On Fri, Mar 11, 2016 at 08:27:57AM +0100, Jan Kiszka wrote:
[...]
> >> @@ -282,7 +288,7 @@ static void vtd_generate_interrupt(IntelIOMMUState *s,
> >> hwaddr mesg_addr_reg,
> >> data = vtd_get_long_raw(s, mesg_data_reg);
> >>
> >> VTD_DPRINTF(FLOG, "msi: addr 0x%"PRIx64 " data 0x%"PRIx
Hi Peter,
On 2016-03-10 06:18, Peter Xu wrote:
> Hi, Jan/Rita,
>
> Have not gone deeper... Got several comments and questions inline.
>
> On Wed, Mar 09, 2016 at 12:58:41AM +0530, Rita Sinha wrote:
>
> [...]
>
>> +static AddressSpace *get_dma_address_space(void)
>> +{
>> +return &PC_MACHIN
Hi, Jan/Rita,
Have not gone deeper... Got several comments and questions inline.
On Wed, Mar 09, 2016 at 12:58:41AM +0530, Rita Sinha wrote:
[...]
> +static AddressSpace *get_dma_address_space(void)
> +{
> +return &PC_MACHINE(qdev_get_machine())->dma_address_space;
> +}
> +
> /* Given the
On 03/08/2016 04:07 PM, Eric Blake wrote:
> On 03/08/2016 12:28 PM, Rita Sinha wrote:
>> From: Jan Kiszka
>>
>> Introduce a DMA default target address space for PCI devices. Catch all
>
> [meta-comment]: When sending a series, both messages should be
> In-Reply-To a 0/2 cover letter, rather than
On 03/08/2016 12:28 PM, Rita Sinha wrote:
> From: Jan Kiszka
>
> Introduce a DMA default target address space for PCI devices. Catch all
[meta-comment]: When sending a series, both messages should be
In-Reply-To a 0/2 cover letter, rather than separate top-level threads.
> interrupt requests to
From: Jan Kiszka
Introduce a DMA default target address space for PCI devices. Catch all
interrupt requests to the front-side bus via an MSI memory region that
is part of that address space. Provide separate address spaces for
IOAPIC and HPET if the IOMMU is active to prepare for adding remapping
