From: Jiri Pirko <jiri@resnulli.us>
To: Eric Dumazet <edumazet@google.com>
Cc: "David S . Miller" <davem@davemloft.net>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
David Ahern <dsahern@kernel.org>,
netdev@vger.kernel.org, eric.dumazet@gmail.com
Subject: Re: [PATCH net-next 13/13] ipv6: use xa_array iterator to implement inet6_netconf_dump_devconf()
Date: Mon, 26 Feb 2024 17:55:47 +0100 [thread overview]
Message-ID: <ZdzCkxLBi_JybU25@nanopsycho> (raw)
In-Reply-To: <20240226155055.1141336-14-edumazet@google.com>
Mon, Feb 26, 2024 at 04:50:55PM CET, edumazet@google.com wrote:
>1) inet6_netconf_dump_devconf() can run under RCU protection
> instead of RTNL.
>
>2) properly return 0 at the end of a dump, avoiding an
> an extra recvmsg() system call.
>
>3) Do not use inet6_base_seq() anymore, for_each_netdev_dump()
> has nice properties. Restarting a GETDEVCONF dump if a device has
> been added/removed or if net->ipv6.dev_addr_genid has changed is moot.
>
>Signed-off-by: Eric Dumazet <edumazet@google.com>
>---
> net/ipv6/addrconf.c | 103 +++++++++++++++++++-------------------------
> 1 file changed, 44 insertions(+), 59 deletions(-)
>
>diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
>index 74c33b132073934290632a953bce8ee6a5124ca9..08b4728733e3ed16d139d2bd4b50328552b3c27f 100644
>--- a/net/ipv6/addrconf.c
>+++ b/net/ipv6/addrconf.c
>@@ -727,17 +727,18 @@ static u32 inet6_base_seq(const struct net *net)
> return res;
> }
>
>-
> static int inet6_netconf_dump_devconf(struct sk_buff *skb,
> struct netlink_callback *cb)
> {
> const struct nlmsghdr *nlh = cb->nlh;
> struct net *net = sock_net(skb->sk);
>- int h, s_h;
>- int idx, s_idx;
>+ struct {
>+ unsigned long ifindex;
>+ unsigned int all_default;
>+ } *ctx = (void *)cb->ctx;
> struct net_device *dev;
> struct inet6_dev *idev;
>- struct hlist_head *head;
>+ int err = 0;
>
> if (cb->strict_check) {
> struct netlink_ext_ack *extack = cb->extack;
>@@ -754,64 +755,47 @@ static int inet6_netconf_dump_devconf(struct sk_buff *skb,
> }
> }
>
>- s_h = cb->args[0];
>- s_idx = idx = cb->args[1];
>-
>- for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
>- idx = 0;
>- head = &net->dev_index_head[h];
>- rcu_read_lock();
>- cb->seq = inet6_base_seq(net);
>- hlist_for_each_entry_rcu(dev, head, index_hlist) {
>- if (idx < s_idx)
>- goto cont;
>- idev = __in6_dev_get(dev);
>- if (!idev)
>- goto cont;
>-
>- if (inet6_netconf_fill_devconf(skb, dev->ifindex,
>- &idev->cnf,
>- NETLINK_CB(cb->skb).portid,
>- nlh->nlmsg_seq,
>- RTM_NEWNETCONF,
>- NLM_F_MULTI,
>- NETCONFA_ALL) < 0) {
>- rcu_read_unlock();
>- goto done;
>- }
>- nl_dump_check_consistent(cb, nlmsg_hdr(skb));
>-cont:
>- idx++;
>- }
>- rcu_read_unlock();
>+ rcu_read_lock();
>+ for_each_netdev_dump(net, dev, ctx->ifindex) {
>+ idev = __in6_dev_get(dev);
>+ if (!idev)
>+ continue;
>+ err = inet6_netconf_fill_devconf(skb, dev->ifindex,
>+ &idev->cnf,
>+ NETLINK_CB(cb->skb).portid,
>+ nlh->nlmsg_seq,
>+ RTM_NEWNETCONF,
>+ NLM_F_MULTI,
>+ NETCONFA_ALL);
>+ if (err < 0)
>+ goto done;
> }
>- if (h == NETDEV_HASHENTRIES) {
>- if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_ALL,
>- net->ipv6.devconf_all,
>- NETLINK_CB(cb->skb).portid,
>- nlh->nlmsg_seq,
>- RTM_NEWNETCONF, NLM_F_MULTI,
>- NETCONFA_ALL) < 0)
>+ if (ctx->all_default == 0) {
>+ err = inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_ALL,
>+ net->ipv6.devconf_all,
>+ NETLINK_CB(cb->skb).portid,
>+ nlh->nlmsg_seq,
>+ RTM_NEWNETCONF, NLM_F_MULTI,
>+ NETCONFA_ALL);
>+ if (err < 0)
> goto done;
>- else
>- h++;
>- }
>- if (h == NETDEV_HASHENTRIES + 1) {
>- if (inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_DEFAULT,
>- net->ipv6.devconf_dflt,
>- NETLINK_CB(cb->skb).portid,
>- nlh->nlmsg_seq,
>- RTM_NEWNETCONF, NLM_F_MULTI,
>- NETCONFA_ALL) < 0)
>+ ctx->all_default++;
>+ }
>+ if (ctx->all_default == 1) {
>+ err = inet6_netconf_fill_devconf(skb, NETCONFA_IFINDEX_DEFAULT,
>+ net->ipv6.devconf_dflt,
>+ NETLINK_CB(cb->skb).portid,
>+ nlh->nlmsg_seq,
>+ RTM_NEWNETCONF, NLM_F_MULTI,
>+ NETCONFA_ALL);
>+ if (err < 0)
> goto done;
>- else
>- h++;
>+ ctx->all_default++;
> }
>-done:
>- cb->args[0] = h;
>- cb->args[1] = idx;
>-
>- return skb->len;
>+done: if (err < 0 && likely(skb->len))
It is common to not mix label and other statement on the same line,
could you split?
Otherwise the patch and the set looks good to me. Thanks!
>+ err = skb->len;
>+ rcu_read_unlock();
>+ return err;
> }
>
> #ifdef CONFIG_SYSCTL
>@@ -7503,7 +7487,8 @@ int __init addrconf_init(void)
> err = rtnl_register_module(THIS_MODULE, PF_INET6, RTM_GETNETCONF,
> inet6_netconf_get_devconf,
> inet6_netconf_dump_devconf,
>- RTNL_FLAG_DOIT_UNLOCKED);
>+ RTNL_FLAG_DOIT_UNLOCKED |
>+ RTNL_FLAG_DUMP_UNLOCKED);
> if (err < 0)
> goto errout;
> err = ipv6_addr_label_rtnl_register();
>--
>2.44.0.rc1.240.g4c46232300-goog
>
>
next prev parent reply other threads:[~2024-02-26 16:55 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-26 15:50 [PATCH net-next 00/13] ipv6: lockless accesses to devconf Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 01/13] ipv6: add ipv6_devconf_read_txrx cacheline_group Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 02/13] ipv6: annotate data-races around cnf.disable_ipv6 Eric Dumazet
2024-02-26 16:09 ` Jiri Pirko
2024-02-26 16:14 ` Eric Dumazet
2024-02-26 16:18 ` Jiri Pirko
2024-02-26 16:24 ` Eric Dumazet
2024-02-26 16:46 ` Jiri Pirko
2024-02-26 15:50 ` [PATCH net-next 03/13] ipv6: annotate data-races around cnf.mtu6 Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 04/13] ipv6: annotate data-races around cnf.hop_limit Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 05/13] ipv6: annotate data-races around cnf.forwarding Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 06/13] ipv6: annotate data-races in ndisc_router_discovery() Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 07/13] ipv6: annotate data-races around idev->cnf.ignore_routes_with_linkdown Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 08/13] ipv6: annotate data-races in rt6_probe() Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 09/13] ipv6: annotate data-races around devconf->proxy_ndp Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 10/13] ipv6: annotate data-races around devconf->disable_policy Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 11/13] ipv6/addrconf: annotate data-races around devconf fields (I) Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 12/13] ipv6/addrconf: annotate data-races around devconf fields (II) Eric Dumazet
2024-02-26 15:50 ` [PATCH net-next 13/13] ipv6: use xa_array iterator to implement inet6_netconf_dump_devconf() Eric Dumazet
2024-02-26 16:55 ` Jiri Pirko [this message]
2024-02-27 11:49 ` Eric Dumazet
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ZdzCkxLBi_JybU25@nanopsycho \
--to=jiri@resnulli.us \
--cc=davem@davemloft.net \
--cc=dsahern@kernel.org \
--cc=edumazet@google.com \
--cc=eric.dumazet@gmail.com \
--cc=kuba@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).