"Frank Ch. Eigler" writes:
> Hi -
>
>> [...]
>> -add_taint(TAINT_DIE);
>> +add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE);
>> [...]
>
> If "UNRELIABLE" a good way to describe it - not DANGEROUS or
> COUNTERPRODUCTIVE or something, then maybe lockdep *can* produce
> reasonable results followi
On Monday, January 21, 2013 10:30:22 AM Rusty Russell wrote:
> Dave Jones writes:
> > On Thu, Jan 17, 2013 at 11:27:27AM +1030, Rusty Russell wrote:
> >
> > > taint: add explicit flag to show whether lock dep is still OK.
> > >
> > > Fix up all callers as they were before, with make one chan
Hi -
> [...]
> - add_taint(TAINT_DIE);
> + add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE);
> [...]
If "UNRELIABLE" a good way to describe it - not DANGEROUS or
COUNTERPRODUCTIVE or something, then maybe lockdep *can* produce
reasonable results following such a taint. If the results are mer
On Thu, Jan 17, 2013 at 11:27:27AM +1030, Rusty Russell wrote:
> taint: add explicit flag to show whether lock dep is still OK.
>
> Fix up all callers as they were before, with make one change: an
> unsigned module taints the kernel, but doesn't turn off lockdep.
>
> Signed-off-by: Rusty
Josh Boyer writes:
>> diff --git a/drivers/iommu/dmar.c b/drivers/iommu/dmar.c
>> index 86e2f4a..7acca64 100644
>> --- a/drivers/iommu/dmar.c
>> +++ b/drivers/iommu/dmar.c
>> @@ -230,7 +230,7 @@ dmar_parse_one_rhsa(struct acpi_dmar_header *header)
>> }
>> }
>> WARN_TAINT(
>>
On Wed, Jan 16, 2013 at 02:04:50PM +1030, Rusty Russell wrote:
> Josh Boyer writes:
> > With module signing enabled but not in enforcing mode, we don't consider
> > unsigned modules to be an error. However, if we encounter an unsigned
> > module we currently taint the module and kernel with TAINT
On Wed, Jan 16, 2013 at 02:04:50PM +1030, Rusty Russell wrote:
> Josh Boyer writes:
> > With module signing enabled but not in enforcing mode, we don't consider
> > unsigned modules to be an error. However, if we encounter an unsigned
> > module we currently taint the module and kernel with TAINT
Josh Boyer writes:
> With module signing enabled but not in enforcing mode, we don't consider
> unsigned modules to be an error. However, if we encounter an unsigned
> module we currently taint the module and kernel with TAINT_FORCED_MODULE.
> That also disables lockdep within the kernel.
How ab
With module signing enabled but not in enforcing mode, we don't consider
unsigned modules to be an error. However, if we encounter an unsigned
module we currently taint the module and kernel with TAINT_FORCED_MODULE.
That also disables lockdep within the kernel.
Given that we aren't in enforcing
9 matches
Mail list logo
