search for: rcv

Mostly out of curiosity: I've seen people advising that we set SO_{RCV,SND}BUF to 8192 as an optimization. I added a call to getsockopt() to grab the old (default) value before setting it to the new, and it turns out (on DG/UX), that the default value is 8760 (or 8704, when you throw in the fact that packets are travelling between different sublans), so the 8192 &quo...

...tuation which I do not understand. I have two directories with few maildirs in it. Each maildir containg some emails. I'm testing scripts which I wrote to manipulate those emails. Each time I want to revert changes in one set of maildirs I've rsync'ing them from backup copy: $ rsync -rcv --delete backup/ mail/ Nothing special about that. However my scripts are modyfing only mail/INBOX maildir, nothing else. To my surprize when I ran rsync command to revert my changes here is what I got (I will use -n in below examples for now): $ rsync -v -n -rc --delete backup/ mail/ sending inc...

Dear LIST: I'm working on SAMBA performance enhancement. Can anyone guide me for a reference about the 4 WATer mark socket options and their effects on TCP/IP performance? Have anyone any experiments about using them in the SAMBA code? -- THX in advance, M. Tajamolian Jun 8

Hi. Between 2 linux (2.6.11 client and 2.6.14 server) machines connected by a 100Mb link I get samba performance copying a file from the client to the server through a smbmounted share of around 4.2MB/s Is this to be expected? Or can it be improved (and if so, how?) I've tried tweaking SO_(SND/RCV)BUF (after reading numerous articles on samba performance...), but it doesn't seem to have any effect. Unfortunately, I can't use jumbo frames at this moment either. I've tried this with shares on ReiserFS and XFS and there is no difference in the transfer rate. Both machines are not e...

...Dec 28, 2016 at 04:09:31PM +0800, Jason Wang wrote: > +static int tun_rx_batched(struct tun_file *tfile, struct sk_buff *skb, > + int more) > +{ > + struct sk_buff_head *queue = &tfile->sk.sk_write_queue; > + struct sk_buff_head process_queue; > + int qlen; > + bool rcv = false; > + > + spin_lock(&queue->lock); Should this be spin_lock_bh()? Below and in tun_get_user() there are explicit local_bh_disable() calls so I guess BHs can interrupt us here and this would deadlock. -------------- next part -------------- A non-text attachment was scrubbed......

...Dec 28, 2016 at 04:09:31PM +0800, Jason Wang wrote: > +static int tun_rx_batched(struct tun_file *tfile, struct sk_buff *skb, > + int more) > +{ > + struct sk_buff_head *queue = &tfile->sk.sk_write_queue; > + struct sk_buff_head process_queue; > + int qlen; > + bool rcv = false; > + > + spin_lock(&queue->lock); Should this be spin_lock_bh()? Below and in tun_get_user() there are explicit local_bh_disable() calls so I guess BHs can interrupt us here and this would deadlock. -------------- next part -------------- A non-text attachment was scrubbed......

Forwarded message: > From full-disclosure-admin@lists.netsys.com Wed Apr 21 11:49:12 2004 > To: full-disclosure@lists.netsys.com > From: Darren Bounds <dbounds@intrusense.com> > Subject: [Full-Disclosure] IETF Draft - Fix for TCP vulnerability > Date: Tue, 20 Apr 2004 18:19:58 -0400 > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > >

...36 @@ static struct sk_buff *tun_alloc_skb(struct tun_file *tfile, return skb; } +static void tun_rx_batched(struct tun_file *tfile, struct sk_buff *skb, + int more) +{ + struct sk_buff_head *queue = &tfile->sk.sk_write_queue; + struct sk_buff_head process_queue; + int qlen; + bool rcv = false; + + spin_lock(&queue->lock); + qlen = skb_queue_len(queue); + __skb_queue_tail(queue, skb); + if (!more || qlen == rx_batched) { + __skb_queue_head_init(&process_queue); + skb_queue_splice_tail_init(queue, &process_queue); + rcv = true; + } + spin_unlock(&queue->lo...

...5,44 @@ static struct sk_buff *tun_alloc_skb(struct tun_file *tfile, return skb; } +static int tun_rx_batched(struct tun_file *tfile, struct sk_buff *skb, + int more) +{ + struct sk_buff_head *queue = &tfile->sk.sk_write_queue; + struct sk_buff_head process_queue; + int qlen; + bool rcv = false; + + spin_lock(&queue->lock); + qlen = skb_queue_len(queue); + if (qlen > rx_batched) + goto drop; + __skb_queue_tail(queue, skb); + if (!more || qlen + 1 > rx_batched) { + __skb_queue_head_init(&process_queue); + skb_queue_splice_tail_init(queue, &process_queue); +...

...lia according to quintum are : CD-Bits ::::::::::::::::::::::: 0001 Invert-Bits ::::::::::::::::::: 0000 DNIS Length ::::::::::::::::::: 9 digits * Answer Tone ::::::::::::::::::: A-6 Send 1st Digit :::::::::::::::: 1 Group B Xmt Idle Tone ::::::::: B-6 Group B Xmt Busy Tone ::::::::: B-3 Group B Rcv Idle Tones :::::::: B-2 & B-3 Group B Rcv Busy Tones :::::::: B-1 & B-2 ANI Request ::::::::::::::::::: Do not request ANI ANI Length ANI Category Request Tone ANI Tone Request ANI Category :::::::::::::::::: I-1 ANI Calling Party Category :::: II-1 Seizure Ack Timeout ::::::::::: 150ms Rel...

...ems to be good. However, since he restarted the network, I do find this in the logs: Dec 10 16:06:01 <server> dhclient[96610]: PRC: Renewing lease on ens3f0. Dec 10 16:06:01 <server> dhclient[96610]: XMT: Renew on ens3f0, interval 10900ms. Dec 10 16:06:01 <server> dhclient[96610]: RCV: Reply message on ens3f0 from fe80::: 222:64ff:fef6:d30c. Dec 10 16:06:01 <server> dhclient[96610]: IA_NA status code NoBinding: "Who are you? Do I know you?" Dec 10 16:06:01 <server> dhclient[96610]: XMT: Request on ens3f0, interval 910ms. Dec 10 16:06:01 <server> dhcli...

Hi: This series tries to implement tx batching support for vhost. This was done by using MSG_MORE as a hint for under layer socket. The backend (e.g tap) can then batch the packets temporarily in a list and submit it all once the number of bacthed exceeds a limitation. Tests shows obvious improvement on guest pktgen over over mlx4(noqueue) on host: Mpps -+%

Hi: This series tries to implement tx batching support for vhost. This was done by using MSG_MORE as a hint for under layer socket. The backend (e.g tap) can then batch the packets temporarily in a list and submit it all once the number of bacthed exceeds a limitation. Tests shows obvious improvement on guest pktgen over over mlx4(noqueue) on host: Mpps -+%

...un_file *tfile, return skb; } +static void tun_rx_batched(struct tun_struct *tun, struct tun_file *tfile, + struct sk_buff *skb, int more) +{ + struct sk_buff_head *queue = &tfile->sk.sk_write_queue; + struct sk_buff_head process_queue; + u32 rx_batched = tun->rx_batched; + bool rcv = false; + + if (!rx_batched || (!more && skb_queue_empty(queue))) { + local_bh_disable(); + netif_receive_skb(skb); + local_bh_enable(); + return; + } + + spin_lock(&queue->lock); + if (!more || skb_queue_len(queue) == rx_batched) { + __skb_queue_head_init(&process_queue);...

...un_file *tfile, return skb; } +static void tun_rx_batched(struct tun_struct *tun, struct tun_file *tfile, + struct sk_buff *skb, int more) +{ + struct sk_buff_head *queue = &tfile->sk.sk_write_queue; + struct sk_buff_head process_queue; + u32 rx_batched = tun->rx_batched; + bool rcv = false; + + if (!rx_batched || (!more && skb_queue_empty(queue))) { + local_bh_disable(); + netif_receive_skb(skb); + local_bh_enable(); + return; + } + + spin_lock(&queue->lock); + if (!more || skb_queue_len(queue) == rx_batched) { + __skb_queue_head_init(&process_queue);...

Hi: This series tries to implement tx batching support for vhost. This was done by using MSG_MORE as a hint for under layer socket. The backend (e.g tap) can then batch the packets temporarily in a list and submit it all once the number of bacthed exceeds a limitation. Tests shows obvious improvement on guest pktgen over over mlx4(noqueue) on host: Mpps -+%

Hi: This series tries to implement tx batching support for vhost. This was done by using MSG_MORE as a hint for under layer socket. The backend (e.g tap) can then batch the packets temporarily in a list and submit it all once the number of bacthed exceeds a limitation. Tests shows obvious improvement on guest pktgen over over mlx4(noqueue) on host: Mpps -+%

Second version, all feedback so far addressed, thanks for the help and interest! The patch to iproute2 has not changed, so I'm not including it this time. Patch 4/4 (the netlink interface) is basically unchanged as well but included for completeness. The other changes have moved forward a bit, to the point where I find them a lot cleaner and am more confident in the code being ready for

Second version, all feedback so far addressed, thanks for the help and interest! The patch to iproute2 has not changed, so I'm not including it this time. Patch 4/4 (the netlink interface) is basically unchanged as well but included for completeness. The other changes have moved forward a bit, to the point where I find them a lot cleaner and am more confident in the code being ready for

Second version, all feedback so far addressed, thanks for the help and interest! The patch to iproute2 has not changed, so I'm not including it this time. Patch 4/4 (the netlink interface) is basically unchanged as well but included for completeness. The other changes have moved forward a bit, to the point where I find them a lot cleaner and am more confident in the code being ready for