Adding another data point: for roughly the last month I’ve been seeing a similar class of errors with an internal LSI 9305-16i / SAS3224 using mpt3sas. I initially thought it might be related to recent Unraid 7.3.x upgrades.

Controller firmware:
mpt3sas_cm0: LSISAS3224: FWVersion(16.00.12.00)

SAS expander:
vendor_id: INTEL; product_id: RES3FV288; FWVersion(B057)

The errors are repeated attempting task abort, timeout 30000 ms, UNKNOWN(0x2003) Result: hostbyte=0x03, I/O error, and Power-on or device reset occurred. The most visible failures are on cache SSDs behind the SAS/expander path, with BTRFS correcting reads afterward. SMART and Unraid disk health do not indicate normal disk failure. I monitor these types of errors aggressively, so a lot of users may not see them as they do not trigger drive health changes or produce normal alerts.

I downgraded to 7.3.0 with no hardware or firmware changes. The system initially booted cleanly, but I later saw the same class of post-boot errors on 7.3.0 as well, including task aborts/resets on the cache SSDs and a BTRFS corrected read. So in my case the downgrade did not fully resolve it, but 7.3.x still not entirely ruled out.

@JorgeB I attempted 7.3.2 -- but the regression is still present. I am going to attempt to file an inquiry on whether they are aware of the issue.

I did a new search and still cannot find any similar reports upstream. If you haven't yet, my suggestion would be to create a Linux kernel bug report, but it's not a general problem, since there was at least another user with the same controller and no issues with 7.3.1, so there may be something else at play.

P.S. just to confirm, when you say:

Tried to add pci_aspm=off to boot statement -- did not fix. The AI wants a couple more tests on the boot statement like:

/bzimage initrd=/bzroot pcie_aspm=off pci=noaer

and nomodeset pci=noaer pcie_aspm=off

and... pcie_port_pm=off pci=noaer

pci_aspm=off would not help for sure; you also tried pcie_aspm=off just by itself, correct?

I did send an email to the slackware submission email box... but do you think this needs to go to the linux main kernel team - any ideas which avenue is best to engage?

Skip

I would try the LSI driver maintainers and the respective kernel list, you can get them from the maintainers list, in this case:

LSILOGIC MPT FUSION DRIVERS (FC/SAS/SPI)

M: Sathya Prakash <[email protected]>

M: Sreekanth Reddy <[email protected]>

M: Suganath Prabu Subramani <[email protected]>

L: [email protected]

L: [email protected]

S: Supported

W: http://www.avagotech.com/support/

F: drivers/message/fusion/

F: drivers/scsi/mpt3sas/

Send to all 3 maintainers (M) and CC to the respective list ([email protected]), mention the kernel it works, the one it doesn't, and include a snippet of the errors.