| a6097e0a | 20-Feb-2025 |
Si-Wei Liu <[email protected]> |
vdpa/mlx5: Fix oversized null mkey longer than 32bit
create_user_mr() has correct code to count the number of null keys
used to fill in a hole for the memory map. However, fill_indir()
does not foll
vdpa/mlx5: Fix oversized null mkey longer than 32bit
create_user_mr() has correct code to count the number of null keys
used to fill in a hole for the memory map. However, fill_indir()
does not follow the same to cap the range up to the 1GB limit
correspondingly. Fill in more null keys for the gaps in between,
so that null keys are correctly populated.
Fixes: 94abbccdf291 ("vdpa/mlx5: Add shared memory registration code")
Cc: [email protected]
Reported-by: Cong Meng <[email protected]>
Signed-off-by: Si-Wei Liu <[email protected]>
Signed-off-by: Dragos Tatulea <[email protected]>
Acked-by: Eugenio Pérez <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Jason Wang <[email protected]>
show more ...
|
| 35025963 | 21-Oct-2024 |
Si-Wei Liu <[email protected]> |
vdpa/mlx5: Fix suboptimal range on iotlb iteration
The starting iova address to iterate iotlb map entry within a range
was set to an irrelevant value when passing to the itree_next()
iterator, altho
vdpa/mlx5: Fix suboptimal range on iotlb iteration
The starting iova address to iterate iotlb map entry within a range
was set to an irrelevant value when passing to the itree_next()
iterator, although luckily it doesn't affect the outcome of finding
out the granule of the smallest iotlb map size. Fix the code to make
it consistent with the following for-loop.
Fixes: 94abbccdf291 ("vdpa/mlx5: Add shared memory registration code")
Signed-off-by: Si-Wei Liu <[email protected]>
Signed-off-by: Dragos Tatulea <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Jason Wang <[email protected]>
show more ...
|
| 29ce8b8a | 21-Oct-2024 |
Si-Wei Liu <[email protected]> |
vdpa/mlx5: Fix PA offset with unaligned starting iotlb map
When calculating the physical address range based on the iotlb and mr
[start,end) ranges, the offset of mr->start relative to map->start
is
vdpa/mlx5: Fix PA offset with unaligned starting iotlb map
When calculating the physical address range based on the iotlb and mr
[start,end) ranges, the offset of mr->start relative to map->start
is not taken into account. This leads to some incorrect and duplicate
mappings.
For the case when mr->start < map->start the code is already correct:
the range in [mr->start, map->start) was handled by a different
iteration.
Fixes: 94abbccdf291 ("vdpa/mlx5: Add shared memory registration code")
Cc: [email protected]
Signed-off-by: Si-Wei Liu <[email protected]>
Signed-off-by: Dragos Tatulea <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Jason Wang <[email protected]>
show more ...
|
| 62111654 | 30-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Postpone MR deletion
Currently, when a new MR is set up, the old MR is deleted. MR deletion
is about 30-40% the time of MR creation. As deleting the old MR is not
important for the proces
vdpa/mlx5: Postpone MR deletion
Currently, when a new MR is set up, the old MR is deleted. MR deletion
is about 30-40% the time of MR creation. As deleting the old MR is not
important for the process of setting up the new MR, this operation
can be postponed.
This series adds a workqueue that does MR garbage collection at a later
point. If the MR lock is taken, the handler will back off and
reschedule. The exception during shutdown: then the handler must
not postpone the work.
Note that this is only a speculative optimization: if there is some
mapping operation that is triggered while the garbage collector handler
has the lock taken, this operation it will have to wait for the handler
to finish.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Cosmin Ratiu <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
show more ...
|
| f30a1232 | 30-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Introduce init/destroy for MR resources
There's currently not a lot of action happening during
the init/destroy of MR resources. But more will be added
in the upcoming patches.
As the mr
vdpa/mlx5: Introduce init/destroy for MR resources
There's currently not a lot of action happening during
the init/destroy of MR resources. But more will be added
in the upcoming patches.
As the mr mutex lock init/destroy has been moved to these
new functions, the lifetime has now shifted away from
mlx5_vdpa_alloc_resources() / mlx5_vdpa_free_resources()
into these new functions. However, the lifetime at the
outer scope remains the same:
mlx5_vdpa_dev_add() / mlx5_vdpa_dev_free()
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Cosmin Ratiu <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
show more ...
|
| 58d4d50e | 30-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Rename mr_mtx -> lock
Now that the mr resources have their own namespace in the
struct, give the lock a clearer name.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Cos
vdpa/mlx5: Rename mr_mtx -> lock
Now that the mr resources have their own namespace in the
struct, give the lock a clearer name.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Cosmin Ratiu <[email protected]>
Acked-by: Eugenio Pérez <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
show more ...
|
| 0071b138 | 30-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Create direct MKEYs in parallel
Use the async interface to issue MTT MKEY creation.
Extra care is taken at the allocation of FW input commands
due to the MTT tables having variable sizes
vdpa/mlx5: Create direct MKEYs in parallel
Use the async interface to issue MTT MKEY creation.
Extra care is taken at the allocation of FW input commands
due to the MTT tables having variable sizes depending on
MR.
The indirect MKEY is still created synchronously at the
end as the direct MKEYs need to be filled in.
This makes create_user_mr() 3-5x faster, depending on
the size of the MR.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Cosmin Ratiu <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
show more ...
|
| 9dba4195 | 16-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Parallelize VQ suspend/resume for CVQ MQ command
change_num_qps() is still suspending/resuming VQs one by one.
This change switches to parallel suspend/resume.
When increasing the number
vdpa/mlx5: Parallelize VQ suspend/resume for CVQ MQ command
change_num_qps() is still suspending/resuming VQs one by one.
This change switches to parallel suspend/resume.
When increasing the number of queues the flow has changed a bit for
simplicity: the setup_vq() function will always be called before
resume_vqs(). If the VQ is initialized, setup_vq() will exit early. If
the VQ is not initialized, setup_vq() will create it and resume_vqs()
will resume it.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Acked-by: Eugenio Pérez <[email protected]>
Tested-by: Lei Yang <[email protected]>
show more ...
|
| 55a7cb05 | 16-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Keep notifiers during suspend but ignore
Unregistering notifiers is a costly operation. Instead of removing
the notifiers during device suspend and adding them back at resume,
simply igno
vdpa/mlx5: Keep notifiers during suspend but ignore
Unregistering notifiers is a costly operation. Instead of removing
the notifiers during device suspend and adding them back at resume,
simply ignore the call when the device is suspended.
At resume time call queue_link_work() to make sure that the device state
is propagated in case there were changes.
For 1 vDPA device x 32 VQs (16 VQPs) attached to a large VM (256 GB RAM,
32 CPUs x 2 threads per core), the device suspend time is reduced from
~13 ms to ~2.5 ms.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Acked-by: Eugenio Pérez <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Tested-by: Lei Yang <[email protected]>
show more ...
|
| 5eb8c7eb | 16-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Parallelize device resume
Currently device resume works on vqs serially. Building up on previous
changes that converted vq operations to the async api, this patch
parallelizes the device
vdpa/mlx5: Parallelize device resume
Currently device resume works on vqs serially. Building up on previous
changes that converted vq operations to the async api, this patch
parallelizes the device resume.
For 1 vDPA device x 32 VQs (16 VQPs) attached to a large VM (256 GB RAM,
32 CPUs x 2 threads per core), the device resume time is reduced from
~16 ms to ~4.5 ms.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Acked-by: Eugenio Pérez <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Tested-by: Lei Yang <[email protected]>
show more ...
|
| dcf3eac0 | 16-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Parallelize device suspend
Currently device suspend works on vqs serially. Building up on previous
changes that converted vq operations to the async api, this patch
parallelizes the devic
vdpa/mlx5: Parallelize device suspend
Currently device suspend works on vqs serially. Building up on previous
changes that converted vq operations to the async api, this patch
parallelizes the device suspend:
1) Suspend all active vqs parallel.
2) Query suspended vqs in parallel.
For 1 vDPA device x 32 VQs (16 VQPs) attached to a large VM (256 GB RAM,
32 CPUs x 2 threads per core), the device suspend time is reduced from
~37 ms to ~13 ms.
A later patch will remove the link unregister operation which will make
it even faster.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Acked-by: Eugenio Pérez <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Tested-by: Lei Yang <[email protected]>
show more ...
|
| 1fcdf43e | 16-Aug-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Use async API for vq query command
Switch firmware vq query command to be issued via the async API to
allow future parallelization.
For now the command is still serial but the infrastruc
vdpa/mlx5: Use async API for vq query command
Switch firmware vq query command to be issued via the async API to
allow future parallelization.
For now the command is still serial but the infrastructure is there
to issue commands in parallel, including ratelimiting the number
of issued async commands to firmware.
A later patch will switch to issuing more commands at a time.
Signed-off-by: Dragos Tatulea <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
Tested-by: Lei Yang <[email protected]>
show more ...
|
| 8e0751af | 26-Jun-2024 |
Dragos Tatulea <[email protected]> |
vdpa/mlx5: Don't enable non-active VQs in .set_vq_ready()
VQ indices in the range [cur_num_qps, max_vqs) represent queues that
have not yet been activated. .set_vq_ready should not activate these
VQ
vdpa/mlx5: Don't enable non-active VQs in .set_vq_ready()
VQ indices in the range [cur_num_qps, max_vqs) represent queues that
have not yet been activated. .set_vq_ready should not activate these
VQs.
Reviewed-by: Cosmin Ratiu <[email protected]>
Acked-by: Eugenio Pérez <[email protected]>
Signed-off-by: Dragos Tatulea <[email protected]>
Message-Id: <[email protected]>
Signed-off-by: Michael S. Tsirkin <[email protected]>
show more ...
|