diff mbox series

[2/2] migration/rdma: advise prefetch write for ODP region

Message ID 20210731140332.8701-3-lizhijian@cn.fujitsu.com (mailing list archive)
State New, archived
Headers show
Series enable fsdax rdma migration | expand

Commit Message

Li, Zhijian July 31, 2021, 2:03 p.m. UTC
The responder mr registering with ODP will sent RNR NAK back to
the requester in the face of the page fault.
---------
ibv_poll_cq wc.status=13 RNR retry counter exceeded!
ibv_poll_cq wrid=WRITE RDMA!
---------
ibv_advise_mr(3) helps to make pages present before the actual IO is
conducted so that the responder does page fault as little as possible.

Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
---
 migration/rdma.c       | 40 ++++++++++++++++++++++++++++++++++++++++
 migration/trace-events |  1 +
 2 files changed, 41 insertions(+)

Comments

Marcel Apfelbaum Aug. 22, 2021, 8:39 a.m. UTC | #1
Hi,

On Sat, Jul 31, 2021 at 5:03 PM Li Zhijian <lizhijian@cn.fujitsu.com> wrote:
>
> The responder mr registering with ODP will sent RNR NAK back to
> the requester in the face of the page fault.
> ---------
> ibv_poll_cq wc.status=13 RNR retry counter exceeded!
> ibv_poll_cq wrid=WRITE RDMA!
> ---------
> ibv_advise_mr(3) helps to make pages present before the actual IO is
> conducted so that the responder does page fault as little as possible.
>
> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
> ---
>  migration/rdma.c       | 40 ++++++++++++++++++++++++++++++++++++++++
>  migration/trace-events |  1 +
>  2 files changed, 41 insertions(+)
>
> diff --git a/migration/rdma.c b/migration/rdma.c
> index 8784b5f22a6..a2ad00d665f 100644
> --- a/migration/rdma.c
> +++ b/migration/rdma.c
> @@ -1117,6 +1117,30 @@ static int qemu_rdma_alloc_qp(RDMAContext *rdma)
>      return 0;
>  }
>
> +/*
> + * ibv_advise_mr to avoid RNR NAK error as far as possible.
> + * The responder mr registering with ODP will sent RNR NAK back to
> + * the requester in the face of the page fault.
> + */
> +static void qemu_rdma_advise_prefetch_write_mr(struct ibv_pd *pd, uint64_t addr,
> +                                               uint32_t len,  uint32_t lkey,
> +                                               const char *name, bool wr)
> +{
> +    int ret;
> +    int advice = wr ? IBV_ADVISE_MR_ADVICE_PREFETCH_WRITE :
> +                 IBV_ADVISE_MR_ADVICE_PREFETCH;
> +    struct ibv_sge sg_list = {.lkey = lkey, .addr = addr, .length = len};
> +
> +    ret = ibv_advise_mr(pd, advice,
> +                        IB_UVERBS_ADVISE_MR_FLAG_FLUSH, &sg_list, 1);
> +    /* ignore the error */
> +    if (ret) {
> +        trace_qemu_rdma_advise_mr(name, len, addr, strerror(errno));
> +    } else {
> +        trace_qemu_rdma_advise_mr(name, len, addr, "successed");
> +    }
> +}
> +
>  static int qemu_rdma_reg_whole_ram_blocks(RDMAContext *rdma)
>  {
>      int i;
> @@ -1140,6 +1164,17 @@ on_demand:
>              perror("Failed to register local dest ram block!\n");
>              break;
>          }
> +
> +        if (access & IBV_ACCESS_ON_DEMAND) {
> +            qemu_rdma_advise_prefetch_write_mr(rdma->pd,
> +                                               (uintptr_t)
> +                                               local->block[i].local_host_addr,
> +                                               local->block[i].length,
> +                                               local->block[i].mr->lkey,
> +                                               local->block[i].block_name,
> +                                               true);
> +        }
> +
>          rdma->total_registrations++;
>      }
>
> @@ -1244,6 +1279,11 @@ on_demand:
>                              rdma->total_registrations);
>              return -1;
>          }
> +        if (access & IBV_ACCESS_ON_DEMAND) {
> +            qemu_rdma_advise_prefetch_write_mr(rdma->pd, (uintptr_t)chunk_start,
> +                                               len, block->pmr[chunk]->lkey,
> +                                               block->block_name, rkey);
> +        }
>          rdma->total_registrations++;
>      }
>
> diff --git a/migration/trace-events b/migration/trace-events
> index 5f6aa580def..901c1d54c12 100644
> --- a/migration/trace-events
> +++ b/migration/trace-events
> @@ -213,6 +213,7 @@ qemu_rdma_poll_other(const char *compstr, int64_t comp, int left) "other complet
>  qemu_rdma_post_send_control(const char *desc) "CONTROL: sending %s.."
>  qemu_rdma_register_and_get_keys(uint64_t len, void *start) "Registering %" PRIu64 " bytes @ %p"
>  qemu_rdma_register_odp_mr(const char *name) "Try to register On-Demand Paging memory region: %s"
> +qemu_rdma_advise_mr(const char *name, uint32_t len, uint64_t addr, const char *res) "Try to advise block %s prefetch write at %" PRIu32 "@0x%" PRIx64 ": %s"
>  qemu_rdma_registration_handle_compress(int64_t length, int index, int64_t offset) "Zapping zero chunk: %" PRId64 " bytes, index %d, offset %" PRId64
>  qemu_rdma_registration_handle_finished(void) ""
>  qemu_rdma_registration_handle_ram_blocks(void) ""
> --
> 2.31.1
>

Following https://github.com/linux-rdma/rdma-core/blob/master/libibverbs/man/ibv_advise_mr.3.md
it looks like it is a best-effort optimization,
I don't see any down-sides to it.
However it seems like it is recommended to use
IBV_ADVISE_MR_FLAG_FLUSH in order to
increase the optimization chances.

Anyway

Reviewed-by: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>

Thanks,
Marcel
lizhijian@fujitsu.com Aug. 23, 2021, 1:21 a.m. UTC | #2
Hi Marcel


On 22/08/2021 16:39, Marcel Apfelbaum wrote:
> Hi,
>
> On Sat, Jul 31, 2021 at 5:03 PM Li Zhijian <lizhijian@cn.fujitsu.com> wrote:
>> The responder mr registering with ODP will sent RNR NAK back to
>> the requester in the face of the page fault.
>> ---------
>> ibv_poll_cq wc.status=13 RNR retry counter exceeded!
>> ibv_poll_cq wrid=WRITE RDMA!
>> ---------
>> ibv_advise_mr(3) helps to make pages present before the actual IO is
>> conducted so that the responder does page fault as little as possible.
>>
>> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com>
>> ---
>>   migration/rdma.c       | 40 ++++++++++++++++++++++++++++++++++++++++
>>   migration/trace-events |  1 +
>>   2 files changed, 41 insertions(+)
>>
>> diff --git a/migration/rdma.c b/migration/rdma.c
>> index 8784b5f22a6..a2ad00d665f 100644
>> --- a/migration/rdma.c
>> +++ b/migration/rdma.c
>> @@ -1117,6 +1117,30 @@ static int qemu_rdma_alloc_qp(RDMAContext *rdma)
>>       return 0;
>>   }
>>
>> +/*
>> + * ibv_advise_mr to avoid RNR NAK error as far as possible.
>> + * The responder mr registering with ODP will sent RNR NAK back to
>> + * the requester in the face of the page fault.
>> + */
>> +static void qemu_rdma_advise_prefetch_write_mr(struct ibv_pd *pd, uint64_t addr,
>> +                                               uint32_t len,  uint32_t lkey,
>> +                                               const char *name, bool wr)
>> +{
>> +    int ret;
>> +    int advice = wr ? IBV_ADVISE_MR_ADVICE_PREFETCH_WRITE :
>> +                 IBV_ADVISE_MR_ADVICE_PREFETCH;
>> +    struct ibv_sge sg_list = {.lkey = lkey, .addr = addr, .length = len};
>> +
>> +    ret = ibv_advise_mr(pd, advice,
>> +                        IB_UVERBS_ADVISE_MR_FLAG_FLUSH, &sg_list, 1);
>> +    /* ignore the error */
> Following https://github.com/linux-rdma/rdma-core/blob/master/libibverbs/man/ibv_advise_mr.3.md
> it looks like it is a best-effort optimization,
> I don't see any down-sides to it.
> However it seems like it is recommended to use
> IBV_ADVISE_MR_FLAG_FLUSH in order to
> increase the optimization chances.
Good catch,  i will update it soon.


Thanks

>
> Anyway
>
> Reviewed-by: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>
>
> Thanks,
> Marcel
>
>
diff mbox series

Patch

diff --git a/migration/rdma.c b/migration/rdma.c
index 8784b5f22a6..a2ad00d665f 100644
--- a/migration/rdma.c
+++ b/migration/rdma.c
@@ -1117,6 +1117,30 @@  static int qemu_rdma_alloc_qp(RDMAContext *rdma)
     return 0;
 }
 
+/*
+ * ibv_advise_mr to avoid RNR NAK error as far as possible.
+ * The responder mr registering with ODP will sent RNR NAK back to
+ * the requester in the face of the page fault.
+ */
+static void qemu_rdma_advise_prefetch_write_mr(struct ibv_pd *pd, uint64_t addr,
+                                               uint32_t len,  uint32_t lkey,
+                                               const char *name, bool wr)
+{
+    int ret;
+    int advice = wr ? IBV_ADVISE_MR_ADVICE_PREFETCH_WRITE :
+                 IBV_ADVISE_MR_ADVICE_PREFETCH;
+    struct ibv_sge sg_list = {.lkey = lkey, .addr = addr, .length = len};
+
+    ret = ibv_advise_mr(pd, advice,
+                        IB_UVERBS_ADVISE_MR_FLAG_FLUSH, &sg_list, 1);
+    /* ignore the error */
+    if (ret) {
+        trace_qemu_rdma_advise_mr(name, len, addr, strerror(errno));
+    } else {
+        trace_qemu_rdma_advise_mr(name, len, addr, "successed");
+    }
+}
+
 static int qemu_rdma_reg_whole_ram_blocks(RDMAContext *rdma)
 {
     int i;
@@ -1140,6 +1164,17 @@  on_demand:
             perror("Failed to register local dest ram block!\n");
             break;
         }
+
+        if (access & IBV_ACCESS_ON_DEMAND) {
+            qemu_rdma_advise_prefetch_write_mr(rdma->pd,
+                                               (uintptr_t)
+                                               local->block[i].local_host_addr,
+                                               local->block[i].length,
+                                               local->block[i].mr->lkey,
+                                               local->block[i].block_name,
+                                               true);
+        }
+
         rdma->total_registrations++;
     }
 
@@ -1244,6 +1279,11 @@  on_demand:
                             rdma->total_registrations);
             return -1;
         }
+        if (access & IBV_ACCESS_ON_DEMAND) {
+            qemu_rdma_advise_prefetch_write_mr(rdma->pd, (uintptr_t)chunk_start,
+                                               len, block->pmr[chunk]->lkey,
+                                               block->block_name, rkey);
+        }
         rdma->total_registrations++;
     }
 
diff --git a/migration/trace-events b/migration/trace-events
index 5f6aa580def..901c1d54c12 100644
--- a/migration/trace-events
+++ b/migration/trace-events
@@ -213,6 +213,7 @@  qemu_rdma_poll_other(const char *compstr, int64_t comp, int left) "other complet
 qemu_rdma_post_send_control(const char *desc) "CONTROL: sending %s.."
 qemu_rdma_register_and_get_keys(uint64_t len, void *start) "Registering %" PRIu64 " bytes @ %p"
 qemu_rdma_register_odp_mr(const char *name) "Try to register On-Demand Paging memory region: %s"
+qemu_rdma_advise_mr(const char *name, uint32_t len, uint64_t addr, const char *res) "Try to advise block %s prefetch write at %" PRIu32 "@0x%" PRIx64 ": %s"
 qemu_rdma_registration_handle_compress(int64_t length, int index, int64_t offset) "Zapping zero chunk: %" PRId64 " bytes, index %d, offset %" PRId64
 qemu_rdma_registration_handle_finished(void) ""
 qemu_rdma_registration_handle_ram_blocks(void) ""