Print ioeventfd, iobus, vmstat and vcpustat information in kvm module.

[SiddhiK29]

No description provided.

[brenns10]

This looks great! A few small nits, but nothing major. Thank you :)

[brenns10]

This corresponds to enum kvm_bus, right? Since there's a nice enum present from UEK5 to mainline, let's use that instead of hard-coding it here.

Notably, this definition here seems to be missing KVM_IOCSR_BUS which I see in the definition, which is a great reason to just use the debuginfo.

enum kvm_bus {
	KVM_MMIO_BUS,
	KVM_PIO_BUS,
	KVM_VIRTIO_CCW_NOTIFY_BUS,
	KVM_FAST_MMIO_BUS,
	KVM_IOCSR_BUS,
	KVM_NR_BUSES
};

[brenns10]

To get the enum name here, we can do this:

Suggested change

	bus_name = KvmIoBus(i).name
	bus_name = enum_name_get(kvm_bus_type, i)

You can declare kvm_bus_type outside all the loops like this:

kvm_bus_type = prog.type("enum kvm_bus")

And you can import enum_name_get() at the top:

from drgn_tools.util import enum_name_get

[brenns10]

Rather than manually managing this counter, you can do:

for i, bus in enumerate(io_bus_iterator):
    ...

[brenns10]

Please don't include the \n here, instead have separate print() above this to put in the newline. Otherwise, the print_table() code considers it a character for alignment purposes, and it misaligns the first row. EG:

VCPU:         0        HALT_SUC:        467     HALT_ATTMPT:    881     HALT_INV:    0
PF_FIXED:      971      PF_GUEST:        0       TLB_FLUSH:      15563   INVLPG:      0
EXITS:         1290959  IO_EXIT:         128538  MMIO_EXIT:      35913   SIG_EXIT:    0
IRQ_WIN_EXIT:  31       NMI_WIN_EXIT:    0       L1D_FLUSH:      0       HALT_EXIT:   2278
REQ_IRQ_EXIT:  0        IRQ_EXITS:       2190    HOST_STATE_RL:  165126  FPU_RL:      164309
INSN_EMUL:     54456    INSN_EMUL_FAIL:  0       HYPERCALLS:     6       IRQ_INJ:     61
NMI_INJ:       0        REQ_EVENT:       46633   PREEMPT_RPT:    18      PREEMT_OTH:  12

[brenns10]

Oh, and one more thing: can you file an Orabug for this (targeting "next" -- see the developer workflow doc), and then include the Orabug statement in the commit message? This will also make sure you get credit in the changelog :)

[SiddhiK29]

Oh, and one more thing: can you file an Orabug for this (targeting "next" -- see the developer workflow doc), and then include the Orabug statement in the commit message? This will also make sure you get credit in the changelog :)

Hi Stephen,
I already have a Orabug filed targeting "next". The Orabug 37713468 is mentioned in commit message.
Can you please confirm if it's fine?

[brenns10]

I already have a Orabug filed targeting "next". The Orabug 37713468 is mentioned in commit message.
Can you please confirm if it's fine?

I don't know why I missed that, my apologies. That's perfect, thank you.

[brenns10]

All I did was rebase on the main branch to get the fix for the v6.16 UEK-next kernels.

With that, this looks good, thank you!