linux-kernel - Re: This is the fourth time I’ve tried to find what led to the regression of outgoing network speed and each time I find the merge commit 8c94ccc7cd691472461448f98e2372c75849406c

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Tue, 6 Feb 2024 02:08:02 +0500
From: Mikhail Gavrilov <mikhail.v.gavrilov@...il.com>
To: "Christian A. Ehrhardt" <lk@...e.de>, niklas.neronin@...ux.intel.com, 
	mathias.nyman@...ux.intel.com
Cc: Linux List Kernel Mailing <linux-kernel@...r.kernel.org>, linux-netdev@...r.kernel.org, 
	Greg KH <gregkh@...uxfoundation.org>
Subject: Re: This is the fourth time I’ve tried to find what led to the regression of outgoing network speed and each time I find the merge commit 8c94ccc7cd691472461448f98e2372c75849406c

On Mon, Feb 5, 2024 at 1:47 AM Christian A. Ehrhardt <lk@...e.de> wrote:
>
>
> Sorry, I was looking at the wrong merge commit and when using
> the commit pinpointed by your bisect your log shows
> that _both_ parents of the bad merge commit are marked as good
> which is somewhat strange.
>
> However, it should be possible to bisect further if you do a rebase
> like this:
>
> $ git cat-file -p
> 8c94ccc7cd691472461448f98e2372c75849406c  | head -n 3
> tree d3907cad2a1fbbbcf71847274fdbdcf5a2aeb9a2
> parent bd736f38c014ba70ba7ec3bdc6af6fe5368d6612
> parent 933bb7b878ddd0f8c094db45551a7daddf806e00
> $ git branch m bd736f38c014ba70ba7ec3bdc6af6fe5368d6612
> $ git branch d933bb7b878ddd0f8c094db45551a7daddf806e000
> $ git checkout d
> Updating files: 100% (11666/11666), done.
> Switched to branch 'd'
> $ git rebase m
> Successfully rebased and updated refs/heads/d.
>
> Now, "m" must be good as per your bisect log and "d" must be bad
> because it is the same tree as the bad merge commit (8c94ccc7cd69).
>
> Due to the rebase there's a liner history between the two, thus
> starting a bisect like this might yield more information:
>
> $ git bisect good m
> $ git bisect bad d
>
>      regards   Christian
>
>

Thanks for real help.
Now I spotted a really bad commit.

57e153dfd0e7a080373fe5853c5609443d97fa5a is the first bad commit
commit 57e153dfd0e7a080373fe5853c5609443d97fa5a
Author: Niklas Neronin <niklas.neronin@...ux.intel.com>
Date:   Fri Dec 1 17:06:40 2023 +0200

    xhci: add handler for only one interrupt line

    Current xHCI driver only supports one "interrupter", meaning we will
    only use one MSI/MSI-X interrupt line. Thus, add handler only to the
    first interrupt line.

    Signed-off-by: Niklas Neronin <niklas.neronin@...ux.intel.com>
    Co-developed-by: Mathias Nyman <mathias.nyman@...ux.intel.com>
    Signed-off-by: Mathias Nyman <mathias.nyman@...ux.intel.com>
    Link: https://lore.kernel.org/r/20231201150647.1307406-13-mathias.nyman@linux.intel.com
    Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>

 drivers/usb/host/xhci-pci.c | 35 ++++++++++-------------------------
 1 file changed, 10 insertions(+), 25 deletions(-)

Niklas, Mathias I spotted decreased network speed on sending when
transferring files via sftp between my workstations in the local
network.
And bisection of issue leads me to this commit.
My motherboard is MPG-B650I-EDGE-WIFI looks like it is related to the
mentioned commit.
https://www.msi.com/Motherboard/MPG-B650I-EDGE-WIFI

-- 
Best Regards,
Mike Gavrilov.

Download attachment "speed-measure-3.zip" of type "application/zip" (1019 bytes)

Download attachment ".config.zip" of type "application/zip" (65390 bytes)

Download attachment "git-bisect-log-regression-network-performance.zip" of type "application/zip" (1033 bytes)

Download attachment "dmesg.zip" of type "application/zip" (58258 bytes)