[PATCHv2 for-next 1/1] RDMA/irdma: Add support for dmabuf pin memory regions

[PATCHv2 for-next 1/1] RDMA/irdma: Add support for dmabuf pin memory regions

[Zhu Yanjun]

From: Zhu Yanjun <yanjun.zhu@linux.dev>

This is a followup to the EFA dmabuf[1]. Irdma driver currently does
not support on-demand-paging(ODP). So it uses habanalabs as the
dmabuf exporter, and irdma as the importer to allow for peer2peer
access through libibverbs.

In this commit, the function ib_umem_dmabuf_get_pinned() is used.
This function is introduced in EFA dmabuf[1] which allows the driver
to get a dmabuf umem which is pinned and does not require move_notify
callback implementation. The returned umem is pinned and DMA mapped
like standard cpu umems, and is released through ib_umem_release().

[1]https://lore.kernel.org/lkml/20211007114018.GD2688930@ziepe.ca/t/

Signed-off-by: Zhu Yanjun <yanjun.zhu@linux.dev>
---
V1->V2: Thanks Shiraz Saleem, he gave me a lot of good suggestions.
        This commit is based on the shared functions from refactored
        irdma_reg_user_mr.
---
 drivers/infiniband/hw/irdma/verbs.c | 43 +++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)

diff --git a/drivers/infiniband/hw/irdma/verbs.c b/drivers/infiniband/hw/irdma/verbs.c
index 6982f38596c8..a638861689c2 100644
--- a/drivers/infiniband/hw/irdma/verbs.c
+++ b/drivers/infiniband/hw/irdma/verbs.c
@@ -2977,6 +2977,48 @@ static struct ib_mr *irdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 len,
 	return ERR_PTR(err);
 }
 
+static struct ib_mr *irdma_reg_user_mr_dmabuf(struct ib_pd *pd, u64 start,
+					      u64 len, u64 virt,
+					      int fd, int access,
+					      struct ib_udata *udata)
+{
+	struct irdma_device *iwdev = to_iwdev(pd->device);
+	struct ib_umem_dmabuf *umem_dmabuf = NULL;
+	struct irdma_mr *iwmr = NULL;
+	int err;
+
+	if (len > iwdev->rf->sc_dev.hw_attrs.max_mr_size)
+		return ERR_PTR(-EINVAL);
+
+	if (udata->inlen < IRDMA_MEM_REG_MIN_REQ_LEN)
+		return ERR_PTR(-EINVAL);
+
+	umem_dmabuf = ib_umem_dmabuf_get_pinned(pd->device, start, len, fd, access);
+	if (IS_ERR(umem_dmabuf)) {
+		err = PTR_ERR(umem_dmabuf);
+		ibdev_dbg(&iwdev->ibdev, "Failed to get dmabuf umem[%d]\n", err);
+		return ERR_PTR(err);
+	}
+
+	iwmr = irdma_alloc_iwmr(&umem_dmabuf->umem, pd, virt, IRDMA_MEMREG_TYPE_MEM);
+	if (IS_ERR(iwmr)) {
+		ib_umem_release(&umem_dmabuf->umem);
+		return (struct ib_mr *)iwmr;
+	}
+
+	err = irdma_reg_user_mr_type_mem(iwmr, access);
+	if (err)
+		goto error;
+
+	return &iwmr->ibmr;
+
+error:
+	irdma_free_iwmr(iwmr);
+	ib_umem_release(&umem_dmabuf->umem);
+
+	return ERR_PTR(err);
+}
+
 /**
  * irdma_reg_phys_mr - register kernel physical memory
  * @pd: ibpd pointer
@@ -4483,6 +4525,7 @@ static const struct ib_device_ops irdma_dev_ops = {
 	.query_port = irdma_query_port,
 	.query_qp = irdma_query_qp,
 	.reg_user_mr = irdma_reg_user_mr,
+	.reg_user_mr_dmabuf = irdma_reg_user_mr_dmabuf,
 	.req_notify_cq = irdma_req_notify_cq,
 	.resize_cq = irdma_resize_cq,
 	INIT_RDMA_OBJ_SIZE(ib_pd, irdma_pd, ibpd),
-- 
2.27.0

Re: [PATCHv2 for-next 1/1] RDMA/irdma: Add support for dmabuf pin memory regions

[Jason Gunthorpe]

On Mon, Jan 30, 2023 at 11:24:07AM +0800, Zhu Yanjun wrote:
> From: Zhu Yanjun <yanjun.zhu@linux.dev>
> 
> This is a followup to the EFA dmabuf[1]. Irdma driver currently does
> not support on-demand-paging(ODP). So it uses habanalabs as the
> dmabuf exporter, and irdma as the importer to allow for peer2peer
> access through libibverbs.
> 
> In this commit, the function ib_umem_dmabuf_get_pinned() is used.
> This function is introduced in EFA dmabuf[1] which allows the driver
> to get a dmabuf umem which is pinned and does not require move_notify
> callback implementation. The returned umem is pinned and DMA mapped
> like standard cpu umems, and is released through ib_umem_release().
> 
> [1]https://lore.kernel.org/lkml/20211007114018.GD2688930@ziepe.ca/t/
> 
> Signed-off-by: Zhu Yanjun <yanjun.zhu@linux.dev>
> ---
> V1->V2: Thanks Shiraz Saleem, he gave me a lot of good suggestions.
>         This commit is based on the shared functions from refactored
>         irdma_reg_user_mr.
> ---
>  drivers/infiniband/hw/irdma/verbs.c | 43 +++++++++++++++++++++++++++++
>  1 file changed, 43 insertions(+)
> 
> diff --git a/drivers/infiniband/hw/irdma/verbs.c b/drivers/infiniband/hw/irdma/verbs.c
> index 6982f38596c8..a638861689c2 100644
> --- a/drivers/infiniband/hw/irdma/verbs.c
> +++ b/drivers/infiniband/hw/irdma/verbs.c
> @@ -2977,6 +2977,48 @@ static struct ib_mr *irdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 len,
>  	return ERR_PTR(err);
>  }
>  
> +static struct ib_mr *irdma_reg_user_mr_dmabuf(struct ib_pd *pd, u64 start,
> +					      u64 len, u64 virt,
> +					      int fd, int access,
> +					      struct ib_udata *udata)
> +{
> +	struct irdma_device *iwdev = to_iwdev(pd->device);
> +	struct ib_umem_dmabuf *umem_dmabuf = NULL;
> +	struct irdma_mr *iwmr = NULL;

No reason to null initialize these

> +	int err;
> +
> +	if (len > iwdev->rf->sc_dev.hw_attrs.max_mr_size)
> +		return ERR_PTR(-EINVAL);
> +
> +	if (udata->inlen < IRDMA_MEM_REG_MIN_REQ_LEN)
> +		return ERR_PTR(-EINVAL);
> +
> +	umem_dmabuf = ib_umem_dmabuf_get_pinned(pd->device, start, len, fd, access);
> +	if (IS_ERR(umem_dmabuf)) {
> +		err = PTR_ERR(umem_dmabuf);
> +		ibdev_dbg(&iwdev->ibdev, "Failed to get dmabuf umem[%d]\n", err);
> +		return ERR_PTR(err);
> +	}
> +
> +	iwmr = irdma_alloc_iwmr(&umem_dmabuf->umem, pd, virt, IRDMA_MEMREG_TYPE_MEM);
> +	if (IS_ERR(iwmr)) {
> +		ib_umem_release(&umem_dmabuf->umem);
> +		return (struct ib_mr *)iwmr;

err = PTR_ERR(iwmr);
goto err_release;

> +	}
> +
> +	err = irdma_reg_user_mr_type_mem(iwmr, access);
> +	if (err)
> +		goto error;
> +
> +	return &iwmr->ibmr;
> +
> +error:
err_iwmr:
> +	irdma_free_iwmr(iwmr);

err_release:

> +	ib_umem_release(&umem_dmabuf->umem);
> +
> +	return ERR_PTR(err);
> +}

Jason

Re: [PATCHv2 for-next 1/1] RDMA/irdma: Add support for dmabuf pin memory regions

[Zhu Yanjun]

在 2023/1/31 22:42, Jason Gunthorpe 写道:
> On Mon, Jan 30, 2023 at 11:24:07AM +0800, Zhu Yanjun wrote:
>> From: Zhu Yanjun <yanjun.zhu@linux.dev>
>>
>> This is a followup to the EFA dmabuf[1]. Irdma driver currently does
>> not support on-demand-paging(ODP). So it uses habanalabs as the
>> dmabuf exporter, and irdma as the importer to allow for peer2peer
>> access through libibverbs.
>>
>> In this commit, the function ib_umem_dmabuf_get_pinned() is used.
>> This function is introduced in EFA dmabuf[1] which allows the driver
>> to get a dmabuf umem which is pinned and does not require move_notify
>> callback implementation. The returned umem is pinned and DMA mapped
>> like standard cpu umems, and is released through ib_umem_release().
>>
>> [1]https://lore.kernel.org/lkml/20211007114018.GD2688930@ziepe.ca/t/
>>
>> Signed-off-by: Zhu Yanjun <yanjun.zhu@linux.dev>
>> ---
>> V1->V2: Thanks Shiraz Saleem, he gave me a lot of good suggestions.
>>          This commit is based on the shared functions from refactored
>>          irdma_reg_user_mr.
>> ---
>>   drivers/infiniband/hw/irdma/verbs.c | 43 +++++++++++++++++++++++++++++
>>   1 file changed, 43 insertions(+)
>>
>> diff --git a/drivers/infiniband/hw/irdma/verbs.c b/drivers/infiniband/hw/irdma/verbs.c
>> index 6982f38596c8..a638861689c2 100644
>> --- a/drivers/infiniband/hw/irdma/verbs.c
>> +++ b/drivers/infiniband/hw/irdma/verbs.c
>> @@ -2977,6 +2977,48 @@ static struct ib_mr *irdma_reg_user_mr(struct ib_pd *pd, u64 start, u64 len,
>>   	return ERR_PTR(err);
>>   }
>>   
>> +static struct ib_mr *irdma_reg_user_mr_dmabuf(struct ib_pd *pd, u64 start,
>> +					      u64 len, u64 virt,
>> +					      int fd, int access,
>> +					      struct ib_udata *udata)
>> +{
>> +	struct irdma_device *iwdev = to_iwdev(pd->device);
>> +	struct ib_umem_dmabuf *umem_dmabuf = NULL;
>> +	struct irdma_mr *iwmr = NULL;
> No reason to null initialize these


Got it.


>
>> +	int err;
>> +
>> +	if (len > iwdev->rf->sc_dev.hw_attrs.max_mr_size)
>> +		return ERR_PTR(-EINVAL);
>> +
>> +	if (udata->inlen < IRDMA_MEM_REG_MIN_REQ_LEN)
>> +		return ERR_PTR(-EINVAL);
>> +
>> +	umem_dmabuf = ib_umem_dmabuf_get_pinned(pd->device, start, len, fd, access);
>> +	if (IS_ERR(umem_dmabuf)) {
>> +		err = PTR_ERR(umem_dmabuf);
>> +		ibdev_dbg(&iwdev->ibdev, "Failed to get dmabuf umem[%d]\n", err);
>> +		return ERR_PTR(err);
>> +	}
>> +
>> +	iwmr = irdma_alloc_iwmr(&umem_dmabuf->umem, pd, virt, IRDMA_MEMREG_TYPE_MEM);
>> +	if (IS_ERR(iwmr)) {
>> +		ib_umem_release(&umem_dmabuf->umem);
>> +		return (struct ib_mr *)iwmr;
> err = PTR_ERR(iwmr);
> goto err_release;


Got it.


>
>> +	}
>> +
>> +	err = irdma_reg_user_mr_type_mem(iwmr, access);
>> +	if (err)
>> +		goto error;
>> +
>> +	return &iwmr->ibmr;
>> +
>> +error:
> err_iwmr:
>> +	irdma_free_iwmr(iwmr);
> err_release:

Got it. I will send the latest commit very soon.

Zhu Yanjun

>
>> +	ib_umem_release(&umem_dmabuf->umem);
>> +
>> +	return ERR_PTR(err);
>> +}
> Jason