[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-07 Thread Yuanhan Liu
On Wed, Apr 06, 2016 at 06:12:07PM +0200, Thomas Monjalon wrote: > 2016-04-07 00:09, Yuanhan Liu: > > On Wed, Apr 06, 2016 at 09:37:53AM +, Loftus, Ciara wrote: > > > > On Wed, Apr 06, 2016 at 03:49:25PM +0900, Tetsuya Mukawa wrote: > > > > > Hi, > > > > > > > > > > I appreciate fixing it. > >

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-07 Thread Yuanhan Liu
On Wed, Apr 06, 2016 at 09:37:53AM +, Loftus, Ciara wrote: > > > > On Wed, Apr 06, 2016 at 03:49:25PM +0900, Tetsuya Mukawa wrote: > > > On 2016/04/06 1:09, Ciara Loftus wrote: > > > > After some testing, it was found that retrieving numa information > > > > about a vhost device via a call to

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Thomas Monjalon
2016-04-06 13:03, Yuanhan Liu: > On Tue, Apr 05, 2016 at 05:09:47PM +0100, Ciara Loftus wrote: > > After some testing, it was found that retrieving numa information > > about a vhost device via a call to get_mempolicy is more > > accurate when performed during the new_device callback versus > >

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Thomas Monjalon
2016-04-07 00:09, Yuanhan Liu: > On Wed, Apr 06, 2016 at 09:37:53AM +, Loftus, Ciara wrote: > > > On Wed, Apr 06, 2016 at 03:49:25PM +0900, Tetsuya Mukawa wrote: > > > > Hi, > > > > > > > > I appreciate fixing it. > > > > Just one worry is that state changed event may be occurred before new >

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Tetsuya Mukawa
On 2016/04/06 16:17, Yuanhan Liu wrote: > On Wed, Apr 06, 2016 at 03:49:25PM +0900, Tetsuya Mukawa wrote: >> On 2016/04/06 1:09, Ciara Loftus wrote: >>> After some testing, it was found that retrieving numa information >>> about a vhost device via a call to get_mempolicy is more >>> accurate when

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Tetsuya Mukawa
On 2016/04/06 1:09, Ciara Loftus wrote: > After some testing, it was found that retrieving numa information > about a vhost device via a call to get_mempolicy is more > accurate when performed during the new_device callback versus > the vring_state_changed callback, in particular upon initial boot

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Tan, Jianfeng
On 4/6/2016 2:17 PM, Yuanhan Liu wrote: > On Wed, Apr 06, 2016 at 02:05:51PM +0800, Tan, Jianfeng wrote: >> >> On 4/6/2016 1:44 PM, Yuanhan Liu wrote: >>> On Wed, Apr 06, 2016 at 01:32:12PM +0800, Tan, Jianfeng wrote: Hi, Just out of interest, seems that the message handling

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Yuanhan Liu
On Wed, Apr 06, 2016 at 02:05:51PM +0800, Tan, Jianfeng wrote: > > > On 4/6/2016 1:44 PM, Yuanhan Liu wrote: > >On Wed, Apr 06, 2016 at 01:32:12PM +0800, Tan, Jianfeng wrote: > >>Hi, > >> > >>Just out of interest, seems that the message handling thread which runs > >>new_device() is

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Tan, Jianfeng
On 4/6/2016 1:44 PM, Yuanhan Liu wrote: > On Wed, Apr 06, 2016 at 01:32:12PM +0800, Tan, Jianfeng wrote: >> Hi, >> >> Just out of interest, seems that the message handling thread which runs >> new_device() is pthread_create() from the thread which calls the >> dev_start(), usually master thread,

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Yuanhan Liu
On Wed, Apr 06, 2016 at 01:32:12PM +0800, Tan, Jianfeng wrote: > Hi, > > Just out of interest, seems that the message handling thread which runs > new_device() is pthread_create() from the thread which calls the > dev_start(), usually master thread, right? But it's not necessary to be the >

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Tan, Jianfeng
Hi, Just out of interest, seems that the message handling thread which runs new_device() is pthread_create() from the thread which calls the dev_start(), usually master thread, right? But it's not necessary to be the master thread to poll pkts from this vhost port, right? So what's the

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Yuanhan Liu
On Tue, Apr 05, 2016 at 05:09:47PM +0100, Ciara Loftus wrote: > After some testing, it was found that retrieving numa information > about a vhost device via a call to get_mempolicy is more > accurate when performed during the new_device callback versus > the vring_state_changed callback, in

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-06 Thread Loftus, Ciara
> > On Wed, Apr 06, 2016 at 03:49:25PM +0900, Tetsuya Mukawa wrote: > > On 2016/04/06 1:09, Ciara Loftus wrote: > > > After some testing, it was found that retrieving numa information > > > about a vhost device via a call to get_mempolicy is more > > > accurate when performed during the

[dpdk-dev] [PATCH] vhost: Fix retrieval of numa information in PMD

2016-04-05 Thread Ciara Loftus
After some testing, it was found that retrieving numa information about a vhost device via a call to get_mempolicy is more accurate when performed during the new_device callback versus the vring_state_changed callback, in particular upon initial boot of the VM. Performing this check during