Received-SPF: pass (google.com: domain of linux-kernel+bounces-164187-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Precedence: bulk
MIME-Version: 1.0
References: <CABXGCsNmEtrN9DK-XmESaPm_1xpXm8A+juE+44Jf6AK5JE0+TQ@mail.gmail.com>
 <874jcl7e83.wl-tiwai@suse.de> <CABXGCsMmRFW3EYJ3UvNd-LO7ZTOyoNqjc_3OAmaCDSL=LuxJqg@mail.gmail.com>
 <87v851e2di.wl-tiwai@suse.de> <CABXGCsNMBRUaY-V8mhUQKdq+CQW5+eGUWL_YCJWXo0cgh9bGJQ@mail.gmail.com>
 <87h6glt9zc.wl-tiwai@suse.de> <CABXGCsMcazRvmiN4XtiHQCE9=dB=M=VsRqB=v+RPmtuhBL29DA@mail.gmail.com>
 <a6f4e20ea2a68f56a7d2c4d76280bca44d6bf421.camel@gmail.com>
 <CABXGCsPdqfXeZUw1ocx8O3NdOEb+h4yQ77+zdNpcwP_4JrYXuQ@mail.gmail.com>
 <CABXGCsMTbmU4CP8CHUqRVXWkGiErFtEVG4COy6RSRWsAuK_-CQ@mail.gmail.com>
 <87pluedgx5.wl-tiwai@suse.de> <87jzkmdghh.wl-tiwai@suse.de>
 <CABXGCsP-e5CjqCFztiym=zjQ=Z00uxYcQPFACEbTwjV=BHEdJQ@mail.gmail.com>
 <87jzkfo0ty.wl-tiwai@suse.de> <CABXGCsMPLWm8xVZ3mk-e8FJht4FjdxEREov0DFKMooMD0BvRyw@mail.gmail.com>
 <87r0enm4m0.wl-tiwai@suse.de>
In-Reply-To: <87r0enm4m0.wl-tiwai@suse.de>
From: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Date: Tue, 30 Apr 2024 19:42:32 +0500
Message-ID: <CABXGCsOxkM+ohzudsTg=e5x=UATArAx0zCLnGHFwHHXs9Rz6VA@mail.gmail.com>
Subject: Re: regression/bisected/6.9 commit 587d67fd929ad89801bcc429675bda90d53f6592
 decrease 30% of gaming performance
To: Takashi Iwai <tiwai@suse.de>
Cc: Linux List Kernel Mailing <linux-kernel@vger.kernel.org>, linux-sound@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Tue, Apr 30, 2024 at 5:12=E2=80=AFPM Takashi Iwai <tiwai@suse.de> wrote:
> Where and how...?  That important piece is missing in you info...
> And, it's likely a false-positive detection by lockdep.
>
> When you disable lockdep, does it still show any performance down?

Of course without "lockdep" performance is even better than the commit
587d67fd929a just reverted.
- mainline kernel without changes with lockdep - 13244 [1]
- mainline kernel with reverted 587d67fd929a with lockdep - 19218 [2]
- mainline kernel without changes without lockdep - 22440 [3]

I think I get what you're saying: after a false-positive detection,
lockdep is partially disabled. Yes?
How do I understand that lockdep has partially turned off?
I would expect to see such a message in the kernel log " INFO: lockdep
is turned off."
But in my case this message didn't appear.
In any case, sorry for wasting your time.

[1] https://postimg.cc/G4NPHMyk
[2] https://postimg.cc/1n0D7sYH
[3] https://postimg.cc/nMYHWRcF

--=20
Best Regards,
Mike Gavrilov.