On Wed, Jun 28, 2023 at 10:33:24AM +0100, David Howells wrote: > Would you consider it a kernel bug, then, if you use sendmsg(MSG_ZEROCOPY) to > send some data from a file mmapping that some other userspace then corrupts by > altering the file before the kernel has managed to send it? I think there's a difference in that sendmsg() will block until it returns (... right?) splice() returns. That implies the copy is done. Then the same thread modifies the file, but the pipe sees the new data, not the old. Doesn't that feel like a bug to you?
