Re: [RFC PATCH -tip 0/5] kprobes: Abolish jprobe APIs

From: Kees Cook
Date: Fri Oct 20 2017 - 12:28:40 EST


On Fri, Oct 20, 2017 at 8:17 AM, Ingo Molnar <mingo@xxxxxxxxxx> wrote:
>
> * Kees Cook <keescook@xxxxxxxxxxxx> wrote:
>
>> On Fri, Oct 20, 2017 at 5:22 AM, Ingo Molnar <mingo@xxxxxxxxxx> wrote:
>> >
>> > Ok, could we now also fix these:
>> >
>> > net/dccp/probe.c:166:2: warning: âregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > net/dccp/probe.c:170:4: warning: âregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > net/dccp/probe.c:190:2: warning: âunregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > drivers/misc/lkdtm_core.c:317:3: warning: âunregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > drivers/misc/lkdtm_core.c:322:2: warning: âregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > drivers/misc/lkdtm_core.c:560:3: warning: âunregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > net/sctp/probe.c:189:2: warning: âregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > net/sctp/probe.c:194:3: warning: âregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > net/sctp/probe.c:240:2: warning: âunregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > net/ipv4/tcp_probe.c:280:2: warning: âregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> > net/ipv4/tcp_probe.c:298:2: warning: âunregister_jprobeâ is deprecated [-Wdeprecated-declarations]
>> >
>> > I'd suggest removing the networking ones and Cc:-ing it to the networking folks -
>> > I strongly doubt anyone is using that functionality for real.
>> >
>> > The LKDTM one Kees had some ideas (patches?) for?
>>
>> Oops, yes, I've sent this patch to Greg now. I had let it sit for a
>> few days of 0-day testing and promptly forgot about it. ;)
>
> Could we, as a special exception, carry this patch in -tip, so that the probes
> related changes stay togher - or is it too distruptive to the flow of other
> in-flight LKDTM patches, creating conflicts, etc.?

I have no problem with that, sure; these were the only outstanding
lkdtm patches (one for jprobes removal, and another for const
clean-ups noticed during the first patch's work). However, Greg pulled
them to his -testing tree already, so I guess it's up to him. Just let
me know what I can do to help. :)

-Kees

--
Kees Cook
Pixel Security