As discussed in the thread https://lists.freebsd.org/pipermail/freebsd-stable/2014-November/081013.html and described by Steven Hartland in https://lists.freebsd.org/pipermail/freebsd-stable/2014-November/081033.html after freebsd-update aacraid no longer provides access to raw drives, but only passX. Reverting a small part of r257847 seems to fix the issue. Should this be in errata for 10.1-RELEASE ?
I do believe that the patch http://codelabs.ru/fbsd/patches/aacraid/aacraid-fix-device-hints-and-explain-raw-devices.diff should add some proper housekeeping to the problem of device hints for aacraid both code- and manual-wise. It will also make default hint files to enable historical 10.0 behaviour for exposing da(4) devices. It will both be good to have some testing from people who really own hardware for aacraid (don't have such systems at hand, only with old cards that are driven by aac) and review/feedback from maintainers. But for me the modified code at least compiles GENERIC on amd64. Thanks!
Right now I still need a cable to test, waiting for it...
I was caught out by this issue last week when upgrading a customer server with an Adaptec 51645 controller. Shouldn't this be in the errata or at least in the bugs section of the aac(4) and aacraid(4) man pages? The Adaptec 51645 in the server is on the supported hardware list at https://www.freebsd.org/releases/10.1R/hardware.html and worked OK in 10.0. I understand that using these controllers isn't ideal, but if they're on the supported hardware list they should work. I'm happy to test a patch with the Adaptec 51645 if that's helpful.
For bugs matching the following conditions: - Status == In Progress - Assignee == "bugs@FreeBSD.org" - Last Modified Year <= 2017 Do - Set Status to "Open"