10:14:29 tsoome: are you currently working on that 4k sector size issue with pcfs and uefi boot? 10:26:59 yes 10:29:59 good 10:30:05 that means I don't have to 10:30:08 :) 10:30:14 :) 10:32:17 the biggest issue i have with this is the silent failure to boot after installing. no error or warning or anything from the installer, or even from "zpool create -B"... 10:34:09 I need to verify my patch for installboot, but its a bit complicated because I'm not really sure about pcfs writes:D 10:35:35 and our mkfs does not support 4k sector size, so other tool is needed to create the fs... 10:36:54 and, it seems we may need to reserve more 4k sectors with zpool create -B... so there are some traps hiding... 10:37:38 why not update mkfs? 10:39:26 thats in process:) but the thing is, need to do with hand in hand of fsck, to be able to verify both created structure on disk and pcfs writes into it:D 10:40:43 so far I have been attempting to verify by using fbsd fsck, but tbh, I'm not quite sure how much it can be trusted either:D 10:42:32 anyhow, need to fix this memlist merge first.... 12:26:49 I see that 1425,5801 is not available in illumos, it's a Chelsio Virtual Function (VF) device from SR-IOV. I was wondering, do we have a device that supports VF? I'd like to have a look at the code and try to implement it in t4nex/cxgbe 21:45:40 In this case the question of differences would be really Chelsio specific. I think the thing to do here is to understand how they handle VFs for this device on other systems. 21:46:10 There shouldn't be anything different in PCIe land here. But the issue will be that I presume you have to skip certain firmware related activities. 21:46:53 Depending on how Chelsio has designed things and what's passed through in the VF bars it'll vary dramatically.