Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <64d41701-55a4-e526-17ae-8936de4bc1ef@suse.de> <20190824051605.GA63850@shbuild999.sh.intel.com>
 <1b897bfe-fd40-3ae3-d867-424d1fc08c44@suse.de> <d114b0b6-6b64-406e-6c3f-a8b8d5502413@intel.com>
 <44029e80-ba00-8246-dec0-fda122d53f5e@suse.de> <90e78ce8-d46a-5154-c324-a05aa1743c98@intel.com>
 <2e1b4d65-d477-f571-845d-fa0a670859af@suse.de> <20190904062716.GC5541@shbuild999.sh.intel.com>
 <72c33bf1-9184-e24a-c084-26d9c8b6f9b7@suse.de> <CAKMK7uGdOtyDHZMSzY8J45bX57EFKo=DWNUi+WL+GVOzoBpUhw@mail.gmail.com>
 <20190904083558.GD5541@shbuild999.sh.intel.com> <CAKMK7uGVKEN=pi4Erc_gtbL3ZFN-b6pm-nXSznjd_rH4H2yn4w@mail.gmail.com>
 <CAPM=9tzDMfRf_VKaiHmnb_KKVwqW3=y=09JO0SJrG6ySe=DbfQ@mail.gmail.com>
In-Reply-To: <CAPM=9tzDMfRf_VKaiHmnb_KKVwqW3=y=09JO0SJrG6ySe=DbfQ@mail.gmail.com>
From:   Daniel Vetter <daniel@ffwll.ch>
Date:   Wed, 4 Sep 2019 13:20:29 +0200
Message-ID: <CAKMK7uGtNu0M74+Ag5-7HJhuHDVv1HoMPz=2XjU6tCkfMScQnA@mail.gmail.com>
Subject: Re: [LKP] [drm/mgag200] 90f479ae51: vm-scalability.median -18.8% regression
To:     Dave Airlie <airlied@gmail.com>
Cc:     Feng Tang <feng.tang@intel.com>,
        Stephen Rothwell <sfr@canb.auug.org.au>,
        Rong Chen <rong.a.chen@intel.com>,
        =?UTF-8?Q?Michel_D=C3=A4nzer?= <michel@daenzer.net>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        dri-devel <dri-devel@lists.freedesktop.org>,
        Thomas Zimmermann <tzimmermann@suse.de>, LKP <lkp@01.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Sep 4, 2019 at 1:15 PM Dave Airlie <airlied@gmail.com> wrote:
>
> On Wed, 4 Sep 2019 at 19:17, Daniel Vetter <daniel@ffwll.ch> wrote:
> >
> > On Wed, Sep 4, 2019 at 10:35 AM Feng Tang <feng.tang@intel.com> wrote:
> > >
> > > Hi Daniel,
> > >
> > > On Wed, Sep 04, 2019 at 10:11:11AM +0200, Daniel Vetter wrote:
> > > > On Wed, Sep 4, 2019 at 8:53 AM Thomas Zimmermann <tzimmermann@suse.de> wrote:
> > > > >
> > > > > Hi
> > > > >
> > > > > Am 04.09.19 um 08:27 schrieb Feng Tang:
> > > > > >> Thank you for testing. But don't get too excited, because the patch
> > > > > >> simulates a bug that was present in the original mgag200 code. A
> > > > > >> significant number of frames are simply skipped. That is apparently the
> > > > > >> reason why it's faster.
> > > > > >
> > > > > > Thanks for the detailed info, so the original code skips time-consuming
> > > > > > work inside atomic context on purpose. Is there any space to optmise it?
> > > > > > If 2 scheduled update worker are handled at almost same time, can one be
> > > > > > skipped?
> > > > >
> > > > > To my knowledge, there's only one instance of the worker. Re-scheduling
> > > > > the worker before a previous instance started, will not create a second
> > > > > instance. The worker's instance will complete all pending updates. So in
> > > > > some way, skipping workers already happens.
> > > >
> > > > So I think that the most often fbcon update from atomic context is the
> > > > blinking cursor. If you disable that one you should be back to the old
> > > > performance level I think, since just writing to dmesg is from process
> > > > context, so shouldn't change.
> > >
> > > Hmm, then for the old driver, it should also do the most update in
> > > non-atomic context?
> > >
> > > One other thing is, I profiled that updating a 3MB shadow buffer needs
> > > 20 ms, which transfer to 150 MB/s bandwidth. Could it be related with
> > > the cache setting of DRM shadow buffer? say the orginal code use a
> > > cachable buffer?
> >
> > Hm, that would indicate the write-combining got broken somewhere. This
> > should definitely be faster. Also we shouldn't transfer the hole
> > thing, except when scrolling ...
>
> First rule of fbcon usage, you are always effectively scrolling.
>
> Also these devices might be on a PCIE 1x piece of wet string, not sure
> if the numbers reflect that.

pcie 1x 1.0 is 250MB/s, so yeah with a bit of inefficiency and
overhead not entirely out of the question that 150MB/s is actually the
hw limit. If it's really pcie 1x 1.0, no idea where to check that.
Also might be worth to double-check that the gpu pci bar is listed as
wc in debugfs/x86/pat_memtype_list.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch