Page MenuHomeFreeBSD

Add Cavium ThunderX xHCI controller PCI ID
ClosedPublic

Authored by emaste on Sep 14 2015, 7:37 PM.
Tags
None
Referenced Files
F106183156: D3665.diff
Thu, Dec 26, 6:38 PM
Unknown Object (File)
Nov 16 2024, 10:12 AM
Unknown Object (File)
Nov 16 2024, 3:40 AM
Unknown Object (File)
Nov 1 2024, 7:32 PM
Unknown Object (File)
Nov 1 2024, 7:10 PM
Unknown Object (File)
Oct 26 2024, 11:09 AM
Unknown Object (File)
Oct 23 2024, 9:32 PM
Unknown Object (File)
Oct 5 2024, 7:51 AM
Subscribers

Details

Summary

See P74 for the full pciconf -lv from the system

none1@pci0:0:16:0:      class=0x0c0330 card=0xa11b177d chip=0xa01b177d rev=0x01 hdr=0x00
    vendor     = 'Cavium, Inc.'
    device     = 'THUNDERX xHCI USB Controller'
    class      = serial bus
    subclass   = USB
none2@pci0:0:17:0:      class=0x0c0330 card=0xa11b177d chip=0xa01b177d rev=0x01 hdr=0x00
    vendor     = 'Cavium, Inc.'
    device     = 'THUNDERX xHCI USB Controller'
    class      = serial bus
 subclass = USB

Diff Detail

Repository
rS FreeBSD src repository - subversion
Lint
Lint Not Applicable
Unit
Tests Not Applicable

Event Timeline

emaste retitled this revision from to Add Cavium ThunderX xHCI controller PCI ID.
emaste updated this object.
emaste edited the test plan for this revision. (Show Details)
emaste added a reviewer: hselasky.
emaste added a subscriber: arm64.
hselasky edited edge metadata.

You are sure this device does not need any port routing quirks?

This revision is now accepted and ready to land.Sep 14 2015, 7:39 PM
sys/dev/usb/controller/xhci_pci.c
116–118 ↗(On Diff #8755)

One question - the sorting here is going to get awkward as we add more xHCI controllers. Should we sort alphabetically by mfgr name and then by device_id? We do want all of the mfgr entries to remain together I'd say, so sorting just by device_id is insufficient. Sorting by "lower 16 bytes of device_id then by upper 16 bits" would work but is awkward.

sys/dev/usb/controller/xhci_pci.c
116–118 ↗(On Diff #8755)

The compiler usually sorts the entries, though sorting first by the lower 16-bits then the higher 16-bits is fine by me. This should be done as a separate commit. It would be really cool if these PCI strings were provided the PCI space, instead of by hand :-)

--HPS

The compiler usually sorts the entries, though sorting first by the lower 16-bits then the higher 16-bits is fine by me.

Right, I'm just talking about sorting for code maintainability. It will get to be a real mess if we have entries ending up in arbitrary order.

This revision was automatically updated to reflect the committed changes.