OFI MR flags selection based on the provider #1077
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -62,14 +62,6 @@ struct fid_cq* shmem_transport_ofi_target_cq; | |
| #if ENABLE_TARGET_CNTR | ||
| struct fid_cntr* shmem_transport_ofi_target_cntrfd; | ||
| #endif | ||
| struct fid_mr* shmem_transport_ofi_target_heap_mrfd; | ||
| struct fid_mr* shmem_transport_ofi_target_data_mrfd; | ||
| uint64_t* shmem_transport_ofi_target_heap_keys; | ||
|
|
@@ -80,7 +72,6 @@ int shmem_transport_ofi_use_absolute_address; | |
| uint8_t** shmem_transport_ofi_target_heap_addrs; | ||
| uint8_t** shmem_transport_ofi_target_data_addrs; | ||
| #endif /* ENABLE_REMOTE_VIRTUAL_ADDRESSING */ | ||
| uint64_t shmem_transport_ofi_max_poll; | ||
| long shmem_transport_ofi_put_poll_limit; | ||
| long shmem_transport_ofi_get_poll_limit; | ||
|
|
@@ -620,6 +611,52 @@ int bind_enable_ep_resources(shmem_transport_ctx_t *ctx) | |
| } | ||
|
|
||
|
|
||
| static inline | ||
| int allocate_separate_heap_data_mr(uint64_t *flags) { | ||
| int ret = 0; | ||
|
|
||
| /* Register separate data and heap segments using keys 0 and 1, | ||
| * respectively. In MR_BASIC_MODE, the keys are ignored and selected by | ||
| * the provider. */ | ||
| ret = fi_mr_reg(shmem_transport_ofi_domainfd, shmem_internal_heap_base, | ||
| shmem_internal_heap_length, | ||
| FI_REMOTE_READ | FI_REMOTE_WRITE, 0, 1ULL, *flags, | ||
| &shmem_transport_ofi_target_heap_mrfd, NULL); | ||
| OFI_CHECK_RETURN_STR(ret, "target memory (heap) registration failed"); | ||
|
|
||
| ret = fi_mr_reg(shmem_transport_ofi_domainfd, shmem_internal_data_base, | ||
| shmem_internal_data_length, | ||
| FI_REMOTE_READ | FI_REMOTE_WRITE, 0, 0ULL, *flags, | ||
| &shmem_transport_ofi_target_data_mrfd, NULL); | ||
| OFI_CHECK_RETURN_STR(ret, "target memory (data) registration failed"); | ||
|
|
||
| /* Bind counter with target memory region for incoming messages */ | ||
| #if ENABLE_TARGET_CNTR | ||
| ret = fi_mr_bind(shmem_transport_ofi_target_heap_mrfd, | ||
| &shmem_transport_ofi_target_cntrfd->fid, | ||
| FI_REMOTE_WRITE); | ||
| OFI_CHECK_RETURN_STR(ret, "target CNTR binding to heap MR failed"); | ||
|
|
||
| ret = fi_mr_bind(shmem_transport_ofi_target_data_mrfd, | ||
| &shmem_transport_ofi_target_cntrfd->fid, | ||
| FI_REMOTE_WRITE); | ||
| OFI_CHECK_RETURN_STR(ret, "target CNTR binding to data MR failed"); | ||
|
|
||
| #ifdef ENABLE_MR_RMA_EVENT | ||
| if (shmem_transport_ofi_mr_rma_event) { | ||
| ret = fi_mr_enable(shmem_transport_ofi_target_data_mrfd); | ||
| OFI_CHECK_RETURN_STR(ret, "target data MR enable failed"); | ||
|
|
||
| ret = fi_mr_enable(shmem_transport_ofi_target_heap_mrfd); | ||
| OFI_CHECK_RETURN_STR(ret, "target heap MR enable failed"); | ||
| } | ||
| #endif /* ENABLE_MR_RMA_EVENT */ | ||
| #endif /* ENABLE_TARGET_CNTR */ | ||
|
|
||
| return ret; | ||
| } | ||
|
|
||
|
|
||
| static inline | ||
| int allocate_recv_cntr_mr(void) | ||
| { | ||
|
|
@@ -653,65 +690,32 @@ int allocate_recv_cntr_mr(void) | |
| } | ||
| #endif | ||
|
|
||
| #if defined(ENABLE_REMOTE_VIRTUAL_ADDRESSING) | ||
| if (shmem_transport_ofi_mr_mode == 0) { | ||
| ret = fi_mr_reg(shmem_transport_ofi_domainfd, 0, UINT64_MAX, | ||
| FI_REMOTE_READ | FI_REMOTE_WRITE, 0, 0ULL, flags, | ||
| &shmem_transport_ofi_target_heap_mrfd, NULL); | ||
| OFI_CHECK_RETURN_STR(ret, "target memory (all) registration failed"); | ||
|
There was a problem hiding this comment. I'm confused why virtual addressing now registers the heap now. Does it not need the data segment? |
||
|
|
||
| /* Bind counter with target memory region for incoming messages */ | ||
| #if ENABLE_TARGET_CNTR | ||
| ret = fi_mr_bind(shmem_transport_ofi_target_heap_mrfd, | ||
| &shmem_transport_ofi_target_cntrfd->fid, | ||
| FI_REMOTE_WRITE); | ||
| OFI_CHECK_RETURN_STR(ret, "target CNTR binding to MR failed"); | ||
|
|
||
| #ifdef ENABLE_MR_RMA_EVENT | ||
| if (shmem_transport_ofi_mr_rma_event) { | ||
| ret = fi_mr_enable(shmem_transport_ofi_target_heap_mrfd); | ||
| OFI_CHECK_RETURN_STR(ret, "target MR enable failed"); | ||
| } | ||
| #endif /* ENABLE_MR_RMA_EVENT */ | ||
| #endif /* ENABLE_TARGET_CNTR */ | ||
| } else { | ||
| ret = allocate_separate_heap_data_mr(&flags); | ||
| } | ||
| #else | ||
| ret = allocate_separate_heap_data_mr(&flags); | ||
| #endif | ||
|
|
||
| return ret; | ||
|
|
@@ -720,7 +724,7 @@ int allocate_recv_cntr_mr(void) | |
| static | ||
| int publish_mr_info(void) | ||
| { | ||
| if (shmem_transport_ofi_mr_mode != 0) | ||
| { | ||
| int err; | ||
| uint64_t heap_key, data_key; | ||
|
|
@@ -744,16 +748,15 @@ int publish_mr_info(void) | |
| RAISE_WARN_STR("Put of data segment key to runtime KVS failed"); | ||
| return 1; | ||
| } | ||
|
|
||
|
|
||
| #ifdef ENABLE_REMOTE_VIRTUAL_ADDRESSING | ||
| if (shmem_transport_ofi_info.p_info->domain_attr->mr_mode & FI_MR_VIRT_ADDR) | ||
| shmem_transport_ofi_use_absolute_address = 1; | ||
| else | ||
| shmem_transport_ofi_use_absolute_address = 0; | ||
| #else /* !ENABLE_REMOTE_VIRTUAL_ADDRESSING */ | ||
|
|
||
| void *heap_base, *data_base; | ||
|
|
||
| if (shmem_transport_ofi_info.p_info->domain_attr->mr_mode & FI_MR_VIRT_ADDR) { | ||
|
|
@@ -775,17 +778,17 @@ int publish_mr_info(void) | |
| RAISE_WARN_STR("Put of data segment address to runtime KVS failed"); | ||
| return 1; | ||
| } | ||
|
|
||
| #endif /* ENABLE_REMOTE_VIRTUAL_ADDRESSING */ | ||
| } | ||
|
|
||
| return 0; | ||
| } | ||
|
|
||
| static | ||
| int populate_mr_tables(void) | ||
| { | ||
| if (shmem_transport_ofi_mr_mode != 0) | ||
| { | ||
| int i, err; | ||
|
|
||
|
|
@@ -818,11 +821,9 @@ int populate_mr_tables(void) | |
| return 1; | ||
| } | ||
| } | ||
|
|
||
|
|
||
| #ifndef ENABLE_REMOTE_VIRTUAL_ADDRESSING | ||
|
|
||
| shmem_transport_ofi_target_heap_addrs = malloc(sizeof(uint8_t*) * shmem_internal_num_pes); | ||
| if (NULL == shmem_transport_ofi_target_heap_addrs) { | ||
|
|
@@ -853,9 +854,9 @@ int populate_mr_tables(void) | |
| return 1; | ||
| } | ||
| } | ||
|
|
||
| #endif /* ENABLE_REMOTE_VIRTUAL_ADDRESSING */ | ||
| } | ||
|
|
||
| return 0; | ||
| } | ||
|
|
@@ -1157,20 +1158,10 @@ int query_for_fabric(struct fabric_info *info) | |
| hints.addr_format = FI_FORMAT_UNSPEC; | ||
| domain_attr.data_progress = FI_PROGRESS_AUTO; | ||
|
There was a problem hiding this comment. Just a note that #1059 possibly motivates a guard warning about and/or preventing the usage of |
||
| domain_attr.resource_mgmt = FI_RM_ENABLED; | ||
|
|
||
| shmem_transport_ofi_mr_mode = set_mr_flag(shmem_transport_ofi_info.prov_name, | ||
| &domain_attr); | ||
|
|
||
| #ifdef ENABLE_THREADS | ||
| if (shmem_internal_thread_level == SHMEM_THREAD_MULTIPLE) { | ||
| #ifdef USE_THREAD_COMPLETION | ||
|
|
@@ -1244,17 +1235,12 @@ int query_for_fabric(struct fabric_info *info) | |
| shmem_transport_ofi_stx_max = 0; | ||
| } | ||
|
|
||
| /* Heap and data use different MR keys, need at least 1 byte of key space | ||
| * if using provider selected keys */ | ||
| if (info->p_info->domain_attr->mr_mode & FI_MR_PROV_KEY) | ||
| info->p_info->domain_attr->mr_key_size = 1; | ||
| else | ||
| info->p_info->domain_attr->mr_key_size = 0; | ||
|
|
||
| shmem_internal_assertp(info->p_info->tx_attr->inject_size >= shmem_transport_ofi_max_buffered_send); | ||
| shmem_transport_ofi_max_buffered_send = info->p_info->tx_attr->inject_size; | ||
|
|
@@ -1417,10 +1403,14 @@ int shmem_transport_init(void) | |
|
|
||
| if (shmem_internal_params.OFI_PROVIDER_provided) | ||
| shmem_transport_ofi_info.prov_name = shmem_internal_params.OFI_PROVIDER; | ||
| else { | ||
| shmem_transport_ofi_info.prov_name = NULL; | ||
| char *fi_provider_provided = getenv("FI_PROVIDER"); | ||
| if (fi_provider_provided != NULL) | ||
| shmem_transport_ofi_info.prov_name = fi_provider_provided; | ||
| else | ||
| shmem_transport_ofi_info.prov_name = NULL; | ||
| } | ||
|
|
||
| if (shmem_internal_params.OFI_FABRIC_provided) | ||
| shmem_transport_ofi_info.fabric_name = shmem_internal_params.OFI_FABRIC; | ||
|
|
@@ -1788,13 +1778,15 @@ int shmem_transport_fini(void) | |
| ret = fi_close(&shmem_transport_ofi_target_cq->fid); | ||
| OFI_CHECK_ERROR_MSG(ret, "Target CQ close failed (%s)\n", fi_strerror(errno)); | ||
|
|
||
| ret = fi_close(&shmem_transport_ofi_target_heap_mrfd->fid); | ||
| OFI_CHECK_ERROR_MSG(ret, "Target heap MR close failed (%s)\n", fi_strerror(errno)); | ||
|
|
||
| #if defined(ENABLE_REMOTE_VIRTUAL_ADDRESSING) | ||
| if (shmem_transport_ofi_mr_mode != 0) { | ||
| ret = fi_close(&shmem_transport_ofi_target_data_mrfd->fid); | ||
|
There was a problem hiding this comment. I don't see a corresponding |
||
| OFI_CHECK_ERROR_MSG(ret, "Target data MR close failed (%s)\n", fi_strerror(errno)); | ||
| } | ||
| #else | ||
| ret = fi_close(&shmem_transport_ofi_target_data_mrfd->fid); | ||
| OFI_CHECK_ERROR_MSG(ret, "Target data MR close failed (%s)\n", fi_strerror(errno)); | ||
| #endif | ||
|
|
||
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Should "none" be called "auto" since it selects MR flags based on the provider?