All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ajay Kaher <akaher@vmware.com>
To: "agraf@csgraf.de" <agraf@csgraf.de>,
	"bhelgaas@google.com" <bhelgaas@google.com>
Cc: "linux-pci@vger.kernel.org" <linux-pci@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
	"linux-hyperv@vger.kernel.org" <linux-hyperv@vger.kernel.org>,
	"Krzysztof Wilczyński" <kw@linux.com>,
	"Ashwin Dayanand Kamat" <kashwindayan@vmware.com>,
	"Alexey Makhalov" <amakhalov@vmware.com>,
	"Srivatsa Bhat" <srivatsab@vmware.com>,
	"Vasavi Sirnapalli" <vsirnapalli@vmware.com>,
	"Nadav Amit" <namit@vmware.com>,
	"Alexander Graf" <graf@amazon.com>
Subject: Follow up questions on LPC 2022: MC: Instant Detection of Virtual Devices
Date: Mon, 19 Sep 2022 17:50:52 +0000	[thread overview]
Message-ID: <6AA2CAD2-C799-46A0-907C-03BB802C4FA1@vmware.com> (raw)


Hi Bjorn, Alexander,

Thanks for attending and raising question during our talk on "Instant Detection of Virtual Devices".
Recoding link for reference: https://youtu.be/0mIO54oQ2zI?t=7151


Bjorn, I hope I understood your questions properly and replied. If some doubts please let us know.

Alexander, you wanted to know about following (please correct if we understood something wrong):
Why we still have ~300 read/write, even after having following progressive approaches:
- MMIO Direct Read
- Skip write to PCIe configs
- Pre-config of PCIe configs

[1] remaining writes:
As per PCI specification, needs to perform following operation to read BAR addr and size:
- Read current value from BAR
- Write all-ones to BAR
- Read new value from BAR
- Write original value back to BAR

Same applies to following configs as well: 
- BARs
- I/O Base/Limit
- Memory Base/Limit
- Prefetchable Memory Base/Limit
- Prefetchable Base Limit/Upper 32 bits
- I/O Base/Limit Upper 16 bits

These write operations couldn't be eliminated with our progressive approaches and major of remaining writes are because of these operations.


[2] remaining reads:
Before initialize MMIO, guest performs very few PCI config reads over PIO. I guess these can't be eliminated. 


--Ajay


                 reply	other threads:[~2022-09-19 17:51 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=6AA2CAD2-C799-46A0-907C-03BB802C4FA1@vmware.com \
    --to=akaher@vmware.com \
    --cc=agraf@csgraf.de \
    --cc=amakhalov@vmware.com \
    --cc=bhelgaas@google.com \
    --cc=graf@amazon.com \
    --cc=kashwindayan@vmware.com \
    --cc=kw@linux.com \
    --cc=linux-hyperv@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=namit@vmware.com \
    --cc=srivatsab@vmware.com \
    --cc=vsirnapalli@vmware.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.