On Mon, 2019-03-18 at 18:01 +0100, Jan Beulich wrote:
> > > > An alternative prototype could be:
> > > > int (*get_func)(char *output);
> > > > if we want the function to write the current parameter value
> > > > into a
> > > > caller-provided buffer, and possibly return error codes.
>
>>> On 19.03.19 at 09:12, wrote:
> On 19/03/2019 09:04, Jan Beulich wrote:
> On 19.03.19 at 07:07, wrote:
>>> Why don't we replace the "*get_func()" with a "char *current_val" being
>>> filled at parameter setting time? How current_val is allocated (static
>>> string or dynamic buffer) just h
On 19/03/2019 09:04, Jan Beulich wrote:
On 19.03.19 at 07:07, wrote:
>> Why don't we replace the "*get_func()" with a "char *current_val" being
>> filled at parameter setting time? How current_val is allocated (static
>> string or dynamic buffer) just has to be known by the custom parameter
>
>>> On 19.03.19 at 07:07, wrote:
> Why don't we replace the "*get_func()" with a "char *current_val" being
> filled at parameter setting time? How current_val is allocated (static
> string or dynamic buffer) just has to be known by the custom parameter
> parsing function.
Could accumulate (over t
On 18/03/2019 18:01, Jan Beulich wrote:
On 18.03.19 at 16:44, wrote:
>> On Mon, 2019-03-18 at 15:02 +0100, Jan Beulich wrote:
> I'm also
> unconvinced this is appropriate if the value is actually
> a signed quantity.
isn't OPT_UINT only for unsigned integers?
>>>
>>> You
>>> On 18.03.19 at 16:44, wrote:
> On Mon, 2019-03-18 at 15:02 +0100, Jan Beulich wrote:
>> > > I'm also
>> > > unconvinced this is appropriate if the value is actually
>> > > a signed quantity.
>> >
>> > isn't OPT_UINT only for unsigned integers?
>>
>> You'll notice that there's no OPT_INT or O
On Mon, 2019-03-18 at 15:02 +0100, Jan Beulich wrote:
> > > > From the return value of strcmp()? I don't think so, because
> > > > you
> may have run past all table entries. Instead it's that property
> that you can use, i.e. checking whether ...
>
> > > > +for ( param = start; param < end
>>> On 18.03.19 at 14:34, wrote:
> On Wed, 2019-03-13 at 17:35 +0100, Jan Beulich wrote:
>> > > > On 06.03.19 at 13:58, wrote:
>> > +found = false;
>>
>> I don't think you need this variable here, or if so, it shouldn't
>> be boolean: Either you mean to support returning data for
>> mult
thanks for the review.
On Wed, 2019-03-13 at 17:35 +0100, Jan Beulich wrote:
> > > > On 06.03.19 at 13:58, wrote:
> > +static int get_params(const char *cmdline, char *values,
> > + const struct kernel_param *start,
> > + const struct kernel_param *end)
>
>>> On 06.03.19 at 13:58, wrote:
> +static int get_params(const char *cmdline, char *values,
> + const struct kernel_param *start,
> + const struct kernel_param *end)
> +{
> +char opt[128], *optkey, *q;
> +const char *p = cmdline, *val = values;
W
Add a sysctl hypercall to support getting hypervisor parameters
at runtime.
Signed-off-by: Vasilis Liaskovitis
---
tools/flask/policy/modules/dom0.te | 2 +-
xen/common/kernel.c | 109
xen/common/sysctl.c | 45
xen/inc
11 matches
Mail list logo
