Compatibility update for 5.14 for blk_alloc_disk()

[ckane]

In Linux 5.14, blk_alloc_queue is no longer exported, and its usage
has been superseded by blk_alloc_disk, which returns a gendisk struct
from which we can still retrieve the struct request_queue* that is
needed in the one place where it is used. This also replaces the call
to alloc_disk(minors), and minors is now set via struct member
assignment.

Motivation and Context

This API change results in a failure to compile on 5.14rc1

Description

How Has This Been Tested?

Tested on my system by booting on kernel 5.14. Unfortunately, wasn't able to test extensively because 5.14 appears to introduce a amdgpu regression that causes drm/Xorg to crash. That said, booted multiple times and tried a few tests from cmdline and it seems to work.

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Performance enhancement (non-breaking change which improves efficiency)
• Code cleanup (non-breaking change which makes code smaller or more readable)
• Breaking change (fix or feature that would cause existing functionality to change)
• Library ABI change (libzfs, libzfs_core, libnvpair, libuutil and libzfsbootenv)
• Documentation (a change to man pages or other documentation)

Checklist:

• My code follows the OpenZFS code style requirements.
• I have updated the documentation accordingly.
• I have read the contributing document.
• I have added tests to cover my changes.
• I have run the ZFS Test Suite with this change applied.
• All commit messages are properly formatted and contain Signed-off-by.

[behlendorf]

@ckane thanks for digging in to this! I put together a very similar patch but ran in problems booting 5.14-rc1 as well so I wasn't able to test it. It looks like we basically have the same fix I just did a little more restructuring to condense the #if conditionals for readability and put the configure test in config/kernel-make-request-fn.m4 instead. It'd be great if you could take a look and let me know what you think of this alternate version. Then we can decide how to proceed.

[ckane]

Thanks @behlendorf - I'll give yours a spin on my system after I go through this bisect exercise on 5.14rc1 to get an idea about what's breaking this kernel for me. I was able to boot to console (I normally get into X11 using startx anyhow), just couldn't go graphical. I think your approach is more complete, so if it works for me I'll just recommend going with your solution.

[ckane]

Ok @behlendorf - I managed to test yours today and it seems to work identically to my version on 5.14, so looks good to me. I don't know what the boot issue is on 5.14rc1 but it sounds like it isn't just me and isn't related to this at all. As well, I tested against code merged for 5.14rc2, works the same, and kernel is still unresponsive for me when I try to use Xorg.

[behlendorf]

@ckane I've gone ahead and opened up PR #12409 with the alternate version of this fix. It should be identical to the one you tested. @ckane @pcd1193182 if you could both re-review it I think we should be able to move forward with getting this kernel compatibility issue resolved.