yfismine

> what is your local OS version? and please check the glibc version as well. python: 3.12.0 os: centos8 lld: ldd (GNU libc) 2.28 ```shell conda install pytorch=2.3.0 cpuonly torchmetrics=1.4.0...

> What's your use case of using to_block()? DGLGraph 

> to_block is to create a bipartite-structured block for message passing. It is more an internal util method than public api of dgl, we expose it just in case users...

> I don't think it's a bug. > > During partition, we make sure all the edges of inner_node are partitioned into current partition. and these edges are called inner...

I used the provided pairtion_graph function to test the segmented subgraph. It is very easy to find that some points are inner_node, but not all the edges directly connected with...

> @Rhett-Ying Why don't we import GraphBolt inside the functions where it is used so that normal DGL users can import DGL. Or the distributed code should not be imported...

Although there are many strange compatibility problems, the version I compiled through the source code can work normally in arm. Thank you very much. @Rhett-Ying

Yes, I know there are ways to avoid this problem, but I think as a framework itself, there should be no assumptions about the user's source data schema. @rudongyu