https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=285415
--- Comment #9 from Mark Johnston <ma...@freebsd.org> --- (In reply to Dimitry Andric from comment #8) Here the assertion seems to be complaining about the bounce buffer length not being a multiple of the page size. The assertions in storvsc_xferbuf_prepare() are disabled on arm64 in commits 6f7b1310b6fe3 and 0af5a0cd2788e with no real explanation given, so I somewhat suspect that the crash is unrelated. (Actually I am not sure how busdma can be forced to make each segment a multiple of the page size, but this certainly appears to be a requirement of the driver, since the callback does not store the segment length anywhere...) -- You are receiving this mail because: You are the assignee for the bug.
