Re: 8.99.41 panic in nvme_poll

[Thomas Klausner <wiz%NetBSD.org@localhost>]

I test-booted FreeBSD 12.
It finishes booting and provides a nvd0 device for the disk.
The ioapic "can't rename" errors are not there either.
 Thomas

On Sat, May 25, 2019 at 11:47:38PM +0200, Thomas Klausner wrote:
> Hi!
> 
> Booting the amd64 USB install image from
> https://nyftp.netbsd.org/pub/NetBSD-daily/HEAD/201905242230Z/images/
> on a new Threadripper system with a Samsung SSD 1TB 970 Evo plus, I
> see (handcopied):
> 
> panic: prevented jump to null instruction pointer (SMEP)
> cpu0: Begin traceback...
> vpanic() at netbsd:vpanic+0x160
> snprintf() at netbsd:snprintf
> trap() at netbsd:trap+0x2ab
> --- trap (number 6) ---
> ?() at 0
> nvme_poll() at netbsd:nvme_poll+0x104
> nvme_attach() at netbsd:nvme_atached+0x927
> nvme_pci_attach() at netbsd_nvme_pci_attach+0x309
> config_attach_loc()...
> config_found_sm_loc()...
> pci_probe_device()...
> pci_enumerate_bus()...
> pcirescan()...
> pciattach()...
> config_attach_loc()...
> config_found_sm_loc()...
> mp_pci_scan()...
> amd64_mainbus_attach()...
> mainbus_attach()...
> config_attach_loc()...
> cpu_configure()...
> main()...
> cpu0: End traceback
> 
> 
> The nvme0 entry looks like this:
> nvme0 at pci0 dev 0 function 0: vendor 144d product a808 (rev. 0x00) (not sure if it's 808 or 000 or something inbetween, bad screenshot)
> nvme0: NVMe 1.3
> nvme0: for admin queue interrupting at msix3 vec 0
> nvme0: for io queue ... interrupting at msix3 vec ... affinity to cpu ... (where ... goes from 0 to 31)
> 
> 
> Is that a bug in our nvme driver?
> 
> 
> 
> Earlier there are a couple:
> ioapic0 at mainbus0 apid 128
> ioapic0: can't remap to apid 128
> 
> (Same for 129, 130). Not sure what those are about. What does ioapic
> do and is this error a problem?
> 
> Thanks,
>  Thomas
>