qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
* [Qemu-devel]  [PATCH V2 0/2] vhost-user: sync backend
@ 2015-07-29 20:06 Marcel Apfelbaum
  2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 1/2] vhost-user: sync backend features Marcel Apfelbaum
  2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 2/2] tests/vhost-user: check vhost-user feature negotiation Marcel Apfelbaum
  0 siblings, 2 replies; 7+ messages in thread
From: Marcel Apfelbaum @ 2015-07-29 20:06 UTC (permalink / raw)
  To: qemu-devel; +Cc: marcel, jasowang, famz, mst

 To be used on top of:
 [PATCH 0/4] vhost-user: protocol updates
 https://lists.gnu.org/archive/html/qemu-devel/2015-07/msg03842.html 

 v1 -> v2:
 - Addressed Michael S. Tsirkin's comments:
   - Prefer a white-list of supported features
   - Add a unit-test to show the problem we are trying to solve
     (run the unit-test before the patch and it will fail)

 Currently the vhost-user supported features are not evaluated.
 The way I see it, and please correct me, the best way to do
 this is to:
  1. Get the backend features on vhost init.
  2. Check them one by one (white-list) against all 
     the currently supported virtio features.
  3. All other code should remain the same.

Marcel Apfelbaum (2):
  vhost-user: sync backend features
  tests/vhost-user: check vhost-user feature negotiation

 hw/virtio/vhost-user.c  | 17 +++++++++++++++++
 tests/vhost-user-test.c | 19 ++++++++++++++++---
 2 files changed, 33 insertions(+), 3 deletions(-)

-- 
2.1.0

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Qemu-devel]  [PATCH V2 1/2] vhost-user: sync backend features
  2015-07-29 20:06 [Qemu-devel] [PATCH V2 0/2] vhost-user: sync backend Marcel Apfelbaum
@ 2015-07-29 20:06 ` Marcel Apfelbaum
  2015-07-30  6:47   ` Michael S. Tsirkin
  2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 2/2] tests/vhost-user: check vhost-user feature negotiation Marcel Apfelbaum
  1 sibling, 1 reply; 7+ messages in thread
From: Marcel Apfelbaum @ 2015-07-29 20:06 UTC (permalink / raw)
  To: qemu-devel; +Cc: marcel, jasowang, famz, mst

Complete vhost-user negotiation by syncing the features
supported by the backend.

Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
---
 hw/virtio/vhost-user.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)

diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index c4428a1..b522437 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -22,6 +22,7 @@
 #include <sys/socket.h>
 #include <sys/un.h>
 #include <linux/vhost.h>
+#include <linux/virtio_net.h>
 
 #define VHOST_MEMORY_MAX_NREGIONS    8
 
@@ -358,6 +359,22 @@ static int vhost_user_init(struct vhost_dev *dev, void *opaque)
         return err;
     }
 
+    if (__virtio_has_feature(msg.u64, VIRTIO_F_ANY_LAYOUT)) {
+        dev->backend_features |= 1ULL << VIRTIO_F_ANY_LAYOUT;
+    }
+
+    if (__virtio_has_feature(msg.u64, VIRTIO_F_VERSION_1)) {
+        dev->backend_features |= 1ULL << VIRTIO_F_VERSION_1;
+    }
+
+    if (__virtio_has_feature(msg.u64, VIRTIO_NET_F_MRG_RXBUF)) {
+        dev->backend_features |= 1ULL << VIRTIO_NET_F_MRG_RXBUF;
+    }
+
+    if (__virtio_has_feature(msg.u64, VIRTIO_NET_F_MQ)) {
+        dev->backend_features |= 1ULL << VIRTIO_NET_F_MQ;
+    }
+
     if (__virtio_has_feature(msg.u64, VHOST_USER_F_PROTOCOL_FEATURES)) {
         dev->backend_features |= 1ULL << VHOST_USER_F_PROTOCOL_FEATURES;
 
-- 
2.1.0

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* [Qemu-devel] [PATCH V2 2/2] tests/vhost-user: check vhost-user feature negotiation
  2015-07-29 20:06 [Qemu-devel] [PATCH V2 0/2] vhost-user: sync backend Marcel Apfelbaum
  2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 1/2] vhost-user: sync backend features Marcel Apfelbaum
@ 2015-07-29 20:06 ` Marcel Apfelbaum
  2015-07-30  6:47   ` Michael S. Tsirkin
  1 sibling, 1 reply; 7+ messages in thread
From: Marcel Apfelbaum @ 2015-07-29 20:06 UTC (permalink / raw)
  To: qemu-devel; +Cc: marcel, jasowang, famz, mst

Check the backend receives all declared virtio features
that are also supported by QEMU virtio.

Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
---
 tests/vhost-user-test.c | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)

diff --git a/tests/vhost-user-test.c b/tests/vhost-user-test.c
index 228acb6..019b880 100644
--- a/tests/vhost-user-test.c
+++ b/tests/vhost-user-test.c
@@ -17,6 +17,7 @@
 #include "sysemu/sysemu.h"
 
 #include <linux/vhost.h>
+#include <linux/virtio_net.h>
 #include <sys/mman.h>
 #include <sys/vfs.h>
 #include <qemu/sockets.h>
@@ -297,14 +298,26 @@ static void chr_read(void *opaque, const uint8_t *buf, int size)
         /* send back features to qemu */
         msg.flags |= VHOST_USER_REPLY_MASK;
         msg.size = sizeof(m.u64);
-        msg.u64 = 0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES;
+        msg.u64 = (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES) |
+                  (0x1ULL << VIRTIO_F_ANY_LAYOUT) |
+                  (0x1ULL << VIRTIO_F_VERSION_1)  |
+                  (0x1ULL << VIRTIO_NET_F_MRG_RXBUF) |
+                  (0x1ULL << VIRTIO_NET_F_MQ);
         p = (uint8_t *) &msg;
         qemu_chr_fe_write_all(chr, p, VHOST_USER_HDR_SIZE + msg.size);
         break;
 
     case VHOST_USER_SET_FEATURES:
-	g_assert_cmpint(msg.u64 & (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES),
-			!=, 0ULL);
+        g_assert_cmpint(msg.u64 & (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES),
+            !=, 0ULL);
+        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_F_ANY_LAYOUT),
+            !=, 0ULL);
+        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_F_VERSION_1),
+            !=, 0ULL);
+        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_NET_F_MRG_RXBUF),
+            !=, 0ULL);
+        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_NET_F_MQ),
+            !=, 0ULL);
         break;
 
     case VHOST_USER_GET_PROTOCOL_FEATURES:
-- 
2.1.0

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* Re: [Qemu-devel] [PATCH V2 1/2] vhost-user: sync backend features
  2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 1/2] vhost-user: sync backend features Marcel Apfelbaum
@ 2015-07-30  6:47   ` Michael S. Tsirkin
  2015-07-30  7:27     ` Marcel Apfelbaum
  0 siblings, 1 reply; 7+ messages in thread
From: Michael S. Tsirkin @ 2015-07-30  6:47 UTC (permalink / raw)
  To: Marcel Apfelbaum; +Cc: jasowang, famz, qemu-devel

On Wed, Jul 29, 2015 at 11:06:48PM +0300, Marcel Apfelbaum wrote:
> Complete vhost-user negotiation by syncing

synching.

> the features
> supported by the backend.
> 
> Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
> ---
>  hw/virtio/vhost-user.c | 17 +++++++++++++++++
>  1 file changed, 17 insertions(+)
> 
> diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
> index c4428a1..b522437 100644
> --- a/hw/virtio/vhost-user.c
> +++ b/hw/virtio/vhost-user.c
> @@ -22,6 +22,7 @@
>  #include <sys/socket.h>
>  #include <sys/un.h>
>  #include <linux/vhost.h>
> +#include <linux/virtio_net.h>
>  
>  #define VHOST_MEMORY_MAX_NREGIONS    8
>  
> @@ -358,6 +359,22 @@ static int vhost_user_init(struct vhost_dev *dev, void *opaque)
>          return err;
>      }
>  
> +    if (__virtio_has_feature(msg.u64, VIRTIO_F_ANY_LAYOUT)) {
> +        dev->backend_features |= 1ULL << VIRTIO_F_ANY_LAYOUT;
> +    }
> +
> +    if (__virtio_has_feature(msg.u64, VIRTIO_F_VERSION_1)) {
> +        dev->backend_features |= 1ULL << VIRTIO_F_VERSION_1;
> +    }
> +
> +    if (__virtio_has_feature(msg.u64, VIRTIO_NET_F_MRG_RXBUF)) {
> +        dev->backend_features |= 1ULL << VIRTIO_NET_F_MRG_RXBUF;
> +    }
> +
> +    if (__virtio_has_feature(msg.u64, VIRTIO_NET_F_MQ)) {
> +        dev->backend_features |= 1ULL << VIRTIO_NET_F_MQ;
> +    }
> +

A bit too early for MQ, isn't it?

>      if (__virtio_has_feature(msg.u64, VHOST_USER_F_PROTOCOL_FEATURES)) {
>          dev->backend_features |= 1ULL << VHOST_USER_F_PROTOCOL_FEATURES;
>  
> -- 
> 2.1.0

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [Qemu-devel] [PATCH V2 2/2] tests/vhost-user: check vhost-user feature negotiation
  2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 2/2] tests/vhost-user: check vhost-user feature negotiation Marcel Apfelbaum
@ 2015-07-30  6:47   ` Michael S. Tsirkin
  2015-07-30  7:30     ` Marcel Apfelbaum
  0 siblings, 1 reply; 7+ messages in thread
From: Michael S. Tsirkin @ 2015-07-30  6:47 UTC (permalink / raw)
  To: Marcel Apfelbaum; +Cc: jasowang, famz, qemu-devel

On Wed, Jul 29, 2015 at 11:06:49PM +0300, Marcel Apfelbaum wrote:
> Check the backend receives all declared virtio features
> that are also supported by QEMU virtio.
> 
> Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
> ---
>  tests/vhost-user-test.c | 19 ++++++++++++++++---
>  1 file changed, 16 insertions(+), 3 deletions(-)
> 
> diff --git a/tests/vhost-user-test.c b/tests/vhost-user-test.c
> index 228acb6..019b880 100644
> --- a/tests/vhost-user-test.c
> +++ b/tests/vhost-user-test.c
> @@ -17,6 +17,7 @@
>  #include "sysemu/sysemu.h"
>  
>  #include <linux/vhost.h>
> +#include <linux/virtio_net.h>
>  #include <sys/mman.h>
>  #include <sys/vfs.h>
>  #include <qemu/sockets.h>
> @@ -297,14 +298,26 @@ static void chr_read(void *opaque, const uint8_t *buf, int size)
>          /* send back features to qemu */
>          msg.flags |= VHOST_USER_REPLY_MASK;
>          msg.size = sizeof(m.u64);
> -        msg.u64 = 0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES;
> +        msg.u64 = (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES) |
> +                  (0x1ULL << VIRTIO_F_ANY_LAYOUT) |
> +                  (0x1ULL << VIRTIO_F_VERSION_1)  |
> +                  (0x1ULL << VIRTIO_NET_F_MRG_RXBUF) |
> +                  (0x1ULL << VIRTIO_NET_F_MQ);
>          p = (uint8_t *) &msg;
>          qemu_chr_fe_write_all(chr, p, VHOST_USER_HDR_SIZE + msg.size);
>          break;
>  

No, this mixes protocol features and virtio features.
It's not what was intended.

>      case VHOST_USER_SET_FEATURES:
> -	g_assert_cmpint(msg.u64 & (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES),
> -			!=, 0ULL);
> +        g_assert_cmpint(msg.u64 & (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES),
> +            !=, 0ULL);
> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_F_ANY_LAYOUT),
> +            !=, 0ULL);
> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_F_VERSION_1),
> +            !=, 0ULL);
> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_NET_F_MRG_RXBUF),
> +            !=, 0ULL);
> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_NET_F_MQ),
> +            !=, 0ULL);
>          break;
>  
>      case VHOST_USER_GET_PROTOCOL_FEATURES:
> -- 
> 2.1.0

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [Qemu-devel] [PATCH V2 1/2] vhost-user: sync backend features
  2015-07-30  6:47   ` Michael S. Tsirkin
@ 2015-07-30  7:27     ` Marcel Apfelbaum
  0 siblings, 0 replies; 7+ messages in thread
From: Marcel Apfelbaum @ 2015-07-30  7:27 UTC (permalink / raw)
  To: Michael S. Tsirkin; +Cc: jasowang, famz, qemu-devel

On 07/30/2015 09:47 AM, Michael S. Tsirkin wrote:
> On Wed, Jul 29, 2015 at 11:06:48PM +0300, Marcel Apfelbaum wrote:
>> Complete vhost-user negotiation by syncing
>
> synching.
OK, thanks
>
>> the features
>> supported by the backend.
>>
>> Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
>> ---
>>   hw/virtio/vhost-user.c | 17 +++++++++++++++++
>>   1 file changed, 17 insertions(+)
>>
>> diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
>> index c4428a1..b522437 100644
>> --- a/hw/virtio/vhost-user.c
>> +++ b/hw/virtio/vhost-user.c
>> @@ -22,6 +22,7 @@
>>   #include <sys/socket.h>
>>   #include <sys/un.h>
>>   #include <linux/vhost.h>
>> +#include <linux/virtio_net.h>
>>
>>   #define VHOST_MEMORY_MAX_NREGIONS    8
>>
>> @@ -358,6 +359,22 @@ static int vhost_user_init(struct vhost_dev *dev, void *opaque)
>>           return err;
>>       }
>>
>> +    if (__virtio_has_feature(msg.u64, VIRTIO_F_ANY_LAYOUT)) {
>> +        dev->backend_features |= 1ULL << VIRTIO_F_ANY_LAYOUT;
>> +    }
>> +
>> +    if (__virtio_has_feature(msg.u64, VIRTIO_F_VERSION_1)) {
>> +        dev->backend_features |= 1ULL << VIRTIO_F_VERSION_1;
>> +    }
>> +
>> +    if (__virtio_has_feature(msg.u64, VIRTIO_NET_F_MRG_RXBUF)) {
>> +        dev->backend_features |= 1ULL << VIRTIO_NET_F_MRG_RXBUF;
>> +    }
>> +
>> +    if (__virtio_has_feature(msg.u64, VIRTIO_NET_F_MQ)) {
>> +        dev->backend_features |= 1ULL << VIRTIO_NET_F_MQ;
>> +    }
>> +
>
> A bit too early for MQ, isn't it?
For a complete solution, until we have it in protocol features

>
>>       if (__virtio_has_feature(msg.u64, VHOST_USER_F_PROTOCOL_FEATURES)) {
>>           dev->backend_features |= 1ULL << VHOST_USER_F_PROTOCOL_FEATURES;
>>
>> --
>> 2.1.0

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [Qemu-devel] [PATCH V2 2/2] tests/vhost-user: check vhost-user feature negotiation
  2015-07-30  6:47   ` Michael S. Tsirkin
@ 2015-07-30  7:30     ` Marcel Apfelbaum
  0 siblings, 0 replies; 7+ messages in thread
From: Marcel Apfelbaum @ 2015-07-30  7:30 UTC (permalink / raw)
  To: Michael S. Tsirkin; +Cc: jasowang, famz, qemu-devel

On 07/30/2015 09:47 AM, Michael S. Tsirkin wrote:
> On Wed, Jul 29, 2015 at 11:06:49PM +0300, Marcel Apfelbaum wrote:
>> Check the backend receives all declared virtio features
>> that are also supported by QEMU virtio.
>>
>> Signed-off-by: Marcel Apfelbaum <marcel@redhat.com>
>> ---
>>   tests/vhost-user-test.c | 19 ++++++++++++++++---
>>   1 file changed, 16 insertions(+), 3 deletions(-)
>>
>> diff --git a/tests/vhost-user-test.c b/tests/vhost-user-test.c
>> index 228acb6..019b880 100644
>> --- a/tests/vhost-user-test.c
>> +++ b/tests/vhost-user-test.c
>> @@ -17,6 +17,7 @@
>>   #include "sysemu/sysemu.h"
>>
>>   #include <linux/vhost.h>
>> +#include <linux/virtio_net.h>
>>   #include <sys/mman.h>
>>   #include <sys/vfs.h>
>>   #include <qemu/sockets.h>
>> @@ -297,14 +298,26 @@ static void chr_read(void *opaque, const uint8_t *buf, int size)
>>           /* send back features to qemu */
>>           msg.flags |= VHOST_USER_REPLY_MASK;
>>           msg.size = sizeof(m.u64);
>> -        msg.u64 = 0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES;
>> +        msg.u64 = (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES) |
>> +                  (0x1ULL << VIRTIO_F_ANY_LAYOUT) |
>> +                  (0x1ULL << VIRTIO_F_VERSION_1)  |
>> +                  (0x1ULL << VIRTIO_NET_F_MRG_RXBUF) |
>> +                  (0x1ULL << VIRTIO_NET_F_MQ);
>>           p = (uint8_t *) &msg;
>>           qemu_chr_fe_write_all(chr, p, VHOST_USER_HDR_SIZE + msg.size);
>>           break;
>>
>
> No, this mixes protocol features and virtio features.
> It's not what was intended.
You asked for a test that shows the problem, here it is.

We can change it as you wish, if you have a better idea, just say,
"no" doesn't help.

Thanks,
Marcel

>
>>       case VHOST_USER_SET_FEATURES:
>> -	g_assert_cmpint(msg.u64 & (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES),
>> -			!=, 0ULL);
>> +        g_assert_cmpint(msg.u64 & (0x1ULL << VHOST_USER_F_PROTOCOL_FEATURES),
>> +            !=, 0ULL);
>> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_F_ANY_LAYOUT),
>> +            !=, 0ULL);
>> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_F_VERSION_1),
>> +            !=, 0ULL);
>> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_NET_F_MRG_RXBUF),
>> +            !=, 0ULL);
>> +        g_assert_cmpint(msg.u64 & (0x1ULL << VIRTIO_NET_F_MQ),
>> +            !=, 0ULL);
>>           break;
>>
>>       case VHOST_USER_GET_PROTOCOL_FEATURES:
>> --
>> 2.1.0

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2015-07-30  7:30 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-07-29 20:06 [Qemu-devel] [PATCH V2 0/2] vhost-user: sync backend Marcel Apfelbaum
2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 1/2] vhost-user: sync backend features Marcel Apfelbaum
2015-07-30  6:47   ` Michael S. Tsirkin
2015-07-30  7:27     ` Marcel Apfelbaum
2015-07-29 20:06 ` [Qemu-devel] [PATCH V2 2/2] tests/vhost-user: check vhost-user feature negotiation Marcel Apfelbaum
2015-07-30  6:47   ` Michael S. Tsirkin
2015-07-30  7:30     ` Marcel Apfelbaum

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).