I'm specifically looking for a device that has four USB 3.0 controllers that can be passed through individually to VMs via KVM. That way each port has the full 5GB/sec USB 3.0 speed all to itself in each VM (up to four). If you look at the RocketRaid 1144c/1144d, the StarTech PEXUSB3S44V or the Sonnet Allegro Pro you will see that they all have four usb controllers each behind a ACS compatible bridge. None of them work properly for individual passthrough under KVM in my experience.
On Sat, Jun 18, 2016 at 1:38 AM, thibaut noah <[email protected]> wrote: > I'm not sure what you mean by passthrough individual controler. > I have a 5ports usb 3 pci controller which i blacklisted on my host (same > way as the graphic card) and i passthrough it to the vm to have plug and > play. It has its own iommu group. If that's what you're looking for i can > search my stuff for the reference > On Jun 18, 2016 05:57, "Nick Rodgers" <[email protected]> wrote: > > Almost a year and two more 4-controller usb 3.0 cards later, I still > haven't found one that does what I'm after. I purchased the StarTech > PEXUSB3S44V and the Sonnet Allegro Pro. Both devices get their own IOMMU > group for each controller without the ACS override patch. The StarTech > device seems to pass through okay, but I get in windows I get a code 10 on > its root hub. The Sonnet Allegro Pro passes through perfectly, until I > restart or shutdown the VM. After the passed through controller's entry in > lspci -vvvv changes from > > 04:00.0 USB controller: Fresco Logic FL1100 USB 3.0 Host Controller (rev > 10) (prog-if 30 [XHCI]) > Subsystem: Sonnet Technologies, Inc. Device 6e31 > Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- > ParErr- Stepping- SERR- FastB2B- DisINTx+ > Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- > <TAbort- <MAbort- >SERR- <PERR- INTx- > Latency: 0, Cache Line Size: 64 bytes > Interrupt: pin A routed to IRQ 26 > NUMA node: 0 > Region 0: Memory at fe500000 (64-bit, non-prefetchable) [size=64K] > Region 2: Memory at fe511000 (64-bit, non-prefetchable) [size=4K] > Region 4: Memory at fe510000 (64-bit, non-prefetchable) [size=4K] > Capabilities: [40] Power Management version 3 > Flags: PMEClk- DSI- D1+ D2- AuxCurrent=0mA > PME(D0+,D1+,D2-,D3hot+,D3cold-) > Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME- > Capabilities: [50] MSI: Enable- Count=1/8 Maskable- 64bit+ > Address: 0000000000000000 Data: 0000 > Capabilities: [70] Express (v2) Endpoint, MSI 00 > DevCap: MaxPayload 512 bytes, PhantFunc 0, Latency L0s > <64ns, L1 <1us > ExtTag- AttnBtn- AttnInd- PwrInd- RBE+ FLReset- > SlotPowerLimit 25.000W > DevCtl: Report errors: Correctable- Non-Fatal- Fatal- > Unsupported- > RlxdOrd- ExtTag- PhantFunc- AuxPwr- NoSnoop+ > MaxPayload 128 bytes, MaxReadReq 512 bytes > DevSta: CorrErr- UncorrErr- FatalErr- UnsuppReq- AuxPwr- > TransPend- > LnkCap: Port #0, Speed 5GT/s, Width x1, ASPM L0s L1, Exit > Latency L0s unlimited, L1 unlimited > ClockPM- Surprise- LLActRep- BwNot- ASPMOptComp- > LnkCtl: ASPM Disabled; RCB 64 bytes Disabled- CommClk- > ExtSynch- ClockPM- AutWidDis- BWInt- AutBWInt- > LnkSta: Speed 5GT/s, Width x1, TrErr- Train- SlotClk+ > DLActive- BWMgmt- ABWMgmt- > DevCap2: Completion Timeout: Not Supported, TimeoutDis+, > LTR-, OBFF Not Supported > DevCtl2: Completion Timeout: 50us to 50ms, TimeoutDis-, > LTR-, OBFF Disabled > LnkCtl2: Target Link Speed: 5GT/s, EnterCompliance- > SpeedDis- > Transmit Margin: Normal Operating Range, > EnterModifiedCompliance- ComplianceSOS- > Compliance De-emphasis: -6dB > LnkSta2: Current De-emphasis Level: -6dB, > EqualizationComplete-, EqualizationPhase1- > EqualizationPhase2-, EqualizationPhase3-, > LinkEqualizationRequest- > Capabilities: [b0] MSI-X: Enable+ Count=8 Masked- > Vector table: BAR=2 offset=00000000 > PBA: BAR=4 offset=00000000 > Capabilities: [100 v1] Advanced Error Reporting > UESta: DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- UnxCmplt- > RxOF- MalfTLP- ECRC- UnsupReq- ACSViol- > UEMsk: DLP- SDES- TLP- FCP- CmpltTO- CmpltAbrt- UnxCmplt- > RxOF- MalfTLP- ECRC- UnsupReq- ACSViol- > UESvrt: DLP+ SDES+ TLP- FCP+ CmpltTO- CmpltAbrt- UnxCmplt- > RxOF+ MalfTLP+ ECRC- UnsupReq- ACSViol- > CESta: RxErr- BadTLP- BadDLLP- Rollover- Timeout- > NonFatalErr- > CEMsk: RxErr- BadTLP- BadDLLP- Rollover- Timeout- > NonFatalErr+ > AERCap: First Error Pointer: 00, GenCap+ CGenEn- ChkCap+ > ChkEn- > Kernel driver in use: vfio-pci > Kernel modules: xhci_pci > > to > > 04:00.0 USB controller: Fresco Logic FL1100 USB 3.0 Host Controller (rev > ff) (prog-if ff) > !!! Unknown header type 7f > Kernel driver in use: vfio-pci > Kernel modules: xhci_pci > > > > Does anyone know of a three or four controller usb 3.0 device that can > pass through individual controllers in kvm? If not I guess I'll have to > stick with the Startech device and defect to esxi, as the passthrough is > reported to work fine there. > > > On Tue, Aug 25, 2015 at 1:38 PM, Alex Williamson < > [email protected]> wrote: > >> On Tue, Aug 25, 2015 at 7:36 AM, Bradley Davis <[email protected]> >> wrote: >> >>> I have a 1144E and was able to pass the individual controllers through >>> to different VMs. My only issue was stability with the USB 3 controllers. >>> They would cause my VMs to crash or hang, when hung I would have to reboot >>> the host. I haven't run into this issue since I removed the card. I updated >>> the firmware on the USB 3 controllers and it didn't help. The SATA >>> controllers passed through just fine, but I didn't have an eSATA to plug >>> into it. >>> >> >> I've also tried to assign an asmedia 1042 USB3 controller and was plagued >> with hangs on reboot, apparently in the OVMF XHCI driver. I switched to an >> NEC chip and it's been working well. In my personal opinion >> </disclaimer>, asmedia is crap. My personal favorite for USB3 is TI, but >> that's only based on bare metal behavior with mass storage devices, I >> haven't actually tried to assign it yet. >> >> Remember folks, device assignment is operating in a mode that's really >> not common for consumer devices and components from value manufacturers do >> not necessarily comply sufficiently with standards to work. We see this on >> realtek nics, asmedia <anything>, marvel SATA controllers, and there will >> surely be more to discover. Thanks, >> >> Alex >> > > > _______________________________________________ > vfio-users mailing list > [email protected] > https://www.redhat.com/mailman/listinfo/vfio-users > >
_______________________________________________ vfio-users mailing list [email protected] https://www.redhat.com/mailman/listinfo/vfio-users
