Re: [PATCH] Use maximum latency when determining L1 ASPM

From: Ian Kumlien <ian.kumlien@gmail.com>
To: Bjorn Helgaas <helgaas@kernel.org>
Cc: Kai-Heng Feng <kai.heng.feng@canonical.com>,
	linux-pci <linux-pci@vger.kernel.org>,
	Alexander Duyck <alexander.duyck@gmail.com>,
	"Saheed O. Bolarinwa" <refactormyself@gmail.com>,
	Puranjay Mohan <puranjay12@gmail.com>
Subject: Re: [PATCH] Use maximum latency when determining L1 ASPM
Date: Sun, 18 Oct 2020 13:35:27 +0200	[thread overview]
Message-ID: <CAA85sZufMEAieVgzxdPrbCzaPV0eM_NYX7idWkLVxQaJrYjC+A@mail.gmail.com> (raw)
In-Reply-To: <CAA85sZuOh8ZtnWj8svxK_9er7hLskTjc0ASUVkvAkv9Rc=Bh_g@mail.gmail.com>

On Sat, Oct 17, 2020 at 12:41 AM Ian Kumlien <ian.kumlien@gmail.com> wrote:
>
> On Fri, Oct 16, 2020 at 11:28 PM Bjorn Helgaas <helgaas@kernel.org> wrote:
> >
> > On Wed, Oct 14, 2020 at 03:33:17PM +0200, Ian Kumlien wrote:
> > > On Wed, Oct 14, 2020 at 10:34 AM Kai-Heng Feng
> > > <kai.heng.feng@canonical.com> wrote:
> > > > > On Oct 12, 2020, at 18:20, Ian Kumlien <ian.kumlien@gmail.com> wrote:
> > > > > On Thu, Oct 8, 2020 at 6:13 PM Bjorn Helgaas <helgaas@kernel.org> wrote:
> >
> > > > >> OK, now we're getting close.  We just need to flesh out the
> > > > >> justification.  We need:
> > > > >>
> > > > >>  - Tidy subject line.  Use "git log --oneline drivers/pci/pcie/aspm.c"
> > > > >>    and follow the example.
> > > > >
> > > > > Will do
> > > > >
> > > > >>  - Description of the problem.  I think it's poor bandwidth on your
> > > > >>    Intel I211 device, but we don't have the complete picture because
> > > > >>    that NIC is 03:00.0, which doesn't appear above at all.
> > > > >
> > > > > I think we'll use Kai-Hengs issue, since it's actually more related to
> > > > > the change itself...
> > > > >
> > > > > Mine is a side effect while Kai-Heng is actually hitting an issue
> > > > > caused by the bug.
> > > >
> > > > I filed a bug here:
> > > > https://bugzilla.kernel.org/show_bug.cgi?id=209671
> > >
> > > Thanks!
> >
> > Sigh.  I feel like I'm just not getting anywhere here.  I still do not
> > have a "before" and "after" set of lspci output.
> >
> > Kai-Heng's bugzilla has two sets of output, but one is a working
> > config with CONFIG_PCIEASPM_DEFAULT=y and the other is a working
> > config with Ian's patch applied and CONFIG_PCIEASPM_POWERSAVE=y.
> >
> > Comparing them doesn't show the effect of Ian's patch; it shows the
> > combined effect of Ian's patch and the CONFIG_PCIEASPM_POWERSAVE=y
> > change.  I'm not really interested in spending a few hours trying to
> > reverse-engineer the important changes.
> >
> > Can you please, please, collect these on your system, Ian?  I assume
> > that you can easily collect it once without your patch, when you see
> > poor I211 NIC performance but the system is otherwise working.  And
> > you can collect it again *with* your patch.  Same Kconfig, same
> > *everything* except adding your patch.
>
> Yeah I can do that, but I would like the changes output from the
> latest patch suggestion
> running on Kai-Heng's system so we can actually see what it does...

Is:
https://bugzilla.kernel.org/show_bug.cgi?id=209725

More to your liking?

> > Bjorn