[v2,4/6] bus/pci: use IOVAs dmak mask check when setting IOVA mode
Checks
Commit Message
Currently the code precludes IOVA mode if IOMMU hardware reports
less addressing bits than necessary for full virtual memory range.
Although VT-d emulation currently only supports 39 bits, it could
be iovas for allocated memlory being within that supported range.
This patch allows IOVA mode in such a case adding a call to
rte_eal_check_dma_mask using the reported addressing bits by the
IOMMU hardware.
Indeed, memory initialization code has been modified for using lower
virtual addresses than those used by the kernel for 64 bits processes
by default, and therefore memsegs iovas can use 39 bits or less for
most systems. And this is likely 100% true for VMs.
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>
---
drivers/bus/pci/linux/pci.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
Comments
On 05-Oct-18 1:06 PM, Alejandro Lucero wrote:
> Currently the code precludes IOVA mode if IOMMU hardware reports
> less addressing bits than necessary for full virtual memory range.
>
> Although VT-d emulation currently only supports 39 bits, it could
> be iovas for allocated memlory being within that supported range.
^^ memory
> This patch allows IOVA mode in such a case adding a call to
> rte_eal_check_dma_mask using the reported addressing bits by the
> IOMMU hardware.
>
> Indeed, memory initialization code has been modified for using lower
> virtual addresses than those used by the kernel for 64 bits processes
> by default, and therefore memsegs iovas can use 39 bits or less for
> most systems. And this is likely 100% true for VMs.
>
> Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>
> ---
> drivers/bus/pci/linux/pci.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/bus/pci/linux/pci.c b/drivers/bus/pci/linux/pci.c
> index a871549..5cf78d7 100644
> --- a/drivers/bus/pci/linux/pci.c
> +++ b/drivers/bus/pci/linux/pci.c
> @@ -588,10 +588,8 @@
> fclose(fp);
>
> mgaw = ((vtd_cap_reg & VTD_CAP_MGAW_MASK) >> VTD_CAP_MGAW_SHIFT) + 1;
> - if (mgaw < X86_VA_WIDTH)
> - return false;
>
> - return true;
> + return rte_eal_check_dma_mask(mgaw) == 0 ? true : false;
Still looks weird but OK, no big deal :)
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
> }
> #elif defined(RTE_ARCH_PPC_64)
> static bool
>
@@ -588,10 +588,8 @@
fclose(fp);
mgaw = ((vtd_cap_reg & VTD_CAP_MGAW_MASK) >> VTD_CAP_MGAW_SHIFT) + 1;
- if (mgaw < X86_VA_WIDTH)
- return false;
- return true;
+ return rte_eal_check_dma_mask(mgaw) == 0 ? true : false;
}
#elif defined(RTE_ARCH_PPC_64)
static bool