Re: [Qemu-devel] [RFC PATCH qemu v2 1/5] vfio: Switch from TARGET_PAGE_MASK to qemu_real_host_page_mask

[Alexey Kardashevskiy <aik@ozlabs.ru>]

On 07/14/2015 04:58 PM, Alexey Kardashevskiy wrote:
> On 07/14/2015 05:13 AM, Alex Williamson wrote:
>> On Tue, 2015-07-14 at 00:56 +1000, Alexey Kardashevskiy wrote:
>>> These started switching from TARGET_PAGE_MASK (hardcoded as 4K) to
>>> a real host page size:
>>> 4e51361d7 "cpu-all: complete "real" host page size API" and
>>> f7ceed190 "vfio: cpu: Use "real" page size API"
>>>
>>> This finished the transition by:
>>> - %s/TARGET_PAGE_MASK/qemu_real_host_page_mask/
>>> - %s/TARGET_PAGE_ALIGN/REAL_HOST_PAGE_ALIGN/
>>> - removing bitfield length for offsets in VFIOQuirk::data as
>>> qemu_real_host_page_mask is not a macro
>>
>> Can we assume that none of the changes to quirks have actually been
>> tested?
>
> No, why? :) I tried it on one of NVIDIAs I got here -
> VGA compatible controller: NVIDIA Corporation GM107GL [Quadro K2200] (rev a2)
> The driver was from NVIDIA (not nouveau) and the test was "acos" (some
> basic CUDA test).

Ufff. My bad. CUDA only works if I remove quirks at all.

Using host page size helps to boot the guest with this Quadro (otherwise 
abort in kvm_set_phys_mem()) but simple CUDA test produces EEH:

aik@aiktest-u14:~$ ./acos
[   47.488858] EEH: Frozen PHB#0-PE#1 detected
[   47.488861] EEH: PE location: vfio_vfio-pci:0000:00:01.0, PHB location: 
vfio_vfio-pci:0000:00:01.0

[   51.559681] Unable to handle kernel paging request for data at address 
0x00000000
[   51.559873] Faulting instruction address: 0xd00000000550d168
[   51.560038] Oops: Kernel access of bad area, sig: 11 [#1]


Does this mean "nack" to %s/TARGET_PAGE_MASK/qemu_real_host_page_mask/ in 
quirks?


-- 
Alexey