Tv

Yup, vmsplice can't do what was thought of here. io_uring could do something similar, but that depends on so much Go infrastructure that it's not really up to us.

Workaround in place: https://github.com/bazil/fuse/commit/7084418246670b17810b8af332b63cdf13440c9b

ENOSYS would mean kernel stops sending Getxattr requests. Putting that in as the default would effectively stop you from implementing xattrs on some nodes. I don't think this should be...

Yeah, I just don't see this as desirable. The typical implementation has different types for different kinds of nodes (dir/file, application-specific types). One of them not implementing Getxattr is not...

`resp.Data` cannot be nil: ``` s := &fuse.ReadResponse{Data: make([]byte, 0, r.Size)} ``` Where do you see a panic in HandleRead? It's routinely called with `data` larger than `resp.Data`.

Not enough information given to help you -> closing this issue. Also, the kernel decides how much data it wants, you can't return more then `ReadRequest.Size` bytes.

In general, I think limiting workers might be a very good idea. However, I don't see how `cp -a` would be able to trigger unlimited handlers. Can you share more...

I have a largish rewrite in progress that does this among many other things. Currently figuring out last bits of the new API.

The current version has pretty miserable memory usage, it'll lose benchmarks pretty easily. There's a bigger refactor under way, but I haven't gotten back to it in a long while....

Serving requests "inline" would eliminate a huge class of useful filesystems, you couldn't do any long-running processing to construct the response. My primary use cases fetches data over the network,...