From: Jiri Olsa <olsajiri@gmail.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>,
Naveen N Rao <naveen@kernel.org>,
"David S. Miller" <davem@davemloft.net>,
Josh Poimboeuf <jpoimboe@kernel.org>,
Jason Baron <jbaron@akamai.com>,
Alice Ryhl <aliceryhl@google.com>,
Steven Rostedt <rostedt@goodmis.org>,
Ard Biesheuvel <ardb@kernel.org>,
Thomas Gleixner <tglx@kernel.org>, Ingo Molnar <mingo@redhat.com>,
Alan Maguire <almagui@kernel.org>,
x86@kernel.org, linux-trace-kernel@vger.kernel.org,
linux-kernel@vger.kernel.org, bpf@vger.kernel.org
Subject: Re: [PATCH 01/13] static_call: Index code text reservation lookups
Date: Thu, 12 Nov 2026 19:33:03 +0100 [thread overview]
Message-ID: <avYGX1UpGlvbvHO5@krava> (raw)
In-Reply-To: <20261112101108.GE2669272@noisy.programming.kicks-ass.net>
On Thu, Nov 12, 2026 at 11:11:08AM +0100, Peter Zijlstra wrote:
> On Thu, Nov 12, 2026 at 10:07:34AM +0100, Peter Zijlstra wrote:
> > Looking at this in a little more detail; I'm confused.
> >
> > You're literally adding an index, so that you can find the
> > static_call_size, while all you really need is static_call_addr().
>
> /size/site/ obviously :-)
>
> > That is, you could have just created an array of relocated site->addr
> > and sorted that, that would safe you a ton of indirection all around.
> >
> > That is:
> >
> > index->entries[i] = (long)static_call_addr(start + i) - (long)&index->entries[i];
>
> Should be:
>
> index->entries[i] = (long)static_call_addr(start + i) - (long)index->entries;
>
> They have to all be relative to one location in order for sorting to
> make sense.
>
> > That will make your lookup take at least half the amount of loads.
hum, but we still need the site’s init flag for static_call_is_init
I could store site pointer directly instead of the index and see how
much faster it is
> >
> >
> > Also, you really should support modules, there is no point being fast
> > until you load a module.
I had that, but decided to remove it because my follow up changes
did not support modules as well.. I will include it
jirka
next prev parent reply other threads:[~2026-11-12 18:33 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-09 19:49 [PATCH 00/13] kprobes: Reduce registration and optimization lookup costs Jiri Olsa
2026-11-09 19:49 ` [PATCH 01/13] static_call: Index code text reservation lookups Jiri Olsa
2026-11-12 8:27 ` Peter Zijlstra
2026-11-12 18:33 ` Jiri Olsa
2026-11-12 8:34 ` Peter Zijlstra
2026-11-12 18:33 ` Jiri Olsa
2026-11-12 9:07 ` Peter Zijlstra
2026-11-12 10:11 ` Peter Zijlstra
2026-11-12 18:33 ` Jiri Olsa [this message]
2026-11-09 19:49 ` [PATCH 02/13] jump_label: Index " Jiri Olsa
2026-11-12 8:33 ` Peter Zijlstra
2026-11-12 18:33 ` Jiri Olsa
2026-11-09 19:49 ` [PATCH 03/13] jump_label/x86: Avoid decoding instructions for jump entry size Jiri Olsa
2026-11-12 8:39 ` Peter Zijlstra
2026-11-12 18:33 ` Jiri Olsa
2026-11-09 19:55 ` [PATCH 04/13] kprobes: Maintain an rbtree of registered probes Jiri Olsa
2026-11-10 8:12 ` Masami Hiramatsu (Google)
2026-11-10 10:37 ` Jiri Olsa
2026-11-10 15:38 ` Masami Hiramatsu (Google)
2026-11-11 10:49 ` Jiri Olsa
2026-11-12 8:50 ` Peter Zijlstra
2026-11-09 19:56 ` [PATCH 05/13] kprobes: Add an iterator for address ranges Jiri Olsa
2026-11-09 19:56 ` [PATCH 06/13] kprobes: Add an address iterator for probe lookup Jiri Olsa
2026-11-09 19:56 ` [PATCH 07/13] kprobes: Use kprobe iterator in get_optimized_kprobe Jiri Olsa
2026-11-09 19:57 ` [PATCH 08/13] x86/kprobes: Use kprobe iterator in __recover_optprobed_insn Jiri Olsa
2026-11-09 19:57 ` [PATCH 09/13] x86/kprobes: Use kprobe iterator in arch_check_optimized_kprobe Jiri Olsa
2026-11-09 19:57 ` [PATCH 10/13] x86/kprobes: Use kprobe address iterator in can_optimize Jiri Olsa
2026-11-09 19:57 ` [PATCH 11/13] x86/kprobes: Use kprobe address iterator in can_probe Jiri Olsa
2026-11-09 19:58 ` [PATCH 12/13] x86/kprobes: Use kprobe address iterator in in copy_optimized_instructions Jiri Olsa
2026-11-09 19:58 ` [PATCH 13/13] samples/kprobes: Add batch probe example Jiri Olsa
2026-11-11 2:20 ` kernel test robot
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=avYGX1UpGlvbvHO5@krava \
--to=olsajiri@gmail.com \
--cc=aliceryhl@google.com \
--cc=almagui@kernel.org \
--cc=ardb@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=davem@davemloft.net \
--cc=jbaron@akamai.com \
--cc=jpoimboe@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-trace-kernel@vger.kernel.org \
--cc=mhiramat@kernel.org \
--cc=mingo@redhat.com \
--cc=naveen@kernel.org \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=tglx@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
all inboxes | Powered by JetHome®