From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-13.4 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,MAILING_LIST_MULTI,MENTIONS_GIT_HOSTING, NICE_REPLY_A,SPF_HELO_NONE,SPF_PASS autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9E15AC433FE for ; Tue, 14 Sep 2021 00:39:00 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 7DA7C61130 for ; Tue, 14 Sep 2021 00:39:00 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S238815AbhINAkP (ORCPT ); Mon, 13 Sep 2021 20:40:15 -0400 Received: from mail.kernel.org ([198.145.29.99]:56836 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229549AbhINAkN (ORCPT ); Mon, 13 Sep 2021 20:40:13 -0400 Received: by mail.kernel.org (Postfix) with ESMTPSA id B028A60FBF; Tue, 14 Sep 2021 00:38:53 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1631579937; bh=NROVtiiKAsV1pJjMRHzw3+WInkMNdoRv0LJPub82ur4=; h=Date:From:To:Cc:Subject:In-Reply-To:References:From; b=pw5UzlRTmjvlPG7uuxHo3+/UOJAGIW6kz3+awtrN1NrUxsNZn+gGKpIk83+V2uTfM GMysa5dCDuA9UklKyUyKxlaPvUiUQDUCXvcw2HqZ+2N4h49td7Ud5HMvA+5N54Tj/+ S2saNKRu1KsvZTVFdysMegCmTdTXV2qSNb54vpwT2j+kLI9koxY0couTG9UiQmJoKI b6ZVZCevqX2BoGdTcZfRo++7O9GH4thr3jpWB6QiSWSugUwxAsB9NV7zW3gEBHFx+n tWW3hVG4hsPlw+M2bAlsxuIFarks7kpSVQzvfNGcMC/NQDhUGbZFlDvwPOCnYz2qa1 UCZnm5vBe5AHQ== Date: Tue, 14 Sep 2021 09:38:52 +0900 From: Masami Hiramatsu To: Andrii Nakryiko Cc: Josh Poimboeuf , Ingo Molnar , Steven Rostedt , X86 ML , Daniel Xu , open list , bpf , Jakub Kicinski , Ingo Molnar , Alexei Starovoitov , Thomas Gleixner , Borislav Petkov , Peter Zijlstra , Kernel Team , Yonghong Song , linux-ia64@vger.kernel.org, Abhishek Sagar , "Paul E . McKenney" Subject: Re: [PATCH -tip v10 00/16] kprobes: Fix stacktrace with kretprobes on x86 Message-Id: <20210914093852.9ed9c70cbc414c0aa3ae5304@kernel.org> In-Reply-To: References: <162756755600.301564.4957591913842010341.stgit@devnote2> <20210730083549.4e36df1cba88e408dc60b031@kernel.org> <20210824143242.a0558b6632eef0407282364e@kernel.org> X-Mailer: Sylpheed 3.7.0 (GTK+ 2.24.32; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 13 Sep 2021 10:14:55 -0700 Andrii Nakryiko wrote: > On Mon, Aug 23, 2021 at 10:32 PM Masami Hiramatsu wrote: > > > > On Mon, 23 Aug 2021 22:12:06 -0700 > > Andrii Nakryiko wrote: > > > > > On Thu, Jul 29, 2021 at 4:35 PM Masami Hiramatsu wrote: > > > > > > > > On Thu, 29 Jul 2021 23:05:56 +0900 > > > > Masami Hiramatsu wrote: > > > > > > > > > Hello, > > > > > > > > > > This is the 10th version of the series to fix the stacktrace with kretprobe on x86. > > > > > > > > > > The previous version is here; > > > > > > > > > > https://lore.kernel.org/bpf/162601048053.1318837.1550594515476777588.stgit@devnote2/ > > > > > > > > > > This version is rebased on top of new kprobes cleanup series(*1) and merging > > > > > Josh's objtool update series (*2)(*3) as [6/16] and [7/16]. > > > > > > > > > > (*1) https://lore.kernel.org/bpf/162748615977.59465.13262421617578791515.stgit@devnote2/ > > > > > (*2) https://lore.kernel.org/bpf/20210710192433.x5cgjsq2ksvaqnss@treble/ > > > > > (*3) https://lore.kernel.org/bpf/20210710192514.ghvksi3ozhez4lvb@treble/ > > > > > > > > > > Changes from v9: > > > > > - Add Josh's objtool update patches with a build error fix as [6/16] and [7/16]. > > > > > - Add a API document for kretprobe_find_ret_addr() and check cur != NULL in [5/16]. > > > > > > > > > > With this series, unwinder can unwind stack correctly from ftrace as below; > > > > > > > > > > # cd /sys/kernel/debug/tracing > > > > > # echo > trace > > > > > # echo 1 > options/sym-offset > > > > > # echo r vfs_read >> kprobe_events > > > > > # echo r full_proxy_read >> kprobe_events > > > > > # echo traceoff:1 > events/kprobes/r_vfs_read_0/trigger > > > > > # echo stacktrace:1 > events/kprobes/r_full_proxy_read_0/trigger > > > > > # echo 1 > events/kprobes/enable > > > > > # cat /sys/kernel/debug/kprobes/list > > > > > ffffffff8133b740 r full_proxy_read+0x0 [FTRACE] > > > > > ffffffff812560b0 r vfs_read+0x0 [FTRACE] > > > > > # echo 0 > events/kprobes/enable > > > > > # cat trace > > > > > # tracer: nop > > > > > # > > > > > # entries-in-buffer/entries-written: 3/3 #P:8 > > > > > # > > > > > # _-----=> irqs-off > > > > > # / _----=> need-resched > > > > > # | / _---=> hardirq/softirq > > > > > # || / _--=> preempt-depth > > > > > # ||| / delay > > > > > # TASK-PID CPU# |||| TIMESTAMP FUNCTION > > > > > # | | | |||| | | > > > > > <...>-134 [007] ...1 16.185877: r_full_proxy_read_0: (vfs_read+0x98/0x180 <- full_proxy_read) > > > > > <...>-134 [007] ...1 16.185901: > > > > > => kretprobe_trace_func+0x209/0x300 > > > > > => kretprobe_dispatcher+0x4a/0x70 > > > > > => __kretprobe_trampoline_handler+0xd4/0x170 > > > > > => trampoline_handler+0x43/0x60 > > > > > => kretprobe_trampoline+0x2a/0x50 > > > > > => vfs_read+0x98/0x180 > > > > > => ksys_read+0x5f/0xe0 > > > > > => do_syscall_64+0x37/0x90 > > > > > => entry_SYSCALL_64_after_hwframe+0x44/0xae > > > > > <...>-134 [007] ...1 16.185902: r_vfs_read_0: (ksys_read+0x5f/0xe0 <- vfs_read) > > > > > > > > > > This shows the double return probes (vfs_read() and full_proxy_read()) on the stack > > > > > correctly unwinded. (vfs_read() returns to 'ksys_read+0x5f' and full_proxy_read() > > > > > returns to 'vfs_read+0x98') > > > > > > > > > > This also changes the kretprobe behavisor a bit, now the instraction pointer in > > > > > the 'pt_regs' passed to kretprobe user handler is correctly set the real return > > > > > address. So user handlers can get it via instruction_pointer() API, and can use > > > > > stack_trace_save_regs(). > > > > > > > > > > You can also get this series from > > > > > git://git.kernel.org/pub/scm/linux/kernel/git/mhiramat/linux.git kprobes/kretprobe-stackfix-v9 > > > > > > > > Oops, this is of course 'kprobes/kretprobe-stackfix-v10'. And this branch includes above (*1) series. > > > > > > Hi Masami, > > > > > > Was this ever merged/applied? This is a very important functionality > > > for BPF kretprobes, so I hope this won't slip through the cracks. > > > > No, not yet as far as I know. > > I'm waiting for any comment on this series. Since this is basically > > x86 ORC unwinder improvement, this series should be merged to -tip tree. > > > > Hey Masami, > > It's been a while since you posted v10. It seems like this series > doesn't apply cleanly anymore. Do you mind rebasing and resubmitting > it again to refresh the series and make it easier for folks to review > and test it? Yes, I'm planning to do that this week soon. Thank you for ping me :) > > Also, do I understand correctly that [0] is a dependency of this > series? If yes, please rebase and resubmit that one as well. Not sure > on the status of Josh's patches you have dependency on as well. Can > you please coordinate with him and maybe incorporate them into your > series? Sorry I can not see [0], could you tell me another URL or title? Or is that Kees's patch [1]? [1] https://lore.kernel.org/all/20210903021326.206548-1-keescook@chromium.org/T/#u > > Please also cc Paul McKenney for the future > revisions so he can follow along as well? Thanks! OK! > > > [0] https://patchwork.kernel.org/project/netdevbpf/list/?series=522757&state=* > > > > > Ingo and Josh, > > > > Could you give me any comment, please? > > > > Thank you, > > > > > > > Thanks! > > > > > > > > > > > Thank you, > > > > > > > > -- > > > > Masami Hiramatsu > > > > > > -- > > Masami Hiramatsu -- Masami Hiramatsu