On 11/12/21 at 10:08am, David Hildenbrand wrote:
> > > "that allows supervisor mode programs to optionally set user-space
> > > memory mappings so that access to those mappings from supervisor mode
> > > will cause a trap. This makes it harder for malicious programs to
> > > "trick" the kernel
> > "that allows supervisor mode programs to optionally set user-space
> > memory mappings so that access to those mappings from supervisor mode
> > will cause a trap. This makes it harder for malicious programs to
> > "trick" the kernel into using instructions or data from a user-space
> >
On 11/12/21 at 09:16am, David Hildenbrand wrote:
> On 12.11.21 08:01, Baoquan He wrote:
> > On 11/11/21 at 08:18pm, David Hildenbrand wrote:
> >> To clear a user buffer we cannot simply use memset, we have to use
> >> clear_user(). Using a kernel config based on rawhide Fedora and a
> >>
On 12.11.21 08:01, Baoquan He wrote:
> On 11/11/21 at 08:18pm, David Hildenbrand wrote:
>> To clear a user buffer we cannot simply use memset, we have to use
>> clear_user(). Using a kernel config based on rawhide Fedora and a
>> virtio-mem device that registers a vmcore_cb, I can easily trigger:
On 11/11/21 at 08:18pm, David Hildenbrand wrote:
> To clear a user buffer we cannot simply use memset, we have to use
> clear_user(). Using a kernel config based on rawhide Fedora and a
> virtio-mem device that registers a vmcore_cb, I can easily trigger:
>
> [ 11.327580] systemd[1]: Starting
To clear a user buffer we cannot simply use memset, we have to use
clear_user(). Using a kernel config based on rawhide Fedora and a
virtio-mem device that registers a vmcore_cb, I can easily trigger:
[ 11.327580] systemd[1]: Starting Kdump Vmcore Save Service...
[ 11.339697] kdump[420]: