[Bug 241774] FreeBSD 11.3 & 12.0 has broken SCSI & Networking on KVM/QEMU Q35 with OVMF

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=241774

--- Comment #12 from John Hartley <drum at graphica.com.au> ---
Hi Tommy T,

I am a bit puzzled, you our posting...

Your pciconf (below):


<<START>>

none1 at pci0:1:0:0:       class=0x020000 card=0x11001af4 chip=0x10411af4 rev=0x01
hdr=0x00
    vendor     = 'Red Hat, Inc.'
    device     = 'Virtio network device'
    class      = network
    subclass   = ethernet
    bar   [14] = type Memory, range 32, base rx98e00000, size 4096, enabled
    bar   [20] = type Prefetchable Memory, range 64, base rx800000000, size
16384, enabled
    cap 11[dc] = MSI-X supports 3 messages
                 Table in map 0x14[0x0], PBA in map 0x14[0x800]
    cap 09[c8] = vendor (length 20)
    cap 09[b4] = vendor (length 20)
    cap 09[a4] = vendor (length 16)
    cap 09[94] = vendor (length 16)
    cap 09[84] = vendor (length 16)
    cap 01[7c] = powerspec 3  supports D0 D3  current D0
    cap 10[40] = PCI-Express 2 endpoint max data 128(128)
                 link x1(x1) speed 2.5(2.5) ASPM disabled(L0s)
sym0 at pci0:2:0:0:        class=0x010000 card=0x10000000 chip=0x00121000 rev=0x00
hdr=0x00
    vendor     = 'LSI Logic / Symbios Logic'
    device     = '53c895a'
    class      = mass storage
    subclass   = SCSI
    bar   [10] = type I/O Port, range 32, base rxb000, size 256, enabled
    bar   [14] = type Memory, range 32, base rx98c02000, size 1024, enabled
    bar   [18] = type Memory, range 32, base rx98c00000, size 8192, enabled
none2 at pci0:3:0:0:       class=0x010000 card=0x11001af4 chip=0x10481af4 rev=0x01
hdr=0x00
    vendor     = 'Red Hat, Inc.'
    device     = 'Virtio SCSI'
    class      = mass storage
    subclass   = SCSI
    bar   [14] = type Memory, range 32, base rx98a00000, size 4096, enabled
    bar   [20] = type Prefetchable Memory, range 64, base rx800100000, size
16384, enabled
    cap 11[dc] = MSI-X supports 4 messages
                 Table in map 0x14[0x0], PBA in map 0x14[0x800]
    cap 09[c8] = vendor (length 20)
    cap 09[b4] = vendor (length 20)
    cap 09[a4] = vendor (length 16)
    cap 09[94] = vendor (length 16)
    cap 09[84] = vendor (length 16)
    cap 01[7c] = powerspec 3  supports D0 D3  current D0
    cap 10[40] = PCI-Express 2 endpoint max data 128(128)
                 link x1(x1) speed 2.5(2.5) ASPM disabled(L0s)

<<END>>

This shows VirtIO SCSI & VirtIO Network... but you state that you cannot get
VirtIO going (which is same problem I have).

Is this pciconf report from i440FX QEMU/KVM machine rather than a Q35 one ?

I have not managed to flip my test VM from Q35 to i440FX as per discussion in
your bug thread yet as I got into large regression build.

I am now part way through building new kernel that is 11.3 with mutant mostly
11.2 networking sub-system.

I say mostly 11.2 networking, as I believe that the VirtIO network version ID
bug you identified in combination of possible introduction of defect with
changes in the netmap 11.3 code. This is why your fix was not enough and why
all the network interfaces are broken not just VirtIO.

My mutant 11.3 Kernel + 11.2 Network build got much bigger than expected as the
11.3 netmap code results in ripple through impact down into all of the driver
code and also up into the network abstraction layers :-(

So when I wound back to 11.2 netmap I have to also wind back on these impacts.

I am almost there I believe this this "regressed 11.3" custom kernel.

If I am right about netmap problem (I hope I am ...), then I can home on the
specific issue with this code section.

I hope to test soon, after many many builds...

Thanks again for your tips and insights.

Cheers,


John

-- 
You are receiving this mail because:
You are the assignee for the bug.