From: "Michael S. Tsirkin" <mst@redhat.com>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Jason Wang <jasowang@redhat.com>,
davem@davemloft.net, edumazet@google.com, hkchu@google.com,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [net-next rfc 1/3] net: avoid high order memory allocation for queues by using flex array
Date: Wed, 19 Jun 2013 21:07:09 +0300 [thread overview]
Message-ID: <20130619180709.GA15017@redhat.com> (raw)
In-Reply-To: <1371657511.3252.324.camel@edumazet-glaptop>
On Wed, Jun 19, 2013 at 08:58:31AM -0700, Eric Dumazet wrote:
> On Wed, 2013-06-19 at 18:40 +0300, Michael S. Tsirkin wrote:
>
> > That's a good trick too - vmalloc memory is a bit slower
> > on x86 since it's not using a huge page, but that's only
> > when we have lots of CPUs/queues...
> >
> > Short term - how about switching to vmalloc if > 32 queues?
>
> You do not have to switch "if > 32 queues"
>
> diff --git a/net/core/dev.c b/net/core/dev.c
> index fa007db..473cc9e 100644
> --- a/net/core/dev.c
> +++ b/net/core/dev.c
> @@ -130,6 +130,7 @@
> #include <linux/cpu_rmap.h>
> #include <linux/static_key.h>
> #include <linux/hashtable.h>
> +#include <linux/vmalloc.h>
>
> #include "net-sysfs.h"
>
> @@ -5253,6 +5254,14 @@ static void netdev_init_one_queue(struct net_device *dev,
> #endif
> }
>
> +static void netif_free_tx_queues(struct net_device *dev)
> +{
> + if (is_vmalloc_addr(dev->_tx))
> + vfree(dev->_tx);
> + else
> + kfree(dev->_tx);
> +}
> +
> static int netif_alloc_netdev_queues(struct net_device *dev)
> {
> unsigned int count = dev->num_tx_queues;
> @@ -5261,9 +5270,11 @@ static int netif_alloc_netdev_queues(struct net_device *dev)
> BUG_ON(count < 1);
>
> tx = kcalloc(count, sizeof(struct netdev_queue), GFP_KERNEL);
As we have explicit code to recover, maybe set __GFP_NOWARN?
If we are out of memory, vzalloc will warn too, we don't
need two warnings.
> - if (!tx)
> - return -ENOMEM;
> -
> + if (!tx) {
> + tx = vzalloc(count * sizeof(struct netdev_queue));
> + if (!tx)
> + return -ENOMEM;
> + }
> dev->_tx = tx;
>
> netdev_for_each_tx_queue(dev, netdev_init_one_queue, NULL);
> @@ -5811,7 +5822,7 @@ free_all:
>
> free_pcpu:
> free_percpu(dev->pcpu_refcnt);
> - kfree(dev->_tx);
> + netif_free_tx_queues(dev);
> #ifdef CONFIG_RPS
> kfree(dev->_rx);
> #endif
> @@ -5836,7 +5847,7 @@ void free_netdev(struct net_device *dev)
>
> release_net(dev_net(dev));
>
> - kfree(dev->_tx);
> + netif_free_tx_queues(dev);
> #ifdef CONFIG_RPS
> kfree(dev->_rx);
> #endif
>
next prev parent reply other threads:[~2013-06-19 18:06 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-19 5:40 [net-next rfc 0/3] increase the limit of tuntap queues Jason Wang
2013-06-19 5:40 ` [net-next rfc 1/3] net: avoid high order memory allocation for queues by using flex array Jason Wang
2013-06-19 6:31 ` Eric Dumazet
2013-06-19 7:14 ` Jason Wang
2013-06-19 9:11 ` Michael S. Tsirkin
2013-06-19 9:56 ` Eric Dumazet
2013-06-19 12:22 ` Michael S. Tsirkin
2013-06-19 15:40 ` Michael S. Tsirkin
2013-06-19 15:58 ` Eric Dumazet
2013-06-19 16:06 ` David Laight
2013-06-19 16:28 ` Eric Dumazet
2013-06-19 18:07 ` Michael S. Tsirkin [this message]
2013-06-20 5:14 ` Jason Wang
2013-06-20 6:05 ` Eric Dumazet
2013-06-19 5:40 ` [net-next rfc 2/3] tuntap: reduce the size of tun_struct " Jason Wang
2013-06-19 5:40 ` [net-next rfc 3/3] tuntap: increase the max queues to 16 Jason Wang
2013-06-19 6:34 ` Eric Dumazet
2013-06-19 7:15 ` Jason Wang
2013-06-19 19:16 ` Jerry Chu
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=20130619180709.GA15017@redhat.com \
--to=mst@redhat.com \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=eric.dumazet@gmail.com \
--cc=hkchu@google.com \
--cc=jasowang@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
/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).