I regularly track -STABLE on this small form factor machine I use as a router. After upgrading on April 29 (last was April 23) the machine hung on boot just before polling the ISA bus. I tracked the commit down to https://cgit.freebsd.org/src/commit/?h=stable/13&id=7c0af1af0a8b467b8c61ee9b3aef0e527536deaa. Specifically, adding support for the Intel Sunrise Point Controller 0 (which this machine has). Removing this line restores normal booting. Attached are my dmesg and pciconf output. I'm not sure if perhaps this controller needs a quirk...?
Created attachment 241882 [details] dmesg
Created attachment 241883 [details] pciconf
(In reply to Joe Marcus Clarke from comment #0) Thanks for the report, sorry for the breakage. What happens if you change the final member, `2`, to a `0` on that line? Is this UART controller broken out in a way that you can test the result and see if it actually works in addition to not hanging your boot?
Thanks, Kyle. My research said either that or the clock (or both) might be worth trying. Unfortunately, I just went out of town, and I don't want to risk taking out the router from remote. I'll try this weekend. And, yes, I'll be able to test the UART.
Unfortunately, the shift reg change didn't work. The hang was the same (just after igb5 is probed and before ISA). I tried the 24x clock, too (with both 0 and 2 for shift reg) and the behavior was the same. I even disabled the two serial ports in the BIOS, but the boot still hung. The only thing that works is not to probe that PCI device. I noticed in -CURRENT there was some MSI exclusion support added (if I read that right). Not sure if that would be helpful or not here.
A commit in branch main references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=20d8d9809a8c0a3ddd1d0156e77ec53a1ed95747 commit 20d8d9809a8c0a3ddd1d0156e77ec53a1ed95747 Author: Kyle Evans <kevans@FreeBSD.org> AuthorDate: 2023-05-09 06:38:32 +0000 Commit: Kyle Evans <kevans@FreeBSD.org> CommitDate: 2023-05-09 06:38:32 +0000 Revert "uart(4): add Sunrise Point UART controllers" This reverts commit d1b6271118188dd25a18f2372ab1d3004335ea3c. I've received multiple reports of machines failing to boot with this hardware; back it out for now until we can fix it. PR: 271147 MFC after: Morning coffee sys/dev/uart/uart_bus_pci.c | 2 -- 1 file changed, 2 deletions(-)
A commit in branch stable/13 references this bug: URL: https://cgit.FreeBSD.org/src/commit/?id=780466256114fccaed8ae2547675c74d5f82d99c commit 780466256114fccaed8ae2547675c74d5f82d99c Author: Kyle Evans <kevans@FreeBSD.org> AuthorDate: 2023-05-09 06:38:32 +0000 Commit: Kyle Evans <kevans@FreeBSD.org> CommitDate: 2023-05-09 17:33:45 +0000 Revert "uart(4): add Sunrise Point UART controllers" This reverts commit d1b6271118188dd25a18f2372ab1d3004335ea3c. I've received multiple reports of machines failing to boot with this hardware; back it out for now until we can fix it. PR: 271147 (cherry picked from commit 20d8d9809a8c0a3ddd1d0156e77ec53a1ed95747) sys/dev/uart/uart_bus_pci.c | 2 -- 1 file changed, 2 deletions(-)