08:52:34 <endlich> I have a Lenovo Thinkcentre m710q with i5-7500T (see specs here: https://psref.lenovo.com/syspool/Sys/PDF/ThinkCentre/ThinkCentre_M710_Tiny/ThinkCentre_M710_Tiny_Spec.pdf ), running OmniOS right now. I have attempted to install simply via the usb image dd'ed on a USB3 flashdrive, connected to the machine via a USB3 port, but while I managed to boot into the omniOS bootloader, I never managed to boot into 
08:52:37 <endlich> the installer, It compleined about the USB. I assume there is a compatibility issue somewhere in the XHCI support. I had to connect the flashdrive via an additional USB2 hub. Doing so the bootloader managed to sucessfully find the flashdrive and booted into the installer.
08:53:12 <endlich> Interestingly after the install I was able to use USB3 disks with OmniOS, but not to boot from.
10:14:57 <tsoome_> yep, there appears to be a problem with "bios handover". if boot loader can see the stick fine (so the firmware IO functions are ok), but kernel has trouble with accessing this disk,  but it appears wot work ok when you plug it once system is up and running, then it has to be handover issue of some sort.
10:29:01 <jclulow> tsoome_: I'm not sure that it's the handover stuff
10:29:15 <jclulow> I've been tinkering with: https://github.com/oxidecomputer/illumos-gate/commit/38227f41e8ecc81dca17725c3b3edc41b5f421c6#diff-08e98d08351c08736a52fb22e2c3211e874efb8d3aa26f290c2cbed85e32ab03R2231
10:33:37 <tsoome_> well, the handover function itself is plain and simple, but well, sure, it can be some sort of bug in hw etc, but then again, if you do not have this problem when BIOS is not involved, then it is something about hw setup we are doing.
10:35:39 <tsoome_> of course there could be more of it too as there are reports  about not working USB3 sticks/disks ...
10:35:58 <jclulow> There are very few reports with any details unfortunately
10:36:04 <tsoome_> yep
10:36:44 <jclulow> I finally found a system that has the device-not-working-after-boot problem though, and doing a reset of the ports after we bring the controller online appears to kick the firmware out of whatever hole it gets stuck in
10:37:14 <jclulow> There is at least one Intel errata that covers an adjacent stuck port behaviour
10:37:40 <jclulow> XHCI is a mess of unfortunate and underspecified firmware, unfortunately
10:37:58 <wiedi_> I have some supermicro boxes here where usb sticks don't show up after boot until we replug them. If you want me to get some data or test things I'd be happy to
10:38:20 <tsoome_> and it really does not help to have this mixture of minor versions too:)
10:38:22 <jclulow> wiedi_: If you apply that patch above, it would be interesting to know if your problem goes away
10:38:51 <wiedi_> cool, will do that
10:39:31 <jclulow> It's 0240 though so I need to go to bed haha
10:40:35 <tsoome_> g'night!
12:01:29 <sjorge> andyf was it you that s end me a illumos binary of openSearChest at some point ?
12:01:44 <sjorge> I lost it and I need to low level format a new EXOS drive to 4k logical
14:10:07 <andyf> That doesn't ring any bells, sorry.
15:48:35 <sjorge> https://github.com/Seagate/openSeaChest upstream does
15:48:41 <sjorge> List illumos suppory
15:48:55 <sjorge> I wonder if it will just compile without issue