From: peterz@infradead.org
To: Masami Hiramatsu <mhiramat@kernel.org>
Cc: linux-kernel@vger.kernel.org, Eddy_Wu@trendmicro.com,
x86@kernel.org, davem@davemloft.net, rostedt@goodmis.org,
naveen.n.rao@linux.ibm.com, anil.s.keshavamurthy@intel.com,
linux-arch@vger.kernel.org, cameron@moodycamel.com,
oleg@redhat.com, will@kernel.org, paulmck@kernel.org
Subject: Re: [RFC][PATCH 3/7] kprobes: Remove kretprobe hash
Date: Fri, 28 Aug 2020 10:44:11 +0200 [thread overview]
Message-ID: <20200828084411.GP1362448@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20200828030059.d6618caf5b0214c424b941df@kernel.org>
On Fri, Aug 28, 2020 at 03:00:59AM +0900, Masami Hiramatsu wrote:
> On Thu, 27 Aug 2020 18:12:40 +0200
> Peter Zijlstra <peterz@infradead.org> wrote:
>
> > +static void invalidate_rp_inst(struct task_struct *t, struct kretprobe *rp)
> > +{
> > + struct invl_rp_ipi iri = {
> > + .task = t,
> > + .rp = rp,
> > + .done = false
> > + };
> > +
> > + for (;;) {
> > + if (try_invoke_on_locked_down_task(t, __invalidate_rp_inst, rp))
> > + return;
> > +
> > + smp_call_function_single(task_cpu(t), __invalidate_rp_ipi, &iri, 1);
> > + if (iri.done)
> > + return;
> > + }
>
> Hmm, what about making a status place holder and point it from
> each instance to tell it is valid or not?
>
> struct kretprobe_holder {
> atomic_t refcnt;
> struct kretprobe *rp;
> };
>
> struct kretprobe {
> ...
> struct kretprobe_holder *rph; // allocate at register
> ...
> };
>
> struct kretprobe_instance {
> ...
> struct kretprobe_holder *rph; // free if refcnt == 0
> ...
> };
>
> cleanup_rp_inst(struct kretprobe *rp)
> {
> rp->rph->rp = NULL;
> }
>
> kretprobe_trampoline_handler()
> {
> ...
> rp = READ_ONCE(ri->rph-rp);
> if (likely(rp)) {
> // call rp->handler
> } else
> rcu_call(ri, free_rp_inst_rcu);
> ...
> }
>
> free_rp_inst_rcu()
> {
> if (!atomic_dec_return(ri->rph->refcnt))
> kfree(ri->rph);
> kfree(ri);
> }
>
> This increase kretprobe_instance a bit, but make things simpler.
> (and still keep lockless, atomic op is in the rcu callback).
Yes, much better.
Although I'd _love_ to get rid of rp->data_size, then we can simplify
all of this even more. I was thinking we could then have a single global
freelist thing and add some per-cpu cache to it (say 4-8 entries) to
avoid the worst contention.
And then make function-graph use this, instead of the other way around
:-)
next prev parent reply other threads:[~2020-08-28 8:45 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-27 16:12 [RFC][PATCH 0/7] kprobes: Make kretprobes lockless Peter Zijlstra
2020-08-27 16:12 ` [RFC][PATCH 1/7] llist: Add nonatomic __llist_add() Peter Zijlstra
2020-08-27 16:12 ` [RFC][PATCH 2/7] sched: Fix try_invoke_on_locked_down_task() semantics Peter Zijlstra
2020-08-27 16:12 ` [RFC][PATCH 3/7] kprobes: Remove kretprobe hash Peter Zijlstra
2020-08-27 18:00 ` Masami Hiramatsu
2020-08-28 8:44 ` peterz [this message]
2020-08-28 9:07 ` Masami Hiramatsu
2020-08-28 4:44 ` Masami Hiramatsu
2020-08-28 13:11 ` Eddy_Wu
2020-08-28 13:38 ` peterz
2020-08-28 13:51 ` Masami Hiramatsu
2020-08-28 13:58 ` peterz
2020-08-28 14:19 ` Masami Hiramatsu
2020-08-28 14:11 ` Eddy_Wu
2020-08-28 14:19 ` peterz
2020-08-28 14:41 ` Masami Hiramatsu
2020-08-28 14:49 ` Masami Hiramatsu
2020-08-27 16:12 ` [RFC][PATCH 4/7] kprobe: Dont kfree() from breakpoint context Peter Zijlstra
2020-08-27 16:12 ` [RFC][PATCH 5/7] asm-generic/atomic: Add try_cmpxchg() fallbacks Peter Zijlstra
2020-08-27 16:12 ` [RFC][PATCH 6/7] freelist: Lock less freelist Peter Zijlstra
2020-08-27 16:37 ` peterz
[not found] ` <CAFCw3doX6KK5DwpG_OB331Mdw8uYeVqn8YPTjKh_a-m7ZB9+3A@mail.gmail.com>
2020-08-27 16:56 ` peterz
2020-08-27 17:00 ` Cameron
2020-08-27 19:08 ` Boqun Feng
2020-08-27 19:57 ` Cameron
2020-08-28 1:34 ` Boqun Feng
2020-08-28 4:03 ` Lai Jiangshan
2020-08-28 14:46 ` Oleg Nesterov
2020-08-28 15:29 ` peterz
2020-08-29 3:05 ` Cameron
2020-08-27 16:12 ` [RFC][PATCH 7/7] kprobes: Replace rp->free_instance with freelist Peter Zijlstra
2020-08-28 8:48 ` peterz
2020-08-28 9:13 ` Masami Hiramatsu
2020-08-28 9:18 ` peterz
2020-08-28 10:44 ` Masami Hiramatsu
2020-08-29 2:29 ` Cameron
2020-08-29 2:31 ` Cameron
2020-08-29 9:15 ` Masami Hiramatsu
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200828084411.GP1362448@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=Eddy_Wu@trendmicro.com \
--cc=anil.s.keshavamurthy@intel.com \
--cc=cameron@moodycamel.com \
--cc=davem@davemloft.net \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mhiramat@kernel.org \
--cc=naveen.n.rao@linux.ibm.com \
--cc=oleg@redhat.com \
--cc=paulmck@kernel.org \
--cc=rostedt@goodmis.org \
--cc=will@kernel.org \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).