* garbage collect old memory registration code
@ 2015-11-15 18:05 Christoph Hellwig
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
0 siblings, 1 reply; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:05 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
This series removes huge chunks of code related to old memory
registration methods that we don't use anymore.
This expects my "IB: merge struct ib_device_attr into struct ib_device"
patch to be already applied.
Also available as a git tree:
http://git.infradead.org/users/hch/rdma.git/shortlog/refs/heads/rdma-mr-cleanups
git://git.infradead.org/users/hch/rdma.git rdma-mr-cleanups
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* [PATCH 1/9] IB: start documenting device capabilities
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
@ 2015-11-15 18:05 ` Christoph Hellwig
2015-11-15 18:05 ` [PATCH 2/9] IB: remove ib_query_mr Christoph Hellwig
` (8 subsequent siblings)
9 siblings, 0 replies; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:05 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
Just IB_DEVICE_LOCAL_DMA_LKEY and IB_DEVICE_MEM_MGT_EXTENSIONS for now
as I'm most familar with those.
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
Reviewed-by: Sagi Grimberg <sagig-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
Reviewed-By: Jason Gunthorpe <jgunthorpe-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
---
include/rdma/ib_verbs.h | 18 ++++++++++++++++++
1 file changed, 18 insertions(+)
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 45ce36e..6034f92 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -120,6 +120,14 @@ enum ib_device_cap_flags {
IB_DEVICE_RC_RNR_NAK_GEN = (1<<12),
IB_DEVICE_SRQ_RESIZE = (1<<13),
IB_DEVICE_N_NOTIFY_CQ = (1<<14),
+
+ /*
+ * This device supports a per-device lkey or stag that can be
+ * used without performing a memory registration for the local
+ * memory. Note that ULPs should never check this flag, but
+ * instead of use the local_dma_lkey flag in the ib_pd structure,
+ * which will always contain a usable lkey.
+ */
IB_DEVICE_LOCAL_DMA_LKEY = (1<<15),
IB_DEVICE_RESERVED = (1<<16), /* old SEND_W_INV */
IB_DEVICE_MEM_WINDOW = (1<<17),
@@ -133,6 +141,16 @@ enum ib_device_cap_flags {
IB_DEVICE_UD_IP_CSUM = (1<<18),
IB_DEVICE_UD_TSO = (1<<19),
IB_DEVICE_XRC = (1<<20),
+
+ /*
+ * This device supports the IB "base memory management extension",
+ * which includes support for fast registrations (IB_WR_REG_MR,
+ * IB_WR_LOCAL_INV and IB_WR_SEND_WITH_INV verbs). This flag should
+ * also be set by any iWarp device which must support FRs to comply
+ * to the iWarp verbs spec. iWarp devices also support the
+ * IB_WR_RDMA_READ_WITH_INV verb for RDMA READs that invalidate the
+ * stag.
+ */
IB_DEVICE_MEM_MGT_EXTENSIONS = (1<<21),
IB_DEVICE_BLOCK_MULTICAST_LOOPBACK = (1<<22),
IB_DEVICE_MEM_WINDOW_TYPE_2A = (1<<23),
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 2/9] IB: remove ib_query_mr
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-15 18:05 ` [PATCH 1/9] IB: start documenting device capabilities Christoph Hellwig
@ 2015-11-15 18:05 ` Christoph Hellwig
2015-11-15 18:05 ` [PATCH 3/9] IB: remove support for phys MRs Christoph Hellwig
` (7 subsequent siblings)
9 siblings, 0 replies; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:05 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
This functionality has no users and was only supported by the staged out
EHCA driver.
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
drivers/infiniband/core/verbs.c | 7 -----
drivers/staging/rdma/ehca/ehca_iverbs.h | 2 --
drivers/staging/rdma/ehca/ehca_main.c | 1 -
drivers/staging/rdma/ehca/ehca_mrmw.c | 49 ---------------------------------
include/rdma/ib_verbs.h | 18 ------------
5 files changed, 77 deletions(-)
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 45c197a..0e21367 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1216,13 +1216,6 @@ struct ib_mr *ib_get_dma_mr(struct ib_pd *pd, int mr_access_flags)
}
EXPORT_SYMBOL(ib_get_dma_mr);
-int ib_query_mr(struct ib_mr *mr, struct ib_mr_attr *mr_attr)
-{
- return mr->device->query_mr ?
- mr->device->query_mr(mr, mr_attr) : -ENOSYS;
-}
-EXPORT_SYMBOL(ib_query_mr);
-
int ib_dereg_mr(struct ib_mr *mr)
{
struct ib_pd *pd;
diff --git a/drivers/staging/rdma/ehca/ehca_iverbs.h b/drivers/staging/rdma/ehca/ehca_iverbs.h
index 75c9876..4a45ca3 100644
--- a/drivers/staging/rdma/ehca/ehca_iverbs.h
+++ b/drivers/staging/rdma/ehca/ehca_iverbs.h
@@ -94,8 +94,6 @@ int ehca_rereg_phys_mr(struct ib_mr *mr,
struct ib_phys_buf *phys_buf_array,
int num_phys_buf, int mr_access_flags, u64 *iova_start);
-int ehca_query_mr(struct ib_mr *mr, struct ib_mr_attr *mr_attr);
-
int ehca_dereg_mr(struct ib_mr *mr);
struct ib_mw *ehca_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
diff --git a/drivers/staging/rdma/ehca/ehca_main.c b/drivers/staging/rdma/ehca/ehca_main.c
index 285e560..0be7959 100644
--- a/drivers/staging/rdma/ehca/ehca_main.c
+++ b/drivers/staging/rdma/ehca/ehca_main.c
@@ -513,7 +513,6 @@ static int ehca_init_device(struct ehca_shca *shca)
shca->ib_device.get_dma_mr = ehca_get_dma_mr;
shca->ib_device.reg_phys_mr = ehca_reg_phys_mr;
shca->ib_device.reg_user_mr = ehca_reg_user_mr;
- shca->ib_device.query_mr = ehca_query_mr;
shca->ib_device.dereg_mr = ehca_dereg_mr;
shca->ib_device.rereg_phys_mr = ehca_rereg_phys_mr;
shca->ib_device.alloc_mw = ehca_alloc_mw;
diff --git a/drivers/staging/rdma/ehca/ehca_mrmw.c b/drivers/staging/rdma/ehca/ehca_mrmw.c
index f914b30..eb274c1 100644
--- a/drivers/staging/rdma/ehca/ehca_mrmw.c
+++ b/drivers/staging/rdma/ehca/ehca_mrmw.c
@@ -589,55 +589,6 @@ rereg_phys_mr_exit0:
return ret;
} /* end ehca_rereg_phys_mr() */
-/*----------------------------------------------------------------------*/
-
-int ehca_query_mr(struct ib_mr *mr, struct ib_mr_attr *mr_attr)
-{
- int ret = 0;
- u64 h_ret;
- struct ehca_shca *shca =
- container_of(mr->device, struct ehca_shca, ib_device);
- struct ehca_mr *e_mr = container_of(mr, struct ehca_mr, ib.ib_mr);
- unsigned long sl_flags;
- struct ehca_mr_hipzout_parms hipzout;
-
- if ((e_mr->flags & EHCA_MR_FLAG_FMR)) {
- ehca_err(mr->device, "not supported for FMR, mr=%p e_mr=%p "
- "e_mr->flags=%x", mr, e_mr, e_mr->flags);
- ret = -EINVAL;
- goto query_mr_exit0;
- }
-
- memset(mr_attr, 0, sizeof(struct ib_mr_attr));
- spin_lock_irqsave(&e_mr->mrlock, sl_flags);
-
- h_ret = hipz_h_query_mr(shca->ipz_hca_handle, e_mr, &hipzout);
- if (h_ret != H_SUCCESS) {
- ehca_err(mr->device, "hipz_mr_query failed, h_ret=%lli mr=%p "
- "hca_hndl=%llx mr_hndl=%llx lkey=%x",
- h_ret, mr, shca->ipz_hca_handle.handle,
- e_mr->ipz_mr_handle.handle, mr->lkey);
- ret = ehca2ib_return_code(h_ret);
- goto query_mr_exit1;
- }
- mr_attr->pd = mr->pd;
- mr_attr->device_virt_addr = hipzout.vaddr;
- mr_attr->size = hipzout.len;
- mr_attr->lkey = hipzout.lkey;
- mr_attr->rkey = hipzout.rkey;
- ehca_mrmw_reverse_map_acl(&hipzout.acl, &mr_attr->mr_access_flags);
-
-query_mr_exit1:
- spin_unlock_irqrestore(&e_mr->mrlock, sl_flags);
-query_mr_exit0:
- if (ret)
- ehca_err(mr->device, "ret=%i mr=%p mr_attr=%p",
- ret, mr, mr_attr);
- return ret;
-} /* end ehca_query_mr() */
-
-/*----------------------------------------------------------------------*/
-
int ehca_dereg_mr(struct ib_mr *mr)
{
int ret = 0;
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 6034f92..83d6ee8 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -1165,15 +1165,6 @@ struct ib_phys_buf {
u64 size;
};
-struct ib_mr_attr {
- struct ib_pd *pd;
- u64 device_virt_addr;
- u64 size;
- int mr_access_flags;
- u32 lkey;
- u32 rkey;
-};
-
enum ib_mr_rereg_flags {
IB_MR_REREG_TRANS = 1,
IB_MR_REREG_PD = (1<<1),
@@ -1709,8 +1700,6 @@ struct ib_device {
int mr_access_flags,
struct ib_pd *pd,
struct ib_udata *udata);
- int (*query_mr)(struct ib_mr *mr,
- struct ib_mr_attr *mr_attr);
int (*dereg_mr)(struct ib_mr *mr);
struct ib_mr * (*alloc_mr)(struct ib_pd *pd,
enum ib_mr_type mr_type,
@@ -2850,13 +2839,6 @@ static inline void ib_dma_free_coherent(struct ib_device *dev,
}
/**
- * ib_query_mr - Retrieves information about a specific memory region.
- * @mr: The memory region to retrieve information about.
- * @mr_attr: The attributes of the specified memory region.
- */
-int ib_query_mr(struct ib_mr *mr, struct ib_mr_attr *mr_attr);
-
-/**
* ib_dereg_mr - Deregisters a memory region and removes it from the
* HCA translation table.
* @mr: The memory region to deregister.
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 3/9] IB: remove support for phys MRs
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-15 18:05 ` [PATCH 1/9] IB: start documenting device capabilities Christoph Hellwig
2015-11-15 18:05 ` [PATCH 2/9] IB: remove ib_query_mr Christoph Hellwig
@ 2015-11-15 18:05 ` Christoph Hellwig
[not found] ` <1447610762-3781-4-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-15 18:05 ` [PATCH 4/9] IB: remove in-kernel support for memory windows Christoph Hellwig
` (6 subsequent siblings)
9 siblings, 1 reply; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:05 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
We have stopped using phys MRs in the kernel a while ago, so let's
remove all the cruft used to implement them.
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
drivers/infiniband/hw/cxgb3/iwch_mem.c | 31 ---
drivers/infiniband/hw/cxgb3/iwch_provider.c | 69 ------
drivers/infiniband/hw/cxgb3/iwch_provider.h | 4 -
drivers/infiniband/hw/cxgb4/iw_cxgb4.h | 11 -
drivers/infiniband/hw/cxgb4/mem.c | 248 ---------------------
drivers/infiniband/hw/cxgb4/provider.c | 2 -
drivers/infiniband/hw/mthca/mthca_provider.c | 84 -------
drivers/infiniband/hw/nes/nes_cm.c | 7 +-
drivers/infiniband/hw/nes/nes_verbs.c | 3 +-
drivers/infiniband/hw/nes/nes_verbs.h | 5 +
drivers/infiniband/hw/ocrdma/ocrdma_main.c | 1 -
drivers/infiniband/hw/ocrdma/ocrdma_verbs.c | 163 --------------
drivers/infiniband/hw/ocrdma/ocrdma_verbs.h | 3 -
drivers/infiniband/hw/qib/qib_mr.c | 51 +----
drivers/infiniband/hw/qib/qib_verbs.c | 1 -
drivers/infiniband/hw/qib/qib_verbs.h | 4 -
drivers/staging/rdma/amso1100/c2_provider.c | 1 -
drivers/staging/rdma/ehca/ehca_iverbs.h | 11 -
drivers/staging/rdma/ehca/ehca_main.c | 2 -
drivers/staging/rdma/ehca/ehca_mrmw.c | 321 ---------------------------
drivers/staging/rdma/ehca/ehca_mrmw.h | 5 -
drivers/staging/rdma/hfi1/mr.c | 51 +----
drivers/staging/rdma/hfi1/verbs.c | 1 -
drivers/staging/rdma/hfi1/verbs.h | 4 -
drivers/staging/rdma/ipath/ipath_mr.c | 55 -----
drivers/staging/rdma/ipath/ipath_verbs.c | 1 -
drivers/staging/rdma/ipath/ipath_verbs.h | 4 -
include/rdma/ib_verbs.h | 16 +-
28 files changed, 15 insertions(+), 1144 deletions(-)
diff --git a/drivers/infiniband/hw/cxgb3/iwch_mem.c b/drivers/infiniband/hw/cxgb3/iwch_mem.c
index 5c36ee2..3a5e27d 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_mem.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_mem.c
@@ -75,37 +75,6 @@ int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php,
return ret;
}
-int iwch_reregister_mem(struct iwch_dev *rhp, struct iwch_pd *php,
- struct iwch_mr *mhp,
- int shift,
- int npages)
-{
- u32 stag;
- int ret;
-
- /* We could support this... */
- if (npages > mhp->attr.pbl_size)
- return -ENOMEM;
-
- stag = mhp->attr.stag;
- if (cxio_reregister_phys_mem(&rhp->rdev,
- &stag, mhp->attr.pdid,
- mhp->attr.perms,
- mhp->attr.zbva,
- mhp->attr.va_fbo,
- mhp->attr.len,
- shift - 12,
- mhp->attr.pbl_size, mhp->attr.pbl_addr))
- return -ENOMEM;
-
- ret = iwch_finish_mem_reg(mhp, stag);
- if (ret)
- cxio_dereg_mem(&rhp->rdev, mhp->attr.stag, mhp->attr.pbl_size,
- mhp->attr.pbl_addr);
-
- return ret;
-}
-
int iwch_alloc_pbl(struct iwch_mr *mhp, int npages)
{
mhp->attr.pbl_addr = cxio_hal_pblpool_alloc(&mhp->rhp->rdev,
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c
index 1567b5b..9576e15 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c
@@ -556,73 +556,6 @@ err:
}
-static int iwch_reregister_phys_mem(struct ib_mr *mr,
- int mr_rereg_mask,
- struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf,
- int acc, u64 * iova_start)
-{
-
- struct iwch_mr mh, *mhp;
- struct iwch_pd *php;
- struct iwch_dev *rhp;
- __be64 *page_list = NULL;
- int shift = 0;
- u64 total_size;
- int npages = 0;
- int ret;
-
- PDBG("%s ib_mr %p ib_pd %p\n", __func__, mr, pd);
-
- /* There can be no memory windows */
- if (atomic_read(&mr->usecnt))
- return -EINVAL;
-
- mhp = to_iwch_mr(mr);
- rhp = mhp->rhp;
- php = to_iwch_pd(mr->pd);
-
- /* make sure we are on the same adapter */
- if (rhp != php->rhp)
- return -EINVAL;
-
- memcpy(&mh, mhp, sizeof *mhp);
-
- if (mr_rereg_mask & IB_MR_REREG_PD)
- php = to_iwch_pd(pd);
- if (mr_rereg_mask & IB_MR_REREG_ACCESS)
- mh.attr.perms = iwch_ib_to_tpt_access(acc);
- if (mr_rereg_mask & IB_MR_REREG_TRANS) {
- ret = build_phys_page_list(buffer_list, num_phys_buf,
- iova_start,
- &total_size, &npages,
- &shift, &page_list);
- if (ret)
- return ret;
- }
-
- ret = iwch_reregister_mem(rhp, php, &mh, shift, npages);
- kfree(page_list);
- if (ret) {
- return ret;
- }
- if (mr_rereg_mask & IB_MR_REREG_PD)
- mhp->attr.pdid = php->pdid;
- if (mr_rereg_mask & IB_MR_REREG_ACCESS)
- mhp->attr.perms = iwch_ib_to_tpt_access(acc);
- if (mr_rereg_mask & IB_MR_REREG_TRANS) {
- mhp->attr.zbva = 0;
- mhp->attr.va_fbo = *iova_start;
- mhp->attr.page_size = shift - 12;
- mhp->attr.len = (u32) total_size;
- mhp->attr.pbl_size = npages;
- }
-
- return 0;
-}
-
-
static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt, int acc, struct ib_udata *udata)
{
@@ -1414,8 +1347,6 @@ int iwch_register_device(struct iwch_dev *dev)
dev->ibdev.resize_cq = iwch_resize_cq;
dev->ibdev.poll_cq = iwch_poll_cq;
dev->ibdev.get_dma_mr = iwch_get_dma_mr;
- dev->ibdev.reg_phys_mr = iwch_register_phys_mem;
- dev->ibdev.rereg_phys_mr = iwch_reregister_phys_mem;
dev->ibdev.reg_user_mr = iwch_reg_user_mr;
dev->ibdev.dereg_mr = iwch_dereg_mr;
dev->ibdev.alloc_mw = iwch_alloc_mw;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h
index 2ac85b8..f4fa6d6 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.h
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h
@@ -341,10 +341,6 @@ void iwch_unregister_device(struct iwch_dev *dev);
void stop_read_rep_timer(struct iwch_qp *qhp);
int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php,
struct iwch_mr *mhp, int shift);
-int iwch_reregister_mem(struct iwch_dev *rhp, struct iwch_pd *php,
- struct iwch_mr *mhp,
- int shift,
- int npages);
int iwch_alloc_pbl(struct iwch_mr *mhp, int npages);
void iwch_free_pbl(struct iwch_mr *mhp);
int iwch_write_pbl(struct iwch_mr *mhp, __be64 *pages, int npages, int offset);
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index 00e55fa..dd00cf2 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -968,17 +968,6 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start,
u64 length, u64 virt, int acc,
struct ib_udata *udata);
struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc);
-struct ib_mr *c4iw_register_phys_mem(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf,
- int acc,
- u64 *iova_start);
-int c4iw_reregister_phys_mem(struct ib_mr *mr,
- int mr_rereg_mask,
- struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf,
- int acc, u64 *iova_start);
int c4iw_dereg_mr(struct ib_mr *ib_mr);
int c4iw_destroy_cq(struct ib_cq *ib_cq);
struct ib_cq *c4iw_create_cq(struct ib_device *ibdev,
diff --git a/drivers/infiniband/hw/cxgb4/mem.c b/drivers/infiniband/hw/cxgb4/mem.c
index e1629ab..1eb833a 100644
--- a/drivers/infiniband/hw/cxgb4/mem.c
+++ b/drivers/infiniband/hw/cxgb4/mem.c
@@ -392,32 +392,6 @@ static int register_mem(struct c4iw_dev *rhp, struct c4iw_pd *php,
return ret;
}
-static int reregister_mem(struct c4iw_dev *rhp, struct c4iw_pd *php,
- struct c4iw_mr *mhp, int shift, int npages)
-{
- u32 stag;
- int ret;
-
- if (npages > mhp->attr.pbl_size)
- return -ENOMEM;
-
- stag = mhp->attr.stag;
- ret = write_tpt_entry(&rhp->rdev, 0, &stag, 1, mhp->attr.pdid,
- FW_RI_STAG_NSMR, mhp->attr.perms,
- mhp->attr.mw_bind_enable, mhp->attr.zbva,
- mhp->attr.va_fbo, mhp->attr.len, shift - 12,
- mhp->attr.pbl_size, mhp->attr.pbl_addr);
- if (ret)
- return ret;
-
- ret = finish_mem_reg(mhp, stag);
- if (ret)
- dereg_mem(&rhp->rdev, mhp->attr.stag, mhp->attr.pbl_size,
- mhp->attr.pbl_addr);
-
- return ret;
-}
-
static int alloc_pbl(struct c4iw_mr *mhp, int npages)
{
mhp->attr.pbl_addr = c4iw_pblpool_alloc(&mhp->rhp->rdev,
@@ -431,228 +405,6 @@ static int alloc_pbl(struct c4iw_mr *mhp, int npages)
return 0;
}
-static int build_phys_page_list(struct ib_phys_buf *buffer_list,
- int num_phys_buf, u64 *iova_start,
- u64 *total_size, int *npages,
- int *shift, __be64 **page_list)
-{
- u64 mask;
- int i, j, n;
-
- mask = 0;
- *total_size = 0;
- for (i = 0; i < num_phys_buf; ++i) {
- if (i != 0 && buffer_list[i].addr & ~PAGE_MASK)
- return -EINVAL;
- if (i != 0 && i != num_phys_buf - 1 &&
- (buffer_list[i].size & ~PAGE_MASK))
- return -EINVAL;
- *total_size += buffer_list[i].size;
- if (i > 0)
- mask |= buffer_list[i].addr;
- else
- mask |= buffer_list[i].addr & PAGE_MASK;
- if (i != num_phys_buf - 1)
- mask |= buffer_list[i].addr + buffer_list[i].size;
- else
- mask |= (buffer_list[i].addr + buffer_list[i].size +
- PAGE_SIZE - 1) & PAGE_MASK;
- }
-
- if (*total_size > 0xFFFFFFFFULL)
- return -ENOMEM;
-
- /* Find largest page shift we can use to cover buffers */
- for (*shift = PAGE_SHIFT; *shift < 27; ++(*shift))
- if ((1ULL << *shift) & mask)
- break;
-
- buffer_list[0].size += buffer_list[0].addr & ((1ULL << *shift) - 1);
- buffer_list[0].addr &= ~0ull << *shift;
-
- *npages = 0;
- for (i = 0; i < num_phys_buf; ++i)
- *npages += (buffer_list[i].size +
- (1ULL << *shift) - 1) >> *shift;
-
- if (!*npages)
- return -EINVAL;
-
- *page_list = kmalloc(sizeof(u64) * *npages, GFP_KERNEL);
- if (!*page_list)
- return -ENOMEM;
-
- n = 0;
- for (i = 0; i < num_phys_buf; ++i)
- for (j = 0;
- j < (buffer_list[i].size + (1ULL << *shift) - 1) >> *shift;
- ++j)
- (*page_list)[n++] = cpu_to_be64(buffer_list[i].addr +
- ((u64) j << *shift));
-
- PDBG("%s va 0x%llx mask 0x%llx shift %d len %lld pbl_size %d\n",
- __func__, (unsigned long long)*iova_start,
- (unsigned long long)mask, *shift, (unsigned long long)*total_size,
- *npages);
-
- return 0;
-
-}
-
-int c4iw_reregister_phys_mem(struct ib_mr *mr, int mr_rereg_mask,
- struct ib_pd *pd, struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start)
-{
-
- struct c4iw_mr mh, *mhp;
- struct c4iw_pd *php;
- struct c4iw_dev *rhp;
- __be64 *page_list = NULL;
- int shift = 0;
- u64 total_size;
- int npages;
- int ret;
-
- PDBG("%s ib_mr %p ib_pd %p\n", __func__, mr, pd);
-
- /* There can be no memory windows */
- if (atomic_read(&mr->usecnt))
- return -EINVAL;
-
- mhp = to_c4iw_mr(mr);
- rhp = mhp->rhp;
- php = to_c4iw_pd(mr->pd);
-
- /* make sure we are on the same adapter */
- if (rhp != php->rhp)
- return -EINVAL;
-
- memcpy(&mh, mhp, sizeof *mhp);
-
- if (mr_rereg_mask & IB_MR_REREG_PD)
- php = to_c4iw_pd(pd);
- if (mr_rereg_mask & IB_MR_REREG_ACCESS) {
- mh.attr.perms = c4iw_ib_to_tpt_access(acc);
- mh.attr.mw_bind_enable = (acc & IB_ACCESS_MW_BIND) ==
- IB_ACCESS_MW_BIND;
- }
- if (mr_rereg_mask & IB_MR_REREG_TRANS) {
- ret = build_phys_page_list(buffer_list, num_phys_buf,
- iova_start,
- &total_size, &npages,
- &shift, &page_list);
- if (ret)
- return ret;
- }
-
- if (mr_exceeds_hw_limits(rhp, total_size)) {
- kfree(page_list);
- return -EINVAL;
- }
-
- ret = reregister_mem(rhp, php, &mh, shift, npages);
- kfree(page_list);
- if (ret)
- return ret;
- if (mr_rereg_mask & IB_MR_REREG_PD)
- mhp->attr.pdid = php->pdid;
- if (mr_rereg_mask & IB_MR_REREG_ACCESS)
- mhp->attr.perms = c4iw_ib_to_tpt_access(acc);
- if (mr_rereg_mask & IB_MR_REREG_TRANS) {
- mhp->attr.zbva = 0;
- mhp->attr.va_fbo = *iova_start;
- mhp->attr.page_size = shift - 12;
- mhp->attr.len = (u32) total_size;
- mhp->attr.pbl_size = npages;
- }
-
- return 0;
-}
-
-struct ib_mr *c4iw_register_phys_mem(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start)
-{
- __be64 *page_list;
- int shift;
- u64 total_size;
- int npages;
- struct c4iw_dev *rhp;
- struct c4iw_pd *php;
- struct c4iw_mr *mhp;
- int ret;
-
- PDBG("%s ib_pd %p\n", __func__, pd);
- php = to_c4iw_pd(pd);
- rhp = php->rhp;
-
- mhp = kzalloc(sizeof(*mhp), GFP_KERNEL);
- if (!mhp)
- return ERR_PTR(-ENOMEM);
-
- mhp->rhp = rhp;
-
- /* First check that we have enough alignment */
- if ((*iova_start & ~PAGE_MASK) != (buffer_list[0].addr & ~PAGE_MASK)) {
- ret = -EINVAL;
- goto err;
- }
-
- if (num_phys_buf > 1 &&
- ((buffer_list[0].addr + buffer_list[0].size) & ~PAGE_MASK)) {
- ret = -EINVAL;
- goto err;
- }
-
- ret = build_phys_page_list(buffer_list, num_phys_buf, iova_start,
- &total_size, &npages, &shift,
- &page_list);
- if (ret)
- goto err;
-
- if (mr_exceeds_hw_limits(rhp, total_size)) {
- kfree(page_list);
- ret = -EINVAL;
- goto err;
- }
-
- ret = alloc_pbl(mhp, npages);
- if (ret) {
- kfree(page_list);
- goto err;
- }
-
- ret = write_pbl(&mhp->rhp->rdev, page_list, mhp->attr.pbl_addr,
- npages);
- kfree(page_list);
- if (ret)
- goto err_pbl;
-
- mhp->attr.pdid = php->pdid;
- mhp->attr.zbva = 0;
-
- mhp->attr.perms = c4iw_ib_to_tpt_access(acc);
- mhp->attr.va_fbo = *iova_start;
- mhp->attr.page_size = shift - 12;
-
- mhp->attr.len = (u32) total_size;
- mhp->attr.pbl_size = npages;
- ret = register_mem(rhp, php, mhp, shift);
- if (ret)
- goto err_pbl;
-
- return &mhp->ibmr;
-
-err_pbl:
- c4iw_pblpool_free(&mhp->rhp->rdev, mhp->attr.pbl_addr,
- mhp->attr.pbl_size << 3);
-
-err:
- kfree(mhp);
- return ERR_PTR(ret);
-
-}
-
struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc)
{
struct c4iw_dev *rhp;
diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c
index b7703bc..186319e 100644
--- a/drivers/infiniband/hw/cxgb4/provider.c
+++ b/drivers/infiniband/hw/cxgb4/provider.c
@@ -509,8 +509,6 @@ int c4iw_register_device(struct c4iw_dev *dev)
dev->ibdev.resize_cq = c4iw_resize_cq;
dev->ibdev.poll_cq = c4iw_poll_cq;
dev->ibdev.get_dma_mr = c4iw_get_dma_mr;
- dev->ibdev.reg_phys_mr = c4iw_register_phys_mem;
- dev->ibdev.rereg_phys_mr = c4iw_reregister_phys_mem;
dev->ibdev.reg_user_mr = c4iw_reg_user_mr;
dev->ibdev.dereg_mr = c4iw_dereg_mr;
dev->ibdev.alloc_mw = c4iw_alloc_mw;
diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c
index 28d7a8b..6d0a1db 100644
--- a/drivers/infiniband/hw/mthca/mthca_provider.c
+++ b/drivers/infiniband/hw/mthca/mthca_provider.c
@@ -892,89 +892,6 @@ static struct ib_mr *mthca_get_dma_mr(struct ib_pd *pd, int acc)
return &mr->ibmr;
}
-static struct ib_mr *mthca_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf,
- int acc,
- u64 *iova_start)
-{
- struct mthca_mr *mr;
- u64 *page_list;
- u64 total_size;
- unsigned long mask;
- int shift;
- int npages;
- int err;
- int i, j, n;
-
- mask = buffer_list[0].addr ^ *iova_start;
- total_size = 0;
- for (i = 0; i < num_phys_buf; ++i) {
- if (i != 0)
- mask |= buffer_list[i].addr;
- if (i != num_phys_buf - 1)
- mask |= buffer_list[i].addr + buffer_list[i].size;
-
- total_size += buffer_list[i].size;
- }
-
- if (mask & ~PAGE_MASK)
- return ERR_PTR(-EINVAL);
-
- shift = __ffs(mask | 1 << 31);
-
- buffer_list[0].size += buffer_list[0].addr & ((1ULL << shift) - 1);
- buffer_list[0].addr &= ~0ull << shift;
-
- mr = kmalloc(sizeof *mr, GFP_KERNEL);
- if (!mr)
- return ERR_PTR(-ENOMEM);
-
- npages = 0;
- for (i = 0; i < num_phys_buf; ++i)
- npages += (buffer_list[i].size + (1ULL << shift) - 1) >> shift;
-
- if (!npages)
- return &mr->ibmr;
-
- page_list = kmalloc(npages * sizeof *page_list, GFP_KERNEL);
- if (!page_list) {
- kfree(mr);
- return ERR_PTR(-ENOMEM);
- }
-
- n = 0;
- for (i = 0; i < num_phys_buf; ++i)
- for (j = 0;
- j < (buffer_list[i].size + (1ULL << shift) - 1) >> shift;
- ++j)
- page_list[n++] = buffer_list[i].addr + ((u64) j << shift);
-
- mthca_dbg(to_mdev(pd->device), "Registering memory at %llx (iova %llx) "
- "in PD %x; shift %d, npages %d.\n",
- (unsigned long long) buffer_list[0].addr,
- (unsigned long long) *iova_start,
- to_mpd(pd)->pd_num,
- shift, npages);
-
- err = mthca_mr_alloc_phys(to_mdev(pd->device),
- to_mpd(pd)->pd_num,
- page_list, shift, npages,
- *iova_start, total_size,
- convert_access(acc), mr);
-
- if (err) {
- kfree(page_list);
- kfree(mr);
- return ERR_PTR(err);
- }
-
- kfree(page_list);
- mr->umem = NULL;
-
- return &mr->ibmr;
-}
-
static struct ib_mr *mthca_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt, int acc, struct ib_udata *udata)
{
@@ -1339,7 +1256,6 @@ int mthca_register_device(struct mthca_dev *dev)
dev->ib_dev.destroy_cq = mthca_destroy_cq;
dev->ib_dev.poll_cq = mthca_poll_cq;
dev->ib_dev.get_dma_mr = mthca_get_dma_mr;
- dev->ib_dev.reg_phys_mr = mthca_reg_phys_mr;
dev->ib_dev.reg_user_mr = mthca_reg_user_mr;
dev->ib_dev.dereg_mr = mthca_dereg_mr;
dev->ib_dev.get_port_immutable = mthca_port_immutable;
diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c
index 8a3ad17..242c87d 100644
--- a/drivers/infiniband/hw/nes/nes_cm.c
+++ b/drivers/infiniband/hw/nes/nes_cm.c
@@ -3319,10 +3319,9 @@ int nes_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
ibphysbuf.addr = nesqp->ietf_frame_pbase + mpa_frame_offset;
ibphysbuf.size = buff_len;
tagged_offset = (u64)(unsigned long)*start_buff;
- ibmr = nesibdev->ibdev.reg_phys_mr((struct ib_pd *)nespd,
- &ibphysbuf, 1,
- IB_ACCESS_LOCAL_WRITE,
- &tagged_offset);
+ ibmr = nes_reg_phys_mr(&nespd->ibpd, &ibphysbuf, 1,
+ IB_ACCESS_LOCAL_WRITE,
+ &tagged_offset);
if (!ibmr) {
nes_debug(NES_DBG_CM, "Unable to register memory region"
"for lSMM for cm_node = %p \n",
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index 2bad036..453ebc2 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -2019,7 +2019,7 @@ static int nes_reg_mr(struct nes_device *nesdev, struct nes_pd *nespd,
/**
* nes_reg_phys_mr
*/
-static struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
+struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
struct ib_phys_buf *buffer_list, int num_phys_buf, int acc,
u64 * iova_start)
{
@@ -3832,7 +3832,6 @@ struct nes_ib_device *nes_init_ofa_device(struct net_device *netdev)
nesibdev->ibdev.destroy_cq = nes_destroy_cq;
nesibdev->ibdev.poll_cq = nes_poll_cq;
nesibdev->ibdev.get_dma_mr = nes_get_dma_mr;
- nesibdev->ibdev.reg_phys_mr = nes_reg_phys_mr;
nesibdev->ibdev.reg_user_mr = nes_reg_user_mr;
nesibdev->ibdev.dereg_mr = nes_dereg_mr;
nesibdev->ibdev.alloc_mw = nes_alloc_mw;
diff --git a/drivers/infiniband/hw/nes/nes_verbs.h b/drivers/infiniband/hw/nes/nes_verbs.h
index a204b67..38e38cf 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.h
+++ b/drivers/infiniband/hw/nes/nes_verbs.h
@@ -190,4 +190,9 @@ struct nes_qp {
u8 pau_state;
__u64 nesuqp_addr;
};
+
+struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
+ struct ib_phys_buf *buffer_list, int num_phys_buf, int acc,
+ u64 * iova_start);
+
#endif /* NES_VERBS_H */
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_main.c b/drivers/infiniband/hw/ocrdma/ocrdma_main.c
index 963be66..f91131f 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_main.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_main.c
@@ -174,7 +174,6 @@ static int ocrdma_register_device(struct ocrdma_dev *dev)
dev->ibdev.req_notify_cq = ocrdma_arm_cq;
dev->ibdev.get_dma_mr = ocrdma_get_dma_mr;
- dev->ibdev.reg_phys_mr = ocrdma_reg_kernel_mr;
dev->ibdev.dereg_mr = ocrdma_dereg_mr;
dev->ibdev.reg_user_mr = ocrdma_reg_user_mr;
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
index 74cf67a..3cffab7 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c
@@ -3017,169 +3017,6 @@ pl_err:
return ERR_PTR(-ENOMEM);
}
-#define MAX_KERNEL_PBE_SIZE 65536
-static inline int count_kernel_pbes(struct ib_phys_buf *buf_list,
- int buf_cnt, u32 *pbe_size)
-{
- u64 total_size = 0;
- u64 buf_size = 0;
- int i;
- *pbe_size = roundup(buf_list[0].size, PAGE_SIZE);
- *pbe_size = roundup_pow_of_two(*pbe_size);
-
- /* find the smallest PBE size that we can have */
- for (i = 0; i < buf_cnt; i++) {
- /* first addr may not be page aligned, so ignore checking */
- if ((i != 0) && ((buf_list[i].addr & ~PAGE_MASK) ||
- (buf_list[i].size & ~PAGE_MASK))) {
- return 0;
- }
-
- /* if configured PBE size is greater then the chosen one,
- * reduce the PBE size.
- */
- buf_size = roundup(buf_list[i].size, PAGE_SIZE);
- /* pbe_size has to be even multiple of 4K 1,2,4,8...*/
- buf_size = roundup_pow_of_two(buf_size);
- if (*pbe_size > buf_size)
- *pbe_size = buf_size;
-
- total_size += buf_size;
- }
- *pbe_size = *pbe_size > MAX_KERNEL_PBE_SIZE ?
- (MAX_KERNEL_PBE_SIZE) : (*pbe_size);
-
- /* num_pbes = total_size / (*pbe_size); this is implemented below. */
-
- return total_size >> ilog2(*pbe_size);
-}
-
-static void build_kernel_pbes(struct ib_phys_buf *buf_list, int ib_buf_cnt,
- u32 pbe_size, struct ocrdma_pbl *pbl_tbl,
- struct ocrdma_hw_mr *hwmr)
-{
- int i;
- int idx;
- int pbes_per_buf = 0;
- u64 buf_addr = 0;
- int num_pbes;
- struct ocrdma_pbe *pbe;
- int total_num_pbes = 0;
-
- if (!hwmr->num_pbes)
- return;
-
- pbe = (struct ocrdma_pbe *)pbl_tbl->va;
- num_pbes = 0;
-
- /* go through the OS phy regions & fill hw pbe entries into pbls. */
- for (i = 0; i < ib_buf_cnt; i++) {
- buf_addr = buf_list[i].addr;
- pbes_per_buf =
- roundup_pow_of_two(roundup(buf_list[i].size, PAGE_SIZE)) /
- pbe_size;
- hwmr->len += buf_list[i].size;
- /* number of pbes can be more for one OS buf, when
- * buffers are of different sizes.
- * split the ib_buf to one or more pbes.
- */
- for (idx = 0; idx < pbes_per_buf; idx++) {
- /* we program always page aligned addresses,
- * first unaligned address is taken care by fbo.
- */
- if (i == 0) {
- /* for non zero fbo, assign the
- * start of the page.
- */
- pbe->pa_lo =
- cpu_to_le32((u32) (buf_addr & PAGE_MASK));
- pbe->pa_hi =
- cpu_to_le32((u32) upper_32_bits(buf_addr));
- } else {
- pbe->pa_lo =
- cpu_to_le32((u32) (buf_addr & 0xffffffff));
- pbe->pa_hi =
- cpu_to_le32((u32) upper_32_bits(buf_addr));
- }
- buf_addr += pbe_size;
- num_pbes += 1;
- total_num_pbes += 1;
- pbe++;
-
- if (total_num_pbes == hwmr->num_pbes)
- goto mr_tbl_done;
- /* if the pbl is full storing the pbes,
- * move to next pbl.
- */
- if (num_pbes == (hwmr->pbl_size/sizeof(u64))) {
- pbl_tbl++;
- pbe = (struct ocrdma_pbe *)pbl_tbl->va;
- num_pbes = 0;
- }
- }
- }
-mr_tbl_done:
- return;
-}
-
-struct ib_mr *ocrdma_reg_kernel_mr(struct ib_pd *ibpd,
- struct ib_phys_buf *buf_list,
- int buf_cnt, int acc, u64 *iova_start)
-{
- int status = -ENOMEM;
- struct ocrdma_mr *mr;
- struct ocrdma_pd *pd = get_ocrdma_pd(ibpd);
- struct ocrdma_dev *dev = get_ocrdma_dev(ibpd->device);
- u32 num_pbes;
- u32 pbe_size = 0;
-
- if ((acc & IB_ACCESS_REMOTE_WRITE) && !(acc & IB_ACCESS_LOCAL_WRITE))
- return ERR_PTR(-EINVAL);
-
- mr = kzalloc(sizeof(*mr), GFP_KERNEL);
- if (!mr)
- return ERR_PTR(status);
-
- num_pbes = count_kernel_pbes(buf_list, buf_cnt, &pbe_size);
- if (num_pbes == 0) {
- status = -EINVAL;
- goto pbl_err;
- }
- status = ocrdma_get_pbl_info(dev, mr, num_pbes);
- if (status)
- goto pbl_err;
-
- mr->hwmr.pbe_size = pbe_size;
- mr->hwmr.fbo = *iova_start - (buf_list[0].addr & PAGE_MASK);
- mr->hwmr.va = *iova_start;
- mr->hwmr.local_rd = 1;
- mr->hwmr.remote_wr = (acc & IB_ACCESS_REMOTE_WRITE) ? 1 : 0;
- mr->hwmr.remote_rd = (acc & IB_ACCESS_REMOTE_READ) ? 1 : 0;
- mr->hwmr.local_wr = (acc & IB_ACCESS_LOCAL_WRITE) ? 1 : 0;
- mr->hwmr.remote_atomic = (acc & IB_ACCESS_REMOTE_ATOMIC) ? 1 : 0;
- mr->hwmr.mw_bind = (acc & IB_ACCESS_MW_BIND) ? 1 : 0;
-
- status = ocrdma_build_pbl_tbl(dev, &mr->hwmr);
- if (status)
- goto pbl_err;
- build_kernel_pbes(buf_list, buf_cnt, pbe_size, mr->hwmr.pbl_table,
- &mr->hwmr);
- status = ocrdma_reg_mr(dev, &mr->hwmr, pd->id, acc);
- if (status)
- goto mbx_err;
-
- mr->ibmr.lkey = mr->hwmr.lkey;
- if (mr->hwmr.remote_wr || mr->hwmr.remote_rd)
- mr->ibmr.rkey = mr->hwmr.lkey;
- return &mr->ibmr;
-
-mbx_err:
- ocrdma_free_mr_pbl_tbl(dev, &mr->hwmr);
-pbl_err:
- kfree(mr);
- return ERR_PTR(status);
-}
-
static int ocrdma_set_page(struct ib_mr *ibmr, u64 addr)
{
struct ocrdma_mr *mr = get_ocrdma_mr(ibmr);
diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.h b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.h
index f2ce048..82f476f 100644
--- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.h
+++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.h
@@ -115,9 +115,6 @@ int ocrdma_post_srq_recv(struct ib_srq *, struct ib_recv_wr *,
int ocrdma_dereg_mr(struct ib_mr *);
struct ib_mr *ocrdma_get_dma_mr(struct ib_pd *, int acc);
-struct ib_mr *ocrdma_reg_kernel_mr(struct ib_pd *,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start);
struct ib_mr *ocrdma_reg_user_mr(struct ib_pd *, u64 start, u64 length,
u64 virt, int acc, struct ib_udata *);
struct ib_mr *ocrdma_alloc_mr(struct ib_pd *pd,
diff --git a/drivers/infiniband/hw/qib/qib_mr.c b/drivers/infiniband/hw/qib/qib_mr.c
index 294f5c7..5f53304 100644
--- a/drivers/infiniband/hw/qib/qib_mr.c
+++ b/drivers/infiniband/hw/qib/qib_mr.c
@@ -150,10 +150,7 @@ static struct qib_mr *alloc_mr(int count, struct ib_pd *pd)
rval = init_qib_mregion(&mr->mr, pd, count);
if (rval)
goto bail;
- /*
- * ib_reg_phys_mr() will initialize mr->ibmr except for
- * lkey and rkey.
- */
+
rval = qib_alloc_lkey(&mr->mr, 0);
if (rval)
goto bail_mregion;
@@ -171,52 +168,6 @@ bail:
}
/**
- * qib_reg_phys_mr - register a physical memory region
- * @pd: protection domain for this memory region
- * @buffer_list: pointer to the list of physical buffers to register
- * @num_phys_buf: the number of physical buffers to register
- * @iova_start: the starting address passed over IB which maps to this MR
- *
- * Returns the memory region on success, otherwise returns an errno.
- */
-struct ib_mr *qib_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start)
-{
- struct qib_mr *mr;
- int n, m, i;
- struct ib_mr *ret;
-
- mr = alloc_mr(num_phys_buf, pd);
- if (IS_ERR(mr)) {
- ret = (struct ib_mr *)mr;
- goto bail;
- }
-
- mr->mr.user_base = *iova_start;
- mr->mr.iova = *iova_start;
- mr->mr.access_flags = acc;
-
- m = 0;
- n = 0;
- for (i = 0; i < num_phys_buf; i++) {
- mr->mr.map[m]->segs[n].vaddr = (void *) buffer_list[i].addr;
- mr->mr.map[m]->segs[n].length = buffer_list[i].size;
- mr->mr.length += buffer_list[i].size;
- n++;
- if (n == QIB_SEGSZ) {
- m++;
- n = 0;
- }
- }
-
- ret = &mr->ibmr;
-
-bail:
- return ret;
-}
-
-/**
* qib_reg_user_mr - register a userspace memory region
* @pd: protection domain for this memory region
* @start: starting userspace address
diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c
index 9e1af0b..e56bd65 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.c
+++ b/drivers/infiniband/hw/qib/qib_verbs.c
@@ -2206,7 +2206,6 @@ int qib_register_ib_device(struct qib_devdata *dd)
ibdev->poll_cq = qib_poll_cq;
ibdev->req_notify_cq = qib_req_notify_cq;
ibdev->get_dma_mr = qib_get_dma_mr;
- ibdev->reg_phys_mr = qib_reg_phys_mr;
ibdev->reg_user_mr = qib_reg_user_mr;
ibdev->dereg_mr = qib_dereg_mr;
ibdev->alloc_mr = qib_alloc_mr;
diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h
index 2baf5ad..2d538c8c 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.h
+++ b/drivers/infiniband/hw/qib/qib_verbs.h
@@ -1032,10 +1032,6 @@ int qib_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata);
struct ib_mr *qib_get_dma_mr(struct ib_pd *pd, int acc);
-struct ib_mr *qib_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start);
-
struct ib_mr *qib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt_addr, int mr_access_flags,
struct ib_udata *udata);
diff --git a/drivers/staging/rdma/amso1100/c2_provider.c b/drivers/staging/rdma/amso1100/c2_provider.c
index ee2ff87..6c3e9cc 100644
--- a/drivers/staging/rdma/amso1100/c2_provider.c
+++ b/drivers/staging/rdma/amso1100/c2_provider.c
@@ -831,7 +831,6 @@ int c2_register_device(struct c2_dev *dev)
dev->ibdev.destroy_cq = c2_destroy_cq;
dev->ibdev.poll_cq = c2_poll_cq;
dev->ibdev.get_dma_mr = c2_get_dma_mr;
- dev->ibdev.reg_phys_mr = c2_reg_phys_mr;
dev->ibdev.reg_user_mr = c2_reg_user_mr;
dev->ibdev.dereg_mr = c2_dereg_mr;
dev->ibdev.get_port_immutable = c2_port_immutable;
diff --git a/drivers/staging/rdma/ehca/ehca_iverbs.h b/drivers/staging/rdma/ehca/ehca_iverbs.h
index 4a45ca3..219f635 100644
--- a/drivers/staging/rdma/ehca/ehca_iverbs.h
+++ b/drivers/staging/rdma/ehca/ehca_iverbs.h
@@ -79,21 +79,10 @@ int ehca_destroy_ah(struct ib_ah *ah);
struct ib_mr *ehca_get_dma_mr(struct ib_pd *pd, int mr_access_flags);
-struct ib_mr *ehca_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *phys_buf_array,
- int num_phys_buf,
- int mr_access_flags, u64 *iova_start);
-
struct ib_mr *ehca_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt, int mr_access_flags,
struct ib_udata *udata);
-int ehca_rereg_phys_mr(struct ib_mr *mr,
- int mr_rereg_mask,
- struct ib_pd *pd,
- struct ib_phys_buf *phys_buf_array,
- int num_phys_buf, int mr_access_flags, u64 *iova_start);
-
int ehca_dereg_mr(struct ib_mr *mr);
struct ib_mw *ehca_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
diff --git a/drivers/staging/rdma/ehca/ehca_main.c b/drivers/staging/rdma/ehca/ehca_main.c
index 0be7959..ba023426 100644
--- a/drivers/staging/rdma/ehca/ehca_main.c
+++ b/drivers/staging/rdma/ehca/ehca_main.c
@@ -511,10 +511,8 @@ static int ehca_init_device(struct ehca_shca *shca)
shca->ib_device.req_notify_cq = ehca_req_notify_cq;
/* shca->ib_device.req_ncomp_notif = ehca_req_ncomp_notif; */
shca->ib_device.get_dma_mr = ehca_get_dma_mr;
- shca->ib_device.reg_phys_mr = ehca_reg_phys_mr;
shca->ib_device.reg_user_mr = ehca_reg_user_mr;
shca->ib_device.dereg_mr = ehca_dereg_mr;
- shca->ib_device.rereg_phys_mr = ehca_rereg_phys_mr;
shca->ib_device.alloc_mw = ehca_alloc_mw;
shca->ib_device.bind_mw = ehca_bind_mw;
shca->ib_device.dealloc_mw = ehca_dealloc_mw;
diff --git a/drivers/staging/rdma/ehca/ehca_mrmw.c b/drivers/staging/rdma/ehca/ehca_mrmw.c
index eb274c1..1c1a8dd 100644
--- a/drivers/staging/rdma/ehca/ehca_mrmw.c
+++ b/drivers/staging/rdma/ehca/ehca_mrmw.c
@@ -196,120 +196,6 @@ get_dma_mr_exit0:
/*----------------------------------------------------------------------*/
-struct ib_mr *ehca_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *phys_buf_array,
- int num_phys_buf,
- int mr_access_flags,
- u64 *iova_start)
-{
- struct ib_mr *ib_mr;
- int ret;
- struct ehca_mr *e_mr;
- struct ehca_shca *shca =
- container_of(pd->device, struct ehca_shca, ib_device);
- struct ehca_pd *e_pd = container_of(pd, struct ehca_pd, ib_pd);
-
- u64 size;
-
- if ((num_phys_buf <= 0) || !phys_buf_array) {
- ehca_err(pd->device, "bad input values: num_phys_buf=%x "
- "phys_buf_array=%p", num_phys_buf, phys_buf_array);
- ib_mr = ERR_PTR(-EINVAL);
- goto reg_phys_mr_exit0;
- }
- if (((mr_access_flags & IB_ACCESS_REMOTE_WRITE) &&
- !(mr_access_flags & IB_ACCESS_LOCAL_WRITE)) ||
- ((mr_access_flags & IB_ACCESS_REMOTE_ATOMIC) &&
- !(mr_access_flags & IB_ACCESS_LOCAL_WRITE))) {
- /*
- * Remote Write Access requires Local Write Access
- * Remote Atomic Access requires Local Write Access
- */
- ehca_err(pd->device, "bad input values: mr_access_flags=%x",
- mr_access_flags);
- ib_mr = ERR_PTR(-EINVAL);
- goto reg_phys_mr_exit0;
- }
-
- /* check physical buffer list and calculate size */
- ret = ehca_mr_chk_buf_and_calc_size(phys_buf_array, num_phys_buf,
- iova_start, &size);
- if (ret) {
- ib_mr = ERR_PTR(ret);
- goto reg_phys_mr_exit0;
- }
- if ((size == 0) ||
- (((u64)iova_start + size) < (u64)iova_start)) {
- ehca_err(pd->device, "bad input values: size=%llx iova_start=%p",
- size, iova_start);
- ib_mr = ERR_PTR(-EINVAL);
- goto reg_phys_mr_exit0;
- }
-
- e_mr = ehca_mr_new();
- if (!e_mr) {
- ehca_err(pd->device, "out of memory");
- ib_mr = ERR_PTR(-ENOMEM);
- goto reg_phys_mr_exit0;
- }
-
- /* register MR on HCA */
- if (ehca_mr_is_maxmr(size, iova_start)) {
- e_mr->flags |= EHCA_MR_FLAG_MAXMR;
- ret = ehca_reg_maxmr(shca, e_mr, iova_start, mr_access_flags,
- e_pd, &e_mr->ib.ib_mr.lkey,
- &e_mr->ib.ib_mr.rkey);
- if (ret) {
- ib_mr = ERR_PTR(ret);
- goto reg_phys_mr_exit1;
- }
- } else {
- struct ehca_mr_pginfo pginfo;
- u32 num_kpages;
- u32 num_hwpages;
- u64 hw_pgsize;
-
- num_kpages = NUM_CHUNKS(((u64)iova_start % PAGE_SIZE) + size,
- PAGE_SIZE);
- /* for kernel space we try most possible pgsize */
- hw_pgsize = ehca_get_max_hwpage_size(shca);
- num_hwpages = NUM_CHUNKS(((u64)iova_start % hw_pgsize) + size,
- hw_pgsize);
- memset(&pginfo, 0, sizeof(pginfo));
- pginfo.type = EHCA_MR_PGI_PHYS;
- pginfo.num_kpages = num_kpages;
- pginfo.hwpage_size = hw_pgsize;
- pginfo.num_hwpages = num_hwpages;
- pginfo.u.phy.num_phys_buf = num_phys_buf;
- pginfo.u.phy.phys_buf_array = phys_buf_array;
- pginfo.next_hwpage =
- ((u64)iova_start & ~PAGE_MASK) / hw_pgsize;
-
- ret = ehca_reg_mr(shca, e_mr, iova_start, size, mr_access_flags,
- e_pd, &pginfo, &e_mr->ib.ib_mr.lkey,
- &e_mr->ib.ib_mr.rkey, EHCA_REG_MR);
- if (ret) {
- ib_mr = ERR_PTR(ret);
- goto reg_phys_mr_exit1;
- }
- }
-
- /* successful registration of all pages */
- return &e_mr->ib.ib_mr;
-
-reg_phys_mr_exit1:
- ehca_mr_delete(e_mr);
-reg_phys_mr_exit0:
- if (IS_ERR(ib_mr))
- ehca_err(pd->device, "h_ret=%li pd=%p phys_buf_array=%p "
- "num_phys_buf=%x mr_access_flags=%x iova_start=%p",
- PTR_ERR(ib_mr), pd, phys_buf_array,
- num_phys_buf, mr_access_flags, iova_start);
- return ib_mr;
-} /* end ehca_reg_phys_mr() */
-
-/*----------------------------------------------------------------------*/
-
struct ib_mr *ehca_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt, int mr_access_flags,
struct ib_udata *udata)
@@ -437,158 +323,6 @@ reg_user_mr_exit0:
/*----------------------------------------------------------------------*/
-int ehca_rereg_phys_mr(struct ib_mr *mr,
- int mr_rereg_mask,
- struct ib_pd *pd,
- struct ib_phys_buf *phys_buf_array,
- int num_phys_buf,
- int mr_access_flags,
- u64 *iova_start)
-{
- int ret;
-
- struct ehca_shca *shca =
- container_of(mr->device, struct ehca_shca, ib_device);
- struct ehca_mr *e_mr = container_of(mr, struct ehca_mr, ib.ib_mr);
- u64 new_size;
- u64 *new_start;
- u32 new_acl;
- struct ehca_pd *new_pd;
- u32 tmp_lkey, tmp_rkey;
- unsigned long sl_flags;
- u32 num_kpages = 0;
- u32 num_hwpages = 0;
- struct ehca_mr_pginfo pginfo;
-
- if (!(mr_rereg_mask & IB_MR_REREG_TRANS)) {
- /* TODO not supported, because PHYP rereg hCall needs pages */
- ehca_err(mr->device, "rereg without IB_MR_REREG_TRANS not "
- "supported yet, mr_rereg_mask=%x", mr_rereg_mask);
- ret = -EINVAL;
- goto rereg_phys_mr_exit0;
- }
-
- if (mr_rereg_mask & IB_MR_REREG_PD) {
- if (!pd) {
- ehca_err(mr->device, "rereg with bad pd, pd=%p "
- "mr_rereg_mask=%x", pd, mr_rereg_mask);
- ret = -EINVAL;
- goto rereg_phys_mr_exit0;
- }
- }
-
- if ((mr_rereg_mask &
- ~(IB_MR_REREG_TRANS | IB_MR_REREG_PD | IB_MR_REREG_ACCESS)) ||
- (mr_rereg_mask == 0)) {
- ret = -EINVAL;
- goto rereg_phys_mr_exit0;
- }
-
- /* check other parameters */
- if (e_mr == shca->maxmr) {
- /* should be impossible, however reject to be sure */
- ehca_err(mr->device, "rereg internal max-MR impossible, mr=%p "
- "shca->maxmr=%p mr->lkey=%x",
- mr, shca->maxmr, mr->lkey);
- ret = -EINVAL;
- goto rereg_phys_mr_exit0;
- }
- if (mr_rereg_mask & IB_MR_REREG_TRANS) { /* transl., i.e. addr/size */
- if (e_mr->flags & EHCA_MR_FLAG_FMR) {
- ehca_err(mr->device, "not supported for FMR, mr=%p "
- "flags=%x", mr, e_mr->flags);
- ret = -EINVAL;
- goto rereg_phys_mr_exit0;
- }
- if (!phys_buf_array || num_phys_buf <= 0) {
- ehca_err(mr->device, "bad input values mr_rereg_mask=%x"
- " phys_buf_array=%p num_phys_buf=%x",
- mr_rereg_mask, phys_buf_array, num_phys_buf);
- ret = -EINVAL;
- goto rereg_phys_mr_exit0;
- }
- }
- if ((mr_rereg_mask & IB_MR_REREG_ACCESS) && /* change ACL */
- (((mr_access_flags & IB_ACCESS_REMOTE_WRITE) &&
- !(mr_access_flags & IB_ACCESS_LOCAL_WRITE)) ||
- ((mr_access_flags & IB_ACCESS_REMOTE_ATOMIC) &&
- !(mr_access_flags & IB_ACCESS_LOCAL_WRITE)))) {
- /*
- * Remote Write Access requires Local Write Access
- * Remote Atomic Access requires Local Write Access
- */
- ehca_err(mr->device, "bad input values: mr_rereg_mask=%x "
- "mr_access_flags=%x", mr_rereg_mask, mr_access_flags);
- ret = -EINVAL;
- goto rereg_phys_mr_exit0;
- }
-
- /* set requested values dependent on rereg request */
- spin_lock_irqsave(&e_mr->mrlock, sl_flags);
- new_start = e_mr->start;
- new_size = e_mr->size;
- new_acl = e_mr->acl;
- new_pd = container_of(mr->pd, struct ehca_pd, ib_pd);
-
- if (mr_rereg_mask & IB_MR_REREG_TRANS) {
- u64 hw_pgsize = ehca_get_max_hwpage_size(shca);
-
- new_start = iova_start; /* change address */
- /* check physical buffer list and calculate size */
- ret = ehca_mr_chk_buf_and_calc_size(phys_buf_array,
- num_phys_buf, iova_start,
- &new_size);
- if (ret)
- goto rereg_phys_mr_exit1;
- if ((new_size == 0) ||
- (((u64)iova_start + new_size) < (u64)iova_start)) {
- ehca_err(mr->device, "bad input values: new_size=%llx "
- "iova_start=%p", new_size, iova_start);
- ret = -EINVAL;
- goto rereg_phys_mr_exit1;
- }
- num_kpages = NUM_CHUNKS(((u64)new_start % PAGE_SIZE) +
- new_size, PAGE_SIZE);
- num_hwpages = NUM_CHUNKS(((u64)new_start % hw_pgsize) +
- new_size, hw_pgsize);
- memset(&pginfo, 0, sizeof(pginfo));
- pginfo.type = EHCA_MR_PGI_PHYS;
- pginfo.num_kpages = num_kpages;
- pginfo.hwpage_size = hw_pgsize;
- pginfo.num_hwpages = num_hwpages;
- pginfo.u.phy.num_phys_buf = num_phys_buf;
- pginfo.u.phy.phys_buf_array = phys_buf_array;
- pginfo.next_hwpage =
- ((u64)iova_start & ~PAGE_MASK) / hw_pgsize;
- }
- if (mr_rereg_mask & IB_MR_REREG_ACCESS)
- new_acl = mr_access_flags;
- if (mr_rereg_mask & IB_MR_REREG_PD)
- new_pd = container_of(pd, struct ehca_pd, ib_pd);
-
- ret = ehca_rereg_mr(shca, e_mr, new_start, new_size, new_acl,
- new_pd, &pginfo, &tmp_lkey, &tmp_rkey);
- if (ret)
- goto rereg_phys_mr_exit1;
-
- /* successful reregistration */
- if (mr_rereg_mask & IB_MR_REREG_PD)
- mr->pd = pd;
- mr->lkey = tmp_lkey;
- mr->rkey = tmp_rkey;
-
-rereg_phys_mr_exit1:
- spin_unlock_irqrestore(&e_mr->mrlock, sl_flags);
-rereg_phys_mr_exit0:
- if (ret)
- ehca_err(mr->device, "ret=%i mr=%p mr_rereg_mask=%x pd=%p "
- "phys_buf_array=%p num_phys_buf=%x mr_access_flags=%x "
- "iova_start=%p",
- ret, mr, mr_rereg_mask, pd, phys_buf_array,
- num_phys_buf, mr_access_flags, iova_start);
- return ret;
-} /* end ehca_rereg_phys_mr() */
-
int ehca_dereg_mr(struct ib_mr *mr)
{
int ret = 0;
@@ -1713,61 +1447,6 @@ ehca_dereg_internal_maxmr_exit0:
/*----------------------------------------------------------------------*/
-/*
- * check physical buffer array of MR verbs for validness and
- * calculates MR size
- */
-int ehca_mr_chk_buf_and_calc_size(struct ib_phys_buf *phys_buf_array,
- int num_phys_buf,
- u64 *iova_start,
- u64 *size)
-{
- struct ib_phys_buf *pbuf = phys_buf_array;
- u64 size_count = 0;
- u32 i;
-
- if (num_phys_buf == 0) {
- ehca_gen_err("bad phys buf array len, num_phys_buf=0");
- return -EINVAL;
- }
- /* check first buffer */
- if (((u64)iova_start & ~PAGE_MASK) != (pbuf->addr & ~PAGE_MASK)) {
- ehca_gen_err("iova_start/addr mismatch, iova_start=%p "
- "pbuf->addr=%llx pbuf->size=%llx",
- iova_start, pbuf->addr, pbuf->size);
- return -EINVAL;
- }
- if (((pbuf->addr + pbuf->size) % PAGE_SIZE) &&
- (num_phys_buf > 1)) {
- ehca_gen_err("addr/size mismatch in 1st buf, pbuf->addr=%llx "
- "pbuf->size=%llx", pbuf->addr, pbuf->size);
- return -EINVAL;
- }
-
- for (i = 0; i < num_phys_buf; i++) {
- if ((i > 0) && (pbuf->addr % PAGE_SIZE)) {
- ehca_gen_err("bad address, i=%x pbuf->addr=%llx "
- "pbuf->size=%llx",
- i, pbuf->addr, pbuf->size);
- return -EINVAL;
- }
- if (((i > 0) && /* not 1st */
- (i < (num_phys_buf - 1)) && /* not last */
- (pbuf->size % PAGE_SIZE)) || (pbuf->size == 0)) {
- ehca_gen_err("bad size, i=%x pbuf->size=%llx",
- i, pbuf->size);
- return -EINVAL;
- }
- size_count += pbuf->size;
- pbuf++;
- }
-
- *size = size_count;
- return 0;
-} /* end ehca_mr_chk_buf_and_calc_size() */
-
-/*----------------------------------------------------------------------*/
-
/* check page list of map FMR verb for validness */
int ehca_fmr_check_page_list(struct ehca_mr *e_fmr,
u64 *page_list,
diff --git a/drivers/staging/rdma/ehca/ehca_mrmw.h b/drivers/staging/rdma/ehca/ehca_mrmw.h
index 50d8b51..52bfa95 100644
--- a/drivers/staging/rdma/ehca/ehca_mrmw.h
+++ b/drivers/staging/rdma/ehca/ehca_mrmw.h
@@ -98,11 +98,6 @@ int ehca_reg_maxmr(struct ehca_shca *shca,
int ehca_dereg_internal_maxmr(struct ehca_shca *shca);
-int ehca_mr_chk_buf_and_calc_size(struct ib_phys_buf *phys_buf_array,
- int num_phys_buf,
- u64 *iova_start,
- u64 *size);
-
int ehca_fmr_check_page_list(struct ehca_mr *e_fmr,
u64 *page_list,
int list_len);
diff --git a/drivers/staging/rdma/hfi1/mr.c b/drivers/staging/rdma/hfi1/mr.c
index 568f185..a3f8b88 100644
--- a/drivers/staging/rdma/hfi1/mr.c
+++ b/drivers/staging/rdma/hfi1/mr.c
@@ -167,10 +167,7 @@ static struct hfi1_mr *alloc_mr(int count, struct ib_pd *pd)
rval = init_mregion(&mr->mr, pd, count);
if (rval)
goto bail;
- /*
- * ib_reg_phys_mr() will initialize mr->ibmr except for
- * lkey and rkey.
- */
+
rval = hfi1_alloc_lkey(&mr->mr, 0);
if (rval)
goto bail_mregion;
@@ -188,52 +185,6 @@ bail:
}
/**
- * hfi1_reg_phys_mr - register a physical memory region
- * @pd: protection domain for this memory region
- * @buffer_list: pointer to the list of physical buffers to register
- * @num_phys_buf: the number of physical buffers to register
- * @iova_start: the starting address passed over IB which maps to this MR
- *
- * Returns the memory region on success, otherwise returns an errno.
- */
-struct ib_mr *hfi1_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start)
-{
- struct hfi1_mr *mr;
- int n, m, i;
- struct ib_mr *ret;
-
- mr = alloc_mr(num_phys_buf, pd);
- if (IS_ERR(mr)) {
- ret = (struct ib_mr *)mr;
- goto bail;
- }
-
- mr->mr.user_base = *iova_start;
- mr->mr.iova = *iova_start;
- mr->mr.access_flags = acc;
-
- m = 0;
- n = 0;
- for (i = 0; i < num_phys_buf; i++) {
- mr->mr.map[m]->segs[n].vaddr = (void *) buffer_list[i].addr;
- mr->mr.map[m]->segs[n].length = buffer_list[i].size;
- mr->mr.length += buffer_list[i].size;
- n++;
- if (n == HFI1_SEGSZ) {
- m++;
- n = 0;
- }
- }
-
- ret = &mr->ibmr;
-
-bail:
- return ret;
-}
-
-/**
* hfi1_reg_user_mr - register a userspace memory region
* @pd: protection domain for this memory region
* @start: starting userspace address
diff --git a/drivers/staging/rdma/hfi1/verbs.c b/drivers/staging/rdma/hfi1/verbs.c
index cb5b346..1b3d8d9 100644
--- a/drivers/staging/rdma/hfi1/verbs.c
+++ b/drivers/staging/rdma/hfi1/verbs.c
@@ -2011,7 +2011,6 @@ int hfi1_register_ib_device(struct hfi1_devdata *dd)
ibdev->poll_cq = hfi1_poll_cq;
ibdev->req_notify_cq = hfi1_req_notify_cq;
ibdev->get_dma_mr = hfi1_get_dma_mr;
- ibdev->reg_phys_mr = hfi1_reg_phys_mr;
ibdev->reg_user_mr = hfi1_reg_user_mr;
ibdev->dereg_mr = hfi1_dereg_mr;
ibdev->alloc_mr = hfi1_alloc_mr;
diff --git a/drivers/staging/rdma/hfi1/verbs.h b/drivers/staging/rdma/hfi1/verbs.h
index 041ad07..255792a 100644
--- a/drivers/staging/rdma/hfi1/verbs.h
+++ b/drivers/staging/rdma/hfi1/verbs.h
@@ -1012,10 +1012,6 @@ int hfi1_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata);
struct ib_mr *hfi1_get_dma_mr(struct ib_pd *pd, int acc);
-struct ib_mr *hfi1_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start);
-
struct ib_mr *hfi1_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt_addr, int mr_access_flags,
struct ib_udata *udata);
diff --git a/drivers/staging/rdma/ipath/ipath_mr.c b/drivers/staging/rdma/ipath/ipath_mr.c
index c7278f6..b76b0ce 100644
--- a/drivers/staging/rdma/ipath/ipath_mr.c
+++ b/drivers/staging/rdma/ipath/ipath_mr.c
@@ -98,10 +98,6 @@ static struct ipath_mr *alloc_mr(int count,
}
mr->mr.mapsz = m;
- /*
- * ib_reg_phys_mr() will initialize mr->ibmr except for
- * lkey and rkey.
- */
if (!ipath_alloc_lkey(lk_table, &mr->mr))
goto bail;
mr->ibmr.rkey = mr->ibmr.lkey = mr->mr.lkey;
@@ -121,57 +117,6 @@ done:
}
/**
- * ipath_reg_phys_mr - register a physical memory region
- * @pd: protection domain for this memory region
- * @buffer_list: pointer to the list of physical buffers to register
- * @num_phys_buf: the number of physical buffers to register
- * @iova_start: the starting address passed over IB which maps to this MR
- *
- * Returns the memory region on success, otherwise returns an errno.
- */
-struct ib_mr *ipath_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start)
-{
- struct ipath_mr *mr;
- int n, m, i;
- struct ib_mr *ret;
-
- mr = alloc_mr(num_phys_buf, &to_idev(pd->device)->lk_table);
- if (mr == NULL) {
- ret = ERR_PTR(-ENOMEM);
- goto bail;
- }
-
- mr->mr.pd = pd;
- mr->mr.user_base = *iova_start;
- mr->mr.iova = *iova_start;
- mr->mr.length = 0;
- mr->mr.offset = 0;
- mr->mr.access_flags = acc;
- mr->mr.max_segs = num_phys_buf;
- mr->umem = NULL;
-
- m = 0;
- n = 0;
- for (i = 0; i < num_phys_buf; i++) {
- mr->mr.map[m]->segs[n].vaddr = (void *) buffer_list[i].addr;
- mr->mr.map[m]->segs[n].length = buffer_list[i].size;
- mr->mr.length += buffer_list[i].size;
- n++;
- if (n == IPATH_SEGSZ) {
- m++;
- n = 0;
- }
- }
-
- ret = &mr->ibmr;
-
-bail:
- return ret;
-}
-
-/**
* ipath_reg_user_mr - register a userspace memory region
* @pd: protection domain for this memory region
* @start: starting userspace address
diff --git a/drivers/staging/rdma/ipath/ipath_verbs.c b/drivers/staging/rdma/ipath/ipath_verbs.c
index 7ab1520..02d8834 100644
--- a/drivers/staging/rdma/ipath/ipath_verbs.c
+++ b/drivers/staging/rdma/ipath/ipath_verbs.c
@@ -2149,7 +2149,6 @@ int ipath_register_ib_device(struct ipath_devdata *dd)
dev->poll_cq = ipath_poll_cq;
dev->req_notify_cq = ipath_req_notify_cq;
dev->get_dma_mr = ipath_get_dma_mr;
- dev->reg_phys_mr = ipath_reg_phys_mr;
dev->reg_user_mr = ipath_reg_user_mr;
dev->dereg_mr = ipath_dereg_mr;
dev->alloc_fmr = ipath_alloc_fmr;
diff --git a/drivers/staging/rdma/ipath/ipath_verbs.h b/drivers/staging/rdma/ipath/ipath_verbs.h
index 0a90a56..6c70a89 100644
--- a/drivers/staging/rdma/ipath/ipath_verbs.h
+++ b/drivers/staging/rdma/ipath/ipath_verbs.h
@@ -828,10 +828,6 @@ int ipath_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata);
struct ib_mr *ipath_get_dma_mr(struct ib_pd *pd, int acc);
-struct ib_mr *ipath_reg_phys_mr(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 *iova_start);
-
struct ib_mr *ipath_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt_addr, int mr_access_flags,
struct ib_udata *udata);
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 83d6ee8..a3402af 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -1165,6 +1165,10 @@ struct ib_phys_buf {
u64 size;
};
+/*
+ * XXX: these are apparently used for ->rereg_user_mr, no idea why they
+ * are hidden here instead of a uapi header!
+ */
enum ib_mr_rereg_flags {
IB_MR_REREG_TRANS = 1,
IB_MR_REREG_PD = (1<<1),
@@ -1683,11 +1687,6 @@ struct ib_device {
int wc_cnt);
struct ib_mr * (*get_dma_mr)(struct ib_pd *pd,
int mr_access_flags);
- struct ib_mr * (*reg_phys_mr)(struct ib_pd *pd,
- struct ib_phys_buf *phys_buf_array,
- int num_phys_buf,
- int mr_access_flags,
- u64 *iova_start);
struct ib_mr * (*reg_user_mr)(struct ib_pd *pd,
u64 start, u64 length,
u64 virt_addr,
@@ -1707,13 +1706,6 @@ struct ib_device {
int (*map_mr_sg)(struct ib_mr *mr,
struct scatterlist *sg,
int sg_nents);
- int (*rereg_phys_mr)(struct ib_mr *mr,
- int mr_rereg_mask,
- struct ib_pd *pd,
- struct ib_phys_buf *phys_buf_array,
- int num_phys_buf,
- int mr_access_flags,
- u64 *iova_start);
struct ib_mw * (*alloc_mw)(struct ib_pd *pd,
enum ib_mw_type type);
int (*bind_mw)(struct ib_qp *qp,
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 4/9] IB: remove in-kernel support for memory windows
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
` (2 preceding siblings ...)
2015-11-15 18:05 ` [PATCH 3/9] IB: remove support for phys MRs Christoph Hellwig
@ 2015-11-15 18:05 ` Christoph Hellwig
[not found] ` <1447610762-3781-5-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-15 18:05 ` [PATCH 5/9] cxgb3: simplify iwch_get_dma_wr Christoph Hellwig
` (5 subsequent siblings)
9 siblings, 1 reply; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:05 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
Remove the unused ib_allow_mw and ib_bind_mw functions, remove the
unused IB_WR_BIND_MW and IB_WC_BIND_MW opcodes and move ib_dealloc_mw
into the uverbs module.
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
Documentation/infiniband/core_locking.txt | 2 -
drivers/infiniband/core/uverbs.h | 2 +
drivers/infiniband/core/uverbs_cmd.c | 4 +-
drivers/infiniband/core/uverbs_main.c | 13 ++++-
drivers/infiniband/core/verbs.c | 36 -------------
drivers/infiniband/hw/cxgb3/iwch_cq.c | 4 --
drivers/infiniband/hw/cxgb3/iwch_provider.c | 1 -
drivers/infiniband/hw/cxgb3/iwch_provider.h | 3 --
drivers/infiniband/hw/cxgb3/iwch_qp.c | 82 -----------------------------
drivers/infiniband/hw/cxgb4/cq.c | 3 --
drivers/infiniband/hw/cxgb4/iw_cxgb4.h | 2 -
drivers/infiniband/hw/cxgb4/provider.c | 1 -
drivers/infiniband/hw/cxgb4/qp.c | 5 --
drivers/infiniband/hw/mlx4/cq.c | 3 --
drivers/infiniband/hw/mlx4/main.c | 1 -
drivers/infiniband/hw/mlx4/mlx4_ib.h | 2 -
drivers/infiniband/hw/mlx4/mr.c | 22 --------
drivers/infiniband/hw/mlx4/qp.c | 27 ----------
drivers/infiniband/hw/mlx5/cq.c | 3 --
drivers/infiniband/hw/mthca/mthca_cq.c | 3 --
drivers/infiniband/hw/nes/nes_verbs.c | 75 --------------------------
drivers/staging/rdma/amso1100/c2_cq.c | 3 --
drivers/staging/rdma/ehca/ehca_iverbs.h | 3 --
drivers/staging/rdma/ehca/ehca_main.c | 1 -
drivers/staging/rdma/ehca/ehca_mrmw.c | 12 -----
drivers/staging/rdma/ehca/ehca_reqs.c | 1 -
include/rdma/ib_verbs.h | 54 -------------------
27 files changed, 16 insertions(+), 352 deletions(-)
diff --git a/Documentation/infiniband/core_locking.txt b/Documentation/infiniband/core_locking.txt
index e167854..4b1f36b 100644
--- a/Documentation/infiniband/core_locking.txt
+++ b/Documentation/infiniband/core_locking.txt
@@ -15,7 +15,6 @@ Sleeping and interrupt context
modify_ah
query_ah
destroy_ah
- bind_mw
post_send
post_recv
poll_cq
@@ -31,7 +30,6 @@ Sleeping and interrupt context
ib_modify_ah
ib_query_ah
ib_destroy_ah
- ib_bind_mw
ib_post_send
ib_post_recv
ib_req_notify_cq
diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h
index 94bbd8c..612ccfd 100644
--- a/drivers/infiniband/core/uverbs.h
+++ b/drivers/infiniband/core/uverbs.h
@@ -204,6 +204,8 @@ void ib_uverbs_event_handler(struct ib_event_handler *handler,
struct ib_event *event);
void ib_uverbs_dealloc_xrcd(struct ib_uverbs_device *dev, struct ib_xrcd *xrcd);
+int uverbs_dealloc_mw(struct ib_mw *mw);
+
struct ib_uverbs_flow_spec {
union {
union {
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 29c1413..b0f7872 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -1240,7 +1240,7 @@ err_copy:
idr_remove_uobj(&ib_uverbs_mw_idr, uobj);
err_unalloc:
- ib_dealloc_mw(mw);
+ uverbs_dealloc_mw(mw);
err_put:
put_pd_read(pd);
@@ -1269,7 +1269,7 @@ ssize_t ib_uverbs_dealloc_mw(struct ib_uverbs_file *file,
mw = uobj->object;
- ret = ib_dealloc_mw(mw);
+ ret = uverbs_dealloc_mw(mw);
if (!ret)
uobj->live = 0;
diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c
index e3ef288..39680ae 100644
--- a/drivers/infiniband/core/uverbs_main.c
+++ b/drivers/infiniband/core/uverbs_main.c
@@ -133,6 +133,17 @@ static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file,
static void ib_uverbs_add_one(struct ib_device *device);
static void ib_uverbs_remove_one(struct ib_device *device, void *client_data);
+int uverbs_dealloc_mw(struct ib_mw *mw)
+{
+ struct ib_pd *pd = mw->pd;
+ int ret;
+
+ ret = mw->device->dealloc_mw(mw);
+ if (!ret)
+ atomic_dec(&pd->usecnt);
+ return ret;
+}
+
static void ib_uverbs_release_dev(struct kobject *kobj)
{
struct ib_uverbs_device *dev =
@@ -224,7 +235,7 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
struct ib_mw *mw = uobj->object;
idr_remove_uobj(&ib_uverbs_mw_idr, uobj);
- ib_dealloc_mw(mw);
+ uverbs_dealloc_mw(mw);
kfree(uobj);
}
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 0e21367..b31beb1 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1267,42 +1267,6 @@ struct ib_mr *ib_alloc_mr(struct ib_pd *pd,
}
EXPORT_SYMBOL(ib_alloc_mr);
-/* Memory windows */
-
-struct ib_mw *ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type)
-{
- struct ib_mw *mw;
-
- if (!pd->device->alloc_mw)
- return ERR_PTR(-ENOSYS);
-
- mw = pd->device->alloc_mw(pd, type);
- if (!IS_ERR(mw)) {
- mw->device = pd->device;
- mw->pd = pd;
- mw->uobject = NULL;
- mw->type = type;
- atomic_inc(&pd->usecnt);
- }
-
- return mw;
-}
-EXPORT_SYMBOL(ib_alloc_mw);
-
-int ib_dealloc_mw(struct ib_mw *mw)
-{
- struct ib_pd *pd;
- int ret;
-
- pd = mw->pd;
- ret = mw->device->dealloc_mw(mw);
- if (!ret)
- atomic_dec(&pd->usecnt);
-
- return ret;
-}
-EXPORT_SYMBOL(ib_dealloc_mw);
-
/* "Fast" memory regions */
struct ib_fmr *ib_alloc_fmr(struct ib_pd *pd,
diff --git a/drivers/infiniband/hw/cxgb3/iwch_cq.c b/drivers/infiniband/hw/cxgb3/iwch_cq.c
index cfe4049..97fbfd2 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_cq.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_cq.c
@@ -115,10 +115,6 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp,
case T3_SEND_WITH_SE_INV:
wc->opcode = IB_WC_SEND;
break;
- case T3_BIND_MW:
- wc->opcode = IB_WC_BIND_MW;
- break;
-
case T3_LOCAL_INV:
wc->opcode = IB_WC_LOCAL_INV;
break;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c
index 9576e15..b184933 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c
@@ -1350,7 +1350,6 @@ int iwch_register_device(struct iwch_dev *dev)
dev->ibdev.reg_user_mr = iwch_reg_user_mr;
dev->ibdev.dereg_mr = iwch_dereg_mr;
dev->ibdev.alloc_mw = iwch_alloc_mw;
- dev->ibdev.bind_mw = iwch_bind_mw;
dev->ibdev.dealloc_mw = iwch_dealloc_mw;
dev->ibdev.alloc_mr = iwch_alloc_mr;
dev->ibdev.map_mr_sg = iwch_map_mr_sg;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h
index f4fa6d6..f24df44 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.h
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h
@@ -330,9 +330,6 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
struct ib_send_wr **bad_wr);
int iwch_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
struct ib_recv_wr **bad_wr);
-int iwch_bind_mw(struct ib_qp *qp,
- struct ib_mw *mw,
- struct ib_mw_bind *mw_bind);
int iwch_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc);
int iwch_post_terminate(struct iwch_qp *qhp, struct respQ_msg_t *rsp_msg);
int iwch_post_zb_read(struct iwch_ep *ep);
diff --git a/drivers/infiniband/hw/cxgb3/iwch_qp.c b/drivers/infiniband/hw/cxgb3/iwch_qp.c
index d0548fc..d939980 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_qp.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_qp.c
@@ -526,88 +526,6 @@ out:
return err;
}
-int iwch_bind_mw(struct ib_qp *qp,
- struct ib_mw *mw,
- struct ib_mw_bind *mw_bind)
-{
- struct iwch_dev *rhp;
- struct iwch_mw *mhp;
- struct iwch_qp *qhp;
- union t3_wr *wqe;
- u32 pbl_addr;
- u8 page_size;
- u32 num_wrs;
- unsigned long flag;
- struct ib_sge sgl;
- int err=0;
- enum t3_wr_flags t3_wr_flags;
- u32 idx;
- struct t3_swsq *sqp;
-
- qhp = to_iwch_qp(qp);
- mhp = to_iwch_mw(mw);
- rhp = qhp->rhp;
-
- spin_lock_irqsave(&qhp->lock, flag);
- if (qhp->attr.state > IWCH_QP_STATE_RTS) {
- spin_unlock_irqrestore(&qhp->lock, flag);
- return -EINVAL;
- }
- num_wrs = Q_FREECNT(qhp->wq.sq_rptr, qhp->wq.sq_wptr,
- qhp->wq.sq_size_log2);
- if (num_wrs == 0) {
- spin_unlock_irqrestore(&qhp->lock, flag);
- return -ENOMEM;
- }
- idx = Q_PTR2IDX(qhp->wq.wptr, qhp->wq.size_log2);
- PDBG("%s: idx 0x%0x, mw 0x%p, mw_bind 0x%p\n", __func__, idx,
- mw, mw_bind);
- wqe = (union t3_wr *) (qhp->wq.queue + idx);
-
- t3_wr_flags = 0;
- if (mw_bind->send_flags & IB_SEND_SIGNALED)
- t3_wr_flags = T3_COMPLETION_FLAG;
-
- sgl.addr = mw_bind->bind_info.addr;
- sgl.lkey = mw_bind->bind_info.mr->lkey;
- sgl.length = mw_bind->bind_info.length;
- wqe->bind.reserved = 0;
- wqe->bind.type = TPT_VATO;
-
- /* TBD: check perms */
- wqe->bind.perms = iwch_ib_to_tpt_bind_access(
- mw_bind->bind_info.mw_access_flags);
- wqe->bind.mr_stag = cpu_to_be32(mw_bind->bind_info.mr->lkey);
- wqe->bind.mw_stag = cpu_to_be32(mw->rkey);
- wqe->bind.mw_len = cpu_to_be32(mw_bind->bind_info.length);
- wqe->bind.mw_va = cpu_to_be64(mw_bind->bind_info.addr);
- err = iwch_sgl2pbl_map(rhp, &sgl, 1, &pbl_addr, &page_size);
- if (err) {
- spin_unlock_irqrestore(&qhp->lock, flag);
- return err;
- }
- wqe->send.wrid.id0.hi = qhp->wq.sq_wptr;
- sqp = qhp->wq.sq + Q_PTR2IDX(qhp->wq.sq_wptr, qhp->wq.sq_size_log2);
- sqp->wr_id = mw_bind->wr_id;
- sqp->opcode = T3_BIND_MW;
- sqp->sq_wptr = qhp->wq.sq_wptr;
- sqp->complete = 0;
- sqp->signaled = (mw_bind->send_flags & IB_SEND_SIGNALED);
- wqe->bind.mr_pbl_addr = cpu_to_be32(pbl_addr);
- wqe->bind.mr_pagesz = page_size;
- build_fw_riwrh((void *)wqe, T3_WR_BIND, t3_wr_flags,
- Q_GENBIT(qhp->wq.wptr, qhp->wq.size_log2), 0,
- sizeof(struct t3_bind_mw_wr) >> 3, T3_SOPEOP);
- ++(qhp->wq.wptr);
- ++(qhp->wq.sq_wptr);
- spin_unlock_irqrestore(&qhp->lock, flag);
-
- if (cxio_wq_db_enabled(&qhp->wq))
- ring_doorbell(qhp->wq.doorbell, qhp->wq.qpid);
-
- return err;
-}
-
static inline void build_term_codes(struct respQ_msg_t *rsp_msg,
u8 *layer_type, u8 *ecode)
{
diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c
index de9cd69..cf21df4 100644
--- a/drivers/infiniband/hw/cxgb4/cq.c
+++ b/drivers/infiniband/hw/cxgb4/cq.c
@@ -744,9 +744,6 @@ static int c4iw_poll_cq_one(struct c4iw_cq *chp, struct ib_wc *wc)
case FW_RI_SEND_WITH_SE:
wc->opcode = IB_WC_SEND;
break;
- case FW_RI_BIND_MW:
- wc->opcode = IB_WC_BIND_MW;
- break;
case FW_RI_LOCAL_INV:
wc->opcode = IB_WC_LOCAL_INV;
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index dd00cf2..fb2de75 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -947,8 +947,6 @@ int c4iw_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
struct ib_send_wr **bad_wr);
int c4iw_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
struct ib_recv_wr **bad_wr);
-int c4iw_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
- struct ib_mw_bind *mw_bind);
int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param);
int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog);
int c4iw_destroy_listen(struct iw_cm_id *cm_id);
diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c
index 186319e..5544adc 100644
--- a/drivers/infiniband/hw/cxgb4/provider.c
+++ b/drivers/infiniband/hw/cxgb4/provider.c
@@ -512,7 +512,6 @@ int c4iw_register_device(struct c4iw_dev *dev)
dev->ibdev.reg_user_mr = c4iw_reg_user_mr;
dev->ibdev.dereg_mr = c4iw_dereg_mr;
dev->ibdev.alloc_mw = c4iw_alloc_mw;
- dev->ibdev.bind_mw = c4iw_bind_mw;
dev->ibdev.dealloc_mw = c4iw_dealloc_mw;
dev->ibdev.alloc_mr = c4iw_alloc_mr;
dev->ibdev.map_mr_sg = c4iw_map_mr_sg;
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c
index aa515af..e99345e 100644
--- a/drivers/infiniband/hw/cxgb4/qp.c
+++ b/drivers/infiniband/hw/cxgb4/qp.c
@@ -933,11 +933,6 @@ int c4iw_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
return err;
}
-int c4iw_bind_mw(struct ib_qp *qp, struct ib_mw *mw, struct ib_mw_bind *mw_bind)
-{
- return -ENOSYS;
-}
-
static inline void build_term_codes(struct t4_cqe *err_cqe, u8 *layer_type,
u8 *ecode)
{
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index b88fc8f..9f8b516 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -811,9 +811,6 @@ repoll:
wc->opcode = IB_WC_MASKED_FETCH_ADD;
wc->byte_len = 8;
break;
- case MLX4_OPCODE_BIND_MW:
- wc->opcode = IB_WC_BIND_MW;
- break;
case MLX4_OPCODE_LSO:
wc->opcode = IB_WC_LSO;
break;
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 7026580..4b1c40f 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2313,7 +2313,6 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
if (dev->caps.flags & MLX4_DEV_CAP_FLAG_MEM_WINDOW ||
dev->caps.bmme_flags & MLX4_BMME_FLAG_TYPE_2_WIN) {
ibdev->ib_dev.alloc_mw = mlx4_ib_alloc_mw;
- ibdev->ib_dev.bind_mw = mlx4_ib_bind_mw;
ibdev->ib_dev.dealloc_mw = mlx4_ib_dealloc_mw;
ibdev->ib_dev.uverbs_cmd_mask |=
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 1caa11e..8916e9b 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -704,8 +704,6 @@ struct ib_mr *mlx4_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
struct ib_udata *udata);
int mlx4_ib_dereg_mr(struct ib_mr *mr);
struct ib_mw *mlx4_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
-int mlx4_ib_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
- struct ib_mw_bind *mw_bind);
int mlx4_ib_dealloc_mw(struct ib_mw *mw);
struct ib_mr *mlx4_ib_alloc_mr(struct ib_pd *pd,
enum ib_mr_type mr_type,
diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c
index 4d1e1c6..242b94e 100644
--- a/drivers/infiniband/hw/mlx4/mr.c
+++ b/drivers/infiniband/hw/mlx4/mr.c
@@ -366,28 +366,6 @@ err_free:
return ERR_PTR(err);
}
-int mlx4_ib_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
- struct ib_mw_bind *mw_bind)
-{
- struct ib_bind_mw_wr wr;
- struct ib_send_wr *bad_wr;
- int ret;
-
- memset(&wr, 0, sizeof(wr));
- wr.wr.opcode = IB_WR_BIND_MW;
- wr.wr.wr_id = mw_bind->wr_id;
- wr.wr.send_flags = mw_bind->send_flags;
- wr.mw = mw;
- wr.bind_info = mw_bind->bind_info;
- wr.rkey = ib_inc_rkey(mw->rkey);
-
- ret = mlx4_ib_post_send(qp, &wr.wr, &bad_wr);
- if (!ret)
- mw->rkey = wr.rkey;
-
- return ret;
-}
-
int mlx4_ib_dealloc_mw(struct ib_mw *ibmw)
{
struct mlx4_ib_mw *mw = to_mmw(ibmw);
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index a2e4ca5..9781ac3 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -114,7 +114,6 @@ static const __be32 mlx4_ib_opcode[] = {
[IB_WR_REG_MR] = cpu_to_be32(MLX4_OPCODE_FMR),
[IB_WR_MASKED_ATOMIC_CMP_AND_SWP] = cpu_to_be32(MLX4_OPCODE_MASKED_ATOMIC_CS),
[IB_WR_MASKED_ATOMIC_FETCH_AND_ADD] = cpu_to_be32(MLX4_OPCODE_MASKED_ATOMIC_FA),
- [IB_WR_BIND_MW] = cpu_to_be32(MLX4_OPCODE_BIND_MW),
};
static struct mlx4_ib_sqp *to_msqp(struct mlx4_ib_qp *mqp)
@@ -2521,25 +2520,6 @@ static void set_reg_seg(struct mlx4_wqe_fmr_seg *fseg,
fseg->reserved[1] = 0;
}
-static void set_bind_seg(struct mlx4_wqe_bind_seg *bseg,
- struct ib_bind_mw_wr *wr)
-{
- bseg->flags1 =
- convert_access(wr->bind_info.mw_access_flags) &
- cpu_to_be32(MLX4_WQE_FMR_AND_BIND_PERM_REMOTE_READ |
- MLX4_WQE_FMR_AND_BIND_PERM_REMOTE_WRITE |
- MLX4_WQE_FMR_AND_BIND_PERM_ATOMIC);
- bseg->flags2 = 0;
- if (wr->mw->type == IB_MW_TYPE_2)
- bseg->flags2 |= cpu_to_be32(MLX4_WQE_BIND_TYPE_2);
- if (wr->bind_info.mw_access_flags & IB_ZERO_BASED)
- bseg->flags2 |= cpu_to_be32(MLX4_WQE_BIND_ZERO_BASED);
- bseg->new_rkey = cpu_to_be32(wr->rkey);
- bseg->lkey = cpu_to_be32(wr->bind_info.mr->lkey);
- bseg->addr = cpu_to_be64(wr->bind_info.addr);
- bseg->length = cpu_to_be64(wr->bind_info.length);
-}
-
static void set_local_inv_seg(struct mlx4_wqe_local_inval_seg *iseg, u32 rkey)
{
memset(iseg, 0, sizeof(*iseg));
@@ -2860,13 +2840,6 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
size += sizeof(struct mlx4_wqe_fmr_seg) / 16;
break;
- case IB_WR_BIND_MW:
- ctrl->srcrb_flags |=
- cpu_to_be32(MLX4_WQE_CTRL_STRONG_ORDER);
- set_bind_seg(wqe, bind_mw_wr(wr));
- wqe += sizeof(struct mlx4_wqe_bind_seg);
- size += sizeof(struct mlx4_wqe_bind_seg) / 16;
- break;
default:
/* No extra segments required for sends */
break;
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 3dfd287..82f86ed 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -154,9 +154,6 @@ static void handle_good_req(struct ib_wc *wc, struct mlx5_cqe64 *cqe,
wc->opcode = IB_WC_MASKED_FETCH_ADD;
wc->byte_len = 8;
break;
- case MLX5_OPCODE_BIND_MW:
- wc->opcode = IB_WC_BIND_MW;
- break;
case MLX5_OPCODE_UMR:
wc->opcode = get_umr_comp(wq, idx);
break;
diff --git a/drivers/infiniband/hw/mthca/mthca_cq.c b/drivers/infiniband/hw/mthca/mthca_cq.c
index 40ba833..a6531ff 100644
--- a/drivers/infiniband/hw/mthca/mthca_cq.c
+++ b/drivers/infiniband/hw/mthca/mthca_cq.c
@@ -608,9 +608,6 @@ static inline int mthca_poll_one(struct mthca_dev *dev,
entry->opcode = IB_WC_FETCH_ADD;
entry->byte_len = MTHCA_ATOMIC_BYTE_LEN;
break;
- case MTHCA_OPCODE_BIND_MW:
- entry->opcode = IB_WC_BIND_MW;
- break;
default:
entry->opcode = MTHCA_OPCODE_INVALID;
break;
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index 453ebc2..640f68f 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -206,80 +206,6 @@ static int nes_dealloc_mw(struct ib_mw *ibmw)
}
-/**
- * nes_bind_mw
- */
-static int nes_bind_mw(struct ib_qp *ibqp, struct ib_mw *ibmw,
- struct ib_mw_bind *ibmw_bind)
-{
- u64 u64temp;
- struct nes_vnic *nesvnic = to_nesvnic(ibqp->device);
- struct nes_device *nesdev = nesvnic->nesdev;
- /* struct nes_mr *nesmr = to_nesmw(ibmw); */
- struct nes_qp *nesqp = to_nesqp(ibqp);
- struct nes_hw_qp_wqe *wqe;
- unsigned long flags = 0;
- u32 head;
- u32 wqe_misc = 0;
- u32 qsize;
-
- if (nesqp->ibqp_state > IB_QPS_RTS)
- return -EINVAL;
-
- spin_lock_irqsave(&nesqp->lock, flags);
-
- head = nesqp->hwqp.sq_head;
- qsize = nesqp->hwqp.sq_tail;
-
- /* Check for SQ overflow */
- if (((head + (2 * qsize) - nesqp->hwqp.sq_tail) % qsize) == (qsize - 1)) {
- spin_unlock_irqrestore(&nesqp->lock, flags);
- return -ENOMEM;
- }
-
- wqe = &nesqp->hwqp.sq_vbase[head];
- /* nes_debug(NES_DBG_MR, "processing sq wqe at %p, head = %u.\n", wqe, head); */
- nes_fill_init_qp_wqe(wqe, nesqp, head);
- u64temp = ibmw_bind->wr_id;
- set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_COMP_SCRATCH_LOW_IDX, u64temp);
- wqe_misc = NES_IWARP_SQ_OP_BIND;
-
- wqe_misc |= NES_IWARP_SQ_WQE_LOCAL_FENCE;
-
- if (ibmw_bind->send_flags & IB_SEND_SIGNALED)
- wqe_misc |= NES_IWARP_SQ_WQE_SIGNALED_COMPL;
-
- if (ibmw_bind->bind_info.mw_access_flags & IB_ACCESS_REMOTE_WRITE)
- wqe_misc |= NES_CQP_STAG_RIGHTS_REMOTE_WRITE;
- if (ibmw_bind->bind_info.mw_access_flags & IB_ACCESS_REMOTE_READ)
- wqe_misc |= NES_CQP_STAG_RIGHTS_REMOTE_READ;
-
- set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_MISC_IDX, wqe_misc);
- set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_MR_IDX,
- ibmw_bind->bind_info.mr->lkey);
- set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_MW_IDX, ibmw->rkey);
- set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_LENGTH_LOW_IDX,
- ibmw_bind->bind_info.length);
- wqe->wqe_words[NES_IWARP_SQ_BIND_WQE_LENGTH_HIGH_IDX] = 0;
- u64temp = (u64)ibmw_bind->bind_info.addr;
- set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_VA_FBO_LOW_IDX, u64temp);
-
- head++;
- if (head >= qsize)
- head = 0;
-
- nesqp->hwqp.sq_head = head;
- barrier();
-
- nes_write32(nesdev->regs+NES_WQE_ALLOC,
- (1 << 24) | 0x00800000 | nesqp->hwqp.qp_id);
-
- spin_unlock_irqrestore(&nesqp->lock, flags);
-
- return 0;
-}
-
-
/*
* nes_alloc_fast_mr
*/
@@ -3836,7 +3762,6 @@ struct nes_ib_device *nes_init_ofa_device(struct net_device *netdev)
nesibdev->ibdev.dereg_mr = nes_dereg_mr;
nesibdev->ibdev.alloc_mw = nes_alloc_mw;
nesibdev->ibdev.dealloc_mw = nes_dealloc_mw;
- nesibdev->ibdev.bind_mw = nes_bind_mw;
nesibdev->ibdev.alloc_mr = nes_alloc_mr;
nesibdev->ibdev.map_mr_sg = nes_map_mr_sg;
diff --git a/drivers/staging/rdma/amso1100/c2_cq.c b/drivers/staging/rdma/amso1100/c2_cq.c
index 3ef881f..7ad0c08 100644
--- a/drivers/staging/rdma/amso1100/c2_cq.c
+++ b/drivers/staging/rdma/amso1100/c2_cq.c
@@ -173,9 +173,6 @@ static inline int c2_poll_one(struct c2_dev *c2dev,
case C2_WR_TYPE_RDMA_READ:
entry->opcode = IB_WC_RDMA_READ;
break;
- case C2_WR_TYPE_BIND_MW:
- entry->opcode = IB_WC_BIND_MW;
- break;
case C2_WR_TYPE_RECV:
entry->byte_len = be32_to_cpu(ce->bytes_rcvd);
entry->opcode = IB_WC_RECV;
diff --git a/drivers/staging/rdma/ehca/ehca_iverbs.h b/drivers/staging/rdma/ehca/ehca_iverbs.h
index 219f635..630416d 100644
--- a/drivers/staging/rdma/ehca/ehca_iverbs.h
+++ b/drivers/staging/rdma/ehca/ehca_iverbs.h
@@ -87,9 +87,6 @@ int ehca_dereg_mr(struct ib_mr *mr);
struct ib_mw *ehca_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
-int ehca_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
- struct ib_mw_bind *mw_bind);
-
int ehca_dealloc_mw(struct ib_mw *mw);
struct ib_fmr *ehca_alloc_fmr(struct ib_pd *pd,
diff --git a/drivers/staging/rdma/ehca/ehca_main.c b/drivers/staging/rdma/ehca/ehca_main.c
index ba023426..915e924 100644
--- a/drivers/staging/rdma/ehca/ehca_main.c
+++ b/drivers/staging/rdma/ehca/ehca_main.c
@@ -514,7 +514,6 @@ static int ehca_init_device(struct ehca_shca *shca)
shca->ib_device.reg_user_mr = ehca_reg_user_mr;
shca->ib_device.dereg_mr = ehca_dereg_mr;
shca->ib_device.alloc_mw = ehca_alloc_mw;
- shca->ib_device.bind_mw = ehca_bind_mw;
shca->ib_device.dealloc_mw = ehca_dealloc_mw;
shca->ib_device.alloc_fmr = ehca_alloc_fmr;
shca->ib_device.map_phys_fmr = ehca_map_phys_fmr;
diff --git a/drivers/staging/rdma/ehca/ehca_mrmw.c b/drivers/staging/rdma/ehca/ehca_mrmw.c
index 1c1a8dd..c6e3245 100644
--- a/drivers/staging/rdma/ehca/ehca_mrmw.c
+++ b/drivers/staging/rdma/ehca/ehca_mrmw.c
@@ -413,18 +413,6 @@ alloc_mw_exit0:
/*----------------------------------------------------------------------*/
-int ehca_bind_mw(struct ib_qp *qp,
- struct ib_mw *mw,
- struct ib_mw_bind *mw_bind)
-{
- /* TODO: not supported up to now */
- ehca_gen_err("bind MW currently not supported by HCAD");
-
- return -EPERM;
-} /* end ehca_bind_mw() */
-
-/*----------------------------------------------------------------------*/
-
int ehca_dealloc_mw(struct ib_mw *mw)
{
u64 h_ret;
diff --git a/drivers/staging/rdma/ehca/ehca_reqs.c b/drivers/staging/rdma/ehca/ehca_reqs.c
index 10e2074..11813b8 100644
--- a/drivers/staging/rdma/ehca/ehca_reqs.c
+++ b/drivers/staging/rdma/ehca/ehca_reqs.c
@@ -614,7 +614,6 @@ int ehca_post_srq_recv(struct ib_srq *srq,
static const u8 ib_wc_opcode[255] = {
[0x01] = IB_WC_RECV+1,
[0x02] = IB_WC_RECV_RDMA_WITH_IMM+1,
- [0x04] = IB_WC_BIND_MW+1,
[0x08] = IB_WC_FETCH_ADD+1,
[0x10] = IB_WC_COMP_SWAP+1,
[0x20] = IB_WC_RDMA_WRITE+1,
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index a3402af..0bbbd4b 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -704,7 +704,6 @@ enum ib_wc_opcode {
IB_WC_RDMA_READ,
IB_WC_COMP_SWAP,
IB_WC_FETCH_ADD,
- IB_WC_BIND_MW,
IB_WC_LSO,
IB_WC_LOCAL_INV,
IB_WC_REG_MR,
@@ -997,7 +996,6 @@ enum ib_wr_opcode {
IB_WR_REG_MR,
IB_WR_MASKED_ATOMIC_CMP_AND_SWP,
IB_WR_MASKED_ATOMIC_FETCH_AND_ADD,
- IB_WR_BIND_MW,
IB_WR_REG_SIG_MR,
/* reserve values for low level drivers' internal use.
* These values will not be used at all in the ib core layer.
@@ -1117,19 +1115,6 @@ static inline struct ib_reg_wr *reg_wr(struct ib_send_wr *wr)
return container_of(wr, struct ib_reg_wr, wr);
}
-struct ib_bind_mw_wr {
- struct ib_send_wr wr;
- struct ib_mw *mw;
- /* The new rkey for the memory window. */
- u32 rkey;
- struct ib_mw_bind_info bind_info;
-};
-
-static inline struct ib_bind_mw_wr *bind_mw_wr(struct ib_send_wr *wr)
-{
- return container_of(wr, struct ib_bind_mw_wr, wr);
-}
-
struct ib_sig_handover_wr {
struct ib_send_wr wr;
struct ib_sig_attrs *sig_attrs;
@@ -1708,9 +1693,6 @@ struct ib_device {
int sg_nents);
struct ib_mw * (*alloc_mw)(struct ib_pd *pd,
enum ib_mw_type type);
- int (*bind_mw)(struct ib_qp *qp,
- struct ib_mw *mw,
- struct ib_mw_bind *mw_bind);
int (*dealloc_mw)(struct ib_mw *mw);
struct ib_fmr * (*alloc_fmr)(struct ib_pd *pd,
int mr_access_flags,
@@ -2867,42 +2849,6 @@ static inline u32 ib_inc_rkey(u32 rkey)
}
/**
- * ib_alloc_mw - Allocates a memory window.
- * @pd: The protection domain associated with the memory window.
- * @type: The type of the memory window (1 or 2).
- */
-struct ib_mw *ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
-
-/**
- * ib_bind_mw - Posts a work request to the send queue of the specified
- * QP, which binds the memory window to the given address range and
- * remote access attributes.
- * @qp: QP to post the bind work request on.
- * @mw: The memory window to bind.
- * @mw_bind: Specifies information about the memory window, including
- * its address range, remote access rights, and associated memory region.
- *
- * If there is no immediate error, the function will update the rkey member
- * of the mw parameter to its new value. The bind operation can still fail
- * asynchronously.
- */
-static inline int ib_bind_mw(struct ib_qp *qp,
- struct ib_mw *mw,
- struct ib_mw_bind *mw_bind)
-{
- /* XXX reference counting in corresponding MR? */
- return mw->device->bind_mw ?
- mw->device->bind_mw(qp, mw, mw_bind) :
- -ENOSYS;
-}
-
-/**
- * ib_dealloc_mw - Deallocates a memory window.
- * @mw: The memory window to deallocate.
- */
-int ib_dealloc_mw(struct ib_mw *mw);
-
-/**
* ib_alloc_fmr - Allocates a unmapped fast memory region.
* @pd: The protection domain associated with the unmapped region.
* @mr_access_flags: Specifies the memory access rights.
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 5/9] cxgb3: simplify iwch_get_dma_wr
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
` (3 preceding siblings ...)
2015-11-15 18:05 ` [PATCH 4/9] IB: remove in-kernel support for memory windows Christoph Hellwig
@ 2015-11-15 18:05 ` Christoph Hellwig
[not found] ` <1447610762-3781-6-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-15 18:05 ` [PATCH 6/9] nes: simplify nes_reg_phys_mr calling conventions Christoph Hellwig
` (4 subsequent siblings)
9 siblings, 1 reply; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:05 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
Fold simplified versions of build_phys_page_list and
iwch_register_phys_mem into iwch_get_dma_wr now that no other callers
are left.
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
drivers/infiniband/hw/cxgb3/iwch_mem.c | 71 ---------------------------
drivers/infiniband/hw/cxgb3/iwch_provider.c | 75 ++++++++++++-----------------
drivers/infiniband/hw/cxgb3/iwch_provider.h | 8 ---
3 files changed, 30 insertions(+), 124 deletions(-)
diff --git a/drivers/infiniband/hw/cxgb3/iwch_mem.c b/drivers/infiniband/hw/cxgb3/iwch_mem.c
index 3a5e27d..1d04c87 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_mem.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_mem.c
@@ -99,74 +99,3 @@ int iwch_write_pbl(struct iwch_mr *mhp, __be64 *pages, int npages, int offset)
return cxio_write_pbl(&mhp->rhp->rdev, pages,
mhp->attr.pbl_addr + (offset << 3), npages);
}
-
-int build_phys_page_list(struct ib_phys_buf *buffer_list,
- int num_phys_buf,
- u64 *iova_start,
- u64 *total_size,
- int *npages,
- int *shift,
- __be64 **page_list)
-{
- u64 mask;
- int i, j, n;
-
- mask = 0;
- *total_size = 0;
- for (i = 0; i < num_phys_buf; ++i) {
- if (i != 0 && buffer_list[i].addr & ~PAGE_MASK)
- return -EINVAL;
- if (i != 0 && i != num_phys_buf - 1 &&
- (buffer_list[i].size & ~PAGE_MASK))
- return -EINVAL;
- *total_size += buffer_list[i].size;
- if (i > 0)
- mask |= buffer_list[i].addr;
- else
- mask |= buffer_list[i].addr & PAGE_MASK;
- if (i != num_phys_buf - 1)
- mask |= buffer_list[i].addr + buffer_list[i].size;
- else
- mask |= (buffer_list[i].addr + buffer_list[i].size +
- PAGE_SIZE - 1) & PAGE_MASK;
- }
-
- if (*total_size > 0xFFFFFFFFULL)
- return -ENOMEM;
-
- /* Find largest page shift we can use to cover buffers */
- for (*shift = PAGE_SHIFT; *shift < 27; ++(*shift))
- if ((1ULL << *shift) & mask)
- break;
-
- buffer_list[0].size += buffer_list[0].addr & ((1ULL << *shift) - 1);
- buffer_list[0].addr &= ~0ull << *shift;
-
- *npages = 0;
- for (i = 0; i < num_phys_buf; ++i)
- *npages += (buffer_list[i].size +
- (1ULL << *shift) - 1) >> *shift;
-
- if (!*npages)
- return -EINVAL;
-
- *page_list = kmalloc(sizeof(u64) * *npages, GFP_KERNEL);
- if (!*page_list)
- return -ENOMEM;
-
- n = 0;
- for (i = 0; i < num_phys_buf; ++i)
- for (j = 0;
- j < (buffer_list[i].size + (1ULL << *shift) - 1) >> *shift;
- ++j)
- (*page_list)[n++] = cpu_to_be64(buffer_list[i].addr +
- ((u64) j << *shift));
-
- PDBG("%s va 0x%llx mask 0x%llx shift %d len %lld pbl_size %d\n",
- __func__, (unsigned long long) *iova_start,
- (unsigned long long) mask, *shift, (unsigned long long) *total_size,
- *npages);
-
- return 0;
-
-}
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c
index b184933..9642ec4a 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c
@@ -479,24 +479,26 @@ static int iwch_dereg_mr(struct ib_mr *ib_mr)
return 0;
}
-static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf,
- int acc,
- u64 *iova_start)
+static struct ib_mr *iwch_get_dma_mr(struct ib_pd *pd, int acc)
{
+ const u64 total_size = 0xffffffff;
+ const u64 mask = (total_size + PAGE_SIZE - 1) & PAGE_MASK;
+ struct iwch_pd *php = to_iwch_pd(pd);
+ struct iwch_dev *rhp = php->rhp;
__be64 *page_list;
int shift;
- u64 total_size;
- int npages;
- struct iwch_dev *rhp;
- struct iwch_pd *php;
struct iwch_mr *mhp;
- int ret;
+ int ret, npages, i;
PDBG("%s ib_pd %p\n", __func__, pd);
- php = to_iwch_pd(pd);
- rhp = php->rhp;
+
+ /*
+ * T3 only supports 32 bits of size.
+ */
+ if (sizeof(phys_addr_t) > 4) {
+ pr_warn_once(MOD "Cannot support dma_mrs on this platform.\n");
+ return ERR_PTR(-ENOTSUPP);
+ }
mhp = kzalloc(sizeof(*mhp), GFP_KERNEL);
if (!mhp)
@@ -504,22 +506,28 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd,
mhp->rhp = rhp;
- /* First check that we have enough alignment */
- if ((*iova_start & ~PAGE_MASK) != (buffer_list[0].addr & ~PAGE_MASK)) {
+ /* Find largest page shift we can use to cover buffers */
+ for (shift = PAGE_SHIFT; shift < 27; ++(shift))
+ if ((1ULL << shift) & mask)
+ break;
+
+ npages = (total_size + (1ULL << shift) - 1) >> shift;
+ if (!npages) {
ret = -EINVAL;
goto err;
}
- if (num_phys_buf > 1 &&
- ((buffer_list[0].addr + buffer_list[0].size) & ~PAGE_MASK)) {
- ret = -EINVAL;
+ page_list = kmalloc_array(npages, sizeof(u64), GFP_KERNEL);
+ if (!page_list) {
+ ret = -ENOMEM;
goto err;
}
- ret = build_phys_page_list(buffer_list, num_phys_buf, iova_start,
- &total_size, &npages, &shift, &page_list);
- if (ret)
- goto err;
+ for (i = 0; i < npages; i++)
+ page_list[i] = cpu_to_be64((u64)i << shift);
+
+ PDBG("%s mask 0x%llx shift %d len %lld pbl_size %d\n",
+ __func__, mask, shift, total_size, npages);
ret = iwch_alloc_pbl(mhp, npages);
if (ret) {
@@ -536,7 +544,7 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd,
mhp->attr.zbva = 0;
mhp->attr.perms = iwch_ib_to_tpt_access(acc);
- mhp->attr.va_fbo = *iova_start;
+ mhp->attr.va_fbo = 0;
mhp->attr.page_size = shift - 12;
mhp->attr.len = (u32) total_size;
@@ -553,7 +561,6 @@ err_pbl:
err:
kfree(mhp);
return ERR_PTR(ret);
-
}
static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
@@ -659,28 +666,6 @@ err:
return ERR_PTR(err);
}
-static struct ib_mr *iwch_get_dma_mr(struct ib_pd *pd, int acc)
-{
- struct ib_phys_buf bl;
- u64 kva;
- struct ib_mr *ibmr;
-
- PDBG("%s ib_pd %p\n", __func__, pd);
-
- /*
- * T3 only supports 32 bits of size.
- */
- if (sizeof(phys_addr_t) > 4) {
- pr_warn_once(MOD "Cannot support dma_mrs on this platform.\n");
- return ERR_PTR(-ENOTSUPP);
- }
- bl.size = 0xffffffff;
- bl.addr = 0;
- kva = 0;
- ibmr = iwch_register_phys_mem(pd, &bl, 1, acc, &kva);
- return ibmr;
-}
-
static struct ib_mw *iwch_alloc_mw(struct ib_pd *pd, enum ib_mw_type type)
{
struct iwch_dev *rhp;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h
index f24df44..252c464 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.h
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h
@@ -341,14 +341,6 @@ int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php,
int iwch_alloc_pbl(struct iwch_mr *mhp, int npages);
void iwch_free_pbl(struct iwch_mr *mhp);
int iwch_write_pbl(struct iwch_mr *mhp, __be64 *pages, int npages, int offset);
-int build_phys_page_list(struct ib_phys_buf *buffer_list,
- int num_phys_buf,
- u64 *iova_start,
- u64 *total_size,
- int *npages,
- int *shift,
- __be64 **page_list);
-
#define IWCH_NODE_DESC "cxgb3 Chelsio Communications"
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 6/9] nes: simplify nes_reg_phys_mr calling conventions
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
` (4 preceding siblings ...)
2015-11-15 18:05 ` [PATCH 5/9] cxgb3: simplify iwch_get_dma_wr Christoph Hellwig
@ 2015-11-15 18:05 ` Christoph Hellwig
2015-11-15 18:06 ` [PATCH 7/9] amso1100: fold c2_reg_phys_mr into c2_get_dma_mr Christoph Hellwig
` (3 subsequent siblings)
9 siblings, 0 replies; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:05 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
Just pass and address/size pair instead of an ib_phys_buf array.
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
drivers/infiniband/hw/nes/nes_cm.c | 10 +--
drivers/infiniband/hw/nes/nes_verbs.c | 140 ++++++++--------------------------
drivers/infiniband/hw/nes/nes_verbs.h | 3 +-
3 files changed, 37 insertions(+), 116 deletions(-)
diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c
index 242c87d..bc37adb 100644
--- a/drivers/infiniband/hw/nes/nes_cm.c
+++ b/drivers/infiniband/hw/nes/nes_cm.c
@@ -3232,7 +3232,6 @@ int nes_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
int passive_state;
struct nes_ib_device *nesibdev;
struct ib_mr *ibmr = NULL;
- struct ib_phys_buf ibphysbuf;
struct nes_pd *nespd;
u64 tagged_offset;
u8 mpa_frame_offset = 0;
@@ -3316,12 +3315,11 @@ int nes_accept(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
u64temp = (unsigned long)nesqp;
nesibdev = nesvnic->nesibdev;
nespd = nesqp->nespd;
- ibphysbuf.addr = nesqp->ietf_frame_pbase + mpa_frame_offset;
- ibphysbuf.size = buff_len;
tagged_offset = (u64)(unsigned long)*start_buff;
- ibmr = nes_reg_phys_mr(&nespd->ibpd, &ibphysbuf, 1,
- IB_ACCESS_LOCAL_WRITE,
- &tagged_offset);
+ ibmr = nes_reg_phys_mr(&nespd->ibpd,
+ nesqp->ietf_frame_pbase + mpa_frame_offset,
+ buff_len, IB_ACCESS_LOCAL_WRITE,
+ &tagged_offset);
if (!ibmr) {
nes_debug(NES_DBG_CM, "Unable to register memory region"
"for lSMM for cm_node = %p \n",
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index 640f68f..4e57bf0 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -1945,9 +1945,8 @@ static int nes_reg_mr(struct nes_device *nesdev, struct nes_pd *nespd,
/**
* nes_reg_phys_mr
*/
-struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
- struct ib_phys_buf *buffer_list, int num_phys_buf, int acc,
- u64 * iova_start)
+struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd, u64 addr, u64 size,
+ int acc, u64 *iova_start)
{
u64 region_length;
struct nes_pd *nespd = to_nespd(ib_pd);
@@ -1959,13 +1958,10 @@ struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
struct nes_vpbl vpbl;
struct nes_root_vpbl root_vpbl;
u32 stag;
- u32 i;
unsigned long mask;
u32 stag_index = 0;
u32 next_stag_index = 0;
u32 driver_key = 0;
- u32 root_pbl_index = 0;
- u32 cur_pbl_index = 0;
int err = 0;
int ret = 0;
u16 pbl_count = 0;
@@ -1984,11 +1980,8 @@ struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
next_stag_index >>= 8;
next_stag_index %= nesadapter->max_mr;
- if (num_phys_buf > (1024*512)) {
- return ERR_PTR(-E2BIG);
- }
- if ((buffer_list[0].addr ^ *iova_start) & ~PAGE_MASK)
+ if ((addr ^ *iova_start) & ~PAGE_MASK)
return ERR_PTR(-EINVAL);
err = nes_alloc_resource(nesadapter, nesadapter->allocated_mrs, nesadapter->max_mr,
@@ -2003,84 +1996,33 @@ struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
return ERR_PTR(-ENOMEM);
}
- for (i = 0; i < num_phys_buf; i++) {
+ /* Allocate a 4K buffer for the PBL */
+ vpbl.pbl_vbase = pci_alloc_consistent(nesdev->pcidev, 4096,
+ &vpbl.pbl_pbase);
+ nes_debug(NES_DBG_MR, "Allocating leaf PBL, va = %p, pa = 0x%016lX\n",
+ vpbl.pbl_vbase, (unsigned long)vpbl.pbl_pbase);
+ if (!vpbl.pbl_vbase) {
+ nes_free_resource(nesadapter, nesadapter->allocated_mrs, stag_index);
+ ibmr = ERR_PTR(-ENOMEM);
+ kfree(nesmr);
+ goto reg_phys_err;
+ }
- if ((i & 0x01FF) == 0) {
- if (root_pbl_index == 1) {
- /* Allocate the root PBL */
- root_vpbl.pbl_vbase = pci_alloc_consistent(nesdev->pcidev, 8192,
- &root_vpbl.pbl_pbase);
- nes_debug(NES_DBG_MR, "Allocating root PBL, va = %p, pa = 0x%08X\n",
- root_vpbl.pbl_vbase, (unsigned int)root_vpbl.pbl_pbase);
- if (!root_vpbl.pbl_vbase) {
- pci_free_consistent(nesdev->pcidev, 4096, vpbl.pbl_vbase,
- vpbl.pbl_pbase);
- nes_free_resource(nesadapter, nesadapter->allocated_mrs, stag_index);
- kfree(nesmr);
- return ERR_PTR(-ENOMEM);
- }
- root_vpbl.leaf_vpbl = kzalloc(sizeof(*root_vpbl.leaf_vpbl)*1024, GFP_KERNEL);
- if (!root_vpbl.leaf_vpbl) {
- pci_free_consistent(nesdev->pcidev, 8192, root_vpbl.pbl_vbase,
- root_vpbl.pbl_pbase);
- pci_free_consistent(nesdev->pcidev, 4096, vpbl.pbl_vbase,
- vpbl.pbl_pbase);
- nes_free_resource(nesadapter, nesadapter->allocated_mrs, stag_index);
- kfree(nesmr);
- return ERR_PTR(-ENOMEM);
- }
- root_vpbl.pbl_vbase[0].pa_low = cpu_to_le32((u32)vpbl.pbl_pbase);
- root_vpbl.pbl_vbase[0].pa_high =
- cpu_to_le32((u32)((((u64)vpbl.pbl_pbase) >> 32)));
- root_vpbl.leaf_vpbl[0] = vpbl;
- }
- /* Allocate a 4K buffer for the PBL */
- vpbl.pbl_vbase = pci_alloc_consistent(nesdev->pcidev, 4096,
- &vpbl.pbl_pbase);
- nes_debug(NES_DBG_MR, "Allocating leaf PBL, va = %p, pa = 0x%016lX\n",
- vpbl.pbl_vbase, (unsigned long)vpbl.pbl_pbase);
- if (!vpbl.pbl_vbase) {
- nes_free_resource(nesadapter, nesadapter->allocated_mrs, stag_index);
- ibmr = ERR_PTR(-ENOMEM);
- kfree(nesmr);
- goto reg_phys_err;
- }
- /* Fill in the root table */
- if (1 <= root_pbl_index) {
- root_vpbl.pbl_vbase[root_pbl_index].pa_low =
- cpu_to_le32((u32)vpbl.pbl_pbase);
- root_vpbl.pbl_vbase[root_pbl_index].pa_high =
- cpu_to_le32((u32)((((u64)vpbl.pbl_pbase) >> 32)));
- root_vpbl.leaf_vpbl[root_pbl_index] = vpbl;
- }
- root_pbl_index++;
- cur_pbl_index = 0;
- }
- mask = !buffer_list[i].size;
- if (i != 0)
- mask |= buffer_list[i].addr;
- if (i != num_phys_buf - 1)
- mask |= buffer_list[i].addr + buffer_list[i].size;
-
- if (mask & ~PAGE_MASK) {
- nes_free_resource(nesadapter, nesadapter->allocated_mrs, stag_index);
- nes_debug(NES_DBG_MR, "Invalid buffer addr or size\n");
- ibmr = ERR_PTR(-EINVAL);
- kfree(nesmr);
- goto reg_phys_err;
- }
+ mask = !size;
- region_length += buffer_list[i].size;
- if ((i != 0) && (single_page)) {
- if ((buffer_list[i-1].addr+PAGE_SIZE) != buffer_list[i].addr)
- single_page = 0;
- }
- vpbl.pbl_vbase[cur_pbl_index].pa_low = cpu_to_le32((u32)buffer_list[i].addr & PAGE_MASK);
- vpbl.pbl_vbase[cur_pbl_index++].pa_high =
- cpu_to_le32((u32)((((u64)buffer_list[i].addr) >> 32)));
+ if (mask & ~PAGE_MASK) {
+ nes_free_resource(nesadapter, nesadapter->allocated_mrs, stag_index);
+ nes_debug(NES_DBG_MR, "Invalid buffer addr or size\n");
+ ibmr = ERR_PTR(-EINVAL);
+ kfree(nesmr);
+ goto reg_phys_err;
}
+ region_length += size;
+ vpbl.pbl_vbase[0].pa_low = cpu_to_le32((u32)addr & PAGE_MASK);
+ vpbl.pbl_vbase[0].pa_high = cpu_to_le32((u32)((((u64)addr) >> 32)));
+
stag = stag_index << 8;
stag |= driver_key;
stag += (u32)stag_key;
@@ -2090,17 +2032,15 @@ struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
stag, (unsigned long)*iova_start, (unsigned long)region_length, stag_index);
/* Make the leaf PBL the root if only one PBL */
- if (root_pbl_index == 1) {
- root_vpbl.pbl_pbase = vpbl.pbl_pbase;
- }
+ root_vpbl.pbl_pbase = vpbl.pbl_pbase;
if (single_page) {
pbl_count = 0;
} else {
- pbl_count = root_pbl_index;
+ pbl_count = 1;
}
ret = nes_reg_mr(nesdev, nespd, stag, region_length, &root_vpbl,
- buffer_list[0].addr, pbl_count, (u16)cur_pbl_index, acc, iova_start,
+ addr, pbl_count, 1, acc, iova_start,
&nesmr->pbls_used, &nesmr->pbl_4k);
if (ret == 0) {
@@ -2113,21 +2053,9 @@ struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
ibmr = ERR_PTR(-ENOMEM);
}
- reg_phys_err:
- /* free the resources */
- if (root_pbl_index == 1) {
- /* single PBL case */
- pci_free_consistent(nesdev->pcidev, 4096, vpbl.pbl_vbase, vpbl.pbl_pbase);
- } else {
- for (i=0; i<root_pbl_index; i++) {
- pci_free_consistent(nesdev->pcidev, 4096, root_vpbl.leaf_vpbl[i].pbl_vbase,
- root_vpbl.leaf_vpbl[i].pbl_pbase);
- }
- kfree(root_vpbl.leaf_vpbl);
- pci_free_consistent(nesdev->pcidev, 8192, root_vpbl.pbl_vbase,
- root_vpbl.pbl_pbase);
- }
-
+reg_phys_err:
+ /* single PBL case */
+ pci_free_consistent(nesdev->pcidev, 4096, vpbl.pbl_vbase, vpbl.pbl_pbase);
return ibmr;
}
@@ -2137,17 +2065,13 @@ struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
*/
static struct ib_mr *nes_get_dma_mr(struct ib_pd *pd, int acc)
{
- struct ib_phys_buf bl;
u64 kva = 0;
nes_debug(NES_DBG_MR, "\n");
- bl.size = (u64)0xffffffffffULL;
- bl.addr = 0;
- return nes_reg_phys_mr(pd, &bl, 1, acc, &kva);
+ return nes_reg_phys_mr(pd, 0, 0xffffffffffULL, acc, &kva);
}
-
/**
* nes_reg_user_mr
*/
diff --git a/drivers/infiniband/hw/nes/nes_verbs.h b/drivers/infiniband/hw/nes/nes_verbs.h
index 38e38cf..7029088 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.h
+++ b/drivers/infiniband/hw/nes/nes_verbs.h
@@ -192,7 +192,6 @@ struct nes_qp {
};
struct ib_mr *nes_reg_phys_mr(struct ib_pd *ib_pd,
- struct ib_phys_buf *buffer_list, int num_phys_buf, int acc,
- u64 * iova_start);
+ u64 addr, u64 size, int acc, u64 *iova_start);
#endif /* NES_VERBS_H */
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 7/9] amso1100: fold c2_reg_phys_mr into c2_get_dma_mr
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
` (5 preceding siblings ...)
2015-11-15 18:05 ` [PATCH 6/9] nes: simplify nes_reg_phys_mr calling conventions Christoph Hellwig
@ 2015-11-15 18:06 ` Christoph Hellwig
[not found] ` <1447610762-3781-8-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-15 18:06 ` [PATCH 8/9] ehca: stop using struct ib_phys_buf Christoph Hellwig
` (2 subsequent siblings)
9 siblings, 1 reply; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:06 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
drivers/staging/rdma/amso1100/c2_provider.c | 67 ++++++-----------------------
1 file changed, 12 insertions(+), 55 deletions(-)
diff --git a/drivers/staging/rdma/amso1100/c2_provider.c b/drivers/staging/rdma/amso1100/c2_provider.c
index 6c3e9cc..2de5a8f 100644
--- a/drivers/staging/rdma/amso1100/c2_provider.c
+++ b/drivers/staging/rdma/amso1100/c2_provider.c
@@ -323,43 +323,21 @@ static inline u32 c2_convert_access(int acc)
C2_ACF_LOCAL_READ | C2_ACF_WINDOW_BIND;
}
-static struct ib_mr *c2_reg_phys_mr(struct ib_pd *ib_pd,
- struct ib_phys_buf *buffer_list,
- int num_phys_buf, int acc, u64 * iova_start)
+static struct ib_mr *c2_get_dma_mr(struct ib_pd *pd, int acc)
{
struct c2_mr *mr;
u64 *page_list;
- u32 total_len;
- int err, i, j, k, page_shift, pbl_depth;
+ const u32 total_len = 0xffffffff; /* AMSO1100 limit */
+ int err, page_shift, pbl_depth, i;
+ u64 kva = 0;
- pbl_depth = 0;
- total_len = 0;
+ pr_debug("%s:%u\n", __func__, __LINE__);
- page_shift = PAGE_SHIFT;
/*
- * If there is only 1 buffer we assume this could
- * be a map of all phy mem...use a 32k page_shift.
+ * This is a map of all phy mem...use a 32k page_shift.
*/
- if (num_phys_buf == 1)
- page_shift += 3;
-
- for (i = 0; i < num_phys_buf; i++) {
-
- if (offset_in_page(buffer_list[i].addr)) {
- pr_debug("Unaligned Memory Buffer: 0x%x\n",
- (unsigned int) buffer_list[i].addr);
- return ERR_PTR(-EINVAL);
- }
-
- if (!buffer_list[i].size) {
- pr_debug("Invalid Buffer Size\n");
- return ERR_PTR(-EINVAL);
- }
-
- total_len += buffer_list[i].size;
- pbl_depth += ALIGN(buffer_list[i].size,
- BIT(page_shift)) >> page_shift;
- }
+ page_shift = PAGE_SHIFT + 3;
+ pbl_depth = ALIGN(total_len, BIT(page_shift)) >> page_shift;
page_list = vmalloc(sizeof(u64) * pbl_depth);
if (!page_list) {
@@ -368,16 +346,8 @@ static struct ib_mr *c2_reg_phys_mr(struct ib_pd *ib_pd,
return ERR_PTR(-ENOMEM);
}
- for (i = 0, j = 0; i < num_phys_buf; i++) {
-
- int naddrs;
-
- naddrs = ALIGN(buffer_list[i].size,
- BIT(page_shift)) >> page_shift;
- for (k = 0; k < naddrs; k++)
- page_list[j++] = (buffer_list[i].addr +
- (k << page_shift));
- }
+ for (i = 0; i < pbl_depth; i++)
+ page_list[i++] = (i << page_shift);
mr = kmalloc(sizeof(*mr), GFP_KERNEL);
if (!mr) {
@@ -390,12 +360,12 @@ static struct ib_mr *c2_reg_phys_mr(struct ib_pd *ib_pd,
pr_debug("%s - page shift %d, pbl_depth %d, total_len %u, "
"*iova_start %llx, first pa %llx, last pa %llx\n",
__func__, page_shift, pbl_depth, total_len,
- (unsigned long long) *iova_start,
+ (unsigned long long) kva,
(unsigned long long) page_list[0],
(unsigned long long) page_list[pbl_depth-1]);
err = c2_nsmr_register_phys_kern(to_c2dev(ib_pd->device), page_list,
BIT(page_shift), pbl_depth,
- total_len, 0, iova_start,
+ total_len, 0, &kva,
c2_convert_access(acc), mr);
vfree(page_list);
if (err) {
@@ -406,19 +376,6 @@ static struct ib_mr *c2_reg_phys_mr(struct ib_pd *ib_pd,
return &mr->ibmr;
}
-static struct ib_mr *c2_get_dma_mr(struct ib_pd *pd, int acc)
-{
- struct ib_phys_buf bl;
- u64 kva = 0;
-
- pr_debug("%s:%u\n", __func__, __LINE__);
-
- /* AMSO1100 limit */
- bl.size = 0xffffffff;
- bl.addr = 0;
- return c2_reg_phys_mr(pd, &bl, 1, acc, &kva);
-}
-
static struct ib_mr *c2_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
u64 virt, int acc, struct ib_udata *udata)
{
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 8/9] ehca: stop using struct ib_phys_buf
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
` (6 preceding siblings ...)
2015-11-15 18:06 ` [PATCH 7/9] amso1100: fold c2_reg_phys_mr into c2_get_dma_mr Christoph Hellwig
@ 2015-11-15 18:06 ` Christoph Hellwig
2015-11-15 18:06 ` [PATCH 9/9] IB: remove the struct ib_phys_buf definition Christoph Hellwig
2015-11-23 17:58 ` garbage collect old memory registration code Jason Gunthorpe
9 siblings, 0 replies; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:06 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
And simplify the calling convention for full-memory registrations.
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
drivers/staging/rdma/ehca/ehca_classes.h | 5 +-
drivers/staging/rdma/ehca/ehca_mrmw.c | 94 +++++++++++++++-----------------
2 files changed, 46 insertions(+), 53 deletions(-)
diff --git a/drivers/staging/rdma/ehca/ehca_classes.h b/drivers/staging/rdma/ehca/ehca_classes.h
index bd45e0f..e8c3387 100644
--- a/drivers/staging/rdma/ehca/ehca_classes.h
+++ b/drivers/staging/rdma/ehca/ehca_classes.h
@@ -316,9 +316,8 @@ struct ehca_mr_pginfo {
union {
struct { /* type EHCA_MR_PGI_PHYS section */
- int num_phys_buf;
- struct ib_phys_buf *phys_buf_array;
- u64 next_buf;
+ u64 addr;
+ u16 size;
} phy;
struct { /* type EHCA_MR_PGI_USER section */
struct ib_umem *region;
diff --git a/drivers/staging/rdma/ehca/ehca_mrmw.c b/drivers/staging/rdma/ehca/ehca_mrmw.c
index c6e3245..1814af7 100644
--- a/drivers/staging/rdma/ehca/ehca_mrmw.c
+++ b/drivers/staging/rdma/ehca/ehca_mrmw.c
@@ -1289,7 +1289,6 @@ int ehca_reg_internal_maxmr(
u64 *iova_start;
u64 size_maxmr;
struct ehca_mr_pginfo pginfo;
- struct ib_phys_buf ib_pbuf;
u32 num_kpages;
u32 num_hwpages;
u64 hw_pgsize;
@@ -1310,8 +1309,6 @@ int ehca_reg_internal_maxmr(
/* register internal max-MR on HCA */
size_maxmr = ehca_mr_len;
iova_start = (u64 *)ehca_map_vaddr((void *)(KERNELBASE + PHYSICAL_START));
- ib_pbuf.addr = 0;
- ib_pbuf.size = size_maxmr;
num_kpages = NUM_CHUNKS(((u64)iova_start % PAGE_SIZE) + size_maxmr,
PAGE_SIZE);
hw_pgsize = ehca_get_max_hwpage_size(shca);
@@ -1323,8 +1320,8 @@ int ehca_reg_internal_maxmr(
pginfo.num_kpages = num_kpages;
pginfo.num_hwpages = num_hwpages;
pginfo.hwpage_size = hw_pgsize;
- pginfo.u.phy.num_phys_buf = 1;
- pginfo.u.phy.phys_buf_array = &ib_pbuf;
+ pginfo.u.phy.addr = 0;
+ pginfo.u.phy.size = size_maxmr;
ret = ehca_reg_mr(shca, e_mr, iova_start, size_maxmr, 0, e_pd,
&pginfo, &e_mr->ib.ib_mr.lkey,
@@ -1620,57 +1617,54 @@ static int ehca_set_pagebuf_phys(struct ehca_mr_pginfo *pginfo,
u32 number, u64 *kpage)
{
int ret = 0;
- struct ib_phys_buf *pbuf;
+ u64 addr = pginfo->u.phy.addr;
+ u64 size = pginfo->u.phy.size;
u64 num_hw, offs_hw;
u32 i = 0;
- /* loop over desired phys_buf_array entries */
- while (i < number) {
- pbuf = pginfo->u.phy.phys_buf_array + pginfo->u.phy.next_buf;
- num_hw = NUM_CHUNKS((pbuf->addr % pginfo->hwpage_size) +
- pbuf->size, pginfo->hwpage_size);
- offs_hw = (pbuf->addr & ~(pginfo->hwpage_size - 1)) /
- pginfo->hwpage_size;
- while (pginfo->next_hwpage < offs_hw + num_hw) {
- /* sanity check */
- if ((pginfo->kpage_cnt >= pginfo->num_kpages) ||
- (pginfo->hwpage_cnt >= pginfo->num_hwpages)) {
- ehca_gen_err("kpage_cnt >= num_kpages, "
- "kpage_cnt=%llx num_kpages=%llx "
- "hwpage_cnt=%llx "
- "num_hwpages=%llx i=%x",
- pginfo->kpage_cnt,
- pginfo->num_kpages,
- pginfo->hwpage_cnt,
- pginfo->num_hwpages, i);
- return -EFAULT;
- }
- *kpage = (pbuf->addr & ~(pginfo->hwpage_size - 1)) +
- (pginfo->next_hwpage * pginfo->hwpage_size);
- if ( !(*kpage) && pbuf->addr ) {
- ehca_gen_err("pbuf->addr=%llx pbuf->size=%llx "
- "next_hwpage=%llx", pbuf->addr,
- pbuf->size, pginfo->next_hwpage);
- return -EFAULT;
- }
- (pginfo->hwpage_cnt)++;
- (pginfo->next_hwpage)++;
- if (PAGE_SIZE >= pginfo->hwpage_size) {
- if (pginfo->next_hwpage %
- (PAGE_SIZE / pginfo->hwpage_size) == 0)
- (pginfo->kpage_cnt)++;
- } else
- pginfo->kpage_cnt += pginfo->hwpage_size /
- PAGE_SIZE;
- kpage++;
- i++;
- if (i >= number) break;
+ num_hw = NUM_CHUNKS((addr % pginfo->hwpage_size) + size,
+ pginfo->hwpage_size);
+ offs_hw = (addr & ~(pginfo->hwpage_size - 1)) / pginfo->hwpage_size;
+
+ while (pginfo->next_hwpage < offs_hw + num_hw) {
+ /* sanity check */
+ if ((pginfo->kpage_cnt >= pginfo->num_kpages) ||
+ (pginfo->hwpage_cnt >= pginfo->num_hwpages)) {
+ ehca_gen_err("kpage_cnt >= num_kpages, "
+ "kpage_cnt=%llx num_kpages=%llx "
+ "hwpage_cnt=%llx "
+ "num_hwpages=%llx i=%x",
+ pginfo->kpage_cnt,
+ pginfo->num_kpages,
+ pginfo->hwpage_cnt,
+ pginfo->num_hwpages, i);
+ return -EFAULT;
}
- if (pginfo->next_hwpage >= offs_hw + num_hw) {
- (pginfo->u.phy.next_buf)++;
- pginfo->next_hwpage = 0;
+ *kpage = (addr & ~(pginfo->hwpage_size - 1)) +
+ (pginfo->next_hwpage * pginfo->hwpage_size);
+ if ( !(*kpage) && addr ) {
+ ehca_gen_err("addr=%llx size=%llx "
+ "next_hwpage=%llx", addr,
+ size, pginfo->next_hwpage);
+ return -EFAULT;
}
+ (pginfo->hwpage_cnt)++;
+ (pginfo->next_hwpage)++;
+ if (PAGE_SIZE >= pginfo->hwpage_size) {
+ if (pginfo->next_hwpage %
+ (PAGE_SIZE / pginfo->hwpage_size) == 0)
+ (pginfo->kpage_cnt)++;
+ } else
+ pginfo->kpage_cnt += pginfo->hwpage_size /
+ PAGE_SIZE;
+ kpage++;
+ i++;
+ if (i >= number) break;
+ }
+ if (pginfo->next_hwpage >= offs_hw + num_hw) {
+ pginfo->next_hwpage = 0;
}
+
return ret;
}
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* [PATCH 9/9] IB: remove the struct ib_phys_buf definition
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
` (7 preceding siblings ...)
2015-11-15 18:06 ` [PATCH 8/9] ehca: stop using struct ib_phys_buf Christoph Hellwig
@ 2015-11-15 18:06 ` Christoph Hellwig
2015-11-23 17:58 ` garbage collect old memory registration code Jason Gunthorpe
9 siblings, 0 replies; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-15 18:06 UTC (permalink / raw)
To: Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
---
include/rdma/ib_verbs.h | 5 -----
1 file changed, 5 deletions(-)
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 0bbbd4b..3253563 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -1145,11 +1145,6 @@ enum ib_access_flags {
IB_ACCESS_ON_DEMAND = (1<<6),
};
-struct ib_phys_buf {
- u64 addr;
- u64 size;
-};
-
/*
* XXX: these are apparently used for ->rereg_user_mr, no idea why they
* are hidden here instead of a uapi header!
--
1.9.1
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related [flat|nested] 18+ messages in thread
* Re: [PATCH 3/9] IB: remove support for phys MRs
[not found] ` <1447610762-3781-4-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
@ 2015-11-16 16:41 ` Steve Wise
0 siblings, 0 replies; 18+ messages in thread
From: Steve Wise @ 2015-11-16 16:41 UTC (permalink / raw)
To: Christoph Hellwig, Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
On 11/15/2015 12:05 PM, Christoph Hellwig wrote:
> We have stopped using phys MRs in the kernel a while ago, so let's
> remove all the cruft used to implement them.
>
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
>
For the cxgb3/4 parts:
Reviewed-by: Steve Wise <swise-7bPotxP6k4+P2YhJcF5u+vpXobYPEAuW@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH 5/9] cxgb3: simplify iwch_get_dma_wr
[not found] ` <1447610762-3781-6-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
@ 2015-11-16 16:52 ` Steve Wise
0 siblings, 0 replies; 18+ messages in thread
From: Steve Wise @ 2015-11-16 16:52 UTC (permalink / raw)
To: Christoph Hellwig, Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
On 11/15/2015 12:05 PM, Christoph Hellwig wrote:
> Fold simplified versions of build_phys_page_list and
> iwch_register_phys_mem into iwch_get_dma_wr now that no other callers
> are left.
>
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
> ---
> drivers/infiniband/hw/cxgb3/iwch_mem.c | 71 ---------------------------
> drivers/infiniband/hw/cxgb3/iwch_provider.c | 75 ++++++++++++-----------------
> drivers/infiniband/hw/cxgb3/iwch_provider.h | 8 ---
> 3 files changed, 30 insertions(+), 124 deletions(-)
>
> diff --git a/drivers/infiniband/hw/cxgb3/iwch_mem.c b/drivers/infiniband/hw/cxgb3/iwch_mem.c
> index 3a5e27d..1d04c87 100644
> --- a/drivers/infiniband/hw/cxgb3/iwch_mem.c
> +++ b/drivers/infiniband/hw/cxgb3/iwch_mem.c
> @@ -99,74 +99,3 @@ int iwch_write_pbl(struct iwch_mr *mhp, __be64 *pages, int npages, int offset)
> return cxio_write_pbl(&mhp->rhp->rdev, pages,
> mhp->attr.pbl_addr + (offset << 3), npages);
> }
> -
> -int build_phys_page_list(struct ib_phys_buf *buffer_list,
> - int num_phys_buf,
> - u64 *iova_start,
> - u64 *total_size,
> - int *npages,
> - int *shift,
> - __be64 **page_list)
> -{
> - u64 mask;
> - int i, j, n;
> -
> - mask = 0;
> - *total_size = 0;
> - for (i = 0; i < num_phys_buf; ++i) {
> - if (i != 0 && buffer_list[i].addr & ~PAGE_MASK)
> - return -EINVAL;
> - if (i != 0 && i != num_phys_buf - 1 &&
> - (buffer_list[i].size & ~PAGE_MASK))
> - return -EINVAL;
> - *total_size += buffer_list[i].size;
> - if (i > 0)
> - mask |= buffer_list[i].addr;
> - else
> - mask |= buffer_list[i].addr & PAGE_MASK;
> - if (i != num_phys_buf - 1)
> - mask |= buffer_list[i].addr + buffer_list[i].size;
> - else
> - mask |= (buffer_list[i].addr + buffer_list[i].size +
> - PAGE_SIZE - 1) & PAGE_MASK;
> - }
> -
> - if (*total_size > 0xFFFFFFFFULL)
> - return -ENOMEM;
> -
> - /* Find largest page shift we can use to cover buffers */
> - for (*shift = PAGE_SHIFT; *shift < 27; ++(*shift))
> - if ((1ULL << *shift) & mask)
> - break;
> -
> - buffer_list[0].size += buffer_list[0].addr & ((1ULL << *shift) - 1);
> - buffer_list[0].addr &= ~0ull << *shift;
> -
> - *npages = 0;
> - for (i = 0; i < num_phys_buf; ++i)
> - *npages += (buffer_list[i].size +
> - (1ULL << *shift) - 1) >> *shift;
> -
> - if (!*npages)
> - return -EINVAL;
> -
> - *page_list = kmalloc(sizeof(u64) * *npages, GFP_KERNEL);
> - if (!*page_list)
> - return -ENOMEM;
> -
> - n = 0;
> - for (i = 0; i < num_phys_buf; ++i)
> - for (j = 0;
> - j < (buffer_list[i].size + (1ULL << *shift) - 1) >> *shift;
> - ++j)
> - (*page_list)[n++] = cpu_to_be64(buffer_list[i].addr +
> - ((u64) j << *shift));
> -
> - PDBG("%s va 0x%llx mask 0x%llx shift %d len %lld pbl_size %d\n",
> - __func__, (unsigned long long) *iova_start,
> - (unsigned long long) mask, *shift, (unsigned long long) *total_size,
> - *npages);
> -
> - return 0;
> -
> -}
> diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c
> index b184933..9642ec4a 100644
> --- a/drivers/infiniband/hw/cxgb3/iwch_provider.c
> +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c
> @@ -479,24 +479,26 @@ static int iwch_dereg_mr(struct ib_mr *ib_mr)
> return 0;
> }
>
> -static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd,
> - struct ib_phys_buf *buffer_list,
> - int num_phys_buf,
> - int acc,
> - u64 *iova_start)
> +static struct ib_mr *iwch_get_dma_mr(struct ib_pd *pd, int acc)
> {
> + const u64 total_size = 0xffffffff;
> + const u64 mask = (total_size + PAGE_SIZE - 1) & PAGE_MASK;
> + struct iwch_pd *php = to_iwch_pd(pd);
> + struct iwch_dev *rhp = php->rhp;
> __be64 *page_list;
> int shift;
> - u64 total_size;
> - int npages;
> - struct iwch_dev *rhp;
> - struct iwch_pd *php;
> struct iwch_mr *mhp;
> - int ret;
> + int ret, npages, i;
>
> PDBG("%s ib_pd %p\n", __func__, pd);
> - php = to_iwch_pd(pd);
> - rhp = php->rhp;
> +
> + /*
> + * T3 only supports 32 bits of size.
> + */
> + if (sizeof(phys_addr_t) > 4) {
> + pr_warn_once(MOD "Cannot support dma_mrs on this platform.\n");
> + return ERR_PTR(-ENOTSUPP);
> + }
>
> mhp = kzalloc(sizeof(*mhp), GFP_KERNEL);
> if (!mhp)
> @@ -504,22 +506,28 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd,
>
> mhp->rhp = rhp;
>
> - /* First check that we have enough alignment */
> - if ((*iova_start & ~PAGE_MASK) != (buffer_list[0].addr & ~PAGE_MASK)) {
> + /* Find largest page shift we can use to cover buffers */
> + for (shift = PAGE_SHIFT; shift < 27; ++(shift))
> + if ((1ULL << shift) & mask)
> + break;
> +
Won't shift always be 26 here?
> + npages = (total_size + (1ULL << shift) - 1) >> shift;
> + if (!npages) {
> ret = -EINVAL;
> goto err;
> }
>
> - if (num_phys_buf > 1 &&
> - ((buffer_list[0].addr + buffer_list[0].size) & ~PAGE_MASK)) {
> - ret = -EINVAL;
> + page_list = kmalloc_array(npages, sizeof(u64), GFP_KERNEL);
> + if (!page_list) {
> + ret = -ENOMEM;
> goto err;
> }
>
> - ret = build_phys_page_list(buffer_list, num_phys_buf, iova_start,
> - &total_size, &npages, &shift, &page_list);
> - if (ret)
> - goto err;
> + for (i = 0; i < npages; i++)
> + page_list[i] = cpu_to_be64((u64)i << shift);
> +
> + PDBG("%s mask 0x%llx shift %d len %lld pbl_size %d\n",
> + __func__, mask, shift, total_size, npages);
>
> ret = iwch_alloc_pbl(mhp, npages);
> if (ret) {
> @@ -536,7 +544,7 @@ static struct ib_mr *iwch_register_phys_mem(struct ib_pd *pd,
> mhp->attr.zbva = 0;
>
> mhp->attr.perms = iwch_ib_to_tpt_access(acc);
> - mhp->attr.va_fbo = *iova_start;
> + mhp->attr.va_fbo = 0;
> mhp->attr.page_size = shift - 12;
>
> mhp->attr.len = (u32) total_size;
> @@ -553,7 +561,6 @@ err_pbl:
> err:
> kfree(mhp);
> return ERR_PTR(ret);
> -
> }
>
> static struct ib_mr *iwch_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
> @@ -659,28 +666,6 @@ err:
> return ERR_PTR(err);
> }
>
> -static struct ib_mr *iwch_get_dma_mr(struct ib_pd *pd, int acc)
> -{
> - struct ib_phys_buf bl;
> - u64 kva;
> - struct ib_mr *ibmr;
> -
> - PDBG("%s ib_pd %p\n", __func__, pd);
> -
> - /*
> - * T3 only supports 32 bits of size.
> - */
> - if (sizeof(phys_addr_t) > 4) {
> - pr_warn_once(MOD "Cannot support dma_mrs on this platform.\n");
> - return ERR_PTR(-ENOTSUPP);
> - }
> - bl.size = 0xffffffff;
> - bl.addr = 0;
> - kva = 0;
> - ibmr = iwch_register_phys_mem(pd, &bl, 1, acc, &kva);
> - return ibmr;
> -}
> -
> static struct ib_mw *iwch_alloc_mw(struct ib_pd *pd, enum ib_mw_type type)
> {
> struct iwch_dev *rhp;
> diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h
> index f24df44..252c464 100644
> --- a/drivers/infiniband/hw/cxgb3/iwch_provider.h
> +++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h
> @@ -341,14 +341,6 @@ int iwch_register_mem(struct iwch_dev *rhp, struct iwch_pd *php,
> int iwch_alloc_pbl(struct iwch_mr *mhp, int npages);
> void iwch_free_pbl(struct iwch_mr *mhp);
> int iwch_write_pbl(struct iwch_mr *mhp, __be64 *pages, int npages, int offset);
> -int build_phys_page_list(struct ib_phys_buf *buffer_list,
> - int num_phys_buf,
> - u64 *iova_start,
> - u64 *total_size,
> - int *npages,
> - int *shift,
> - __be64 **page_list);
> -
>
> #define IWCH_NODE_DESC "cxgb3 Chelsio Communications"
>
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH 7/9] amso1100: fold c2_reg_phys_mr into c2_get_dma_mr
[not found] ` <1447610762-3781-8-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
@ 2015-11-16 16:54 ` Steve Wise
0 siblings, 0 replies; 18+ messages in thread
From: Steve Wise @ 2015-11-16 16:54 UTC (permalink / raw)
To: Christoph Hellwig, Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
I think Doug is removing amso1100 now so this patch isn't needed.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH 4/9] IB: remove in-kernel support for memory windows
[not found] ` <1447610762-3781-5-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
@ 2015-11-16 17:00 ` Sagi Grimberg
[not found] ` <564A0B96.5010703-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
0 siblings, 1 reply; 18+ messages in thread
From: Sagi Grimberg @ 2015-11-16 17:00 UTC (permalink / raw)
To: Christoph Hellwig, Doug Ledford; +Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA
> Remove the unused ib_allow_mw and ib_bind_mw functions, remove the
> unused IB_WR_BIND_MW and IB_WC_BIND_MW opcodes and move ib_dealloc_mw
> into the uverbs module.
>
> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
Will the user-space drivers posting via uverbs (qib, hfi, rxe) need the
post_send interface?
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH 4/9] IB: remove in-kernel support for memory windows
[not found] ` <564A0B96.5010703-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
@ 2015-11-16 17:02 ` Christoph Hellwig
[not found] ` <20151116170258.GA32300-jcswGhMUV9g@public.gmane.org>
0 siblings, 1 reply; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-16 17:02 UTC (permalink / raw)
To: Sagi Grimberg; +Cc: Doug Ledford, linux-rdma-u79uwXL29TY76Z2rM5mHXA
On Mon, Nov 16, 2015 at 07:00:06PM +0200, Sagi Grimberg wrote:
>
>> Remove the unused ib_allow_mw and ib_bind_mw functions, remove the
>> unused IB_WR_BIND_MW and IB_WC_BIND_MW opcodes and move ib_dealloc_mw
>> into the uverbs module.
>>
>> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
>
> Will the user-space drivers posting via uverbs (qib, hfi, rxe) need the
> post_send interface?
BIND_MW hasn't ever been supported through post_send as far as I can
tell, and started explicitly rejecting it a while ago.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH 4/9] IB: remove in-kernel support for memory windows
[not found] ` <20151116170258.GA32300-jcswGhMUV9g@public.gmane.org>
@ 2015-11-16 18:33 ` Sagi Grimberg
[not found] ` <564A2187.3060000-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
0 siblings, 1 reply; 18+ messages in thread
From: Sagi Grimberg @ 2015-11-16 18:33 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Doug Ledford, linux-rdma-u79uwXL29TY76Z2rM5mHXA
On 16/11/2015 19:02, Christoph Hellwig wrote:
> On Mon, Nov 16, 2015 at 07:00:06PM +0200, Sagi Grimberg wrote:
>>
>>> Remove the unused ib_allow_mw and ib_bind_mw functions, remove the
>>> unused IB_WR_BIND_MW and IB_WC_BIND_MW opcodes and move ib_dealloc_mw
>>> into the uverbs module.
>>>
>>> Signed-off-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
>>
>> Will the user-space drivers posting via uverbs (qib, hfi, rxe) need the
>> post_send interface?
>
> BIND_MW hasn't ever been supported through post_send as far as I can
> tell, and started explicitly rejecting it a while ago.
I think there are users for bind_mw in user-space, and I think there is
code that should go upstream soon. The question is weather qib/hfi/rxe
will need it anytime soon (otherwise we can restore them later on as
uverbs support does not exist anyway...)
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: [PATCH 4/9] IB: remove in-kernel support for memory windows
[not found] ` <564A2187.3060000-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
@ 2015-11-16 18:41 ` Christoph Hellwig
0 siblings, 0 replies; 18+ messages in thread
From: Christoph Hellwig @ 2015-11-16 18:41 UTC (permalink / raw)
To: Sagi Grimberg; +Cc: Doug Ledford, linux-rdma-u79uwXL29TY76Z2rM5mHXA
Before we expose any new opcodes to userspace we need to actually
move the defintions of user visible opcodes to a uapi header. Without
that we're doomed to regularly break the ABI.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
* Re: garbage collect old memory registration code
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
` (8 preceding siblings ...)
2015-11-15 18:06 ` [PATCH 9/9] IB: remove the struct ib_phys_buf definition Christoph Hellwig
@ 2015-11-23 17:58 ` Jason Gunthorpe
9 siblings, 0 replies; 18+ messages in thread
From: Jason Gunthorpe @ 2015-11-23 17:58 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: Doug Ledford, linux-rdma-u79uwXL29TY76Z2rM5mHXA
On Sun, Nov 15, 2015 at 07:05:53PM +0100, Christoph Hellwig wrote:
> This series removes huge chunks of code related to old memory
> registration methods that we don't use anymore.
>
> This expects my "IB: merge struct ib_device_attr into struct ib_device"
> patch to be already applied.
>
> Also available as a git tree:
>
> http://git.infradead.org/users/hch/rdma.git/shortlog/refs/heads/rdma-mr-cleanups
> git://git.infradead.org/users/hch/rdma.git rdma-mr-cleanups
Only looked carefully at the core changes here, but they look good to
me. Thanks for doing this.
Reviewed-by: Jason Gunthorpe <jgunthorpe-ePGOBjL8dl3ta4EC/59zMFaTQe2KTcn/@public.gmane.org>
(for core bits)
Driver changes look pretty much trivial too me.
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 18+ messages in thread
end of thread, other threads:[~2015-11-23 17:58 UTC | newest]
Thread overview: 18+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-11-15 18:05 garbage collect old memory registration code Christoph Hellwig
[not found] ` <1447610762-3781-1-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-15 18:05 ` [PATCH 1/9] IB: start documenting device capabilities Christoph Hellwig
2015-11-15 18:05 ` [PATCH 2/9] IB: remove ib_query_mr Christoph Hellwig
2015-11-15 18:05 ` [PATCH 3/9] IB: remove support for phys MRs Christoph Hellwig
[not found] ` <1447610762-3781-4-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-16 16:41 ` Steve Wise
2015-11-15 18:05 ` [PATCH 4/9] IB: remove in-kernel support for memory windows Christoph Hellwig
[not found] ` <1447610762-3781-5-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-16 17:00 ` Sagi Grimberg
[not found] ` <564A0B96.5010703-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
2015-11-16 17:02 ` Christoph Hellwig
[not found] ` <20151116170258.GA32300-jcswGhMUV9g@public.gmane.org>
2015-11-16 18:33 ` Sagi Grimberg
[not found] ` <564A2187.3060000-LDSdmyG8hGV8YrgS2mwiifqBs+8SCbDb@public.gmane.org>
2015-11-16 18:41 ` Christoph Hellwig
2015-11-15 18:05 ` [PATCH 5/9] cxgb3: simplify iwch_get_dma_wr Christoph Hellwig
[not found] ` <1447610762-3781-6-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-16 16:52 ` Steve Wise
2015-11-15 18:05 ` [PATCH 6/9] nes: simplify nes_reg_phys_mr calling conventions Christoph Hellwig
2015-11-15 18:06 ` [PATCH 7/9] amso1100: fold c2_reg_phys_mr into c2_get_dma_mr Christoph Hellwig
[not found] ` <1447610762-3781-8-git-send-email-hch-jcswGhMUV9g@public.gmane.org>
2015-11-16 16:54 ` Steve Wise
2015-11-15 18:06 ` [PATCH 8/9] ehca: stop using struct ib_phys_buf Christoph Hellwig
2015-11-15 18:06 ` [PATCH 9/9] IB: remove the struct ib_phys_buf definition Christoph Hellwig
2015-11-23 17:58 ` garbage collect old memory registration code Jason Gunthorpe
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.