dma/dpaa2: set the vfa bit for rbp with vf
Checks
Commit Message
Set the VFA (Virtual Function Active) bit in
struct qdma_sdd -> rbpcmd_simple in order to use the route-by-port
functionality with PCIe virtual functions.
At the moment, a user wanting to enable route-by-port will call
rte_dpaa2_qdma_vchan_rbp_enable with a rte_dpaa2_qdma_rbp struct.
The struct includes the PCIe Physical and Virtual functions among other
things, which are then copied to qdma_sdd -> rbpcmd_simple, but the vfa
bit is never touched (the bit does exists in rbpcmd_simple),
so route-by-port with virtual functions won't work..
In order to fix this, a vfa bit is added to struct rte_dpaa2_qdma_rbp,
then is copied to qdma_sdd -> rbpcmd_simple.
Fixes: 8caf8427f85a ("dma/dpaa2: introduce driver skeleton")
Signed-off-by: Alvaro Karsz <alvaro.karsz@solid-run.com>
---
drivers/dma/dpaa2/dpaa2_qdma.c | 2 ++
drivers/dma/dpaa2/rte_pmd_dpaa2_qdma.h | 4 +++-
2 files changed, 5 insertions(+), 1 deletion(-)
Comments
Hi,
Any comments on the patch?
Thanks
++Jun
> -----Original Message-----
> From: Alvaro Karsz <alvaro.karsz@solid-run.com>
> Sent: Sunday, April 9, 2023 5:40 PM
> To: Gagandeep Singh <G.Singh@nxp.com>; Hemant Agrawal
> <hemant.agrawal@nxp.com>
> Cc: dev@dpdk.org; David Marchand <david.marchand@redhat.com>
> Subject: Re: [PATCH] dma/dpaa2: set the vfa bit for rbp with vf
>
> Hi,
> Any comments on the patch?
>
> Thanks
Hi,
This patch fixed the issue of qdma route by port for PCIe VF.
"vfa" bit determines if "vfid" is valid or not.
Route by port is a kind of qdam transfer used for PCIe EP to access host by PCIe address and PCIe function. Another kind of transfer to access host is by PCIe address + PCIe outbound address.
-----Original Message-----
From: Gagandeep Singh <G.Singh@nxp.com>
Sent: 2023年4月10日 15:22
To: Alvaro Karsz <alvaro.karsz@solid-run.com>; Hemant Agrawal <hemant.agrawal@nxp.com>; Jun Yang <jun.yang@nxp.com>
Cc: dev@dpdk.org; David Marchand <david.marchand@redhat.com>
Subject: RE: [PATCH] dma/dpaa2: set the vfa bit for rbp with vf
++Jun
> -----Original Message-----
> From: Alvaro Karsz <alvaro.karsz@solid-run.com>
> Sent: Sunday, April 9, 2023 5:40 PM
> To: Gagandeep Singh <G.Singh@nxp.com>; Hemant Agrawal
> <hemant.agrawal@nxp.com>
> Cc: dev@dpdk.org; David Marchand <david.marchand@redhat.com>
> Subject: Re: [PATCH] dma/dpaa2: set the vfa bit for rbp with vf
>
> Hi,
> Any comments on the patch?
>
> Thanks
> Hi,
> This patch fixed the issue of qdma route by port for PCIe VF.
> "vfa" bit determines if "vfid" is valid or not.
> Route by port is a kind of qdam transfer used for PCIe EP to access host by PCIe address and PCIe function. Another kind of transfer to access host is by PCIe address + PCIe outbound address.
Thanks for the confirmation!
Maybe you can now accept/merge the patch?
Thanks
> -----Original Message-----
> From: Alvaro Karsz <alvaro.karsz@solid-run.com>
> Sent: Sunday, April 16, 2023 2:47 PM
> To: Jun Yang <jun.yang@nxp.com>; Gagandeep Singh <G.Singh@nxp.com>;
> Hemant Agrawal <hemant.agrawal@nxp.com>
> Cc: dev@dpdk.org; David Marchand <david.marchand@redhat.com>
> Subject: Re: [PATCH] dma/dpaa2: set the vfa bit for rbp with vf
> Importance: High
>
> > Hi,
> > This patch fixed the issue of qdma route by port for PCIe VF.
> > "vfa" bit determines if "vfid" is valid or not.
> > Route by port is a kind of qdam transfer used for PCIe EP to access host by
> PCIe address and PCIe function. Another kind of transfer to access host is by
> PCIe address + PCIe outbound address.
>
> Thanks for the confirmation!
> Maybe you can now accept/merge the patch?
>
> Thanks
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Thank you.
Do I need to do anything else for this patch to be merged?
> > > Hi,
> > > This patch fixed the issue of qdma route by port for PCIe VF.
> > > "vfa" bit determines if "vfid" is valid or not.
> > > Route by port is a kind of qdam transfer used for PCIe EP to access host by
> > PCIe address and PCIe function. Another kind of transfer to access host is by
> > PCIe address + PCIe outbound address.
> >
> > Thanks for the confirmation!
> > Maybe you can now accept/merge the patch?
> >
> > Thanks
> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Applied, thanks.
@@ -117,6 +117,7 @@ dpaa2_qdma_populate_fle(struct qbman_fle *fle,
/* source */
sdd->read_cmd.portid = rbp->sportid;
sdd->rbpcmd_simple.pfid = rbp->spfid;
+ sdd->rbpcmd_simple.vfa = rbp->vfa;
sdd->rbpcmd_simple.vfid = rbp->svfid;
if (rbp->srbp) {
@@ -129,6 +130,7 @@ dpaa2_qdma_populate_fle(struct qbman_fle *fle,
/* destination */
sdd->write_cmd.portid = rbp->dportid;
sdd->rbpcmd_simple.pfid = rbp->dpfid;
+ sdd->rbpcmd_simple.vfa = rbp->vfa;
sdd->rbpcmd_simple.vfid = rbp->dvfid;
if (rbp->drbp) {
@@ -44,7 +44,9 @@ struct rte_dpaa2_qdma_rbp {
uint32_t svfid:6;
/* using route by port for source */
uint32_t srbp:1;
- uint32_t rsv:4;
+ /* Virtual Function Active */
+ uint32_t vfa:1;
+ uint32_t rsv:3;
};
/** Determines a QDMA job */