On Mon, 2023-10-23 at 16:14 -0700, Si-Wei Liu wrote:
> As subject. There's a vhost_vdpa_reset() done earlier before
> vhost_dev is initialized via vhost_dev_init(), ending up with
> NULL pointer dereference. Fix is to check if vqs is initialized
> before checking backend features and resetting the
On Fri, 2023-10-20 at 18:01 +0200, Eugenio Perez Martin wrote:
> On Wed, Oct 18, 2023 at 7:21 PM Dragos Tatulea wrote:
> >
> > On Wed, 2023-10-18 at 20:14 +0300, Dragos Tatulea wrote:
> > > The current flow for updating an mr works directly on mvdev->mr which
> > > makes it cumbersome to handle m
On Wed, 2023-10-18 at 20:14 +0300, Dragos Tatulea wrote:
> The current flow for updating an mr works directly on mvdev->mr which
> makes it cumbersome to handle multiple new mr structs.
>
> This patch makes the flow more straightforward by having
> mlx5_vdpa_create_mr return a new mr which will up
For the following sequence:
- cvq group is in ASID 0
- .set_map(1, cvq_iotlb)
- .set_group_asid(cvq_group, 1)
... the cvq mapping from ASID 0 will be used. This is not always correct
behaviour.
This patch adds support for the above mentioned flow by saving the iotlb
on each .set_map and updating
They will be used in a follow-up patch.
For dup_iotlb, avoid the src == dst case. This is an error.
Acked-by: Jason Wang
Acked-by: Eugenio Pérez
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mr.c | 19 +++
1 file changed, 11 insertions(+), 8 deletions(-)
diff --git
Vq descriptor mappings are supported in hardware by filling in an
additional mkey which contains the descriptor mappings to the hw vq.
A previous patch in this series added support for hw mkey (mr) creation
for ASID 1.
This patch fills in both the vq data and vq descriptor mkeys based on
group AS
The current flow for updating an mr works directly on mvdev->mr which
makes it cumbersome to handle multiple new mr structs.
This patch makes the flow more straightforward by having
mlx5_vdpa_create_mr return a new mr which will update the old mr (if
any). The old mr will be deleted and unlinked f
Introduce the vq descriptor group and mr per ASID. Until now
.set_map on ASID 1 was only updating the cvq iotlb. From now on it also
creates a mkey for it. The current patch doesn't use it but follow-up
patches will add hardware support for mapping the vq descriptors.
Acked-by: Jason Wang
Acked-b
The mutex is named like it is supposed to protect only the mkey but in
reality it is a global lock for all mr resources.
Shift the mutex to it's rightful location (struct mlx5_vdpa_dev) and
give it a more appropriate name.
Signed-off-by: Dragos Tatulea
Acked-by: Eugenio Pérez
Acked-by: Jason Wa
This patch adapts the mr creation/deletion code to be able to work with
any given mr struct pointer. All the APIs are adapted to take an extra
parameter for the mr.
mlx5_vdpa_create/delete_mr doesn't need a ASID parameter anymore. The
check is done in the caller instead (mlx5_set_map).
This chang
The reslock is taken while refresh is called but iommu_lock is more
specific to this resource. So take the iommu_lock during cvq iotlb
refresh.
Based on Eugenio's patch [0].
[0] https://lore.kernel.org/lkml/20230112142218.725622-4-epere...@redhat.com/
Acked-by: Jason Wang
Suggested-by: Eugenio
The handling of the cvq iotlb is currently coupled with the creation
and destruction of the hardware mkeys (mr).
This patch moves cvq iotlb handling into its own function and shifts it
to a scope that is not related to mr handling. As cvq handling is just a
prune_iotlb + dup_iotlb cycle, put it al
From: Si-Wei Liu
With _F_DESC_ASID backend feature, the device can now support the
VHOST_VDPA_GET_VRING_DESC_GROUP ioctl, and it may expose the descriptor
table (including avail and used ring) in a different group than the
buffers it contains. This new uAPI will fetch the group ID of the
descript
Make mlx5_destroy_mr symmetric to mlx5_create_mr.
Acked-by: Jason Wang
Acked-by: Eugenio Pérez
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 4 ++--
drivers/vdpa/mlx5/core/mr.c| 6 +++---
drivers/vdpa/mlx5/net/mlx5_vnet.c | 12 ++--
3 files changed,
Now that the cvq code is out of mlx5_vdpa_create/destroy_mr, the "dvq"
functions can be folded into their callers.
Having "dvq" in the naming will no longer be accurate in the downstream
patches.
Acked-by: Jason Wang
Acked-by: Eugenio Pérez
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/
From: Si-Wei Liu
Userspace knows if the device has dedicated descriptor group or not
by checking this feature bit.
It's only exposed if the vdpa driver backend implements the
.get_vq_desc_group() operation callback. Userspace trying to negotiate
this feature when it or the dependent _F_IOTLB_ASI
Necessary for upcoming cvq separation from mr allocation.
Acked-by: Jason Wang
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 1 +
drivers/vdpa/mlx5/core/mr.c| 5 +
drivers/vdpa/mlx5/net/mlx5_vnet.c | 4 ++--
3 files changed, 8 insertions(+), 2 deletions(-)
From: Si-Wei Liu
In some cases, the access to the virtqueue's descriptor area, device
and driver areas (precluding indirect descriptor table in guest memory)
may have to be confined to a different address space than where its
buffers reside. Without loss of simplicity and generality with already
Necessary for improved live migration flow. Actual support will be added
in a downstream patch.
Signed-off-by: Dragos Tatulea
Reviewed-by: Gal Pressman
---
include/linux/mlx5/mlx5_ifc.h | 8 +++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/include/linux/mlx5/mlx5_ifc.h b/in
This patch series adds support for vq descriptor table mappings which
are used to improve vdpa live migration downtime. The improvement comes
from using smaller mappings which take less time to create and destroy
in hw.
The first part adds the vdpa core changes from Si-Wei [0].
The second part ad
For the following sequence:
- cvq group is in ASID 0
- .set_map(1, cvq_iotlb)
- .set_group_asid(cvq_group, 1)
... the cvq mapping from ASID 0 will be used. This is not always correct
behaviour.
This patch adds support for the above mentioned flow by saving the iotlb
on each .set_map and updating
They will be used in a follow-up patch.
For dup_iotlb, avoid the src == dst case. This is an error.
Acked-by: Eugenio Pérez
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mr.c | 19 +++
1 file changed, 11 insertions(+), 8 deletions(-)
diff --git a/drivers/vdpa/mlx5/c
Vq descriptor mappings are supported in hardware by filling in an
additional mkey which contains the descriptor mappings to the hw vq.
A previous patch in this series added support for hw mkey (mr) creation
for ASID 1.
This patch fills in both the vq data and vq descriptor mkeys based on
group AS
Introduce the vq descriptor group and mr per ASID. Until now
.set_map on ASID 1 was only updating the cvq iotlb. From now on it also
creates a mkey for it. The current patch doesn't use it but follow-up
patches will add hardware support for mapping the vq descriptors.
Acked-by: Eugenio Pérez
Sign
The current flow for updating an mr works directly on mvdev->mr which
makes it cumbersome to handle multiple new mr structs.
This patch makes the flow more straightforward by having
mlx5_vdpa_create_mr return a new mr which will update the old mr (if
any). The old mr will be deleted and unlinked f
This patch adapts the mr creation/deletion code to be able to work with
any given mr struct pointer. All the APIs are adapted to take an extra
parameter for the mr.
mlx5_vdpa_create/delete_mr doesn't need a ASID parameter anymore. The
check is done in the caller instead (mlx5_set_map).
This chang
The mutex is named like it is supposed to protect only the mkey but in
reality it is a global lock for all mr resources.
Shift the mutex to it's rightful location (struct mlx5_vdpa_dev) and
give it a more appropriate name.
Signed-off-by: Dragos Tatulea
Acked-by: Eugenio Pérez
---
drivers/vdpa/
Make mlx5_destroy_mr symmetric to mlx5_create_mr.
Acked-by: Jason Wang
Acked-by: Eugenio Pérez
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 4 ++--
drivers/vdpa/mlx5/core/mr.c| 6 +++---
drivers/vdpa/mlx5/net/mlx5_vnet.c | 12 ++--
3 files changed,
Now that the cvq code is out of mlx5_vdpa_create/destroy_mr, the "dvq"
functions can be folded into their callers.
Having "dvq" in the naming will no longer be accurate in the downstream
patches.
Acked-by: Jason Wang
Acked-by: Eugenio Pérez
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/
Necessary for upcoming cvq separation from mr allocation.
Acked-by: Jason Wang
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 1 +
drivers/vdpa/mlx5/core/mr.c| 5 +
drivers/vdpa/mlx5/net/mlx5_vnet.c | 4 ++--
3 files changed, 8 insertions(+), 2 deletions(-)
The handling of the cvq iotlb is currently coupled with the creation
and destruction of the hardware mkeys (mr).
This patch moves cvq iotlb handling into its own function and shifts it
to a scope that is not related to mr handling. As cvq handling is just a
prune_iotlb + dup_iotlb cycle, put it al
The reslock is taken while refresh is called but iommu_lock is more
specific to this resource. So take the iommu_lock during cvq iotlb
refresh.
Based on Eugenio's patch [0].
[0] https://lore.kernel.org/lkml/20230112142218.725622-4-epere...@redhat.com/
Acked-by: Jason Wang
Suggested-by: Eugenio
From: Si-Wei Liu
With _F_DESC_ASID backend feature, the device can now support the
VHOST_VDPA_GET_VRING_DESC_GROUP ioctl, and it may expose the descriptor
table (including avail and used ring) in a different group than the
buffers it contains. This new uAPI will fetch the group ID of the
descript
From: Si-Wei Liu
In some cases, the access to the virtqueue's descriptor area, device
and driver areas (precluding indirect descriptor table in guest memory)
may have to be confined to a different address space than where its
buffers reside. Without loss of simplicity and generality with already
From: Si-Wei Liu
Userspace knows if the device has dedicated descriptor group or not
by checking this feature bit.
It's only exposed if the vdpa driver backend implements the
.get_vq_desc_group() operation callback. Userspace trying to negotiate
this feature when it or the dependent _F_IOTLB_ASI
Necessary for improved live migration flow. Actual support will be added
in a downstream patch.
Reviewed-by: Gal Pressman
Signed-off-by: Dragos Tatulea
---
include/linux/mlx5/mlx5_ifc.h | 8 +++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/include/linux/mlx5/mlx5_ifc.h b/in
This patch series adds support for vq descriptor table mappings which
are used to improve vdpa live migration downtime. The improvement comes
from using smaller mappings which take less time to create and destroy
in hw.
The first part adds the vdpa core changes from Si-Wei [0].
The second part ad
On Mon, 2023-10-09 at 14:39 +0800, Jason Wang wrote:
> On Sun, Oct 8, 2023 at 8:05 PM Dragos Tatulea wrote:
> >
> > On Sun, 2023-10-08 at 12:25 +0800, Jason Wang wrote:
> > > On Tue, Sep 26, 2023 at 3:21 PM Dragos Tatulea
> > > wrote:
> > > >
> > > > On Tue, 2023-09-26 at 12:44 +0800, Jason Wan
On Sun, 2023-10-08 at 12:25 +0800, Jason Wang wrote:
> On Tue, Sep 26, 2023 at 3:21 PM Dragos Tatulea wrote:
> >
> > On Tue, 2023-09-26 at 12:44 +0800, Jason Wang wrote:
> > > On Tue, Sep 12, 2023 at 9:02 PM Dragos Tatulea
> > > wrote:
> > > >
> > > > This patch adapts the mr creation/deletion
On Thu, 2023-10-05 at 13:31 -0400, Michael S. Tsirkin wrote:
> On Thu, Sep 28, 2023 at 07:45:11PM +0300, Dragos Tatulea wrote:
> > This patch series adds support for vq descriptor table mappings which
> > are used to improve vdpa live migration downtime. The improvement comes
> > from using smaller
On Thu, 2023-10-05 at 12:41 +0200, Eugenio Perez Martin wrote:
> On Thu, Sep 28, 2023 at 6:50 PM Dragos Tatulea wrote:
> >
> > For the following sequence:
> > - cvq group is in ASID 0
> > - .set_map(1, cvq_iotlb)
> > - .set_group_asid(cvq_group, 1)
> >
> > ... the cvq mapping from ASID 0 will be
On Thu, 2023-10-05 at 12:14 +0200, Eugenio Perez Martin wrote:
> On Thu, Sep 28, 2023 at 6:50 PM Dragos Tatulea wrote:
> >
> > They will be used in a followup patch.
> >
> > Signed-off-by: Dragos Tatulea
> > ---
> > drivers/vdpa/mlx5/core/mr.c | 16
> > 1 file changed, 8 inser
On Thu, 2023-10-05 at 11:42 +0200, Eugenio Perez Martin wrote:
> On Thu, Sep 28, 2023 at 6:50 PM Dragos Tatulea wrote:
> >
> > Vq descriptor mappings are supported in hardware by filling in an
> > additional mkey which contains the descriptor mappings to the hw vq.
> >
> > A previous patch in th
On Wed, 2023-10-04 at 20:53 +0200, Eugenio Perez Martin wrote:
> On Thu, Sep 28, 2023 at 6:50 PM Dragos Tatulea wrote:
> >
> > Introduce the vq descriptor group and ASID 1. Until now .set_map on ASID
>
> s/ASID/vq group/?
>
Oh, indeed.
> > 1 was only updating the cvq iotlb. From now on it also
On Thu, 2023-09-28 at 19:45 +0300, Dragos Tatulea wrote:
> This patch adapts the mr creation/deletion code to be able to work with
> any given mr struct pointer. All the APIs are adapted to take an extra
> parameter for the mr.
>
> mlx5_vdpa_create/delete_mr doesn't need a ASID parameter anymore.
They will be used in a followup patch.
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mr.c | 16
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/vdpa/mlx5/core/mr.c b/drivers/vdpa/mlx5/core/mr.c
index 3dee6d9bed6b..a4135c16b5bf 100644
--- a/driver
For the following sequence:
- cvq group is in ASID 0
- .set_map(1, cvq_iotlb)
- .set_group_asid(cvq_group, 1)
... the cvq mapping from ASID 0 will be used. This is not always correct
behaviour.
This patch adds support for the above mentioned flow by saving the iotlb
on each .set_map and updating
Vq descriptor mappings are supported in hardware by filling in an
additional mkey which contains the descriptor mappings to the hw vq.
A previous patch in this series added support for hw mkey (mr) creation
for ASID 1.
This patch fills in both the vq data and vq descriptor mkeys based on
group AS
Introduce the vq descriptor group and ASID 1. Until now .set_map on ASID
1 was only updating the cvq iotlb. From now on it also creates a mkey
for it. The current patch doesn't use it but follow-up patches will
add hardware support for mapping the vq descriptors.
Signed-off-by: Dragos Tatulea
---
The mutex is named like it is supposed to protect only the mkey but in
reality it is a global lock for all mr resources.
Shift the mutex to it's rightful location (struct mlx5_vdpa_dev) and
give it a more appropriate name.
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h |
Now that the cvq code is out of mlx5_vdpa_create/destroy_mr, the "dvq"
functions can be folded into their callers.
Having "dvq" in the naming will no longer be accurate in the downstream
patches.
Acked-by: Jason Wang
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mr.c | 16 +-
The current flow for updating an mr works directly on mvdev->mr which
makes it cumbersome to handle multiple new mr structs.
This patch makes the flow more straightforward by having
mlx5_vdpa_create_mr return a new mr which will update the old mr (if
any). The old mr will be deleted and unlinked f
The reslock is taken while refresh is called but iommu_lock is more
specific to this resource. So take the iommu_lock during cvq iotlb
refresh.
Based on Eugenio's patch [0].
[0] https://lore.kernel.org/lkml/20230112142218.725622-4-epere...@redhat.com/
Acked-by: Jason Wang
Suggested-by: Eugenio
This patch adapts the mr creation/deletion code to be able to work with
any given mr struct pointer. All the APIs are adapted to take an extra
parameter for the mr.
mlx5_vdpa_create/delete_mr doesn't need a ASID parameter anymore. The
check is done in the caller instead (mlx5_set_map).
This chang
Make mlx5_destroy_mr symmetric to mlx5_create_mr.
Acked-by: Jason Wang
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 4 ++--
drivers/vdpa/mlx5/core/mr.c| 6 +++---
drivers/vdpa/mlx5/net/mlx5_vnet.c | 12 ++--
3 files changed, 11 insertions(+), 11 dele
The handling of the cvq iotlb is currently coupled with the creation
and destruction of the hardware mkeys (mr).
This patch moves cvq iotlb handling into its own function and shifts it
to a scope that is not related to mr handling. As cvq handling is just a
prune_iotlb + dup_iotlb cycle, put it al
From: Si-Wei Liu
With _F_DESC_ASID backend feature, the device can now support the
VHOST_VDPA_GET_VRING_DESC_GROUP ioctl, and it may expose the descriptor
table (including avail and used ring) in a different group than the
buffers it contains. This new uAPI will fetch the group ID of the
descript
Necessary for upcoming cvq separation from mr allocation.
Acked-by: Jason Wang
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 1 +
drivers/vdpa/mlx5/core/mr.c| 5 +
drivers/vdpa/mlx5/net/mlx5_vnet.c | 4 ++--
3 files changed, 8 insertions(+), 2 deletions(-)
From: Si-Wei Liu
Userspace knows if the device has dedicated descriptor group or not
by checking this feature bit.
It's only exposed if the vdpa driver backend implements the
.get_vq_desc_group() operation callback. Userspace trying to negotiate
this feature when it or the dependent _F_IOTLB_ASI
From: Si-Wei Liu
In some cases, the access to the virtqueue's descriptor area, device
and driver areas (precluding indirect descriptor table in guest memory)
may have to be confined to a different address space than where its
buffers reside. Without loss of simplicity and generality with already
Necessary for improved live migration flow. Actual support will be added
in a downstream patch.
Reviewed-by: Gal Pressman
Signed-off-by: Dragos Tatulea
---
include/linux/mlx5/mlx5_ifc.h | 8 +++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/include/linux/mlx5/mlx5_ifc.h b/in
This patch series adds support for vq descriptor table mappings which
are used to improve vdpa live migration downtime. The improvement comes
from using smaller mappings which take less time to create and destroy
in hw.
The first part adds the vdpa core changes from Si-Wei [0].
The second part ad
This patch series adds support for vq descriptor table mappings which
are used to improve vdpa live migration downtime. The improvement comes
from using smaller mappings which take less time to create and destroy
in hw.
The first part adds the vdpa core changes from Si-Wei [0].
The second part ad
On Tue, 2023-09-26 at 05:22 -0700, Si-Wei Liu wrote:
>
>
> On 9/25/2023 12:59 AM, Dragos Tatulea wrote:
> > On Tue, 2023-09-12 at 16:01 +0300, Dragos Tatulea wrote:
> > > This patch series adds support for vq descriptor table mappings which
> > > are used to improve vdpa live migration downtime.
On Tue, 2023-09-26 at 12:44 +0800, Jason Wang wrote:
> On Tue, Sep 12, 2023 at 9:02 PM Dragos Tatulea wrote:
> >
> > This patch adapts the mr creation/deletion code to be able to work with
> > any given mr struct pointer. All the APIs are adapted to take an extra
> > parameter for the mr.
> >
>
As Eli Cohen moved to other work, I'll be the contact point for
mlx5_vdpa.
Acked-by: Jason Wang
Signed-off-by: Dragos Tatulea
---
Changes since v1:
- Fix alphabetical sorting.
- Make naming consistent with other MELLANOX entries.
---
MAINTAINERS | 6 ++
1 file changed, 6 insertions(+)
diff
On Tue, 2023-09-12 at 16:01 +0300, Dragos Tatulea wrote:
> This patch series adds support for vq descriptor table mappings which
> are used to improve vdpa live migration downtime. The improvement comes
> from using smaller mappings which take less time to create and destroy
> in hw.
>
Gentle ping
On Tue, 2023-08-29 at 20:40 +0300, Dragos Tatulea wrote:
> The error path in setup_driver deletes the debugfs entry but doesn't
> clear the pointer. During .dev_del the invalid pointer will be released
> again causing a crash.
>
> This patch fixes the issue by always clearing the debugfs entry in
On Thu, 2023-08-31 at 18:50 +0300, Dragos Tatulea wrote:
> A firmware error is triggered when configuring a 9k MTU on the PF after
> switching to switchdev mode and then using a vdpa device with larger
> (1k) rings:
> mlx5_cmd_out_err: CREATE_GENERAL_OBJECT(0xa00) op_mod(0xd) failed, status bad
> r
Vq descriptor mappings are supported in hardware by filling in an
additional mkey which contains the descriptor mappings to the hw vq.
A previous patch in this series added support for hw mkey (mr) creation
for ASID 1.
This patch fills in both the vq data and vq descriptor mkeys based on
group AS
Introduce the vq descriptor group and ASID 1. Until now .set_map on ASID
1 was only updating the cvq iotlb. From now on it also creates a mkey
for it. The current patch doesn't use it but follow-up patches will
add hardware support for mapping the vq descriptors.
Signed-off-by: Dragos Tatulea
---
The current flow for updating an mr works directly on mvdev->mr which
makes it cumbersome to handle multiple new mr structs.
This patch makes the flow more straightforward by having
mlx5_vdpa_create_mr return a new mr which will update the old mr (if
any). The old mr will be deleted and unlinked f
They will be used in a followup patch.
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mr.c | 16
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/vdpa/mlx5/core/mr.c b/drivers/vdpa/mlx5/core/mr.c
index 3dee6d9bed6b..a4135c16b5bf 100644
--- a/driver
For the following sequence:
- cvq group is in ASID 0
- .set_map(1, cvq_iotlb)
- .set_group_asid(cvq_group, 1)
... the cvq mapping from ASID 0 will be used. This is not always correct
behaviour.
This patch adds support for the above mentioned flow by saving the iotlb
on each .set_map and updating
The mutex is named like it is supposed to protect only the mkey but in
reality it is a global lock for all mr resources.
Shift the mutex to it's rightful location (struct mlx5_vdpa_dev) and
give it a more appropriate name.
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h |
The handling of the cvq iotlb is currently coupled with the creation
and destruction of the hardware mkeys (mr).
This patch moves cvq iotlb handling into its own function and shifts it
to a scope that is not related to mr handling. As cvq handling is just a
prune_iotlb + dup_iotlb cycle, put it al
Make mlx5_destroy_mr symmetric to mlx5_create_mr.
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 4 ++--
drivers/vdpa/mlx5/core/mr.c| 6 +++---
drivers/vdpa/mlx5/net/mlx5_vnet.c | 12 ++--
3 files changed, 11 insertions(+), 11 deletions(-)
diff --git a
Now that the cvq code is out of mlx5_vdpa_create/destroy_mr, the "dvq"
functions can be folded into their callers.
Having "dvq" in the naming will no longer be accurate in the downstream
patches.
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mr.c | 16 +---
1 file changed
This patch adapts the mr creation/deletion code to be able to work with
any given mr struct pointer. All the APIs are adapted to take an extra
parameter for the mr.
mlx5_vdpa_create/delete_mr doesn't need a ASID parameter anymore. The
check is done in the caller instead (mlx5_set_map).
This chang
The reslock is taken while refresh is called but iommu_lock is more
specific to this resource. So take the iommu_lock during cvq iotlb
refresh.
Based on Eugenio's patch [0].
[0] https://lore.kernel.org/lkml/20230112142218.725622-4-epere...@redhat.com/
Suggested-by: Eugenio Pérez
Signed-off-by:
From: Si-Wei Liu
In some cases, the access to the virtqueue's descriptor area, device
and driver areas (precluding indirect descriptor table in guest memory)
may have to be confined to a different address space than where its
buffers reside. Without loss of simplicity and generality with already
From: Si-Wei Liu
Userspace knows if the device has dedicated descriptor group or not
by checking this feature bit.
It's only exposed if the vdpa driver backend implements the
.get_vq_desc_group() operation callback. Userspace trying to negotiate
this feature when it or the dependent _F_IOTLB_ASI
Necessary for upcoming cvq separation from mr allocation.
Signed-off-by: Dragos Tatulea
---
drivers/vdpa/mlx5/core/mlx5_vdpa.h | 1 +
drivers/vdpa/mlx5/core/mr.c| 5 +
drivers/vdpa/mlx5/net/mlx5_vnet.c | 4 ++--
3 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/drivers/
From: Si-Wei Liu
With _F_DESC_ASID backend feature, the device can now support the
VHOST_VDPA_GET_VRING_DESC_GROUP ioctl, and it may expose the descriptor
table (including avail and used ring) in a different group than the
buffers it contains. This new uAPI will fetch the group ID of the
descript
This patch series adds support for vq descriptor table mappings which
are used to improve vdpa live migration downtime. The improvement comes
from using smaller mappings which take less time to create and destroy
in hw.
The first part adds the vdpa core changes from Si-Wei [0].
The second part ad
Hi Jason,
On Mon, 2023-09-11 at 14:43 +0800, Jason Wang wrote:
> On Sat, Sep 9, 2023 at 4:45 PM Si-Wei Liu wrote:
> >
> > Following patchset introduces dedicated group for descriptor table to
> > reduce live migration downtime when passthrough VQ is being switched
> > to shadow VQ. This RFC v2 i
A firmware error is triggered when configuring a 9k MTU on the PF after
switching to switchdev mode and then using a vdpa device with larger
(1k) rings:
mlx5_cmd_out_err: CREATE_GENERAL_OBJECT(0xa00) op_mod(0xd) failed, status bad
resource(0x5), syndrome (0xf6db90), err(-22)
This is due to the fa
A firmware error is triggered when configuring a 9k MTU on the PF after
switching to switchdev mode and then using a vdpa device with larger
(1k) rings:
mlx5_cmd_out_err: CREATE_GENERAL_OBJECT(0xa00) op_mod(0xd) failed, status bad
resource(0x5), syndrome (0xf6db90), err(-22)
This is due to the fa
The error path in setup_driver deletes the debugfs entry but doesn't
clear the pointer. During .dev_del the invalid pointer will be released
again causing a crash.
This patch fixes the issue by always clearing the debugfs entry in
mlx5_vdpa_remove_debugfs. Also, stop removing the debugfs entry in
As Eli Cohen moved to other work, I'll be the contact point for
mlx5_vdpa.
Signed-off-by: Dragos Tatulea
---
MAINTAINERS | 6 ++
1 file changed, 6 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 9a5863f1b016..c9a9259f4d37 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -13555,6 +13
On Mon, 2023-08-14 at 18:43 -0700, Si-Wei Liu wrote:
> This patch is based on top of the "vdpa/mlx5: Fixes
> for ASID handling" series [1].
>
> [1] vdpa/mlx5: Fixes for ASID handling
> https://lore.kernel.org/virtualization/20230802171231.11001-1-dtatu...@nvidia.com/
>
> Signed-off-by: Si-Wei Liu
On Wed, 2023-08-09 at 09:42 +0800, Jason Wang wrote:
> On Tue, Aug 8, 2023 at 3:24 PM Dragos Tatulea wrote:
> >
> > On Tue, 2023-08-08 at 10:57 +0800, Jason Wang wrote:
> > > On Thu, Aug 3, 2023 at 7:40 PM Dragos Tatulea wrote:
> > > >
> > > > On Thu, 2023-08-03 at 16:03 +0800, Jason Wang wrote
On Thu, 2023-08-10 at 04:54 -0400, Michael S. Tsirkin wrote:
> On Wed, Aug 02, 2023 at 08:12:16PM +0300, Dragos Tatulea wrote:
> > This patch series is based on Eugenio's fix for handling CVQs in
> > a different ASID [0].
> >
> > The first patch is the actual fix.
> >
> > The next 2 patches are f
On Tue, 2023-08-08 at 23:13 -0400, Jason Wang wrote:
> Commit 25266128fe16 ("virtio-net: fix race between set queues and
> probe") tries to fix the race between set queues and probe by calling
> _virtnet_set_queues() before DRIVER_OK is set. This violates virtio
> spec. Fixing this by setting queue
On Tue, 2023-07-25 at 03:20 -0400, Jason Wang wrote:
> A race were found where set_channels could be called after registering
> but before virtnet_set_queues() in virtnet_probe(). Fixing this by
> moving the virtnet_set_queues() before netdevice registering. While at
> it, use _virtnet_set_queues()
On Tue, 2023-08-08 at 10:57 +0800, Jason Wang wrote:
> On Thu, Aug 3, 2023 at 7:40 PM Dragos Tatulea wrote:
> >
> > On Thu, 2023-08-03 at 16:03 +0800, Jason Wang wrote:
> > > On Thu, Aug 3, 2023 at 1:13 AM Dragos Tatulea wrote:
> > > >
> > > > The mr->initialized flag is shared between the cont
The ndev was accessed on shutdown without a check if it actually exists.
This triggered the crash pasted below.
Instead of doing the ndev check, delete the shutdown handler altogether.
The irqs will be released at the parent VF level (mlx5_core).
BUG: kernel NULL pointer dereference, address: 00
On Wed, 2023-08-02 at 09:56 +0200, Dragos Tatulea wrote:
> On Wed, 2023-08-02 at 10:51 +0800, Jason Wang wrote:
> > On Tue, Aug 1, 2023 at 4:17 PM Dragos Tatulea wrote:
> > >
> > > On Tue, 2023-08-01 at 11:59 +0800, Jason Wang wrote:
> > > > On Mon, Jul 31, 2023 at 5:08 PM Michael S. Tsirkin
> >
On Thu, 2023-08-03 at 16:03 +0800, Jason Wang wrote:
> On Thu, Aug 3, 2023 at 1:13 AM Dragos Tatulea wrote:
> >
> > The mr->initialized flag is shared between the control vq and data vq
> > part of the mr init/uninit. But if the control vq and data vq get placed
> > in different ASIDs, it can hap
From: Eugenio Pérez
mlx5_vdpa_destroy_mr can be called from .set_map with data ASID after
the control virtqueue ASID iotlb has been populated. The control vq
iotlb must not be cleared, since it will not be populated again.
So call the ASID aware destroy function which makes sure that the
right v
1 - 100 of 134 matches
Mail list logo
