Re: [PATCH v8 1/2] misc: ibmasm: Fix static out-of-bounds MMIO access during probe

[Mingyu Wang]

Hi Greg,


> I do not know.  Do some research into when the hardware was made, what
> it runs on, and try to determine if it really is even used anymore by
> anyone.  If not, then yes, let's drop it!

Thanks for the guidance! I did the research on the IBM RSA hardware,
and it indeed seems to belong in a museum.

Here is what I found regarding its lifecycle:
1. The IBM Remote Supervisor Adapter (RSA) and RSA II were early 2000s
   management PCI/ISA cards for ancient IBM eServer xSeries machines.
2. IBM deprecated the RSA family and replaced it with IMM around 2008.
3. IBM sold its x86 server business to Lenovo in 2014. Modern systems
   use entirely different BMCs.

I also reviewed the recent git history of the driver. Aside from
mechanical treewide VFS API updates, the only recent activity I saw
was another out-of-bounds fix in `command_file_write` a few months ago.

However, just like my patches, that recent fix addresses a security
vulnerability (a userspace-triggered OOB) rather than a functional
bug reported by an active hardware user. Keeping this obsolete driver
in the tree just leaves an unnecessary attack surface that forces core
maintainers to endlessly review security patches for dead hardware.

Given these findings, I believe the driver is essentially dead code.
However, I want to entirely defer to your judgment. If you feel it
is still worth keeping in the kernel, I can continue refining this
OOB patch series. Alternatively, if you agree it is time to retire
it, I would be happy to send a patch to completely remove the
`drivers/misc/ibmasm/` directory.

Please let me know which path you would prefer.

Best regards,
Mingyu Wang