Hello Masa - I was able to apply your patch to the 9.2 tree and built a kernel and pxebooted it. Unfortunately the status now looks like this:
mfii0 at pci5 dev 0 function 0mfii0: firmware fault https://sdf.org/eu (full boot sequence) The machine is 8412km away in a data center and I don't have direct access to it so I cannot unplug the battery, move the board to another slot, et cetera. I can, however, power cycle/reset the machine and access the MegaCLI to look at the two raids. Both appear to be intact and the controller reports no errors, no bad disks or any other issues. I'm hessitant to make any changes physically to the machine until I'm absolutely sure that this is a recovery situation. Thank you for your help on this, I appreciate it! Stephen > I have 9272-8i and it have been working well on netbsd-9 since several > weeks ago. > >> mfii0 at pci1 dev 0 function 0: "LSI MegaRAID SAS 9272-8i", firmware >> 23.34.0-0019, 512MB cache >> mfii0: interrupting at ioapic0 pin 16 > > Could you all who have problem with mfii(4) test with the following diff? > > Index: mfii.c > =================================================================== > RCS file: /cvsroot/src/sys/dev/pci/mfii.c,v > retrieving revision 1.15 > diff -u -p -r1.15 mfii.c > --- mfii.c 13 May 2022 10:44:38 -0000 1.15 > +++ mfii.c 21 Jun 2022 07:46:51 -0000 > @@ -1840,7 +1840,11 @@ mfii_load_mfa(struct mfii_softc *sc, str > static void > mfii_start(struct mfii_softc *sc, struct mfii_ccb *ccb) > { > +#if defined(__LP64__) && 0 > u_long *r = (u_long *)&ccb->ccb_req; > +#else > + uint32_t *r = (uint32_t *)&ccb->ccb_req; > +#endif > > bus_dmamap_sync(sc->sc_dmat, MFII_DMA_MAP(sc->sc_requests), > ccb->ccb_request_offset, MFII_REQUEST_SIZE, > > > > > -- > ----------------------------------------------- > SAITOH Masanobu (msai...@execsw.org > msai...@netbsd.org) >