All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Zhang, Chen" <chen.zhang@intel.com>
To: "Rao, Lei" <lei.rao@intel.com>,
	"lizhijian@cn.fujitsu.com" <lizhijian@cn.fujitsu.com>,
	"jasowang@redhat.com" <jasowang@redhat.com>,
	"quintela@redhat.com" <quintela@redhat.com>,
	"dgilbert@redhat.com" <dgilbert@redhat.com>,
	"pbonzini@redhat.com" <pbonzini@redhat.com>,
	"lukasstraub2@web.de" <lukasstraub2@web.de>
Cc: "qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Subject: RE: [PATCH v5 05/10] Add a function named packet_new_nocopy for COLO.
Date: Thu, 8 Apr 2021 05:30:27 +0000	[thread overview]
Message-ID: <d947308fe00f41ae90b6e2f0241c5c72@intel.com> (raw)
In-Reply-To: <1617263249-54501-6-git-send-email-lei.rao@intel.com>



> -----Original Message-----
> From: Rao, Lei <lei.rao@intel.com>
> Sent: Thursday, April 1, 2021 3:47 PM
> To: Zhang, Chen <chen.zhang@intel.com>; lizhijian@cn.fujitsu.com;
> jasowang@redhat.com; quintela@redhat.com; dgilbert@redhat.com;
> pbonzini@redhat.com; lukasstraub2@web.de
> Cc: qemu-devel@nongnu.org; Rao, Lei <lei.rao@intel.com>
> Subject: [PATCH v5 05/10] Add a function named packet_new_nocopy for
> COLO.
> 
> From: "Rao, Lei" <lei.rao@intel.com>
> 
> Use the packet_new_nocopy instead of packet_new in the filter-rewriter
> module. There will be one less memory copy in the processing of each
> network packet.
> 
> Signed-off-by: Lei Rao <lei.rao@intel.com>
> ---
>  net/colo.c            | 23 +++++++++++++++++++++++
>  net/colo.h            |  1 +
>  net/filter-rewriter.c |  3 +--
>  3 files changed, 25 insertions(+), 2 deletions(-)
> 
> diff --git a/net/colo.c b/net/colo.c
> index ef00609..58106a8 100644
> --- a/net/colo.c
> +++ b/net/colo.c
> @@ -174,6 +174,29 @@ Packet *packet_new(const void *data, int size, int
> vnet_hdr_len)
>      return pkt;
>  }
> 
> +/*
> + * packet_new_nocopy will not copy data, so the caller can't release
> + * the data. And it will be released in packet_destroy.
> + */
> +Packet *packet_new_nocopy(void *data, int size, int vnet_hdr_len) {
> +    Packet *pkt = g_slice_new(Packet);

We can use g_slice_new0() to avoid "pkt->xxx = 0" here.
For the original code also need do this work to optimize code.

Thanks
Chen

> +
> +    pkt->data = data;
> +    pkt->size = size;
> +    pkt->creation_ms = qemu_clock_get_ms(QEMU_CLOCK_HOST);
> +    pkt->vnet_hdr_len = vnet_hdr_len;
> +    pkt->tcp_seq = 0;
> +    pkt->tcp_ack = 0;
> +    pkt->seq_end = 0;
> +    pkt->header_size = 0;
> +    pkt->payload_size = 0;
> +    pkt->offset = 0;
> +    pkt->flags = 0;
> +
> +    return pkt;
> +}
> +
>  void packet_destroy(void *opaque, void *user_data)  {
>      Packet *pkt = opaque;
> diff --git a/net/colo.h b/net/colo.h
> index 573ab91..d91cd24 100644
> --- a/net/colo.h
> +++ b/net/colo.h
> @@ -101,6 +101,7 @@ bool connection_has_tracked(GHashTable
> *connection_track_table,
>                              ConnectionKey *key);  void
> connection_hashtable_reset(GHashTable *connection_track_table);  Packet
> *packet_new(const void *data, int size, int vnet_hdr_len);
> +Packet *packet_new_nocopy(void *data, int size, int vnet_hdr_len);
>  void packet_destroy(void *opaque, void *user_data);  void
> packet_destroy_partial(void *opaque, void *user_data);
> 
> diff --git a/net/filter-rewriter.c b/net/filter-rewriter.c index 10fe393..cb3a96c
> 100644
> --- a/net/filter-rewriter.c
> +++ b/net/filter-rewriter.c
> @@ -270,8 +270,7 @@ static ssize_t
> colo_rewriter_receive_iov(NetFilterState *nf,
>          vnet_hdr_len = nf->netdev->vnet_hdr_len;
>      }
> 
> -    pkt = packet_new(buf, size, vnet_hdr_len);
> -    g_free(buf);
> +    pkt = packet_new_nocopy(buf, size, vnet_hdr_len);
> 
>      /*
>       * if we get tcp packet
> --
> 1.8.3.1



  reply	other threads:[~2021-04-08  5:31 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-01  7:47 [PATCH v5 00/10] Fixed some bugs and optimized some codes for COLO leirao
2021-04-01  7:47 ` [PATCH v5 01/10] Remove some duplicate trace code leirao
2021-04-08  5:21   ` Zhang, Chen
2021-04-01  7:47 ` [PATCH v5 02/10] Fix the qemu crash when guest shutdown during checkpoint leirao
2021-04-08  5:21   ` Zhang, Chen
2021-04-01  7:47 ` [PATCH v5 03/10] Optimize the function of filter_send leirao
2021-04-08  5:23   ` Zhang, Chen
2021-04-01  7:47 ` [PATCH v5 04/10] Remove migrate_set_block_enabled in checkpoint leirao
2021-04-08  5:25   ` Zhang, Chen
2021-04-01  7:47 ` [PATCH v5 05/10] Add a function named packet_new_nocopy for COLO leirao
2021-04-08  5:30   ` Zhang, Chen [this message]
2021-04-01  7:47 ` [PATCH v5 06/10] Add the function of colo_compare_cleanup leirao
2021-04-08  8:48   ` Zhang, Chen
2021-04-01  7:47 ` [PATCH v5 07/10] Reset the auto-converge counter at every checkpoint leirao
2021-04-01  7:47 ` [PATCH v5 08/10] Reduce the PVM stop time during Checkpoint leirao
2021-04-01  7:47 ` [PATCH v5 09/10] Add the function of colo_bitmap_clear_dirty leirao
2021-04-01  7:47 ` [PATCH v5 10/10] Fixed calculation error of pkt->header_size in fill_pkt_tcp_info() leirao
2021-04-08  8:49   ` Zhang, Chen
2021-04-04 10:22 ` [PATCH v5 00/10] Fixed some bugs and optimized some codes for COLO Lukas Straub

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=d947308fe00f41ae90b6e2f0241c5c72@intel.com \
    --to=chen.zhang@intel.com \
    --cc=dgilbert@redhat.com \
    --cc=jasowang@redhat.com \
    --cc=lei.rao@intel.com \
    --cc=lizhijian@cn.fujitsu.com \
    --cc=lukasstraub2@web.de \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=quintela@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.