Existing virtio-blk protocol doesn't have DISCARD/WRITE ZEROES commands
support, this will impact the performance when using SSD backend over
file systems.
Commit 88c85538 "virtio-blk: add discard and write zeroes features to
specification"(see https://github.com/oasis-tcs/virtio-spec) extended
ex
On 2018年06月05日 20:33, Michael S. Tsirkin wrote:
I don't think this is sufficient.
If both primary and standby devices are present, a legacy guest without
support for the feature might see two devices with same mac and get
confused.
I think that we should only make primary visible after guest
On Tue, Jun 5, 2018 at 2:32 PM, Michael S. Tsirkin wrote:
> On Tue, Jun 05, 2018 at 02:16:44PM -0700, Siwei Liu wrote:
>> Good to see this discussion going. I share the same feeling that the
>> decision of plugging the primary (passthrough) should only be made
>> until guest driver acknowledges DR
On Tue, 2018-06-05 at 12:50 -0700, Nick Desaulniers wrote:
> On Tue, Jun 5, 2018 at 12:14 PM Joe Perches wrote:
> >
> > On Tue, 2018-06-05 at 10:23 -0700, Joe Perches wrote:
> > > Perhaps these are simpler as
> > >
> > > #define __inline__inline
> > > #define __inline inline
> >
> > Cu
On Tue, Jun 05, 2018 at 02:16:44PM -0700, Siwei Liu wrote:
> Good to see this discussion going. I share the same feeling that the
> decision of plugging the primary (passthrough) should only be made
> until guest driver acknowledges DRIVER_OK and _F_STANDBY.
> Architecturally this intelligence shou
On Tue, Jun 5, 2018 at 11:28 PM, Arnd Bergmann wrote:
> On Tue, Jun 5, 2018 at 7:05 PM, Nick Desaulniers
> wrote:
>>
>> The semantics of extern inline has changed since gnu89. This means that
>> folks using GCC versions >= 5.1 may see symbol redefinition errors at
>> link time for subdirs that ov
On Tue, Jun 5, 2018 at 7:05 PM, Nick Desaulniers
wrote:
>
> The semantics of extern inline has changed since gnu89. This means that
> folks using GCC versions >= 5.1 may see symbol redefinition errors at
> link time for subdirs that override KBUILD_CFLAGS (making the C standard
> used implicit) re
Good to see this discussion going. I share the same feeling that the
decision of plugging the primary (passthrough) should only be made
until guest driver acknowledges DRIVER_OK and _F_STANDBY.
Architecturally this intelligence should be baken to QEMU itself
rather than moving up to management stac
On Tue, 2018-06-05 at 10:05 -0700, Nick Desaulniers wrote:
> Functions marked extern inline do not emit an externally visible
> function when the gnu89 C standard is used. Some KBUILD Makefiles
> overwrite KBUILD_CFLAGS. This is an issue for GCC 5.1+ users as without
> an explicit C standard specif
On Tue, Jun 05, 2018 at 01:20:33PM -0700, Samudrala, Sridhar wrote:
>
> On 6/5/2018 5:33 AM, Michael S. Tsirkin wrote:
> > I don't think this is sufficient.
>
> Sure. This is not sufficient for a complete solution, but is Qemu the right
> place
> to manage primary/standby interfaces?
>
> I thin
On 6/5/2018 5:33 AM, Michael S. Tsirkin wrote:
I don't think this is sufficient.
Sure. This is not sufficient for a complete solution, but is Qemu the right
place
to manage primary/standby interfaces?
I think the other steps including plugging/unplugging the primary interface
needs
to hand
On Tue, 2018-06-05 at 10:23 -0700, Joe Perches wrote:
> Perhaps these are simpler as
>
> #define __inline__inline
> #define __inline inline
Currently, there are these uses of inline variants in the kernel
$ git grep -w inline | wc -l
68410
$ git grep -w __inline__ | wc -l
503
$ git grep
On 06/05/18 10:52, Nick Desaulniers wrote:
>
> Does the kernel have a different calling convention for 32b x86? How
> does that work? regparm=3? Does that need to be added to the
> declaration?
>
Yes, -mregparm=3. No, doesn't need to be added to the declaration.
>> Something like this added t
On 06/05/18 10:28, H. Peter Anvin wrote:
> On 06/05/18 10:05, Nick Desaulniers wrote:
>> +
>> +/*
>> + * void native_restore_fl(unsigned long flags)
>> + * %rdi: flags
>> + */
>> +ENTRY(native_restore_fl)
>> +push %_ASM_DI
>> +popf
>> +ret
>> +ENDPROC(native_restore_fl)
>> +EXPORT_SYMBO
On 06/05/18 10:05, Nick Desaulniers wrote:
> +
> +/*
> + * void native_restore_fl(unsigned long flags)
> + * %rdi: flags
> + */
> +ENTRY(native_restore_fl)
> + push %_ASM_DI
> + popf
> + ret
> +ENDPROC(native_restore_fl)
> +EXPORT_SYMBOL(native_restore_fl)
>
To work on i386, this woul
I don't think this is sufficient.
If both primary and standby devices are present, a legacy guest without
support for the feature might see two devices with same mac and get
confused.
I think that we should only make primary visible after guest acked the
backup feature bit.
And on reset or when
On Tue, Jun 05, 2018 at 09:36:53AM +0800, Tiwei Bie wrote:
> On Mon, Jun 04, 2018 at 07:32:25PM +0300, Michael S. Tsirkin wrote:
> > On Fri, Jun 01, 2018 at 12:02:39PM +0800, Tiwei Bie wrote:
> > > There is a new feature bit allocated in virtio spec to
> > > support SR-IOV (Single Root I/O Virtuali
Signed-off-by: Tiwei Bie
---
drivers/s390/virtio/virtio_ccw.c | 8
drivers/virtio/virtio_ring.c | 2 ++
2 files changed, 10 insertions(+)
diff --git a/drivers/s390/virtio/virtio_ccw.c b/drivers/s390/virtio/virtio_ccw.c
index 8f5c1d7f751a..ff5b85736d8d 100644
--- a/drivers/s390/virti
This commit introduces the EVENT_IDX support in packed ring.
Signed-off-by: Tiwei Bie
---
drivers/virtio/virtio_ring.c | 74
1 file changed, 67 insertions(+), 7 deletions(-)
diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
index 983ce
This commit introduces the support (without EVENT_IDX) for
packed ring.
Signed-off-by: Tiwei Bie
---
drivers/virtio/virtio_ring.c | 486 ++-
1 file changed, 479 insertions(+), 7 deletions(-)
diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
This commit introduces the support for creating packed ring.
All split ring specific functions are added _split suffix.
Some necessary stubs for packed ring are also added.
Signed-off-by: Tiwei Bie
---
drivers/virtio/virtio_ring.c | 801 +++
include/linux/virtio_r
Signed-off-by: Tiwei Bie
---
include/uapi/linux/virtio_config.h | 5 -
include/uapi/linux/virtio_ring.h | 36 ++
2 files changed, 40 insertions(+), 1 deletion(-)
diff --git a/include/uapi/linux/virtio_config.h
b/include/uapi/linux/virtio_config.h
index 308e209
Hello everyone,
This RFC implements packed ring support in virtio driver.
Some functional tests have been done with Jason's
packed ring implementation in vhost (RFC v5):
https://lwn.net/Articles/755862/
Both of ping and netperf worked as expected.
TODO:
- Refinements (for code and commit log);
23 matches
Mail list logo
