[motor]日志整改-memfabric-hybrid-AtomGit

M.claude/skills+1-1

	@@ -1 +1 @@
1	-../.agents/skills	1	+../.agents/skills

MCLAUDE.md+1-1

	@@ -1 +1 @@
1	-AGENTS.md	1	+AGENTS.md

Msrc/acc_links/csrc/acc_tcp_link_complex_default.cpp+2-2

	@@ -57,8 +57,8 @@ Result AccTcpLinkComplexDefault::EnqueueAndModifyEpoll(const AccMsgHeader &h, co
57	}	57	}
58	auto result = queue_->EnqueueBack(h, d, cbCtx);	58	auto result = queue_->EnqueueBack(h, d, cbCtx);
59	if (UNLIKELY(result != ACC_OK)) {	59	if (UNLIKELY(result != ACC_OK)) {
60	- LOG_WARN("Failed to enqueue message into link " << this->id_ << ", errorCode:" << result	60	+ LOG_ERROR("Failed to enqueue message into link " << this->id_ << ", errorCode:" << result
61	- << ", queue size:" << queue_->GetSize());	61	+ << ", queue size:" << queue_->GetSize());
62	return result;	62	return result;
63	}	63	}
64		64

Msrc/acc_links/csrc/acc_tcp_link_complex_default.h+3-3

Msrc/acc_links/csrc/acc_tcp_link_default.h+4-4

	@@ -113,7 +113,7 @@ public:
113	if (errorNumber == EINTR) { /* interrupted */	113	if (errorNumber == EINTR) { /* interrupted */
114	continue;	114	continue;
115	}	115	}
116	- LOG_WARN("Failed to send data to " << ipPort_ << ", errno " << errorNumber);	116	+ LOG_WARN("Unable to send data to " << ipPort_ << ", reason " << errorNumber);
117	if (g_errnoToReconn.count(errorNumber) > 0) {	117	if (g_errnoToReconn.count(errorNumber) > 0) {
118	return ACC_LINK_NEED_RECONN;	118	return ACC_LINK_NEED_RECONN;
119	}	119	}
	@@ -163,7 +163,7 @@ public:
163	return ACC_OK;	163	return ACC_OK;
164	}	164	}
165	auto errorNumber = errno;	165	auto errorNumber = errno;
166	- LOG_WARN("Failed to send data to " << ipPort_ << ", errno " << errorNumber);	166	+ src LOG_WARN("Unable to send data to " << ipPort_ << ", reason " << errorNumber);
167	return -errorNumber;	167	return -errorNumber;
168	} else {	168	} else {
169	for (int32_t i = 0; i < len; i++) {	169	for (int32_t i = 0; i < len; i++) {
	@@ -194,7 +194,7 @@ public:
194	if (errorNumber == EINTR) { /* interrupted */	194	if (errorNumber == EINTR) { /* interrupted */
195	continue;	195	continue;
196	}	196	}
197	- LOG_WARN("Failed to read data from " << ipPort_ << ", errno " << errorNumber);	197	+ LOG_WARN("Unable to read data from " << ipPort_ << ", reason " << errorNumber);
198	if (g_errnoToReconn.count(errorNumber) > 0) {	198	if (g_errnoToReconn.count(errorNumber) > 0) {
199	return ACC_LINK_NEED_RECONN;	199	return ACC_LINK_NEED_RECONN;
200	}	200	}
	@@ -248,7 +248,7 @@ public:
248	return ACC_OK;	248	return ACC_OK;
249	}	249	}
250	auto errorNumber = errno;	250	auto errorNumber = errno;
251	- LOG_WARN("Failed to receive data from " << ipPort_ << ", errno " << errorNumber);	251	+ LOG_WARN("Unable to receive data from " << ipPort_ << ", reason " << errorNumber);
252	return -errorNumber;	252	return -errorNumber;
253	} else {	253	} else {
254	for (int32_t i = 0; i < len; i++) {	254	for (int32_t i = 0; i < len; i++) {

Msrc/acc_links/csrc/acc_tcp_listener.cpp+3-3

	@@ -158,7 +158,7 @@ void AccTcpListener::RunInThread() noexcept
158	socklen_t len = sizeof(addressIn);	158	socklen_t len = sizeof(addressIn);
159	auto fd = ::accept(listenFd_, reinterpret_cast<struct sockaddr *>(&addressIn), &len);	159	auto fd = ::accept(listenFd_, reinterpret_cast<struct sockaddr *>(&addressIn), &len);
160	if (fd < 0) {	160	if (fd < 0) {
161	- LOG_WARN("Failed to accept on new socket with " << strerror(errno) << ", ignore and continue");	161	+ LOG_WARN("Unable to accept on new socket with " << strerror(errno) << ", ignore and continue");
162	continue;	162	continue;
163	}	163	}
164		164
	@@ -231,7 +231,7 @@ void AccTcpListener::ProcessNewConnection(int fd, struct sockaddr_in addressIn)
231	// newLink作为智能指针异常分支返回时会自动析构释放资源	231	// newLink作为智能指针异常分支返回时会自动析构释放资源
232	auto result = connHandler_(req, newLink.Get());	232	auto result = connHandler_(req, newLink.Get());
233	if (result != ACC_OK) {	233	if (result != ACC_OK) {
234	- LOG_INFO("ProcessNewConnection: connHandler_ non-ok, send error resp result=" << result);	234	+ LOG_WARN("ProcessNewConnection: connHandler_ non-ok, send error resp result=" << result);
235	AccConnResp resp;	235	AccConnResp resp;
236	resp.result = static_cast<int16_t>(result);	236	resp.result = static_cast<int16_t>(result);
237	(void)::send(fd, &resp, sizeof(resp), 0);	237	(void)::send(fd, &resp, sizeof(resp), 0);
	@@ -242,7 +242,7 @@ void AccTcpListener::ProcessNewConnection(int fd, struct sockaddr_in addressIn)
242	resp.result = 0;	242	resp.result = 0;
243	auto sent = newLink->BlockSend(reinterpret_cast<void *>(&resp), sizeof(resp));	243	auto sent = newLink->BlockSend(reinterpret_cast<void *>(&resp), sizeof(resp));
244	if (sent != ACC_OK) {	244	if (sent != ACC_OK) {
245	- LOG_WARN("Failed to connect response to " << ipPort);	245	+ LOG_WARN("Unable to connect response to " << ipPort);
246	}	246	}
247	}	247	}
248	} // namespace acc	248	} // namespace acc

Msrc/acc_links/csrc/acc_tcp_worker.cpp+2-2

	@@ -133,7 +133,7 @@ void AccTcpWorker::SetPropertiesForThread()
133	CPU_ZERO(&(cpuSet));	133	CPU_ZERO(&(cpuSet));
134	CPU_SET(options_.cpuId, &(cpuSet));	134	CPU_SET(options_.cpuId, &(cpuSet));
135	if (pthread_setaffinity_np(pthread_self(), sizeof(cpuSet), &(cpuSet)) != 0) {	135	if (pthread_setaffinity_np(pthread_self(), sizeof(cpuSet), &(cpuSet)) != 0) {
136	- LOG_WARN("Failed to bind worker " << options_.Name() << " to cpu " << options_.cpuId);	136	+ LOG_WARN("Unable to bind worker " << options_.Name() << " to cpu " << options_.cpuId);
137	}	137	}
138	}	138	}
139		139
	@@ -142,7 +142,7 @@ void AccTcpWorker::SetPropertiesForThread()
142		142
143	if (options_.threadPriority != 0) {	143	if (options_.threadPriority != 0) {
144	if (setpriority(PRIO_PROCESS, 0, options_.threadPriority) != 0) {	144	if (setpriority(PRIO_PROCESS, 0, options_.threadPriority) != 0) {
145	- LOG_WARN("Failed to set thread priority of worker " << options_.Name() << ", errno:" << errno);	145	+ LOG_WARN("Unable to set thread priority of worker " << options_.Name() << ":" << errno);
146	}	146	}
147	}	147	}
148	}	148	}

Msrc/acc_links/csrc/security/acc_tcp_ssl_helper.cpp+5-5

	@@ -412,9 +412,9 @@ int AccTcpSslHelper::ProcessCrlAndVerifyCert(std::vector<std::string> paths, X50
412		412
413	auto verifyResult = OpenSslApiWrapper::X509VerifyCert(x509ctx);	413	auto verifyResult = OpenSslApiWrapper::X509VerifyCert(x509ctx);
414	if (verifyResult != 1U) {	414	if (verifyResult != 1U) {
415	- LOG_INFO("Verify failed in callback"	415	+ LOG_ERROR("Verify failed in callback"
416	- << " error: "	416	+ << " error: "
417	- << OpenSslApiWrapper::X509VerifyCertErrorString(OpenSslApiWrapper::X509StoreCtxGetError(x509ctx)));	417	+ << OpenSslApiWrapper::X509VerifyCertErrorString(OpenSslApiWrapper::X509StoreCtxGetError(x509ctx)));
418	return checkFailed;	418	return checkFailed;
419	}	419	}
420		420
	@@ -457,7 +457,7 @@ AccResult AccTcpSslHelper::CertVerify(X509 *cert) const
457	for (const auto &crlPath : tlsCrlPaths) {	457	for (const auto &crlPath : tlsCrlPaths) {
458	X509_CRL *crl = LoadCertRevokeListFile(crlPath.c_str());	458	X509_CRL *crl = LoadCertRevokeListFile(crlPath.c_str());
459	if (crl == nullptr) {	459	if (crl == nullptr) {
460	- LOG_WARN("Failed to load cert revocation list: " << crlPath);	460	+ LOG_WARN("Unable to load cert revocation list: " << crlPath);
461	continue;	461	continue;
462	}	462	}
463		463
	@@ -516,7 +516,7 @@ AccResult AccTcpSslHelper::CheckCertExpiredTask()
516		516
517	auto ret = HandleCertExpiredCheck();	517	auto ret = HandleCertExpiredCheck();
518	if (ret != ACC_OK) {	518	if (ret != ACC_OK) {
519	- LOG_WARN("Failed to handle cert expired check");	519	+ LOG_WARN("Unable to handle cert expired check");
520	}	520	}
521	}	521	}
522	}	522	}

Msrc/hybm/csrc/common/hybm_define.h+6-6

	@@ -158,12 +158,12 @@ typedef struct {
158	} \	158	} \
159	} while (0)	159	} while (0)
160		160
161	-#define DL_LOAD_SYM_OPTIONAL(TARGET_FUNC_VAR, TARGET_FUNC_TYPE, FILE_HANDLE, SYMBOL_NAME) \	161	+#define DL_LOAD_SYM_OPTIONAL(TARGET_FUNC_VAR, TARGET_FUNC_TYPE, FILE_HANDLE, SYMBOL_NAME) \
162	- do { \	162	+ do { \
163	- TARGET_FUNC_VAR = (TARGET_FUNC_TYPE)dlsym(FILE_HANDLE, SYMBOL_NAME); \	163	+ TARGET_FUNC_VAR = (TARGET_FUNC_TYPE)dlsym(FILE_HANDLE, SYMBOL_NAME); \
164	- if ((TARGET_FUNC_VAR) == nullptr) { \	164	+ if ((TARGET_FUNC_VAR) == nullptr) { \
165	- BM_LOG_WARN("Failed to call dlsym to load " << (SYMBOL_NAME) << ", error: " << dlerror()); \	165	+ BM_LOG_WARN("Unable to call dlsym to load " << (SYMBOL_NAME) << ", reason: " << dlerror()); \
166	- } \	166	+ } \
167	} while (0)	167	} while (0)
168		168
169	#define DL_LOAD_SYM_ALT(TARGET_FUNC_VAR, TARGET_FUNC_TYPE, FILE_HANDLE, SYMBOL_NAME, SYMBOL_NAME_ALT) \	169	#define DL_LOAD_SYM_ALT(TARGET_FUNC_VAR, TARGET_FUNC_TYPE, FILE_HANDLE, SYMBOL_NAME, SYMBOL_NAME_ALT) \

Msrc/hybm/csrc/data_operation/host/hybm_data_op_host_rdma.cpp+2-2

	@@ -46,7 +46,7 @@ Result HostDataOpRDMA::Initialize() noexcept
46	rdmaSwapBaseAddr_ =	46	rdmaSwapBaseAddr_ =
47	mmap(nullptr, rdmaSwapSpaceSize_, PROT_READ \| PROT_WRITE, MAP_ANONYMOUS \| MAP_PRIVATE \| MAP_HUGETLB, -1, 0);	47	mmap(nullptr, rdmaSwapSpaceSize_, PROT_READ \| PROT_WRITE, MAP_ANONYMOUS \| MAP_PRIVATE \| MAP_HUGETLB, -1, 0);
48	if (rdmaSwapBaseAddr_ == MAP_FAILED) {	48	if (rdmaSwapBaseAddr_ == MAP_FAILED) {
49	- BM_LOG_WARN("Failed to alloc with huge page, size:" << rdmaSwapSpaceSize_ << " error:" << errno << ", "	49	+ BM_LOG_WARN("Unable to alloc with huge page, size:" << rdmaSwapSpaceSize_ << " error:" << errno << ", "
50	<< SafeStrError(errno)	50	<< SafeStrError(errno)
51	<< ". fallback to mmap with regular pagesize");	51	<< ". fallback to mmap with regular pagesize");
52	rdmaSwapBaseAddr_ =	52	rdmaSwapBaseAddr_ =
	@@ -146,7 +146,7 @@ void HostDataOpRDMA::PreRegisterLocalMr(hybm_copy_params &params, hybm_data_copy
146		146
147	auto ret = transportManager_->RegisterMemoryRegion(mr);	147	auto ret = transportManager_->RegisterMemoryRegion(mr);
148	if (ret != BM_OK) {	148	if (ret != BM_OK) {
149	- BM_LOG_WARN("Failed to pre register host rdma mr, addr:" << (void *)mr.addr << " size: " << mr.size);	149	+ BM_LOG_WARN("Unable to pre register host rdma mr, addr:" << (void *)mr.addr << " size: " << mr.size);
150	}	150	}
151	}	151	}
152		152

Msrc/hybm/csrc/driver/userspace/devmm_svm_gva.cpp+1-1

	@@ -147,7 +147,7 @@ static bool TryUpdateGvaHeap(uint64_t va, size_t len)
147	static int32_t RemoveInGvaHeap(uint64_t va)	147	static int32_t RemoveInGvaHeap(uint64_t va)
148	{	148	{
149	if (!g_gvaHeapMgr.inited) {	149	if (!g_gvaHeapMgr.inited) {
150	- BM_LOG_WARN("remove record in gva heap failed, gva heap not init.");	150	+ BM_LOG_WARN("unable to remove record in gva heap, gva heap not init.");
151	return -1;	151	return -1;
152	}	152	}
153		153

Msrc/hybm/csrc/entity/hybm_entity_default.cpp+1-1

	@@ -747,7 +747,7 @@ int32_t MemEntityDefault::RemoveImported(const std::vector<uint32_t> &ranks) noe
747	if (transportManager_ != nullptr) {	747	if (transportManager_ != nullptr) {
748	auto ret = transportManager_->RemoveRanks(ranks);	748	auto ret = transportManager_->RemoveRanks(ranks);
749	if (ret != BM_OK) {	749	if (ret != BM_OK) {
750	- BM_LOG_WARN("transport remove ranks failed: " << ret);	750	+ BM_LOG_WARN("unable to transport remove ranks : " << ret);
751	}	751	}
752	}	752	}
753		753

Msrc/hybm/csrc/hybm_entry.cpp+1-1

	@@ -99,7 +99,7 @@ HYBM_API int32_t hybm_init(uint16_t deviceId, uint64_t flags)
99	ptracer_config_t config{.tracerType = 1, .dumpFilePath = "/var/log/memfabric_hybrid"};	99	ptracer_config_t config{.tracerType = 1, .dumpFilePath = "/var/log/memfabric_hybrid"};
100	auto ret = ptracer_init(&config);	100	auto ret = ptracer_init(&config);
101	if (ret != BM_OK) {	101	if (ret != BM_OK) {
102	- BM_LOG_WARN("init ptracer module failed, result: " << ret << ", error msg: " << ptracer_get_last_err_msg());	102	+ BM_LOG_WARN("init ptracer module not successful, result: " << ret << ", msg: " << ptracer_get_last_err_msg());
103	}	103	}
104		104
105	ret = hybm_init_hbm_gva(deviceId, flags, g_baseAddr, DlAclApi::GetAscendSocType(), &g_allocHandle);	105	ret = hybm_init_hbm_gva(deviceId, flags, g_baseAddr, DlAclApi::GetAscendSocType(), &g_allocHandle);

Msrc/hybm/csrc/mm/hybm_conn_based_segment.cpp+4-4

Msrc/hybm/csrc/mm/hybm_dev_user_legacy_segment.cpp+2-2

	@@ -295,7 +295,7 @@ void HybmDevUserLegacySegment::RemoveSliceInfo(const uint32_t rankId) noexcept
295	<< ", sliceInfo.rankId=" << sliceInfo.rankId);	295	<< ", sliceInfo.rankId=" << sliceInfo.rankId);
296	auto ret = DlAclApi::RtIpcCloseMemory(address);	296	auto ret = DlAclApi::RtIpcCloseMemory(address);
297	if (ret != 0) {	297	if (ret != 0) {
298	- BM_LOG_WARN("Failed to close memory, address="	298	+ BM_LOG_WARN("Unable to close memory, address="
299	<< address << ", vAddress_"	299	<< address << ", vAddress_"
300	<< reinterpret_cast<void *>(static_cast<ptrdiff_t>(remoteSlice->vAddress_))	300	<< reinterpret_cast<void *>(static_cast<ptrdiff_t>(remoteSlice->vAddress_))
301	<< ", deviceId=" << devicePhyId_ << ", sliceInfo.devicePhyId=" << sliceInfo.devicePhyId	301	<< ", deviceId=" << devicePhyId_ << ", sliceInfo.devicePhyId=" << sliceInfo.devicePhyId
	@@ -452,7 +452,7 @@ void HybmDevUserLegacySegment::CloseMemory() noexcept
452	{	452	{
453	for (auto &addr : registerAddrs_) {	453	for (auto &addr : registerAddrs_) {
454	if (DlAclApi::RtIpcCloseMemory(addr) != 0) {	454	if (DlAclApi::RtIpcCloseMemory(addr) != 0) {
455	- BM_LOG_WARN("Failed to close memory. This may affect future memory registration.");	455	+ BM_LOG_WARN("Unable to close memory. This may affect future memory registration.");
456	}	456	}
457	}	457	}
458	for (auto &it : registerSlices_) {	458	for (auto &it : registerSlices_) {

Msrc/hybm/csrc/mm/hybm_host_shm_segment.cpp+3-3

	@@ -43,7 +43,7 @@ bool IsHugetlbfsMounted() noexcept
43	{	43	{
44	struct statfs fsInfo {};	44	struct statfs fsInfo {};
45	if (statfs(HOST_SHM_HUGEPAGE_DIR, &fsInfo) != 0) {	45	if (statfs(HOST_SHM_HUGEPAGE_DIR, &fsInfo) != 0) {
46	- BM_LOG_WARN("Failed to access hugepage dir " << HOST_SHM_HUGEPAGE_DIR << " error:" << errno << ", "	46	+ BM_LOG_WARN("Unable to access hugepage dir " << HOST_SHM_HUGEPAGE_DIR << " reason:" << errno << ", "
47	<< SafeStrError(errno) << ", will fallback to /dev/shm");	47	<< SafeStrError(errno) << ", will fallback to /dev/shm");
48	return false;	48	return false;
49	}	49	}
	@@ -59,7 +59,7 @@ bool HasAvailableHugePages() noexcept
59	{	59	{
60	std::ifstream memInfoFile("/proc/meminfo");	60	std::ifstream memInfoFile("/proc/meminfo");
61	if (!memInfoFile.is_open()) {	61	if (!memInfoFile.is_open()) {
62	- BM_LOG_WARN("Failed to open /proc/meminfo, will fallback to /dev/shm");	62	+ BM_LOG_WARN("Unable to open /proc/meminfo, will fallback to /dev/shm");
63	return false;	63	return false;
64	}	64	}
65		65
	@@ -84,7 +84,7 @@ bool HasAvailableHugePages() noexcept
84	return true;	84	return true;
85	}	85	}
86		86
87	- BM_LOG_WARN("Failed to parse HugePages_Free from /proc/meminfo, will fallback to /dev/shm");	87	+ BM_LOG_WARN("Unable to parse HugePages_Free from /proc/meminfo, will fallback to /dev/shm");
88	return false;	88	return false;
89	}	89	}
90	} // namespace	90	} // namespace

Msrc/hybm/csrc/mm/hybm_va_manager.cpp+4-4

Msrc/hybm/csrc/mm/hybm_vmm_based_segment.cpp+6-6

Msrc/hybm/csrc/transport/compose/compose_transport_manager.cpp+1-1

	@@ -221,7 +221,7 @@ Result ComposeTransportManager::QueryMemoryKey(uint64_t addr, TransportMemoryKey
221	TransportMemoryKey tmp{};	221	TransportMemoryKey tmp{};
222	auto ret = hostTransportManager_->QueryMemoryKey(addr, tmp);	222	auto ret = hostTransportManager_->QueryMemoryKey(addr, tmp);
223	if (ret != BM_OK) {	223	if (ret != BM_OK) {
224	- BM_LOG_WARN("Failed to query host transport memKey addr:" << std::hex << addr);	224	+ BM_LOG_WARN("Unable to query host transport memKey addr:" << std::hex << addr);
225	// HCOM 无法处理HBM池，这里直接返回，兼容即开启HBM池又要走HCOM通信的场景	225	// HCOM 无法处理HBM池，这里直接返回，兼容即开启HBM池又要走HCOM通信的场景
226	}	226	}
227	WriteHcomMemoryKey(tmp, key);	227	WriteHcomMemoryKey(tmp, key);

Msrc/hybm/csrc/transport/device/device_rdma_indirect_transport_manager.cpp+3-3

	@@ -1121,9 +1121,9 @@ int RdmaIndirectTransportManager::ConnectToRemote(const std::string &nic, uint32
1121	ret = connect(sockfd, serverAddr, addrLen);	1121	ret = connect(sockfd, serverAddr, addrLen);
1122	if (ret < 0) {	1122	if (ret < 0) {
1123	if (i < MAX_RETRIES - 1) {	1123	if (i < MAX_RETRIES - 1) {
1124	- BM_LOG_WARN("connect attempt " << (i + 1) << " failed, localRankId: " << localRankId	1124	+ BM_LOG_WARN("unable to connect attempt " << (i + 1) << " , localRankId: " << localRankId
1125	- << ", remoteRankId: " << remoteRankId << ", retrying in "	1125	+ << ", remoteRankId: " << remoteRankId << ", retrying in "
1126	- << RETRY_DELAY_MS << "ms, nic:" << nic);	1126	+ << RETRY_DELAY_MS << "ms, nic:" << nic);
1127	std::this_thread::sleep_for(std::chrono::milliseconds(RETRY_DELAY_MS));	1127	std::this_thread::sleep_for(std::chrono::milliseconds(RETRY_DELAY_MS));
1128	}	1128	}
1129	}	1129	}

Msrc/hybm/csrc/transport/device/device_rdma_transport_manager.cpp+1-1

	@@ -606,7 +606,7 @@ bool RdmaTransportManager::RaInit(uint32_t deviceId)
606	std::this_thread::sleep_for(WAIT_TIME); // avoid hccl init conflict	606	std::this_thread::sleep_for(WAIT_TIME); // avoid hccl init conflict
607	auto ret = DlHccpApi::RaInit(initConfig);	607	auto ret = DlHccpApi::RaInit(initConfig);
608	if (ret != 0) {	608	if (ret != 0) {
609	- BM_LOG_WARN("Hccp Init RA failed: " << ret << " devid:" << deviceId);	609	+ BM_LOG_ERROR("Hccp Init RA failed: " << ret << " devid:" << deviceId);
610	std::this_thread::sleep_for(WAIT_TIME);	610	std::this_thread::sleep_for(WAIT_TIME);
611	raInitialized = false;	611	raInitialized = false;
612	return false;	612	return false;

Msrc/hybm/csrc/transport/device/fixed_ranks_qp_manager.cpp+1-1

	@@ -572,7 +572,7 @@ void FixedRanksQpManager::CloseConnections(std::unordered_map<uint32_t, AiCoreCo
572	if (it->second.qpHandle != nullptr) {	572	if (it->second.qpHandle != nullptr) {
573	auto ret = DlHccpApi::RaQpDestroy(it->second.qpHandle);	573	auto ret = DlHccpApi::RaQpDestroy(it->second.qpHandle);
574	if (ret != 0) {	574	if (ret != 0) {
575	- BM_LOG_WARN("destroy AI QP to server: " << it->first << " failed: " << ret);	575	+ BM_LOG_WARN("unable to destroy AI QP to server: " << it->first << " ret: " << ret);
576	}	576	}
577	it->second.qpHandle = nullptr;	577	it->second.qpHandle = nullptr;
578	}	578	}

Msrc/hybm/csrc/transport/device/indirect/async_socket_queue.cpp+1-1

	@@ -165,7 +165,7 @@ bool AsyncSocketQueue::Start() noexcept
165	}	165	}
166		166
167	if (CreateEpollSocket() != BM_OK) {	167	if (CreateEpollSocket() != BM_OK) {
168	- BM_LOG_WARN("CreateEpollSocket failed.");	168	+ BM_LOG_ERROR("CreateEpollSocket failed.");
169	return false;	169	return false;
170	}	170	}
171		171

Msrc/hybm/csrc/transport/device/joinable_ranks_qp_manager.cpp+2-2

	@@ -152,7 +152,7 @@ void JoinableRanksQpManager::PutQpHandle(UserQpInfo *qp) const noexcept
152	if (val == 1U) { // 返回减之前的值	152	if (val == 1U) { // 返回减之前的值
153	auto ret = DlHccpApi::RaQpDestroy(qp->qpHandle);	153	auto ret = DlHccpApi::RaQpDestroy(qp->qpHandle);
154	if (ret != 0) {	154	if (ret != 0) {
155	- BM_LOG_WARN("close qp from " << rankId_ << " failed, ret: " << ret);	155	+ BM_LOG_WARN("unable to close qp from " << rankId_ << " , ret: " << ret);
156	}	156	}
157	delete qp;	157	delete qp;
158	}	158	}
	@@ -592,7 +592,7 @@ void JoinableRanksQpManager::RemoveRanksProcess(const std::set<uint32_t> &ranks)
592	}	592	}
593	auto ret = DlHccpApi::RaSocketBatchClose(&closeInfo, 1U);	593	auto ret = DlHccpApi::RaSocketBatchClose(&closeInfo, 1U);
594	if (ret != 0) {	594	if (ret != 0) {
595	- BM_LOG_WARN("close socket from " << rankId_ << " to " << it->first << " failed: " << ret);	595	+ BM_LOG_WARN("unable to close socket from " << rankId_ << " to " << it->first << " : " << ret);
596	} else {	596	} else {
597	BM_LOG_INFO("close socket from " << rankId_ << " to " << it->first << " successful: " << ret);	597	BM_LOG_INFO("close socket from " << rankId_ << " to " << it->first << " successful: " << ret);
598	}	598	}

Msrc/hybm/csrc/transport/device/urma/device_urma_transport_manager.cpp+1-0

	@@ -1532,6 +1532,7 @@ Result DeviceUrmaTransportManager::AsyncConnect()
1532		1532
1533	Result DeviceUrmaTransportManager::WaitForConnected(int64_t timeoutNs)	1533	Result DeviceUrmaTransportManager::WaitForConnected(int64_t timeoutNs)
1534	{	1534	{
		1535	+ (void)timeoutNs;
1535	return BM_OK;	1536	return BM_OK;
1536	}	1537	}
1537		1538

Msrc/hybm/csrc/transport/host/host_hcom_transport_manager.cpp+8-8

	@@ -198,7 +198,7 @@ Result HcomTransportManager::CloseDevice()
198		198
199	auto ret = DlHcomApi::ServiceDestroy(service, HCOM_RPC_SERVICE_NAME);	199	auto ret = DlHcomApi::ServiceDestroy(service, HCOM_RPC_SERVICE_NAME);
200	if (ret != 0) {	200	if (ret != 0) {
201	- BM_LOG_WARN("Failed to destroy hcom service, ret: " << ret);	201	+ BM_LOG_WARN("Unable to destroy hcom service, ret: " << ret);
202	}	202	}
203		203
204	mf::MfTlsUtil::CloseTlsLib();	204	mf::MfTlsUtil::CloseTlsLib();
	@@ -271,7 +271,7 @@ Result HcomTransportManager::RegisterMemoryRegion(const TransportMemoryRegion &m
271	BM_ASSERT_LOG_AND_RETURN(mr.addr != 0 && mr.size != 0, "mr.addr = " << mr.addr << ", " << "mr.size = " << mr.size,	271	BM_ASSERT_LOG_AND_RETURN(mr.addr != 0 && mr.size != 0, "mr.addr = " << mr.addr << ", " << "mr.size = " << mr.size,
272	BM_INVALID_PARAM);	272	BM_INVALID_PARAM);
273	if ((mr.flags & transport::REG_MR_FLAG_DRAM) == 0) {	273	if ((mr.flags & transport::REG_MR_FLAG_DRAM) == 0) {
274	- BM_LOG_WARN("Failed to register hcom mr, mem type flag should be dram.");	274	+ BM_LOG_WARN("Unable to register hcom mr, mem type flag should be dram.");
275	return BM_OK;	275	return BM_OK;
276	}	276	}
277		277
	@@ -529,7 +529,7 @@ Result HcomTransportManager::InnerReadRemote(uint32_t rankId, uint64_t lAddr, ui
529	"size = " << size << " > " << std::numeric_limits<uint32_t>::max(), BM_INVALID_PARAM);	529	"size = " << size << " > " << std::numeric_limits<uint32_t>::max(), BM_INVALID_PARAM);
530	Hcom_Channel channel = channels_[rankId];	530	Hcom_Channel channel = channels_[rankId];
531	if (channel == 0) {	531	if (channel == 0) {
532	- BM_LOG_INFO("Failed to write remote, rankId: " << rankId << " is not connect");	532	+ BM_LOG_WARN("Unable to write remote, rankId: " << rankId << " is not connect");
533	return BM_NOT_CONNECTED;	533	return BM_NOT_CONNECTED;
534	}	534	}
535	Channel_OneSideRequest req;	535	Channel_OneSideRequest req;
	@@ -569,7 +569,7 @@ Result HcomTransportManager::InnerWriteRemote(uint32_t rankId, uint64_t lAddr, u
569	"size = " << size << " > " << std::numeric_limits<uint32_t>::max(), BM_INVALID_PARAM);	569	"size = " << size << " > " << std::numeric_limits<uint32_t>::max(), BM_INVALID_PARAM);
570	Hcom_Channel channel = channels_[rankId];	570	Hcom_Channel channel = channels_[rankId];
571	if (channel == 0) {	571	if (channel == 0) {
572	- BM_LOG_INFO("Failed to write remote, rankId: " << rankId << " is not connect");	572	+ BM_LOG_WARN("Unable to write remote, rankId: " << rankId << " is not connect");
573	return BM_NOT_CONNECTED;	573	return BM_NOT_CONNECTED;
574	}	574	}
575	Channel_OneSideRequest req;	575	Channel_OneSideRequest req;
	@@ -620,7 +620,7 @@ Result HcomTransportManager::ReadRemoteAsync(uint32_t rankId, uint64_t lAddr, ui
620	BM_INVALID_PARAM);	620	BM_INVALID_PARAM);
621	Hcom_Channel channel = channels_[rankId];	621	Hcom_Channel channel = channels_[rankId];
622	if (channel == 0) {	622	if (channel == 0) {
623	- BM_LOG_INFO("Failed to write remote, rankId: " << rankId << " is not connect");	623	+ BM_LOG_WARN("Unable to write remote, rankId: " << rankId << " is not connect");
624	return BM_NOT_CONNECTED;	624	return BM_NOT_CONNECTED;
625	}	625	}
626	Channel_OneSideRequest req;	626	Channel_OneSideRequest req;
	@@ -689,7 +689,7 @@ Result HcomTransportManager::WriteRemoteAsync(uint32_t rankId, uint64_t lAddr, u
689	BM_INVALID_PARAM);	689	BM_INVALID_PARAM);
690	Hcom_Channel channel = channels_[rankId];	690	Hcom_Channel channel = channels_[rankId];
691	if (channel == 0) {	691	if (channel == 0) {
692	- BM_LOG_INFO("Failed to write remote, rankId: " << rankId << " is not connect");	692	+ BM_LOG_WARN("Unable to write remote, rankId: " << rankId << " is not connect");
693	return BM_NOT_CONNECTED;	693	return BM_NOT_CONNECTED;
694	}	694	}
695	Channel_OneSideRequest req;	695	Channel_OneSideRequest req;
	@@ -758,7 +758,7 @@ Result HcomTransportManager::WriteRemoteBatchAsync(uint32_t rankId, const CopyDe
758	BM_INVALID_PARAM);	758	BM_INVALID_PARAM);
759	Hcom_Channel channel = channels_[rankId];	759	Hcom_Channel channel = channels_[rankId];
760	if (channel == 0) {	760	if (channel == 0) {
761	- BM_LOG_INFO("Failed to write remote, rankId: " << rankId << " is not connect");	761	+ BM_LOG_WARN("Unable to write remote, rankId: " << rankId << " is not connect");
762	return BM_NOT_CONNECTED;	762	return BM_NOT_CONNECTED;
763	}	763	}
764		764
	@@ -999,7 +999,7 @@ Result HcomTransportManager::ReadRemoteBatchAsync(uint32_t rankId, const CopyDes
999	BM_INVALID_PARAM);	999	BM_INVALID_PARAM);
1000	Hcom_Channel channel = channels_[rankId];	1000	Hcom_Channel channel = channels_[rankId];
1001	if (channel == 0) {	1001	if (channel == 0) {
1002	- BM_LOG_INFO("Failed to write remote, rankId: " << rankId << " is not connect");	1002	+ BM_LOG_WARN("Unable to write remote, rankId: " << rankId << " is not connect");
1003	return BM_NOT_CONNECTED;	1003	return BM_NOT_CONNECTED;
1004	}	1004	}
1005	uint32_t allBatch = descriptor.counts.size();	1005	uint32_t allBatch = descriptor.counts.size();

Msrc/hybm/csrc/ts_engine/hybm_stream.cpp+4-4

	@@ -130,7 +130,7 @@ int32_t HybmStream::AllocSqcq(uint32_t ssid)
130		130
131	auto ret = DlHalApi::HalSqCqAllocate(deviceId_, &input, &output);	131	auto ret = DlHalApi::HalSqCqAllocate(deviceId_, &input, &output);
132	if (ret != 0) {	132	if (ret != 0) {
133	- BM_LOG_INFO("allocate sq_cq with ts_id:" << tsId_ << " failed: " << ret);	133	+ BM_LOG_ERROR("allocate sq_cq failed, ts_id: " << tsId_ << ", deviceId: " << deviceId_ << ", ret: " << ret);
134	if (stackPtr_ != nullptr) {	134	if (stackPtr_ != nullptr) {
135	DlHalApi::HalMemFree(stackPtr_);	135	DlHalApi::HalMemFree(stackPtr_);
136	stackPtr_ = nullptr;	136	stackPtr_ = nullptr;
	@@ -169,7 +169,7 @@ int32_t HybmStream::AllocLogicCq()
169		169
170	auto ret = DlHalApi::HalSqCqAllocate(deviceId_, &input, &output);	170	auto ret = DlHalApi::HalSqCqAllocate(deviceId_, &input, &output);
171	if (ret != 0) {	171	if (ret != 0) {
172	- BM_LOG_INFO("allocate logic cq with ts_id:" << tsId_ << " failed: " << ret);	172	+ BM_LOG_ERROR("allocate logic cq failed, ts_id: " << tsId_ << ", deviceId: " << deviceId_ << ", ret: " << ret);
173	return ret;	173	return ret;
174	}	174	}
175	logicCq_ = output.cqId;	175	logicCq_ = output.cqId;
	@@ -186,7 +186,7 @@ int32_t HybmStream::AllocLogicCq()
186		186
187	ret = DlHalApi::HalResourceConfig(deviceId_, &in, &configInfo);	187	ret = DlHalApi::HalResourceConfig(deviceId_, &in, &configInfo);
188	if (ret != 0) {	188	if (ret != 0) {
189	- BM_LOG_INFO("bind logic cq with ts_id:" << tsId_ << " failed: " << ret);	189	+ BM_LOG_ERROR("bind logic cq with ts_id:" << tsId_ << " failed: " << ret);
190	halSqCqFreeInfo freeInfo{};	190	halSqCqFreeInfo freeInfo{};
191	freeInfo.type = DRV_LOGIC_TYPE;	191	freeInfo.type = DRV_LOGIC_TYPE;
192	freeInfo.tsId = tsId_;	192	freeInfo.tsId = tsId_;
	@@ -213,7 +213,7 @@ void HybmStream::Destroy()
213		213
214	auto ret = DlHalApi::HalSqCqFree(deviceId_, &info);	214	auto ret = DlHalApi::HalSqCqFree(deviceId_, &info);
215	if (ret != 0) {	215	if (ret != 0) {
216	- BM_LOG_WARN("free sq_cq failed: " << ret);	216	+ BM_LOG_ERROR("free sq_cq failed: " << ret);
217	return;	217	return;
218	}	218	}
219		219

Msrc/hybm/csrc/under_api/dl_hcom_api.cpp+1-1

	@@ -81,7 +81,7 @@ Result DlHcomApi::LoadLibrary()
81	}	81	}
82	hcomHandle = dlopen(hcomLibName, RTLD_NOW \| RTLD_NODELETE);	82	hcomHandle = dlopen(hcomLibName, RTLD_NOW \| RTLD_NODELETE);
83	if (hcomHandle == nullptr) {	83	if (hcomHandle == nullptr) {
84	- BM_LOG_INFO("Failed to open library [" << hcomLibName << "], " << dlerror());	84	+ BM_LOG_WARN("Unable to open library [" << hcomLibName << "], " << dlerror());
85	return BM_DL_FUNCTION_FAILED;	85	return BM_DL_FUNCTION_FAILED;
86	}	86	}
87	DL_LOAD_SYM(gServiceCreate, serviceCreateFunc, hcomHandle, "ubs_hcom_service_create");	87	DL_LOAD_SYM(gServiceCreate, serviceCreateFunc, hcomHandle, "ubs_hcom_service_create");

Msrc/hybm/csrc/under_api/dl_hybm_copy_extend.cpp+1-1

	@@ -50,7 +50,7 @@ Result DlHybmExtendApi::TryLoadLibrary()
50	/* dlopen library */	50	/* dlopen library */
51	libHandle = dlopen(realPath.c_str(), RTLD_NOW \| RTLD_NODELETE);	51	libHandle = dlopen(realPath.c_str(), RTLD_NOW \| RTLD_NODELETE);
52	if (libHandle == nullptr) {	52	if (libHandle == nullptr) {
53	- BM_LOG_INFO("Failed to open library [" << realPath << "], " << dlerror());	53	+ BM_LOG_WARN("Unable to open library [" << realPath << "], " << dlerror());
54	return BM_DL_FUNCTION_FAILED;	54	return BM_DL_FUNCTION_FAILED;
55	}	55	}
56		56

Msrc/smem/csrc/config_store/backend/smem_etcd_client.h+5-5

	@@ -148,8 +148,8 @@ public:
148	size_t rawLen = 0;	148	size_t rawLen = 0;
149	int32_t ret = EtcdApi::EtcdGet(cli, key.data(), &rawBuf, &rawLen);	149	int32_t ret = EtcdApi::EtcdGet(cli, key.data(), &rawBuf, &rawLen);
150	if (ret != 0) {	150	if (ret != 0) {
151	- SM_LOG_WARN("GetValue failed: EtcdGet returned " << ret << ", key: " << key	151	+ SM_LOG_ERROR("GetValue failed: EtcdGet returned " << ret << ", key: " << key
152	- << ", error: " << smem::EtcdApi::EtcdGetLastError(cli));	152	+ << ", error: " << smem::EtcdApi::EtcdGetLastError(cli));
153	return -1;	153	return -1;
154	}	154	}
155	// Handle empty value case (valid scenario in etcd)	155	// Handle empty value case (valid scenario in etcd)
	@@ -175,9 +175,9 @@ public:
175	}	175	}
176	int32_t ret = EtcdApi::EtcdPrefixGet(cli, ctx, flags);	176	int32_t ret = EtcdApi::EtcdPrefixGet(cli, ctx, flags);
177	if (ret != 0) {	177	if (ret != 0) {
178	- SM_LOG_WARN("PrefixGet failed: EtcdGet returned " << ret << ", prefix: " << ctx->prefix	178	+ SM_LOG_ERROR("PrefixGet failed: EtcdGet returned " << ret << ", prefix: " << ctx->prefix
179	- << ", marker:" << ctx->marker	179	+ << ", marker:" << ctx->marker
180	- << ", error: " << smem::EtcdApi::EtcdGetLastError(cli));	180	+ << ", error: " << smem::EtcdApi::EtcdGetLastError(cli));
181	return -1;	181	return -1;
182	}	182	}
183	return 0;	183	return 0;

Msrc/smem/csrc/config_store/backend/smem_etcd_store_backend.cpp+4-4

	@@ -168,8 +168,8 @@ StoreErrorCode SmemEtcdStoreBackend::Get(const std::string &key, std::vector<uin
168	std::string valueStr;	168	std::string valueStr;
169	int32_t ret = EtcdClientV3::GetInstance().GetValue(qualifiedKey, valueStr);	169	int32_t ret = EtcdClientV3::GetInstance().GetValue(qualifiedKey, valueStr);
170	if (ret != 0) {	170	if (ret != 0) {
171	- STORE_LOG_WARN("[ETCD] Get key failed: " << qualifiedKey	171	+ STORE_LOG_ERROR("[ETCD] Get key failed: " << qualifiedKey
172	- << ", msg: " << EtcdClientV3::GetInstance().GetLastError());	172	+ << ", msg: " << EtcdClientV3::GetInstance().GetLastError());
173	return StoreErrorCode::NOT_EXIST;	173	return StoreErrorCode::NOT_EXIST;
174	}	174	}
175		175
	@@ -210,8 +210,8 @@ StoreErrorCode SmemEtcdStoreBackend::Delete(const std::string &key) noexcept
210	const std::string qualifiedKey = BuildClusterQualifiedName(clusterRoot_, key);	210	const std::string qualifiedKey = BuildClusterQualifiedName(clusterRoot_, key);
211	int32_t ret = EtcdClientV3::GetInstance().DeleteKey(qualifiedKey);	211	int32_t ret = EtcdClientV3::GetInstance().DeleteKey(qualifiedKey);
212	if (ret != 0) {	212	if (ret != 0) {
213	- STORE_LOG_WARN("[ETCD] Delete key failed: " << qualifiedKey	213	+ STORE_LOG_ERROR("[ETCD] Delete key failed: " << qualifiedKey
214	- << ", error=" << EtcdClientV3::GetInstance().GetLastError());	214	+ << ", error=" << EtcdClientV3::GetInstance().GetLastError());
215	return StoreErrorCode::ERROR;	215	return StoreErrorCode::ERROR;
216	}	216	}
217		217

Msrc/smem/csrc/config_store/ha/smem_ha_config_store.cpp+2-2

	@@ -156,7 +156,7 @@ bool HaConfigStore::IsLeaderAlive(std::string &leaderAddr) noexcept
156	{	156	{
157	int backendRet = backend_->Get(KEY_LEADER, leaderAddr);	157	int backendRet = backend_->Get(KEY_LEADER, leaderAddr);
158	if (backendRet != SUCCESS) {	158	if (backendRet != SUCCESS) {
159	- SM_LOG_WARN("Backend Get failed, key: " << KEY_LEADER << ", ret: " << backendRet);	159	+ SM_LOG_ERROR("Backend Get failed, key: " << KEY_LEADER << ", ret: " << backendRet);
160	return false;	160	return false;
161	}	161	}
162		162
	@@ -327,7 +327,7 @@ void HaConfigStore::StartServer() noexcept
327	recovered = true;	327	recovered = true;
328	}	328	}
329	} else {	329	} else {
330	- SM_LOG_WARN("Failed to parse worldSize: " << worldSizeStr);	330	+ SM_LOG_WARN("Unable to parse worldSize: " << worldSizeStr);
331	}	331	}
332	}	332	}
333	if (!recovered) {	333	if (!recovered) {

Msrc/smem/csrc/config_store/tcp_store/smem_tcp_config_store_server.cpp+4-4

	@@ -1175,8 +1175,8 @@ bool AccStoreServer::GetStatus() noexcept
1175	std::string status;	1175	std::string status;
1176	auto ret = backend_->Get(KEY_LEADER_STATUS, status);	1176	auto ret = backend_->Get(KEY_LEADER_STATUS, status);
1177	if (ret != 0) {	1177	if (ret != 0) {
1178	- STORE_LOG_WARN("Failed to get leader status from backend, error code: "	1178	+ STORE_LOG_WARN("Unable to get leader status from backend, ret: " << ret << ", attempt: " << (attempt + 1)
1179	- << ret << ", attempt: " << (attempt + 1) << "/" << (retries + 1));	1179	+ << "/" << (retries + 1));
1180	} else if (status == "true") {	1180	} else if (status == "true") {
1181	STORE_LOG_DEBUG("Leader status: active");	1181	STORE_LOG_DEBUG("Leader status: active");
1182	return true;	1182	return true;
	@@ -1187,7 +1187,7 @@ bool AccStoreServer::GetStatus() noexcept
1187	std::this_thread::sleep_for(std::chrono::seconds(retryIntervalSec));	1187	std::this_thread::sleep_for(std::chrono::seconds(retryIntervalSec));
1188	}	1188	}
1189	}	1189	}
1190	- STORE_LOG_WARN("Leader status check failed after " << (retries + 1) << " attempts");	1190	+ STORE_LOG_ERROR("Leader status check failed after " << (retries + 1) << " attempts");
1191	return false;	1191	return false;
1192	}	1192	}
1193		1193
	@@ -1280,7 +1280,7 @@ StoreErrorCode AccStoreServer::RecoverAliveRankIds(std::unordered_set<uint32_t>
1280	std::string rankStr;	1280	std::string rankStr;
1281	auto ret = backend_->Get(KEY_ALIVE_RANK_LIST, rankStr);	1281	auto ret = backend_->Get(KEY_ALIVE_RANK_LIST, rankStr);
1282	if (ret != 0) {	1282	if (ret != 0) {
1283	- STORE_LOG_WARN("Failed to get alive ranks from backend, error code: " << ret);	1283	+ STORE_LOG_WARN("Unable to get alive ranks from backend, ret: " << ret);
1284	return SUCCESS;	1284	return SUCCESS;
1285	}	1285	}
1286	if (rankStr.empty()) {	1286	if (rankStr.empty()) {

Msrc/smem/csrc/net/smem_net_group_engine.cpp+12-12

	@@ -392,7 +392,7 @@ Result SmemNetGroupEngine::RegisterExit(const std::function<void(int)> &exit)
392	{	392	{
393	SM_ASSERT_RETURN(!option_.dynamic, SM_ERROR);	393	SM_ASSERT_RETURN(!option_.dynamic, SM_ERROR);
394	if (globalExitHandler_ != nullptr) {	394	if (globalExitHandler_ != nullptr) {
395	- SM_LOG_WARN("the exit function is not null");	395	+ SM_LOG_ERROR("the exit function is not null");
396	return SM_INVALID_PARAM;	396	return SM_INVALID_PARAM;
397	}	397	}
398	SM_ASSERT_RETURN(exit != nullptr, SM_INVALID_PARAM);	398	SM_ASSERT_RETURN(exit != nullptr, SM_INVALID_PARAM);
	@@ -404,7 +404,7 @@ Result SmemNetGroupEngine::RegisterExit(const std::function<void(int)> &exit)
404	std::placeholders::_2, std::placeholders::_3),	404	std::placeholders::_2, std::placeholders::_3),
405	wid);	405	wid);
406	if (ret != SM_OK) {	406	if (ret != SM_OK) {
407	- SM_LOG_WARN("group watch failed, maybe link down, ret: " << ret);	407	+ SM_LOG_ERROR("group watch failed, maybe link down, ret: " << ret);
408	globalExitHandler_ = nullptr;	408	globalExitHandler_ = nullptr;
409	return ret;	409	return ret;
410	}	410	}
	@@ -422,17 +422,17 @@ void SmemNetGroupEngine::RankExit(int result, const std::string &key, const std:
422	try {	422	try {
423	long tempVal = std::stol(value);	423	long tempVal = std::stol(value);
424	if (tempVal < std::numeric_limits<int>::min() \|\| tempVal > std::numeric_limits<int>::max()) {	424	if (tempVal < std::numeric_limits<int>::min() \|\| tempVal > std::numeric_limits<int>::max()) {
425	- SM_LOG_WARN("value out of int range: " << tempVal << " (value='" << value << "')");	425	+ SM_LOG_ERROR("value out of int range: " << tempVal << " (value='" << value << "')");
426	return;	426	return;
427	}	427	}
428	val = static_cast<int>(tempVal);	428	val = static_cast<int>(tempVal);
429	} catch (...) {	429	} catch (...) {
430	- SM_LOG_WARN("convert string to int failed");	430	+ SM_LOG_ERROR("convert string to int failed");
431	return;	431	return;
432	}	432	}
433	globalExitHandler_(val);	433	globalExitHandler_(val);
434	} else {	434	} else {
435	- SM_LOG_WARN("global exit failed");	435	+ SM_LOG_ERROR("global exit failed");
436	}	436	}
437	}	437	}
438		438
	@@ -655,7 +655,7 @@ int32_t SmemNetGroupEngine::AllocNumber()
655	Result SmemNetGroupEngine::ReleaseNumber(int32_t val)	655	Result SmemNetGroupEngine::ReleaseNumber(int32_t val)
656	{	656	{
657	if (allocedSet_.count(val) == 0) {	657	if (allocedSet_.count(val) == 0) {
658	- SM_LOG_WARN("key(" << val << ") is not exist!");	658	+ SM_LOG_ERROR("key(" << val << ") is not exist!");
659	return SM_OBJECT_NOT_EXISTS;	659	return SM_OBJECT_NOT_EXISTS;
660	}	660	}
661	allocedSet_.erase(val);	661	allocedSet_.erase(val);
	@@ -694,8 +694,8 @@ Result SmemNetGroupEngine::TryRemovePrefixKey(uint32_t rank)
694	prefixKey_.clear();	694	prefixKey_.clear();
695	}	695	}
696	} else {	696	} else {
697	- SM_LOG_INFO("remove key failed, src_rank:" << option_.rank << " key:" << store_->GetCompleteKey(key)	697	+ SM_LOG_WARN("remove key not successful, src_rank:" << option_.rank << " key:" << store_->GetCompleteKey(key)
698	- << " ret:" << ret);	698	+ << " ret:" << ret);
699	}	699	}
700	return ret;	700	return ret;
701	}	701	}
	@@ -837,7 +837,7 @@ Result SmemNetGroupEngine::GroupGatherPrefixKey(uint32_t dstRank, std::string &u
837	std::string ownKey = SMEM_EXCHANGE_INFO_KEY + std::to_string(ownRank);	837	std::string ownKey = SMEM_EXCHANGE_INFO_KEY + std::to_string(ownRank);
838	auto writeRet = store_->Set(ownKey, update);	838	auto writeRet = store_->Set(ownKey, update);
839	if (writeRet != SM_OK) {	839	if (writeRet != SM_OK) {
840	- SM_LOG_WARN("set own prefix_key: " << store_->GetCompleteKey(ownKey) << " failed, ret:" << writeRet);	840	+ SM_LOG_ERROR("set own prefix_key: " << store_->GetCompleteKey(ownKey) << " failed, ret:" << writeRet);
841	}	841	}
842	}	842	}
843	std::string key = SMEM_EXCHANGE_INFO_KEY + std::to_string(dstRank);	843	std::string key = SMEM_EXCHANGE_INFO_KEY + std::to_string(dstRank);
	@@ -885,7 +885,7 @@ uint32_t SmemNetGroupEngine::ReWatchEvent()
885	std::placeholders::_2, std::placeholders::_3),	885	std::placeholders::_2, std::placeholders::_3),
886	wid);	886	wid);
887	if (ret != SM_OK \|\| wid == UINT32_MAX) {	887	if (ret != SM_OK \|\| wid == UINT32_MAX) {
888	- SM_LOG_WARN_LIMIT("group watch failed, ret: " << ret << ", wid: " << wid);	888	+ SM_LOG_ERROR("group watch failed, ret: " << ret << ", wid: " << wid);
889	usleep(SMEM_GROUP_SLEEP_TIMEOUT);	889	usleep(SMEM_GROUP_SLEEP_TIMEOUT);
890	return UINT32_MAX;	890	return UINT32_MAX;
891	}	891	}
	@@ -900,7 +900,7 @@ uint32_t SmemNetGroupEngine::ReWatchLinkDown()
900	WatchRankType::WATCH_RANK_LINK_DOWN,	900	WatchRankType::WATCH_RANK_LINK_DOWN,
901	[this](WatchRankType type, uint32_t downRankId) { RemoteRankLinkDownCb(downRankId); }, wid);	901	[this](WatchRankType type, uint32_t downRankId) { RemoteRankLinkDownCb(downRankId); }, wid);
902	if (ret != SM_OK \|\| wid == UINT32_MAX) {	902	if (ret != SM_OK \|\| wid == UINT32_MAX) {
903	- SM_LOG_WARN_LIMIT("group watch failed, ret: " << ret);	903	+ SM_LOG_ERROR("group watch failed, ret: " << ret);
904	usleep(SMEM_GROUP_SLEEP_TIMEOUT);	904	usleep(SMEM_GROUP_SLEEP_TIMEOUT);
905	return UINT32_MAX;	905	return UINT32_MAX;
906	}	906	}
	@@ -1204,7 +1204,7 @@ void SmemNetGroupEngine::GroupWatchCb(int result, const std::string &key, const
1204	}	1204	}
1205		1205
1206	if (value.length() != SMEM_GROUP_INFO_SIZE) {	1206	if (value.length() != SMEM_GROUP_INFO_SIZE) {
1207	- SM_LOG_WARN("receive group info size is error!");	1207	+ SM_LOG_WARN("received group info size is incorrect.");
1208	ctxRet = SM_ERROR;	1208	ctxRet = SM_ERROR;
1209	}	1209	}
1210		1210

Msrc/smem/csrc/smem_bm/smem_bm.cpp+2-1

	@@ -86,7 +86,8 @@ SMEM_API int32_t smem_bm_init(const char *storeURL, uint32_t worldSize, uint16_t
86		86
87	auto faultInjectionPointStatus = ::ock::mf::FaultInjectionPointRegistry::Register();	87	auto faultInjectionPointStatus = ::ock::mf::FaultInjectionPointRegistry::Register();
88	if (faultInjectionPointStatus != ::ock::mf::FaultInjectionPointStatus::OK) {	88	if (faultInjectionPointStatus != ::ock::mf::FaultInjectionPointStatus::OK) {
89	- SM_LOG_WARN("register fault injection points failed, status: " << static_cast<int>(faultInjectionPointStatus));	89	+ SM_LOG_WARN(
		90	+ "unable to register fault injection points, status: " << static_cast<int>(faultInjectionPointStatus));
90	}	91	}
91		92
92	int32_t ret = SmemBmEntryManager::Instance().Initialize(storeURL, worldSize, deviceId, *config);	93	int32_t ret = SmemBmEntryManager::Instance().Initialize(storeURL, worldSize, deviceId, *config);

Msrc/smem/csrc/smem_bm/smem_bm_entry.cpp+1-1

	@@ -224,7 +224,7 @@ void SmemBmEntry::UnInitalize()
224	if (globalGroup_ != nullptr && globalGroup_->IsJoined()) {	224	if (globalGroup_ != nullptr && globalGroup_->IsJoined()) {
225	auto ret = globalGroup_->GroupLeave();	225	auto ret = globalGroup_->GroupLeave();
226	if (ret != SM_OK) {	226	if (ret != SM_OK) {
227	- SM_LOG_WARN("group leave failed during uninitialize, ret: " << ret);	227	+ SM_LOG_WARN("unable to group leave during uninitialize, ret: " << ret);
228	}	228	}
229	}	229	}
230	// Stop the group engine listen thread before releasing the entity.	230	// Stop the group engine listen thread before releasing the entity.

Msrc/smem/csrc/smem_bm/smem_bm_entry_manager.cpp+1-1

	@@ -173,7 +173,7 @@ Result SmemBmEntryManager::CreateEntryById(uint32_t id, SmemBmEntryPtr &entry /*
173	SM_ASSERT_RETURN(inited_, SM_NOT_STARTED);	173	SM_ASSERT_RETURN(inited_, SM_NOT_STARTED);
174	auto iter = entryIdMap_.find(id);	174	auto iter = entryIdMap_.find(id);
175	if (iter != entryIdMap_.end()) {	175	if (iter != entryIdMap_.end()) {
176	- SM_LOG_WARN("create bm entry failed as already exists, id: " << id);	176	+ SM_LOG_WARN("unable to create bm entry as already exists, id: " << id);
177	return SM_DUPLICATED_OBJECT;	177	return SM_DUPLICATED_OBJECT;
178	}	178	}
179		179

Msrc/smem/csrc/smem_shm/smem_shm_entry.cpp+2-2

	@@ -183,7 +183,7 @@ void SmemShmEntry::InitStepUnreserveMemory()
183	{	183	{
184	auto ret = hybm_unreserve_mem_space(entity_, 0);	184	auto ret = hybm_unreserve_mem_space(entity_, 0);
185	if (ret != 0) {	185	if (ret != 0) {
186	- SM_LOG_WARN("unreserve mem space failed: " << ret);	186	+ SM_LOG_WARN("unable to unreserve mem space: " << ret);
187	}	187	}
188	gva_ = nullptr;	188	gva_ = nullptr;
189	}	189	}
	@@ -205,7 +205,7 @@ void SmemShmEntry::InitStepFreeSlice()
205	auto slice = slice_;	205	auto slice = slice_;
206	auto ret = hybm_free_local_memory(entity_, slice, 0, 0);	206	auto ret = hybm_free_local_memory(entity_, slice, 0, 0);
207	if (ret != 0) {	207	if (ret != 0) {
208	- SM_LOG_WARN("free mem slice failed: " << ret);	208	+ SM_LOG_WARN("unable to free mem slice: " << ret);
209	}	209	}
210	slice_ = nullptr;	210	slice_ = nullptr;
211	}	211	}

Msrc/smem/csrc/smem_shm/smem_shm_entry_manager.cpp+1-1

	@@ -63,7 +63,7 @@ Result SmemShmEntryManager::CreateEntryById(uint32_t id, SmemShmEntryPtr &entry
63	SM_ASSERT_RETURN(inited_, SM_NOT_STARTED);	63	SM_ASSERT_RETURN(inited_, SM_NOT_STARTED);
64	auto iter = entryIdMap_.find(id);	64	auto iter = entryIdMap_.find(id);
65	if (iter != entryIdMap_.end()) {	65	if (iter != entryIdMap_.end()) {
66	- SM_LOG_WARN("create shm entry failed as already exists, id: " << id);	66	+ SM_LOG_WARN("unable to create shm entry as already exists, id: " << id);
67	return SM_DUPLICATED_OBJECT;	67	return SM_DUPLICATED_OBJECT;
68	}	68	}
69		69

	@@ -367,7 +367,7 @@ inline Result AccTcpLinkComplexDefault::HandlePollIn() noexcept
367	/* if errno is eagain is normal, need to continue to receive */	367	/* if errno is eagain is normal, need to continue to receive */
368	/* else meaning failed to read from socket, socket is error */	368	/* else meaning failed to read from socket, socket is error */
369	if (errorNumber != EAGAIN) {	369	if (errorNumber != EAGAIN) {
370	- LOG_WARN("Link " << id_ << " receive header failed, errno " << errorNumber);	370	+ LOG_WARN("Link " << id_ << " receive header not successful, reason " << errorNumber);
371	}	371	}
372		372
373	return (errorNumber == EAGAIN ? ACC_LINK_EAGAIN : ACC_LINK_ERROR);	373	return (errorNumber == EAGAIN ? ACC_LINK_EAGAIN : ACC_LINK_ERROR);
	@@ -390,13 +390,13 @@ inline Result AccTcpLinkComplexDefault::HandlePollIn() noexcept
390	} else { /* ECONNRESET is broken during io, SUCCESS is broken during idle time. */	390	} else { /* ECONNRESET is broken during io, SUCCESS is broken during idle time. */
391	const auto errorNumber = errno; // avoid errno writed by log	391	const auto errorNumber = errno; // avoid errno writed by log
392	if (errorNumber == ECONNRESET \|\| errorNumber == 0) {	392	if (errorNumber == ECONNRESET \|\| errorNumber == 0) {
393	- LOG_INFO("Link " << id_ << " receive body failed, reset by peer, errno " << errorNumber);	393	+ LOG_DEBUG("Link " << id_ << " receive body failed, reset by peer, errno " << errorNumber);
394	return ACC_LINK_ERROR; /* socket is closed by peer, socket is error */	394	return ACC_LINK_ERROR; /* socket is closed by peer, socket is error */
395	}	395	}
396	/* if errno is eagain is normal, need to continue to receive */	396	/* if errno is eagain is normal, need to continue to receive */
397	/* else meaning failed to read from socket,socket is error */	397	/* else meaning failed to read from socket,socket is error */
398	if (errorNumber != EAGAIN) {	398	if (errorNumber != EAGAIN) {
399	- LOG_ERROR("Link " << id_ << " receive body failed, errno " << errorNumber);	399	+ LOG_WARN("Link " << id_ << " receive body not successful, reason " << errorNumber);
400	}	400	}
401		401
402	return (errorNumber == EAGAIN ? ACC_LINK_EAGAIN : ACC_LINK_ERROR);	402	return (errorNumber == EAGAIN ? ACC_LINK_EAGAIN : ACC_LINK_ERROR);

	@@ -437,14 +437,14 @@ void HybmConnBasedSegment::AllocMemory(void sliceAddr, uint64_t lvOffset, uint
437	allocMethod = MemAllocMethod::MMAP;	437	allocMethod = MemAllocMethod::MMAP;
438	return mapped;	438	return mapped;
439	}	439	}
440	- BM_LOG_WARN("Failed to alloc size:" << size << " with hugepage via mmap, error: " << errno << ", "	440	+ BM_LOG_WARN("Unable to alloc size:" << size << " with hugepage via mmap, error: " << errno << ", "
441	<< SafeStrError(errno)	441	<< SafeStrError(errno)
442	<< ". Use 'grep -i huge /proc/meminfo' to check hugepages, "	442	<< ". Use 'grep -i huge /proc/meminfo' to check hugepages, "
443	"and use 'echo <page_num> > /proc/sys/vm/nr_hugepages' to set hugepages.");	443	"and use 'echo <page_num> > /proc/sys/vm/nr_hugepages' to set hugepages.");
444		444
445	// 2. try to alloc DRAM with hugepage via halMemAlloc	445	// 2. try to alloc DRAM with hugepage via halMemAlloc
446	if (options_.enable56BitsGva && options_.shmFd < 0) {	446	if (options_.enable56BitsGva && options_.shmFd < 0) {
447	- BM_LOG_WARN("Trying halMemAlloc for DRAM hugepage allocation. " << "size:" << size);	447	+ BM_LOG_WARN("Trying halMemAlloc for DRAM hugepage allocation. size:" << size);
448		448
449	// Use halMemAlloc to allocate DRAM huge page memory on host	449	// Use halMemAlloc to allocate DRAM huge page memory on host
450	// Flag: MEM_HOST (host memory) \| MEM_TYPE_DDR (DDR/DRAM) \| MEM_PAGE_HUGE (2MB huge page)	450	// Flag: MEM_HOST (host memory) \| MEM_TYPE_DDR (DDR/DRAM) \| MEM_PAGE_HUGE (2MB huge page)
	@@ -457,8 +457,8 @@ void HybmConnBasedSegment::AllocMemory(void sliceAddr, uint64_t lvOffset, uint
457		457
458	int ret = DlHalApi::HalMemAlloc(&halAllocPtr, size, allocFlag);	458	int ret = DlHalApi::HalMemAlloc(&halAllocPtr, size, allocFlag);
459	if (ret != 0 \|\| halAllocPtr == nullptr) {	459	if (ret != 0 \|\| halAllocPtr == nullptr) {
460	- BM_LOG_WARN("halMemAlloc failed, ret:" << ret << " ptr:" << halAllocPtr << ". Cannot allocate " << size	460	+ BM_LOG_WARN("halMemAlloc not successful, ret:" << ret << " ptr:" << halAllocPtr << ". Cannot allocate "
461	- << " bytes DRAM huge page memory");	461	+ << size << " bytes DRAM huge page memory");
462	} else {	462	} else {
463	allocMethod = MemAllocMethod::HAL_MEM_ALLOC;	463	allocMethod = MemAllocMethod::HAL_MEM_ALLOC;
464	BM_LOG_INFO("Successfully allocated DRAM hugepage via halMemAlloc. "	464	BM_LOG_INFO("Successfully allocated DRAM hugepage via halMemAlloc. "

	@@ -123,7 +123,7 @@ void HybmVaManager::RemoveOneVaInfo(uint64_t va, uint32_t type)
123	std::unique_lock<std::shared_mutex> lock(mutex_);	123	std::unique_lock<std::shared_mutex> lock(mutex_);
124	auto it = allocatedMap_[type].find(va);	124	auto it = allocatedMap_[type].find(va);
125	if (it == allocatedMap_[type].end()) {	125	if (it == allocatedMap_[type].end()) {
126	- BM_LOG_WARN("RemoveOneVaInfo failed: address not found. va=" << VaToStr(va) << " type=" << type);	126	+ BM_LOG_WARN("Unable to RemoveOneVaInfo: address not found. va=" << VaToStr(va) << " type=" << type);
127	return;	127	return;
128	}	128	}
129	const AllocatedGvaInfo &info = it->second;	129	const AllocatedGvaInfo &info = it->second;
	@@ -434,13 +434,13 @@ uint64_t HybmVaManager::AllocReserveLvaInner(uint32_t localRankId, uint64_t size
434	void HybmVaManager::FreeReserveGva(uint64_t addr)	434	void HybmVaManager::FreeReserveGva(uint64_t addr)
435	{	435	{
436	if (addr == 0) {	436	if (addr == 0) {
437	- BM_LOG_WARN("FreeReserveGva failed: invalid addr=0");	437	+ BM_LOG_WARN("Unable to FreeReserveGva: invalid addr=0");
438	return;	438	return;
439	}	439	}
440	std::unique_lock<std::shared_mutex> lock(mutex_);	440	std::unique_lock<std::shared_mutex> lock(mutex_);
441	auto it = reservedMap_[HVM_GVA].find(addr);	441	auto it = reservedMap_[HVM_GVA].find(addr);
442	if (it == reservedMap_[HVM_GVA].end()) {	442	if (it == reservedMap_[HVM_GVA].end()) {
443	- BM_LOG_WARN("FreeReserveGva failed: reserved space not found at addr=" << VaToStr(addr));	443	+ BM_LOG_WARN("Unable to FreeReserveGva: reserved space not found at addr=" << VaToStr(addr));
444	return;	444	return;
445	}	445	}
446	const ReservedGvaInfo &info = it->second;	446	const ReservedGvaInfo &info = it->second;
	@@ -454,7 +454,7 @@ void HybmVaManager::FreeReserveGva(uint64_t addr)
454	void HybmVaManager::FreeReserveLva(uint64_t addr, uint32_t type)	454	void HybmVaManager::FreeReserveLva(uint64_t addr, uint32_t type)
455	{	455	{
456	if (addr == 0 \|\| (type != HVM_DVA && type != HVM_HVA)) {	456	if (addr == 0 \|\| (type != HVM_DVA && type != HVM_HVA)) {
457	- BM_LOG_WARN("FreeReserveLva failed: invalid addr:" << addr << ", type:" << type);	457	+ BM_LOG_WARN("Unable to FreeReserveLva: invalid addr:" << addr << ", type:" << type);
458	return;	458	return;
459	}	459	}
460	std::unique_lock<std::shared_mutex> lock(mutex_);	460	std::unique_lock<std::shared_mutex> lock(mutex_);

	@@ -29,9 +29,9 @@ HybmVmmBasedSegment::~HybmVmmBasedSegment()
29		29
30	auto ret = UnReserveMemorySpace();	30	auto ret = UnReserveMemorySpace();
31	if (ret != BM_OK) {	31	if (ret != BM_OK) {
32	- BM_LOG_WARN("Destructor cleanup failed, ret:" << ret	32	+ BM_LOG_WARN("Destructor unable to cleanup, ret:" << ret
33	- << " gva:" << reinterpret_cast<void *>(globalVirtualAddress_)	33	+ << " gva:" << reinterpret_cast<void *>(globalVirtualAddress_)
34	- << " va:" << reinterpret_cast<void *>(localVirtualAddress_));	34	+ << " va:" << reinterpret_cast<void *>(localVirtualAddress_));
35	}	35	}
36	}	36	}
37		37
	@@ -110,7 +110,7 @@ Result HybmVmmBasedSegment::UnReserveMemorySpace() noexcept
110	auto slice = slices_.begin()->second.slice;	110	auto slice = slices_.begin()->second.slice;
111	auto ret = ReleaseSliceMemory(slice);	111	auto ret = ReleaseSliceMemory(slice);
112	if (ret != BM_OK) {	112	if (ret != BM_OK) {
113	- BM_LOG_WARN("ReleaseSliceMemory failed during unreserve, ret:" << ret << " slice:" << slice);	113	+ BM_LOG_WARN("Unable to ReleaseSliceMemory during unreserve, ret:" << ret << " slice:" << slice);
114	}	114	}
115	}	115	}
116	while (!registerSlices_.empty()) {	116	while (!registerSlices_.empty()) {
	@@ -123,7 +123,7 @@ Result HybmVmmBasedSegment::UnReserveMemorySpace() noexcept
123	auto ret = DlHalApi::HalMemAddressFree(lva);	123	auto ret = DlHalApi::HalMemAddressFree(lva);
124	BM_LOG_INFO("free reserved address va:" << lva << " return:" << ret);	124	BM_LOG_INFO("free reserved address va:" << lva << " return:" << ret);
125	if (ret != BM_OK) {	125	if (ret != BM_OK) {
126	- BM_LOG_WARN("HalMemAddressFree failed, keep reserved VA state. ret:"	126	+ BM_LOG_WARN("Unable to HalMemAddressFree, keep reserved VA state. ret:"
127	<< ret << " gva:" << reinterpret_cast<void *>(globalVirtualAddress_) << " va:" << lva);	127	<< ret << " gva:" << reinterpret_cast<void *>(globalVirtualAddress_) << " va:" << lva);
128	}	128	}
129	}	129	}
	@@ -187,7 +187,7 @@ Result HybmVmmBasedSegment::MallocFromHost(size_t size, uint32_t devId, drv_mem_
187	}	187	}
188	int32_t ret = HalMemCreateAdapterFromHost(size, handle, prop);	188	int32_t ret = HalMemCreateAdapterFromHost(size, handle, prop);
189	if (ret == HAL_OUT_OF_MEMORY_ERROR && prop.pg_type == MEM_GIANT_PAGE_TYPE) {	189	if (ret == HAL_OUT_OF_MEMORY_ERROR && prop.pg_type == MEM_GIANT_PAGE_TYPE) {
190	- BM_LOG_INFO("Try HalMemCreate use 1GB page failed, ret:" << ret << ", than try 2MB page");	190	+ BM_LOG_WARN("Try HalMemCreate use 1GB page not successful, ret:" << ret << ", than try 2MB page");
191	prop.pg_type = MEM_HUGE_PAGE_TYPE;	191	prop.pg_type = MEM_HUGE_PAGE_TYPE;
192	ret = HalMemCreateAdapterFromHost(size, handle, prop);	192	ret = HalMemCreateAdapterFromHost(size, handle, prop);
193	}	193	}