nvme(4) related KASSERT ?

[Nicolas Joly <njoly%pasteur.fr@localhost>]

Hi,

Just got a new Dell XPS15 (9550) laptop, and couldn't resist to try
-current on it ... but got hit by a KASSERT which seems to be nvme(4)
related :

nvme0 at pci4 dev 0 function 0: vendor 1c5c product 1284 (rev. 0x00)
nvme0: NVMe 1.2
nvme0: for admin queue interrupting at msix1 vec 0
nvme0: PC300 NVMe SK hynix 512GB, firmware 20005A00, serial FS6BN05021010B61F
nvme0: for io queue 1 interrupting at msix1 vec 1 affinity to cpu0
nvme0: for io queue 2 interrupting at msix1 vec 2 affinity to cpu1
nvme0: for io queue 3 interrupting at msix1 vec 3 affinity to cpu2
nvme0: unable to create io queue
panic: kernel diagnostic assertion "xc->xc_donep < xc->xc_headp" failed: file "
/local/src/NetBSD/src/sys/kern/subr_xcall.c", line 278
fatal breakpoint trap in supervisor mode
trap type 1 code 0 rip ffffffff80215455 cs 8 rflags 246 cr2 0 ilevel 8 rsp ffffffff8189b2f0 
curlwp 0xffffffff81482660 pid 0.1 lowest kstack 0xffffffff818982c0

breakpoint() at netbsd:breakpoint+0x5
vpanic() at netbsd:vpanic+0x140
ch_voltag_convert_in() at netbsd:ch_voltag_convert_in 
xc_broadcast() at netbsd:xc_broadcast+0x254
softint_disestablish() at netbsd:softint_disestablish+0x5c
nvme_pci_intr_disestablish() at netbsd:nvme_pci_intr_disestablish+0x42
nvme_q_delete() at netbsd:nvme_q_delete+0x18b
nvme_attach() at netbsd:nvme_attach+0x7d8
nvme_pci_attach() at netbsd:nvme_pci_attach+0x4dd
config_attach_loc() at netbsd:config_attach_loc+0x17a
config_found_sm_loc() at netbsd:config_found_sm_loc+0x48
pci_probe_device() at netbsd:pci_probe_device+0x58d
pci_enumerate_bus() at netbsd:pci_enumerate_bus+0x171
pcirescan() at netbsd:pcirescan+0x47
pciattach() at netbsd:pciattach+0x193 
config_attach_loc() at netbsd:config_attach_loc+0x17a
config_found_sm_loc() at netbsd:config_found_sm_loc+0x48
ppbattach() at netbsd:ppbattach+0x112 
config_attach_loc() at netbsd:config_attach_loc+0x17a
config_found_sm_loc() at netbsd:config_found_sm_loc+0x48
pci_probe_device() at netbsd:pci_probe_device+0x58d
pci_enumerate_bus() at netbsd:pci_enumerate_bus+0x171
pcirescan() at netbsd:pcirescan+0x47
pciattach() at netbsd:pciattach+0x193
config_attach_loc() at netbsd:config_attach_loc+0x17a
config_found_sm_loc() at netbsd:config_found_sm_loc+0x48
mp_pci_scan() at netbsd:mp_pci_scan+0x9c
mainbus_attach() at netbsd:mainbus_attach+0x317
config_attach_loc() at netbsd:config_attach_loc+0x17a
cpu_configure() at netbsd:cpu_configure+0x26
main() at netbsd:main+0x2b1

With interruptions forced to INTx, it works just fine.

nvme0 at pci4 dev 0 function 0: vendor 1c5c product 1284 (rev. 0x00)
nvme0: NVMe 1.2
nvme0: interrupting at ioapic0 pin 16
nvme0: PC300 NVMe SK hynix 512GB, firmware 20005A00, serial FS6BN05021010B61F
ld0 at nvme0 nsid 1
ld0: 476 GB, 62260 cyl, 255 head, 63 sec, 512 bytes/sect x 1000215216 sectors

Any idea ?

Thanks.

NB: Attached full dmesg in the working case.

-- 
Nicolas Joly

Cluster & Computing Group
Biology IT Center
Institut Pasteur, Paris.