mbox series

[net-next,v4,0/7] enic: Use all the resources configured on VIC

Message ID 20241113-remove_vic_resource_limits-v4-0-a34cf8570c67@cisco.com (mailing list archive)
Headers show
Series enic: Use all the resources configured on VIC | expand

Message

Nelson Escobar (neescoba) Nov. 13, 2024, 11:56 p.m. UTC
Allow users to configure and use more than 8 rx queues and 8 tx queues
on the Cisco VIC.

This series changes the maximum number of tx and rx queues supported
from 8 to the hardware limit of 256, and allocates memory based on the
number of resources configured on the VIC.

Signed-off-by: Nelson Escobar <neescoba@cisco.com>
---
Changes in v4:
- Followed Vadim Fedorenko's suggestions for re-arranging enic_wq and
  adding ____cacheline_aligned to the new structs.
- Link to v3: https://lore.kernel.org/r/20241108-remove_vic_resource_limits-v3-0-3ba8123bcffc@cisco.com

Changes in v3:
- Per Jakub's suggestions, split commit 5 into smaller commits and use
  net_get_num_default_rss_queues() to set the number of RQs used.
- Fixed an issue with commit 2 caught during testing with a missing
  changed needed in enic_init_vnic_resources().
- Link to v2: https://lore.kernel.org/r/20241024-remove_vic_resource_limits-v2-0-039b8cae5fdd@cisco.com

Changes in v2:
- Followed Kalesh's suggestions: removed redundant NULL assigments,
  returning -ENOMEM directly
- Reviewed-by tag for Simon Horman <horms@kernel.org>
- Marked Satish Kharat and John Daley as co-developers to better reflect
  their role in this patch set
- Link to v1: https://lore.kernel.org/r/20241022041707.27402-2-neescoba@cisco.com

---
Nelson Escobar (7):
      enic: Create enic_wq/rq structures to bundle per wq/rq data
      enic: Make MSI-X I/O interrupts come after the other required ones
      enic: Save resource counts we read from HW
      enic: Allocate arrays in enic struct based on VIC config
      enic: Adjust used MSI-X wq/rq/cq/interrupt resources in a more robust way
      enic: Move enic resource adjustments to separate function
      enic: Move kdump check into enic_adjust_resources()

 drivers/net/ethernet/cisco/enic/enic.h         |  62 ++--
 drivers/net/ethernet/cisco/enic/enic_ethtool.c |   8 +-
 drivers/net/ethernet/cisco/enic/enic_main.c    | 386 +++++++++++++++----------
 drivers/net/ethernet/cisco/enic/enic_res.c     |  42 +--
 4 files changed, 299 insertions(+), 199 deletions(-)
---
base-commit: 6f07cd8301706b661776074ddc97c991d107cc91
change-id: 20241023-remove_vic_resource_limits-eaa64f9e65fb

Best regards,

Comments

patchwork-bot+netdevbpf@kernel.org Nov. 15, 2024, 11:50 p.m. UTC | #1
Hello:

This series was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:

On Wed, 13 Nov 2024 23:56:32 +0000 you wrote:
> Allow users to configure and use more than 8 rx queues and 8 tx queues
> on the Cisco VIC.
> 
> This series changes the maximum number of tx and rx queues supported
> from 8 to the hardware limit of 256, and allocates memory based on the
> number of resources configured on the VIC.
> 
> [...]

Here is the summary with links:
  - [net-next,v4,1/7] enic: Create enic_wq/rq structures to bundle per wq/rq data
    https://git.kernel.org/netdev/net-next/c/b67609c93153
  - [net-next,v4,2/7] enic: Make MSI-X I/O interrupts come after the other required ones
    https://git.kernel.org/netdev/net-next/c/231646cb6a8c
  - [net-next,v4,3/7] enic: Save resource counts we read from HW
    https://git.kernel.org/netdev/net-next/c/5aee3324724a
  - [net-next,v4,4/7] enic: Allocate arrays in enic struct based on VIC config
    https://git.kernel.org/netdev/net-next/c/a64e5492ca90
  - [net-next,v4,5/7] enic: Adjust used MSI-X wq/rq/cq/interrupt resources in a more robust way
    https://git.kernel.org/netdev/net-next/c/cc94d6c4d40c
  - [net-next,v4,6/7] enic: Move enic resource adjustments to separate function
    https://git.kernel.org/netdev/net-next/c/374f6c04df8e
  - [net-next,v4,7/7] enic: Move kdump check into enic_adjust_resources()
    https://git.kernel.org/netdev/net-next/c/a28ccf1d6c10

You are awesome, thank you!