Re: svn commit: r280154 - head/usr.sbin/bhyve

2015-03-16 Thread Willem Jan Withagen
On 16-3-2015 22:24, Alexander Motin wrote:
 On 16.03.2015 23:13, Bryan Venteicher wrote:
 On Mon, Mar 16, 2015 at 3:13 PM, Alexander Motin m...@freebsd.org
 mailto:m...@freebsd.org wrote:
 Author: mav
 Date: Mon Mar 16 20:13:25 2015
 New Revision: 280154
 URL: https://svnweb.freebsd.org/changeset/base/280154

 Log:
   Report that we may have write cache, and that we do support FLUSH.

   FreeBSD guest driver does not use that legacy flag, but Linux
 seems does.

 The guest driver calls it by a prior name: VIRTIO_BLK_F_WCE. The write
 cache stuff went through several iterations, and goes away in VirtIO 1.0
 specification.
 
 Yes, I've found that in the driver after the commit. But as I can see
 driver does nothing about it, except reporting via sysctl. I was
 confused by not seeing any use of that flag in data path.
 

There was a point in time that this Flush was in the way of booting
certain Linux versions. I guess that accepting, but ignoring, which what
I did at a certain point in my code, got me going. Until extra code made
it in the code and the command got us further with our progress in Bhyve.

--WjW
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org


Re: svn commit: r280154 - head/usr.sbin/bhyve

2015-03-16 Thread Alexander Motin
On 16.03.2015 23:13, Bryan Venteicher wrote:
 On Mon, Mar 16, 2015 at 3:13 PM, Alexander Motin m...@freebsd.org
 mailto:m...@freebsd.org wrote:
 Author: mav
 Date: Mon Mar 16 20:13:25 2015
 New Revision: 280154
 URL: https://svnweb.freebsd.org/changeset/base/280154
 
 Log:
   Report that we may have write cache, and that we do support FLUSH.
 
   FreeBSD guest driver does not use that legacy flag, but Linux
 seems does.
 
 The guest driver calls it by a prior name: VIRTIO_BLK_F_WCE. The write
 cache stuff went through several iterations, and goes away in VirtIO 1.0
 specification.

Yes, I've found that in the driver after the commit. But as I can see
driver does nothing about it, except reporting via sysctl. I was
confused by not seeing any use of that flag in data path.

-- 
Alexander Motin
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org


Re: svn commit: r280154 - head/usr.sbin/bhyve

2015-03-16 Thread Bryan Venteicher
On Mon, Mar 16, 2015 at 3:13 PM, Alexander Motin m...@freebsd.org wrote:

 Author: mav
 Date: Mon Mar 16 20:13:25 2015
 New Revision: 280154
 URL: https://svnweb.freebsd.org/changeset/base/280154

 Log:
   Report that we may have write cache, and that we do support FLUSH.

   FreeBSD guest driver does not use that legacy flag, but Linux seems does.



The guest driver calls it by a prior name: VIRTIO_BLK_F_WCE. The write
cache stuff went through several iterations, and goes away in VirtIO 1.0
specification.



   MFC after:2 weeks

 Modified:
   head/usr.sbin/bhyve/pci_virtio_block.c

 Modified: head/usr.sbin/bhyve/pci_virtio_block.c

 ==
 --- head/usr.sbin/bhyve/pci_virtio_block.c  Mon Mar 16 20:00:09 2015
   (r280153)
 +++ head/usr.sbin/bhyve/pci_virtio_block.c  Mon Mar 16 20:13:25 2015
   (r280154)
 @@ -64,6 +64,7 @@ __FBSDID($FreeBSD$);
  /* Capability bits */
  #defineVTBLK_F_SEG_MAX (1  2)/* Maximum request
 segments */
  #defineVTBLK_F_BLK_SIZE(1  6)/* cfg block size
 valid */
 +#defineVTBLK_F_FLUSH   (1  9)/* Cache flush
 support */
  #defineVTBLK_F_TOPOLOGY(1  10)   /* Optimal I/O
 alignment */

  /*
 @@ -72,6 +73,7 @@ __FBSDID($FreeBSD$);
  #define VTBLK_S_HOSTCAPS  \
( VTBLK_F_SEG_MAX  | \
  VTBLK_F_BLK_SIZE | \
 +VTBLK_F_FLUSH| \
  VTBLK_F_TOPOLOGY | \
  VIRTIO_RING_F_INDIRECT_DESC )  /* indirect descriptors */



___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org


svn commit: r280154 - head/usr.sbin/bhyve

2015-03-16 Thread Alexander Motin
Author: mav
Date: Mon Mar 16 20:13:25 2015
New Revision: 280154
URL: https://svnweb.freebsd.org/changeset/base/280154

Log:
  Report that we may have write cache, and that we do support FLUSH.
  
  FreeBSD guest driver does not use that legacy flag, but Linux seems does.
  
  MFC after:2 weeks

Modified:
  head/usr.sbin/bhyve/pci_virtio_block.c

Modified: head/usr.sbin/bhyve/pci_virtio_block.c
==
--- head/usr.sbin/bhyve/pci_virtio_block.c  Mon Mar 16 20:00:09 2015
(r280153)
+++ head/usr.sbin/bhyve/pci_virtio_block.c  Mon Mar 16 20:13:25 2015
(r280154)
@@ -64,6 +64,7 @@ __FBSDID($FreeBSD$);
 /* Capability bits */
 #defineVTBLK_F_SEG_MAX (1  2)/* Maximum request 
segments */
 #defineVTBLK_F_BLK_SIZE(1  6)/* cfg block size valid 
*/
+#defineVTBLK_F_FLUSH   (1  9)/* Cache flush support 
*/
 #defineVTBLK_F_TOPOLOGY(1  10)   /* Optimal I/O 
alignment */
 
 /*
@@ -72,6 +73,7 @@ __FBSDID($FreeBSD$);
 #define VTBLK_S_HOSTCAPS  \
   ( VTBLK_F_SEG_MAX  | \
 VTBLK_F_BLK_SIZE | \
+VTBLK_F_FLUSH| \
 VTBLK_F_TOPOLOGY | \
 VIRTIO_RING_F_INDIRECT_DESC )  /* indirect descriptors */
 
___
svn-src-all@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to svn-src-all-unsubscr...@freebsd.org