On Wed, Nov 23, 2022 at 04:15:05PM +0100, Christian König wrote:
> I have not the slightest idea how people got this impression, but I have
> heard it so many time from so many different sources that there must be some
> common cause to this. Is the maybe some book or tutorial how to sophisticate
On Wed, 23 Nov 2022 at 16:15, Christian König wrote:
>
> Am 23.11.22 um 16:08 schrieb Jason Gunthorpe:
> > On Wed, Nov 23, 2022 at 03:34:54PM +0100, Daniel Vetter wrote:
> >>> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> >>> index 1376a47fedeedb..4161241fc3228c 100644
> >>> --- a/virt/
On Wed, 23 Nov 2022 at 16:04, Jason Gunthorpe wrote:
>
> On Wed, Nov 23, 2022 at 03:28:27PM +0100, Daniel Vetter wrote:
>
> > > This patch is known to be broken in so many ways. It also has a major
> > > security hole that it ignores the PTE flags making the page
> > > RO. Ignoring the special bit
Am 23.11.22 um 16:08 schrieb Jason Gunthorpe:
On Wed, Nov 23, 2022 at 03:34:54PM +0100, Daniel Vetter wrote:
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 1376a47fedeedb..4161241fc3228c 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2598,6 +2598,19 @@ static int hv
On Wed, Nov 23, 2022 at 03:34:54PM +0100, Daniel Vetter wrote:
> > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> > index 1376a47fedeedb..4161241fc3228c 100644
> > --- a/virt/kvm/kvm_main.c
> > +++ b/virt/kvm/kvm_main.c
> > @@ -2598,6 +2598,19 @@ static int hva_to_pfn_remapped(struct vm_a
On Wed, Nov 23, 2022 at 03:28:27PM +0100, Daniel Vetter wrote:
> > This patch is known to be broken in so many ways. It also has a major
> > security hole that it ignores the PTE flags making the page
> > RO. Ignoring the special bit is somehow not surprising :(
> >
> > This probably doesn't work,
On Wed, 23 Nov 2022 at 14:28, Jason Gunthorpe wrote:
>
> On Wed, Nov 23, 2022 at 02:12:25PM +0100, Christian König wrote:
> > Am 23.11.22 um 13:53 schrieb Jason Gunthorpe:
> > > On Wed, Nov 23, 2022 at 01:49:41PM +0100, Christian König wrote:
> > > > Am 23.11.22 um 13:46 schrieb Jason Gunthorpe:
>
On Wed, 23 Nov 2022 at 14:28, Jason Gunthorpe wrote:
>
> On Wed, Nov 23, 2022 at 02:12:25PM +0100, Christian König wrote:
> > Am 23.11.22 um 13:53 schrieb Jason Gunthorpe:
> > > On Wed, Nov 23, 2022 at 01:49:41PM +0100, Christian König wrote:
> > > > Am 23.11.22 um 13:46 schrieb Jason Gunthorpe:
>
On Wed, Nov 23, 2022 at 02:12:25PM +0100, Christian König wrote:
> Am 23.11.22 um 13:53 schrieb Jason Gunthorpe:
> > On Wed, Nov 23, 2022 at 01:49:41PM +0100, Christian König wrote:
> > > Am 23.11.22 um 13:46 schrieb Jason Gunthorpe:
> > > > On Wed, Nov 23, 2022 at 11:06:55AM +0100, Daniel Vetter w
Am 23.11.22 um 13:53 schrieb Jason Gunthorpe:
On Wed, Nov 23, 2022 at 01:49:41PM +0100, Christian König wrote:
Am 23.11.22 um 13:46 schrieb Jason Gunthorpe:
On Wed, Nov 23, 2022 at 11:06:55AM +0100, Daniel Vetter wrote:
Maybe a GFP flag to set the page reference count to zero or something
lik
On Wed, Nov 23, 2022 at 01:49:41PM +0100, Christian König wrote:
> Am 23.11.22 um 13:46 schrieb Jason Gunthorpe:
> > On Wed, Nov 23, 2022 at 11:06:55AM +0100, Daniel Vetter wrote:
> >
> > > > Maybe a GFP flag to set the page reference count to zero or something
> > > > like this?
> > > Hm yeah tha
Am 23.11.22 um 13:46 schrieb Jason Gunthorpe:
On Wed, Nov 23, 2022 at 11:06:55AM +0100, Daniel Vetter wrote:
Maybe a GFP flag to set the page reference count to zero or something
like this?
Hm yeah that might work. I'm not sure what it will all break though?
And we'd need to make sure that und
On Wed, Nov 23, 2022 at 11:06:55AM +0100, Daniel Vetter wrote:
> > Maybe a GFP flag to set the page reference count to zero or something
> > like this?
>
> Hm yeah that might work. I'm not sure what it will all break though?
> And we'd need to make sure that underflowing the page refcount dies in
On Wed, 23 Nov 2022 at 10:39, Christian König
wrote:
>
> Am 23.11.22 um 10:30 schrieb Daniel Vetter:
> > On Wed, 23 Nov 2022 at 10:06, Christian König
> > wrote:
> >> Am 22.11.22 um 20:50 schrieb Daniel Vetter:
> >>> On Tue, 22 Nov 2022 at 20:34, Jason Gunthorpe wrote:
> On Tue, Nov 22, 20
Am 23.11.22 um 10:30 schrieb Daniel Vetter:
On Wed, 23 Nov 2022 at 10:06, Christian König wrote:
Am 22.11.22 um 20:50 schrieb Daniel Vetter:
On Tue, 22 Nov 2022 at 20:34, Jason Gunthorpe wrote:
On Tue, Nov 22, 2022 at 08:29:05PM +0100, Daniel Vetter wrote:
You nuke all the ptes. Drivers tha
On Wed, 23 Nov 2022 at 09:07, Thomas Zimmermann wrote:
>
> Hi
>
> Am 22.11.22 um 18:08 schrieb Daniel Vetter:
> > tldr; DMA buffers aren't normal memory, expecting that you can use
> > them like that (like calling get_user_pages works, or that they're
> > accounting like any other normal memory) c
On Wed, 23 Nov 2022 at 10:06, Christian König wrote:
> Am 22.11.22 um 20:50 schrieb Daniel Vetter:
> > On Tue, 22 Nov 2022 at 20:34, Jason Gunthorpe wrote:
> >> On Tue, Nov 22, 2022 at 08:29:05PM +0100, Daniel Vetter wrote:
> >>> You nuke all the ptes. Drivers that move have slightly more than a
Am 22.11.22 um 20:50 schrieb Daniel Vetter:
On Tue, 22 Nov 2022 at 20:34, Jason Gunthorpe wrote:
On Tue, Nov 22, 2022 at 08:29:05PM +0100, Daniel Vetter wrote:
You nuke all the ptes. Drivers that move have slightly more than a
bare struct file, they also have a struct address_space so that
inv
Hi
Am 22.11.22 um 18:08 schrieb Daniel Vetter:
tldr; DMA buffers aren't normal memory, expecting that you can use
them like that (like calling get_user_pages works, or that they're
accounting like any other normal memory) cannot be guaranteed.
Since some userspace only runs on integrated device
On Tue, 22 Nov 2022 at 20:34, Jason Gunthorpe wrote:
> On Tue, Nov 22, 2022 at 08:29:05PM +0100, Daniel Vetter wrote:
> > You nuke all the ptes. Drivers that move have slightly more than a
> > bare struct file, they also have a struct address_space so that
> > invalidate_mapping_range() works.
>
>
On Tue, Nov 22, 2022 at 08:29:05PM +0100, Daniel Vetter wrote:
> You nuke all the ptes. Drivers that move have slightly more than a
> bare struct file, they also have a struct address_space so that
> invalidate_mapping_range() works.
Okay, this is one of the ways that this can be made to work cor
On Tue, 22 Nov 2022 at 19:50, Jason Gunthorpe wrote:
>
> On Tue, Nov 22, 2022 at 07:08:25PM +0100, Daniel Vetter wrote:
> > On Tue, 22 Nov 2022 at 19:04, Jason Gunthorpe wrote:
> > >
> > > On Tue, Nov 22, 2022 at 06:08:00PM +0100, Daniel Vetter wrote:
> > > > tldr; DMA buffers aren't normal memor
On Tue, Nov 22, 2022 at 07:08:25PM +0100, Daniel Vetter wrote:
> On Tue, 22 Nov 2022 at 19:04, Jason Gunthorpe wrote:
> >
> > On Tue, Nov 22, 2022 at 06:08:00PM +0100, Daniel Vetter wrote:
> > > tldr; DMA buffers aren't normal memory, expecting that you can use
> > > them like that (like calling g
On Tue, 22 Nov 2022 at 19:04, Jason Gunthorpe wrote:
>
> On Tue, Nov 22, 2022 at 06:08:00PM +0100, Daniel Vetter wrote:
> > tldr; DMA buffers aren't normal memory, expecting that you can use
> > them like that (like calling get_user_pages works, or that they're
> > accounting like any other normal
On Tue, Nov 22, 2022 at 06:08:00PM +0100, Daniel Vetter wrote:
> tldr; DMA buffers aren't normal memory, expecting that you can use
> them like that (like calling get_user_pages works, or that they're
> accounting like any other normal memory) cannot be guaranteed.
>
> Since some userspace only ru
tldr; DMA buffers aren't normal memory, expecting that you can use
them like that (like calling get_user_pages works, or that they're
accounting like any other normal memory) cannot be guaranteed.
Since some userspace only runs on integrated devices, where all
buffers are actually all resident sys
26 matches
Mail list logo
