net: mana: Record doorbell physical address in PF mode
Subsystems (5)
PATCH v2 1/1 net: mana: Record doorbell physical address in PF mode
From: Long Li <[email protected]>
MANA supports RDMA in PF mode. The driver should record the doorbell physical address when in PF mode.
The doorbell physical address is used by the RDMA driver to map doorbell pages of the device to user-mode applications through RDMA verbs interface. In the past, they have been mapped to user-mode while the device is in VF mode. With the support for PF mode implemented, also expose those pages in PF mode.
---
Changes
v2: add more details in commit message on how the doorbell physical address is used
drivers/net/ethernet/microsoft/mana/gdma_main.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/net/ethernet/microsoft/mana/gdma_main.c +3 -0
@@ -31,6 +31,9 @@ static void mana_gd_init_pf_regs(struct pci_dev *pdev) gc->db_page_base = gc->bar0_va +mana_gd_r64 (gc,GDMA_PF_REG_DB_PAGE_OFF ); + gc->phys_db_page_base = gc->bar0_pa + +mana_gd_r64 (gc,GDMA_PF_REG_DB_PAGE_OFF ); + sriov_base_off =mana_gd_r64 (gc,GDMA_SRIOV_REG_CFG_BASE_OFF ); sriov_base_va = gc->bar0_va + sriov_base_off;
Signature · 1 line
2.25.1
Quoted 35 lines · On Fri, Jun 13, 2025 at 10:46:05AM -0700, [email protected] wrote:
From: Long Li <[email protected]>
MANA supports RDMA in PF mode. The driver should record the doorbell physical address when in PF mode.
The doorbell physical address is used by the RDMA driver to map doorbell pages of the device to user-mode applications through RDMA verbs interface. In the past, they have been mapped to user-mode while the device is in VF mode. With the support for PF mode implemented, also expose those pages in PF mode.
Signed-off-by Long Li <[email protected]>---
Changes
v2: add more details in commit message on how the doorbell physical address is useddrivers/net/ethernet/microsoft/mana/gdma_main.c | 3 +++ 1 file changed, 3 insertions(+)drivers/net/ethernet/microsoft/mana/gdma_main.c +3
-0@@ -31,6 +31,9 @@ static void mana_gd_init_pf_regs(struct pci_dev *pdev) gc->db_page_base = gc->bar0_va +mana_gd_r64 (gc,GDMA_PF_REG_DB_PAGE_OFF ); + gc->phys_db_page_base = gc->bar0_pa + +mana_gd_r64 (gc,GDMA_PF_REG_DB_PAGE_OFF ); + sriov_base_off =mana_gd_r64 (gc,GDMA_SRIOV_REG_CFG_BASE_OFF ); sriov_base_va = gc->bar0_va + sriov_base_off;Signature · 1 line
2.25.1
Jakub Kicinski
maintainerQuoted 8 lines · On Fri, 13 Jun 2025 10:46:05 -0700 [email protected] wrote:
MANA supports RDMA in PF mode. The driver should record the doorbell physical address when in PF mode.
The doorbell physical address is used by the RDMA driver to map doorbell pages of the device to user-mode applications through RDMA verbs interface. In the past, they have been mapped to user-mode while the device is in VF mode. With the support for PF mode implemented, also expose those pages in PF mode.
It'd be good to indicate if the PF mode support is implemented as in ready to be posted to the list, or implemented as in already in the tree? If it's in tree then please quote the commit. If upcoming please rephrase.
This commit msg is much better, thank you, but to a person handling backports it will still not be immediately clear whether this is prep work for a new code path, or a bug fix.
Signature · 1 line
pw-bot: cr
Quoted 17 lines
Subject: Re: [Patch net-next v2] net: mana: Record doorbell physical address in PF mode
Quoted 8 lines · On Fri, 13 Jun 2025 10:46:05 -0700 [email protected] wrote:
MANA supports RDMA in PF mode. The driver should record the doorbell physical address when in PF mode.
The doorbell physical address is used by the RDMA driver to map doorbell pages of the device to user-mode applications through RDMA verbs interface. In the past, they have been mapped to user-mode while the device is in VF mode. With the support for PF mode implemented, also expose those pages in PF mode.
It'd be good to indicate if the PF mode support is implemented as in ready to be posted to the list, or implemented as in already in the tree? If it's in tree then please quote the commit. If upcoming please rephrase.
Quoted 5 lines
This commit msg is much better, thank you, but to a person handling backports it will still not be immediately clear whether this is prep work for a new code path, or a bug fix.
--
pw-bot: cr
Thank you. I'm sending v3.
Long