DPDK patches and discussions
 help / color / mirror / Atom feed
From: Maxime Coquelin <maxime.coquelin@redhat.com>
To: Ilya Maximets <i.maximets@samsung.com>,
	dev@dpdk.org, tiwei.bie@intel.com, zhihong.wang@intel.com,
	jfreimann@redhat.com, nicknickolaev@gmail.com,
	bruce.richardson@intel.com, alejandro.lucero@netronome.com
Cc: dgilbert@redhat.com, stable@dpdk.org
Subject: Re: [dpdk-dev] [PATCH v4 11/19] vhost: introduce postcopy's advise message
Date: Tue, 9 Oct 2018 09:59:10 +0200	[thread overview]
Message-ID: <b5fb1424-f69a-9ca7-a128-ef65edd7d4ff@redhat.com> (raw)
In-Reply-To: <20181008161952eucas1p2f3a63c4bff2e2160b482d865a6a0c9d9~brobzb3311921019210eucas1p2h@eucas1p2.samsung.com>



On 10/08/2018 06:22 PM, Ilya Maximets wrote:
> On 08.10.2018 18:25, Maxime Coquelin wrote:
>> This patch opens a userfaultfd and sends it back to Qemu's
>> VHOST_USER_POSTCOPY_ADVISE request.
>>
>> Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
>> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
>> ---
>>   lib/librte_vhost/vhost.h      |  2 ++
>>   lib/librte_vhost/vhost_user.c | 49 +++++++++++++++++++++++++++++++++++
>>   lib/librte_vhost/vhost_user.h |  3 ++-
>>   3 files changed, 53 insertions(+), 1 deletion(-)
>>
>> diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h
>> index 25ffd7614..21722d8a8 100644
>> --- a/lib/librte_vhost/vhost.h
>> +++ b/lib/librte_vhost/vhost.h
>> @@ -363,6 +363,8 @@ struct virtio_net {
>>   	int			slave_req_fd;
>>   	rte_spinlock_t		slave_req_lock;
>>   
>> +	int			postcopy_ufd;
>> +
>>   	/*
>>   	 * Device id to identify a specific backend device.
>>   	 * It's set to -1 for the default software implementation.
>> diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c
>> index 20f38267d..3cdd2af28 100644
>> --- a/lib/librte_vhost/vhost_user.c
>> +++ b/lib/librte_vhost/vhost_user.c
>> @@ -24,13 +24,19 @@
>>   #include <stdlib.h>
>>   #include <string.h>
>>   #include <unistd.h>
>> +#include <fcntl.h>
>> +#include <sys/ioctl.h>
>>   #include <sys/mman.h>
>>   #include <sys/types.h>
>>   #include <sys/stat.h>
>> +#include <sys/syscall.h>
>>   #include <assert.h>
>>   #ifdef RTE_LIBRTE_VHOST_NUMA
>>   #include <numaif.h>
>>   #endif
>> +#ifdef RTE_LIBRTE_VHOST_POSTCOPY
>> +#include <linux/userfaultfd.h>
>> +#endif
>>   
>>   #include <rte_common.h>
>>   #include <rte_malloc.h>
>> @@ -69,6 +75,7 @@ static const char *vhost_message_str[VHOST_USER_MAX] = {
>>   	[VHOST_USER_IOTLB_MSG]  = "VHOST_USER_IOTLB_MSG",
>>   	[VHOST_USER_CRYPTO_CREATE_SESS] = "VHOST_USER_CRYPTO_CREATE_SESS",
>>   	[VHOST_USER_CRYPTO_CLOSE_SESS] = "VHOST_USER_CRYPTO_CLOSE_SESS",
>> +	[VHOST_USER_POSTCOPY_ADVISE]  = "VHOST_USER_POSTCOPY_ADVISE",
>>   };
>>   
>>   /* The possible results of a message handling function */
>> @@ -130,6 +137,11 @@ vhost_backend_cleanup(struct virtio_net *dev)
>>   		close(dev->slave_req_fd);
>>   		dev->slave_req_fd = -1;
>>   	}
>> +
>> +	if (dev->postcopy_ufd >= 0) {
>> +		close(dev->postcopy_ufd);
>> +		dev->postcopy_ufd = -1;
>> +	}
>>   }
>>   
>>   /*
>> @@ -1508,6 +1520,42 @@ vhost_user_iotlb_msg(struct virtio_net **pdev, struct VhostUserMsg *msg,
>>   	return VH_RESULT_OK;
>>   }
>>   
>> +static int
>> +vhost_user_set_postcopy_advise(struct virtio_net **pdev,
>> +			struct VhostUserMsg *msg,
>> +			int main_fd __rte_unused)
>> +{
>> +	struct virtio_net *dev = *pdev;
>> +#ifdef RTE_LIBRTE_VHOST_POSTCOPY
>> +	struct uffdio_api api_struct;
>> +
>> +	dev->postcopy_ufd = syscall(__NR_userfaultfd, O_CLOEXEC | O_NONBLOCK);
>> +
>> +	if (dev->postcopy_ufd == -1) {
>> +		RTE_LOG(ERR, VHOST_CONFIG, "Userfaultfd not available: %s\n",
>> +				strerror(errno));
>> +		return VH_RESULT_ERR;
>> +	}
>> +	api_struct.api = UFFD_API;
>> +	api_struct.features = 0;
>> +	if (ioctl(dev->postcopy_ufd, UFFDIO_API, &api_struct)) {
>> +		RTE_LOG(ERR, VHOST_CONFIG, "UFFDIO_API ioctl failure: %s\n",
>> +				strerror(errno));
>> +		close(dev->postcopy_ufd);
> 
> Need to clear dev->postcopy_ufd value to avoid accidental closing of some
> other fd on backend cleanup.


Good catch, it will be fixed in v5.

Thanks!
Maxime

  reply	other threads:[~2018-10-09  7:59 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-10-08 15:25 [dpdk-dev] [PATCH v4 00/19] vhost: add postcopy live-migration support Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 01/19] vhost: fix messages error checks Maxime Coquelin
2018-10-09  8:03   ` Tiwei Bie
2018-10-09 12:20     ` Maxime Coquelin
2018-10-09 12:54       ` Tiwei Bie
2018-10-09 17:53         ` Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 02/19] vhost: fix return code of messages requiring replies Maxime Coquelin
2018-10-09  8:05   ` Tiwei Bie
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 03/19] vhost: clarify reply-ack in case a reply was already sent Maxime Coquelin
2018-10-09  8:12   ` Tiwei Bie
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 04/19] vhost: fix payload size of reply Maxime Coquelin
2018-10-09 10:21   ` Tiwei Bie
2018-10-09 10:30     ` Maxime Coquelin
2018-10-09 10:34       ` Maxime Coquelin
2018-10-09 11:02         ` Tiwei Bie
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 05/19] vhost: fix error handling when mem table gets updated Maxime Coquelin
2018-10-09 11:35   ` Tiwei Bie
2018-10-09 11:47     ` Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 06/19] vhost: define postcopy protocol flag Maxime Coquelin
2018-10-09 11:41   ` Tiwei Bie
2018-10-09 11:47     ` Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 07/19] vhost: add number of fds to vhost-user messages and use it Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 08/19] vhost: pass socket fd to message handling callbacks Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 09/19] vhost: enable fds passing when sending vhost-user messages Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 10/19] vhost: add config flag for postcopy feature Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 11/19] vhost: introduce postcopy's advise message Maxime Coquelin
2018-10-08 16:22   ` Ilya Maximets
2018-10-09  7:59     ` Maxime Coquelin [this message]
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 12/19] vhost: add support for postcopy's listen message Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 13/19] vhost: register new regions with userfaultfd Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 14/19] vhost: avoid useless VhostUserMemory copy Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 15/19] vhost: send userfault range addresses back to qemu Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 16/19] vhost: add support to postcopy's end request Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 17/19] vhost: enable postcopy protocol feature Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 18/19] vhost: add flag to enable postcopy live-migration Maxime Coquelin
2018-10-08 15:25 ` [dpdk-dev] [PATCH v4 19/19] net/vhost: add parameter to enable postcopy support Maxime Coquelin

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=b5fb1424-f69a-9ca7-a128-ef65edd7d4ff@redhat.com \
    --to=maxime.coquelin@redhat.com \
    --cc=alejandro.lucero@netronome.com \
    --cc=bruce.richardson@intel.com \
    --cc=dev@dpdk.org \
    --cc=dgilbert@redhat.com \
    --cc=i.maximets@samsung.com \
    --cc=jfreimann@redhat.com \
    --cc=nicknickolaev@gmail.com \
    --cc=stable@dpdk.org \
    --cc=tiwei.bie@intel.com \
    --cc=zhihong.wang@intel.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).