mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Milian Wolff <milian.wolff@kdab.com>
To: alexander.levin@verizon.com
Cc: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"stable@vger.kernel.org" <stable@vger.kernel.org>,
	David Ahern <dsahern@gmail.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com>,
	Yao Jin <yao.jin@linux.intel.com>,
	Arnaldo Carvalho de Melo <acme@redhat.com>
Subject: Re: [PATCH AUTOSEL for 4.14 18/51] perf callchain: Compare symbol name for inlined frames when matching
Date: Thu, 23 Nov 2017 09:55:37 +0100	[thread overview]
Message-ID: <2891438.FCgzeiCqBH@milian-kdab2> (raw)
In-Reply-To: <20171122222526.20021-18-alexander.levin@verizon.com>

[-- Attachment #1: Type: text/plain, Size: 6519 bytes --]

On Wednesday, November 22, 2017 11:25:40 PM CET alexander.levin@verizon.com 
wrote:
> From: Milian Wolff <milian.wolff@kdab.com>
> 
> [ Upstream commit 9856240ad3269f2fdab0b2fa4400ef8aab792061 ]

Hello Alexander,

this is the first time I encounter AUTOSEL. I just want to check: The patch 
below depends on others in a whole series that reworks the handling of inline 
frames. Why is only this one getting selected? I don't even think it can work 
stand-alone?

Thanks

> The fake symbols we create for inlined frames will represent different
> functions but can use the symbol start address. This leads to issues
> when different inline branches all lead to the same function.
> 
> Before:
> ~~~~~
> $ perf report -s sym -i perf.inlining.data --inline --stdio -g function
> ...
>              --38.86%--_start
>                        __libc_start_main
>                        main
> 
>                         --37.57%--std::norm<double> (inlined)
>                                   std::_Norm_helper<true>::_S_do_it<double>
> (inlined)
> 
>                                    --36.36%--std::abs<double> (inlined)
>                                              std::__complex_abs (inlined)
> 
>                                              
> --12.24%--std::linear_congruential_engine<unsigned long, 16807ul, 0ul,
> 2147483647ul>::operator() (inlined) std::__detail::__mod<unsigned long,
> 2147483647ul, 16807ul, 0ul> (inlined) std::__detail::_Mod<unsigned long,
> 2147483647ul, 16807ul, 0ul, true, true>::__calc (inlined) ~~~~~
> 
> Note that this backtrace representation is completely bogus.
> Complex abs does not call the linear congruential engine! It
> is just a side-effect of a longer inlined stack being appended
> to a shorter, different inlined stack, both of which originate
> in the same function (main).
> 
> This patch fixes the issue:
> 
> ~~~~~
> $ perf report -s sym -i perf.inlining.data --inline --stdio -g function
> ...
>              --38.86%--_start
>                        __libc_start_main
>                        main
> 
>                        |--35.59%--std::uniform_real_distribution<double>::op
>                        |erator()<std::linear_congruential_engine<unsigned
>                        |long, 16807ul, 0ul, 2147483647ul> > (inlined)                       
|
>                        |          std::uniform_real_distribution<double>::op
>                        |          erator()<std::linear_congruential_engine<u
>                        |          nsigned long, 16807ul, 0ul, 2147483647ul>
>                        |          > (inlined)                       |          
>                        |           --34.37%--std::__detail::_Adaptor<std::li
>                        |           near_congruential_engine<unsigned long,
>                        |           16807ul, 0ul, 2147483647ul>,
>                        |           double>::operator() (inlined)                       
|           
>                        |                     std::generate_canonical<double,
>                        |                     53ul,
>                        |                     std::linear_congruential_engin
>                        |                     e<unsigned long, 16807ul, 0ul,
>                        |                     2147483647ul> > (inlined)                       
|                     
>                        |                      --12.24%--std::linear_congruen
>                        |                      tial_engine<unsigned long,
>                        |                      16807ul, 0ul,
>                        |                      2147483647ul>::operator()
>                        |                      (inlined)                       
|                      
>                        |                                std::__detail::__mod
>                        |                                <unsigned long,
>                        |                                2147483647ul,
>                        |                                16807ul, 0ul>
>                        |                                (inlined)
>                        |                                std::__detail::_Mod<
>                        |                                unsigned long,
>                        |                                2147483647ul,
>                        |                                16807ul, 0ul, true,
>                        |                                true>::__calc
>                        |                                (inlined)
>                         --1.99%--std::norm<double> (inlined)
>                                   std::_Norm_helper<true>::_S_do_it<double>
> (inlined) std::abs<double> (inlined)
>                                   std::__complex_abs (inlined)
> ~~~~~
> 
> Signed-off-by: Milian Wolff <milian.wolff@kdab.com>
> Reviewed-by: Jiri Olsa <jolsa@redhat.com>
> Reviewed-by: Namhyung Kim <namhyung@kernel.org>
> Cc: David Ahern <dsahern@gmail.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Ravi Bangoria <ravi.bangoria@linux.vnet.ibm.com>
> Cc: Yao Jin <yao.jin@linux.intel.com>
> Link: http://lkml.kernel.org/r/20171009203310.17362-10-milian.wolff@kdab.com
> Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
> [ Fix up conflict with c1fbc0cf81f1 ("perf callchain: Compare dsos (as well)
> for CCKEY_FUNCTION"), remove unneeded hunk ] Signed-off-by: Arnaldo
> Carvalho de Melo <acme@redhat.com>
> 
> Signed-off-by: Sasha Levin <alexander.levin@verizon.com>
> ---
>  tools/perf/util/callchain.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
> index 6031933d811c..3cc51e852cf0 100644
> --- a/tools/perf/util/callchain.c
> +++ b/tools/perf/util/callchain.c
> @@ -697,6 +697,14 @@ static enum match_result match_chain(struct
> callchain_cursor_node *node, }
> 
>  	if (cnode->ms.sym && sym && callchain_param.key == CCKEY_FUNCTION) {
> +		/*
> +		 * Compare inlined frames based on their symbol name because
> +		 * different inlined frames will have the same symbol start
> +		 */
> +		if (cnode->ms.sym->inlined || node->sym->inlined)
> +			return match_chain_strings(cnode->ms.sym->name,
> +						   node->sym->name);
> +
>  		left = cnode->ms.sym->start;
>  		right = sym->start;
>  		left_dso = cnode->ms.map->dso;


-- 
Milian Wolff | milian.wolff@kdab.com | Senior Software Engineer
KDAB (Deutschland) GmbH&Co KG, a KDAB Group company
Tel: +49-30-521325470
KDAB - The Qt Experts

[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 3826 bytes --]

  reply	other threads:[~2017-11-23  8:55 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-11-22 22:25 [PATCH AUTOSEL for 4.14 01/51] ima: fix hash algorithm initialization alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 02/51] s390: vfio-ccw: Do not attempt to free no-op, test and tic cda alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 03/51] PM / Domains: Fix genpd to deal with drivers returning 1 from ->prepare() alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 07/51] staging: greybus: loopback: Fix iteration count on async path alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 05/51] selftests/x86/ldt_get: Add a few additional tests for limits alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 08/51] m68k: fix ColdFire node shift size calculation alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 04/51] s390/pci: do not require AIS facility alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 06/51] selftests/x86/ldt_gdt: Robustify against set_thread_area() and LAR oddities alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 12/51] spi: sh-msiof: Fix DMA transfer size check alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 10/51] staging: rtl8822be: fix wrong dma unmap len alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 11/51] staging: rtl8188eu: avoid a null dereference on pmlmepriv alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 09/51] serial: 8250_fintek: Fix rs485 disablement on invalid ioctl() alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 14/51] mmc: tmio: check mmc_regulator_get_supply return value alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 15/51] mmc: sdhci-msm: fix issue with power irq alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 16/51] hwmon: (pmbus/core) Prevent unintentional setting of page to 0xFF alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 13/51] spi: spi-axi: fix potential use-after-free after deregistration alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 18/51] perf callchain: Compare symbol name for inlined frames when matching alexander.levin
2017-11-23  8:55   ` Milian Wolff [this message]
2017-11-28 15:12     ` alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 20/51] usb: dwc2: Error out of dwc2_hsotg_ep_disable() if we're in host mode alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 17/51] perf/core: Fix __perf_read_group_add() locking alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 19/51] usb: dwc2: Fix UDC state tracking alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 22/51] PCI: dra7xx: Create functional dependency between PCIe and PHY alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 21/51] usb: phy: tahvo: fix error handling in tahvo_usb_probe() alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 23/51] x86/intel_rdt: Initialize bitmask of shareable resource if CDP enabled alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 24/51] x86/intel_rdt: Fix potential deadlock during resctrl mount alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 26/51] kprobes: Use synchronize_rcu_tasks() for optprobe with CONFIG_PREEMPT=y alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 27/51] x86/entry: Use SYSCALL_DEFINE() macros for sys_modify_ldt() alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 25/51] serial: 8250: Preserve DLD[7:4] for PORT_XR17V35X alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 30/51] irqchip/gic: Make quirks matching conditional on init return value alexander.levin
2017-11-23  8:50   ` Marc Zyngier
2017-11-28 15:13     ` alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 29/51] dt-bindings: timer: renesas, cmt: Fix SoC-specific compatible values alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 28/51] clocksource/drivers/arm_arch_timer: Validate CNTFRQ after enabling frame alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 31/51] EDAC, sb_edac: Fix missing break in switch alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 33/51] staging: fsl-dpaa2/eth: Account for Rx FD buffers on error path alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 36/51] usb: xhci: Return error when host is dead in xhci_disable_slot() alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 34/51] staging: rtl8822be: Keep array subscript no lower than zero alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 32/51] usb: mtu3: fix error return code in ssusb_gadget_init() alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 35/51] ARM: cpuidle: Correct driver unregistration if init fails alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 37/51] sysrq : fix Show Regs call trace on ARM alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 39/51] usbip: tools: Install all headers needed for libusbip development alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 40/51] serial: imx: Update cached mctrl value when changing RTS alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 38/51] serial: sh-sci: suppress warning for ports without dma channels alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 43/51] perf test attr: Fix python error on empty result alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 44/51] kprobes/x86: Disable preemption in ftrace-based jprobes alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 42/51] perf test attr: Fix ignored test case result alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 41/51] staging: fsl-mc/dpio: Fix incorrect comparison alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 45/51] locking/refcounts, x86/asm: Use unique .text section for refcount exceptions alexander.levin
2017-11-22 23:19   ` Kees Cook
2017-11-23  1:13     ` alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 47/51] tools include: Do not use poison with C++ alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 48/51] perf tools: Fix leaking rec_argv in error cases alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 46/51] s390/ptrace: fix guarded storage regset handling alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 51/51] iio: multiplexer: add NULL check on devm_kzalloc() and devm_kmemdup() return values alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 50/51] iio: adc: ti-ads1015: add 10% to conversion wait time alexander.levin
2017-11-22 22:25 ` [PATCH AUTOSEL for 4.14 49/51] mm, x86/mm: Fix performance regression in get_user_pages_fast() alexander.levin

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=2891438.FCgzeiCqBH@milian-kdab2 \
    --to=milian.wolff@kdab.com \
    --cc=acme@redhat.com \
    --cc=alexander.levin@verizon.com \
    --cc=dsahern@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peterz@infradead.org \
    --cc=ravi.bangoria@linux.vnet.ibm.com \
    --cc=stable@vger.kernel.org \
    --cc=yao.jin@linux.intel.com \
    /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®