On Wed, Nov 07, 2018 at 04:17:09PM +, Robin Murphy wrote:
> FWIW it looks like it *has* always been possible to hit this crash by
> allocating a domain and freeing it again without attaching any devices, it's
> just highly improbable for any sane code to do that explicitly, so the real
> latent
On 2018-11-07 4:03 pm, Joerg Roedel wrote:
On Wed, Nov 07, 2018 at 04:50:40PM +0100, Geert Uytterhoeven wrote:
It only got triggered by the combination of commits 6c2fb2ea76361da9
("of/device: Set bus DMA mask as appropriate") and b4ebe6063204da58
("dma-direct: implement complete bus_dma_mask ha
On Wed, Nov 07, 2018 at 02:18:50PM +0100, Geert Uytterhoeven wrote:
> Fix this by checking if the domain's context already exists, before
> trying to destroy it.
>
> Signed-off-by: Geert Uytterhoeven
> ---
> drivers/iommu/ipmmu-vmsa.c | 3 +++
> 1 file changed, 3 insertions(+)
Applied, thanks
On Wed, Nov 07, 2018 at 04:50:40PM +0100, Geert Uytterhoeven wrote:
> It only got triggered by the combination of commits 6c2fb2ea76361da9
> ("of/device: Set bus DMA mask as appropriate") and b4ebe6063204da58
> ("dma-direct: implement complete bus_dma_mask handling"), which is being
> fixed by "of/
Hi Jörg,
On Wed, Nov 7, 2018 at 4:34 PM Joerg Roedel wrote:
> On Wed, Nov 07, 2018 at 01:22:52PM +, Robin Murphy wrote:
> > On 2018-11-07 1:18 pm, Geert Uytterhoeven wrote:
> > > Fix this by checking if the domain's context already exists, before
> > > trying to destroy it.
> >
> > Reviewed-b
On Wed, Nov 07, 2018 at 01:22:52PM +, Robin Murphy wrote:
> On 2018-11-07 1:18 pm, Geert Uytterhoeven wrote:
> > Fix this by checking if the domain's context already exists, before
> > trying to destroy it.
>
> Reviewed-by: Robin Murphy
Does this need a Fixes-tag? If so, which patch should b
On 2018-11-07 1:18 pm, Geert Uytterhoeven wrote:
If iommu_ops.add_device() fails, iommu_ops.domain_free() is still
called, leading to a crash, as the domain was only partially
initialized:
ipmmu-vmsa e67b.mmu: Cannot accommodate DMA translation for IOMMU page
tables
sata_rcar ee30
If iommu_ops.add_device() fails, iommu_ops.domain_free() is still
called, leading to a crash, as the domain was only partially
initialized:
ipmmu-vmsa e67b.mmu: Cannot accommodate DMA translation for IOMMU page
tables
sata_rcar ee30.sata: Unable to initialize IPMMU context
iom
