DPDK patches and discussions
 help / color / mirror / Atom feed
From: Maxime Coquelin <maxime.coquelin@redhat.com>
To: "Xia, Chenbo" <chenbo.xia@intel.com>,
	"dev@dpdk.org" <dev@dpdk.org>, "Ding, Xuan" <xuan.ding@intel.com>
Subject: Re: [dpdk-dev] [PATCH 21.02 1/3] vhost: refactor postcopy region registration
Date: Tue, 22 Dec 2020 11:51:10 +0100
Message-ID: <f11ff2bc-bc5f-08af-a8c9-e7649d6cdee5@redhat.com> (raw)
In-Reply-To: <MN2PR11MB40632C4488E3DF162326158B9CCC0@MN2PR11MB4063.namprd11.prod.outlook.com>

Hi Chenbo,

On 12/9/20 3:16 PM, Xia, Chenbo wrote:
> Hi Maxime,
> 
>> -----Original Message-----
>> From: Maxime Coquelin <maxime.coquelin@redhat.com>
>> Sent: Monday, November 16, 2020 7:36 PM
>> To: dev@dpdk.org; Xia, Chenbo <chenbo.xia@intel.com>; Ding, Xuan
>> <xuan.ding@intel.com>
>> Cc: Maxime Coquelin <maxime.coquelin@redhat.com>
>> Subject: [PATCH 21.02 1/3] vhost: refactor postcopy region registration
>>
>> This patch moves the registration of memory regions to
>> userfaultfd to a dedicated function, with the goal of
>> simplifying VHOST_USER_SET_MEM_TABLE request handling
>> function.
>>
>> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
>> ---
>>  lib/librte_vhost/vhost_user.c | 77 +++++++++++++++++++++--------------
>>  1 file changed, 46 insertions(+), 31 deletions(-)
>>
>> diff --git a/lib/librte_vhost/vhost_user.c b/lib/librte_vhost/vhost_user.c
>> index 45c8ac09da..b8a9e41a2d 100644
>> --- a/lib/librte_vhost/vhost_user.c
>> +++ b/lib/librte_vhost/vhost_user.c
>> @@ -998,6 +998,49 @@ vhost_memory_changed(struct VhostUserMemory *new,
>>  	return false;
>>  }
>>
>> +#ifdef RTE_LIBRTE_VHOST_POSTCOPY
>> +static int
>> +vhost_user_postcopy_region_register(struct virtio_net *dev,
>> +		struct rte_vhost_mem_region *reg)
>> +{
>> +	struct uffdio_register reg_struct;
>> +
>> +	/*
>> +	 * Let's register all the mmap'ed area to ensure
>> +	 * alignment on page boundary.
>> +	 */
>> +	reg_struct.range.start = (uint64_t)(uintptr_t)reg->mmap_addr;
>> +	reg_struct.range.len = reg->mmap_size;
>> +	reg_struct.mode = UFFDIO_REGISTER_MODE_MISSING;
>> +
>> +	if (ioctl(dev->postcopy_ufd, UFFDIO_REGISTER,
>> +				&reg_struct)) {
>> +		VHOST_LOG_CONFIG(ERR, "Failed to register ufd for region "
>> +				"%" PRIx64 " - %" PRIx64 " (ufd = %d) %s\n",
>> +				(uint64_t)reg_struct.range.start,
>> +				(uint64_t)reg_struct.range.start +
>> +				(uint64_t)reg_struct.range.len - 1,
>> +				dev->postcopy_ufd,
>> +				strerror(errno));
>> +		return -1;
>> +	}
>> +
>> +	VHOST_LOG_CONFIG(INFO, "\t userfaultfd registered for range : %"
>> PRIx64 " - %" PRIx64 "\n",
>> +			(uint64_t)reg_struct.range.start,
>> +			(uint64_t)reg_struct.range.start +
>> +			(uint64_t)reg_struct.range.len - 1);
>> +
>> +	return 0;
>> +}
>> +#else
>> +static int
>> +vhost_user_postcopy_region_register(struct virtio_net *dev,
>> +		struct rte_vhost_mem_region *reg)
>> +{
>> +	return -1;
>> +}
> 
> Better add __rte_unused here to avoid warnings when postcopy not configured 😊

Good catch!
I will post a v2 with adding these.

Thanks,
Maxime

> Thanks,
> Chenbo
> 
>> +#endif
>> +
>>  static int
>>  vhost_user_set_mem_table(struct virtio_net **pdev, struct VhostUserMsg
>> *msg,
>>  			int main_fd)
>> @@ -1209,38 +1252,10 @@ vhost_user_set_mem_table(struct virtio_net **pdev,
>> struct VhostUserMsg *msg,
>>  		}
>>
>>  		/* Now userfault register and we can use the memory */
>> -		for (i = 0; i < memory->nregions; i++) {
>> -#ifdef RTE_LIBRTE_VHOST_POSTCOPY
>> -			reg = &dev->mem->regions[i];
>> -			struct uffdio_register reg_struct;
>> -
>> -			/*
>> -			 * Let's register all the mmap'ed area to ensure
>> -			 * alignment on page boundary.
>> -			 */
>> -			reg_struct.range.start =
>> -				(uint64_t)(uintptr_t)reg->mmap_addr;
>> -			reg_struct.range.len = reg->mmap_size;
>> -			reg_struct.mode = UFFDIO_REGISTER_MODE_MISSING;
>> -
>> -			if (ioctl(dev->postcopy_ufd, UFFDIO_REGISTER,
>> -						&reg_struct)) {
>> -				VHOST_LOG_CONFIG(ERR,
>> -					"Failed to register ufd for region %d: (ufd
>> = %d) %s\n",
>> -					i, dev->postcopy_ufd,
>> -					strerror(errno));
>> +		for (i = 0; i < memory->nregions; i++)
>> +			if (vhost_user_postcopy_region_register(dev,
>> +						&dev->mem->regions[i]) < 0)
>>  				goto free_mem_table;
>> -			}
>> -			VHOST_LOG_CONFIG(INFO,
>> -				"\t userfaultfd registered for range : "
>> -				"%" PRIx64 " - %" PRIx64 "\n",
>> -				(uint64_t)reg_struct.range.start,
>> -				(uint64_t)reg_struct.range.start +
>> -				(uint64_t)reg_struct.range.len - 1);
>> -#else
>> -			goto free_mem_table;
>> -#endif
>> -		}
>>  	}
>>
>>  	for (i = 0; i < dev->nr_vring; i++) {
>> --
>> 2.26.2
> 


  reply	other threads:[~2020-12-22 10:51 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-16 11:36 [dpdk-dev] [PATCH 21.02 0/3] vhost: vhost_user_set_mem_table refactoring Maxime Coquelin
2020-11-16 11:36 ` [dpdk-dev] [PATCH 21.02 1/3] vhost: refactor postcopy region registration Maxime Coquelin
2020-12-09 14:16   ` Xia, Chenbo
2020-12-22 10:51     ` Maxime Coquelin [this message]
2020-11-16 11:36 ` [dpdk-dev] [PATCH 21.02 2/3] vhost: refactor postcopy registration Maxime Coquelin
2020-12-09 14:16   ` Xia, Chenbo
2020-11-16 11:36 ` [dpdk-dev] [PATCH 21.02 3/3] vhost: refactor memory regions mapping Maxime Coquelin
2020-12-09 14:16   ` Xia, Chenbo

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=f11ff2bc-bc5f-08af-a8c9-e7649d6cdee5@redhat.com \
    --to=maxime.coquelin@redhat.com \
    --cc=chenbo.xia@intel.com \
    --cc=dev@dpdk.org \
    --cc=xuan.ding@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

DPDK patches and discussions

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://inbox.dpdk.org/dev/0 dev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 dev dev/ https://inbox.dpdk.org/dev \
		dev@dpdk.org
	public-inbox-index dev

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.dev


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git