On 05/11, Lucas De Marchi wrote:
>
> On Fri, May 10, 2013 at 2:10 PM, Oleg Nesterov wrote:
> >
> >> > -char modprobe_path[KMOD_PATH_LEN] = "/sbin/modprobe";
> >> > +char modprobe_path[KMOD_PATH_LEN] = "/sbin/modprobe -q --";
> >
> > No. This is incompatible change, we shouldn't do this.
>
> But th
On 05/10, Lucas De Marchi wrote:
>
> On Fri, May 10, 2013 at 2:35 PM, Oleg Nesterov wrote:
> > On 05/10, Oleg Nesterov wrote:
> >>
> >> > > -char modprobe_path[KMOD_PATH_LEN] = "/sbin/modprobe";
> >> > > +char modprobe_path[KMOD_PATH_LEN] = "/sbin/modprobe -q --";
> >>
> >> No. This is incompatibl
On Fri, May 10, 2013 at 2:10 PM, Oleg Nesterov wrote:
> On 05/10, Lucas De Marchi wrote:
>>
>> On Fri, May 10, 2013 at 12:36 PM, Oleg Nesterov wrote:
>> > Well, personally I think it would be better to use kasprintf(), see the
>> > patch I sent (it is actually wrong, needs kfree(args) before retu
On Fri, May 10, 2013 at 2:35 PM, Oleg Nesterov wrote:
> On 05/10, Oleg Nesterov wrote:
>>
>> On 05/10, Lucas De Marchi wrote:
>> >
>> > but I think it's a good tradeoff and covers other use cases as you
>> > pointed out as well.
>>
>> OK, good.
>
> Yes, perhaps this makes sense anyway but...
>
>>
On 05/10, Oleg Nesterov wrote:
>
> On 05/10, Lucas De Marchi wrote:
> >
> > but I think it's a good tradeoff and covers other use cases as you
> > pointed out as well.
>
> OK, good.
Yes, perhaps this makes sense anyway but...
> > Ok. I'll give it a try.
>
> Please wait a bit, I'll send v2. See be
On 05/10, Lucas De Marchi wrote:
>
> On Fri, May 10, 2013 at 12:36 PM, Oleg Nesterov wrote:
> > Well, personally I think it would be better to use kasprintf(), see the
> > patch I sent (it is actually wrong, needs kfree(args) before return).
> >
> > Or. How about the patch below? It should be spli
On Fri, May 10, 2013 at 12:36 PM, Oleg Nesterov wrote:
> On 05/10, Lucas De Marchi wrote:
>>
>> Oh, right. Forgot about that. And this patch set should have been sent
>> as RFC, since I'm interested in feedback about the idea. What do you
>> think?
>
> Well, personally I think it would be better t
On 05/10, Lucas De Marchi wrote:
>
> Oh, right. Forgot about that. And this patch set should have been sent
> as RFC, since I'm interested in feedback about the idea. What do you
> think?
Well, personally I think it would be better to use kasprintf(), see the
patch I sent (it is actually wrong, ne
On 05/10, Oleg Nesterov wrote:
>
> On 05/10, Lucas De Marchi wrote:
> >
> > -char modprobe_path[KMOD_PATH_LEN] = "/sbin/modprobe -q --";
> > +char modprobe_path[KMOD_PATH_LEN] = CONFIG_DEFAULT_MODULE_LOAD_BIN;
>
> But even after 1/3 and 2/3 this can break free_modprobe_argv() ?
> It assumes that ar
On Fri, May 10, 2013 at 9:58 AM, Oleg Nesterov wrote:
>
> On 05/10, Lucas De Marchi wrote:
> >
> > -char modprobe_path[KMOD_PATH_LEN] = "/sbin/modprobe -q --";
> > +char modprobe_path[KMOD_PATH_LEN] = CONFIG_DEFAULT_MODULE_LOAD_BIN;
>
> But even after 1/3 and 2/3 this can break free_modprobe_argv(
On 05/10, Lucas De Marchi wrote:
>
> -char modprobe_path[KMOD_PATH_LEN] = "/sbin/modprobe -q --";
> +char modprobe_path[KMOD_PATH_LEN] = CONFIG_DEFAULT_MODULE_LOAD_BIN;
But even after 1/3 and 2/3 this can break free_modprobe_argv() ?
It assumes that argv[3] is module_name.
Oleg.
--
To unsubscrib
The hard-coded path of modprobe doesn't allow distros to put the binary
in another place without add links from one place to another. For
example, in recent versions of distros like Fedora, Arch and Suse, kmod
is the default tool to load modules and they have to create a link from
modprobe to kmod
12 matches
Mail list logo