Michael Brown

@loood Thank you for the debug output. The values being read from the card are very strange: PF 7, port 3, queues 0x7ff-0x7ff, BAR 3. All of these values represent...

@loood Could you please also try (in UEFI mode) `DEBUG=intelxl:3,pcimsix,efi_pci:3`?

Commentary inline below: > ``` > EFIPCI 0000:44:00.0 (8086:15ff class 020000) has driver "x710-10gt-b" > EFIPCI 0000:44:00.0 found range [c9000000,ca900000) -> [c9000000,ca900000) > EFIPCI 0000:44:00.0 found range [c4000000,c8100000) -> [c4000000,c8100000)...

> Same here. We're using `master` (I know), and it worked till a couple of weeks ago. So some commits in between must have broken it. In that case, https://ipxe.org/howto/bisect...

@DimanKuskov Thank you for the detailed description. > 1. Update `--keep` argument description and add information that image must be reused by its base name. This feels user-unfriendly, and I'd...

The code is one day old and someone has already found a use for it :slightly_smiling_face: I'm hoping to get as far as implementing `kexec` within iPXE, so that the...

@antonym I haven't had a chance to implement kexec yet. From the (sparse) kexec documentation, it looks as though the easiest route will be via the `kexec_file_load()` syscall, passing in...

This seems to be a less useful variant of the existing "imgtrust" command, since it does not allow for the possibility of updating the image in future without also updating...

(My last comment should of course say "imgverify command" instead of "imgtrust command"; sorry).

The issue is almost certainly that the existing VLAN device (created by the platform firmware) is not being deleted when iPXE disconnects existing drivers. UEFI enforces uniqueness of device paths...