Merge branch 'main' into move-oci

Author: mlefeb01

cpp/tensorrt_llm/batch_manager/cacheFormatter.cpp

@@ -761,10 +761,11 @@ void CacheFormatter::unformat(tensorrt_llm::batch_manager::TransferSession& sess
                 {
                     cacheBufferId = mCacheTransBufferManager->assignBufferIndexForRecv();
                 }
-                TLLM_CHECK(cacheBufferId.has_value());
                 auto [recvSplitCachestmp, bufferCoverTargetNumtmp, onlyUseDynamicBuffer]
                     = mCacheTransBufferManager->getOrAllocateRecvBuffers(
                         cacheBufferId, static_cast<int>(targetNum), bufferEleSizes, bufferManager);
+                TLLM_CHECK(cacheBufferId.has_value() || onlyUseDynamicBuffer);
+
                 bufferCoverTargetNum = bufferCoverTargetNumtmp;
                 remainNoCoverTargetNum = targetNum > bufferCoverTargetNum ? targetNum - bufferCoverTargetNum : 0;
 

cpp/tensorrt_llm/batch_manager/cacheTransBuffer.cpp

@@ -394,10 +394,11 @@ std::tuple<std::vector<runtime::ITensor::SharedPtr>, size_t, bool> CacheTransBuf
         {
             TLLM_LOG_WARNING(
                 "CacheTransceiver getOrAllocateBuffers: bufferCoverTargetNum:%d < targetNum:%d, may use dynamic "
-                "buffer, "
-                "it's better to increase MaxTokensInBuffer in cacheTransceiverConfig, otherwise, the performance may "
-                "be degraded, requestedNumberOfElements.size():%ld, mNumberOfElements:%ld, "
-                "requestedNumberOfElements[0]:%ld",
+                "buffer which will fail with NIXL backend. It is recommended to set "
+                "cacheTransceiverConfig.MaxTokensInBuffer (cache_transceiver_config.max_tokens_in_buffer in config "
+                "YAML file) to a value greater than the maximum ISL of the processed requests. Otherwise, performance "
+                "may be degraded or transfer may fail.  requestedNumberOfElements.size():%ld, "
+                "mNumberOfElements:%ld, requestedNumberOfElements[0]:%ld",
                 bufferCoverTargetNum, targetNum, requestedNumberOfElements.size(), mNumberOfElements,
                 requestedNumberOfElements[0]);
         }