15,566 bytes)
16. Re: [patch] TCP throughput after 2.2.17-pre1 (score: 1): Author: Andi Kleen <ak@xxxxxx>; Date: Mon, 19 Jun 2000 13:33:25 +0200; It does actually. The schedule_timeout(1) will be woken earlier with the kfree_skb wakeup. -Andi; /archives/netdev/2000-06/msg00508.html (8,609 bytes)

Current List: 1 - 16
Page: [1]