Re: Use of floppy disks and parallel ATA on amd64

[Thor Lancelot Simon <tls%panix.com@localhost>]

On Tue, Oct 15, 2024 at 07:04:25AM -0700, Jason Thorpe wrote:
> 
> Problems here include:
> 
> 1. No way to use this from other I/O paths, like file system I/O ??? there are lots of other random limits that get involved there (e.g. UBC_WINSIZE), none of which are directly related to the capabilities of the underlying device.
> 
> 2. No way to consider limits further up the chain.  The capabilities of the end block device are only just one aspect.  You need to consider limits of e.g. an intervening IOMMU, or maybe a buggy revision of a PCI bridge.  Right now, we have ???leaf node??? and ???root node??? (platform constraints - hello sun2!).
> 
> 3. There???s probably more, but I???m still on my first cup of coffee.

FWIW, the old tls-maxphys branch largely addresses these issues, albeit not
elegantly in all cases.  I don't have time to resuscitate it; though jak@
just found one of the stumpers that originally stalled me out years ago
(line 431 of ufs_readwrite.c, which clamps writes to 64K).  The other thing
we would confront if we removed the MAXPHYS limit is over-aggressive
readahead when the (correctly propagated, per-device) max-io-size constraint
is very large; nothing but MAXPHYS really holds that back now.

Thor