qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Laszlo Ersek <lersek@redhat.com>
To: Marcel Apfelbaum <marcel@redhat.com>
Cc: "Gabriel L. Somlo (GMail)" <gsomlo@gmail.com>,
	edk2-devel@lists.sourceforge.net,
	qemu devel list <qemu-devel@nongnu.org>,
	"Michael S. Tsirkin" <mst@redhat.com>
Subject: Re: [Qemu-devel] [RFC] edk2 support for a new QEMU device - PXB (PCI Expander Device)
Date: Thu, 04 Jun 2015 18:27:09 +0200	[thread overview]
Message-ID: <55707C5D.9060000@redhat.com> (raw)
In-Reply-To: <55706C59.8000908@redhat.com>

On 06/04/15 17:18, Marcel Apfelbaum wrote:
> On 06/04/2015 04:04 PM, Laszlo Ersek wrote:
>> On 06/04/15 11:42, Marcel Apfelbaum wrote:
>>> On 06/04/2015 02:11 AM, Laszlo Ersek wrote:
>>
>>>> What element type do you propose for the array in the new fw_cfg file?
>>>> (And what name for the fw_cfg file itself?)
>>>>
>>>> "etc/extra-pci-roots" uses uint64_t, little endian, for the number of
>>>> extra root buses. (In fact if you expose the explicit list in a
>>>> separate
>>>> file, then the element count is not even necessary separately, because
>>>> file sizes are available in the fw_cfg directory, and I can divide the
>>>> file size with the element size.)
>>
>>> I can prepare another file.
>>
>> As long as we're crossing neither a QEMU nor a SeaBIOS release boundary,
>> I think we could just change the contents of the same file, with the
>> existing name.
> The extra-roots file was existing before PXB.
> I am afraid to break some other thing.
> This is why I prefer another file.

Noted.

>>> Regarding the new  array, each element
>>> should be
>>> a number between 0x0 and 0xff, so a uint8_t seems fair.
>>
>> Hm. The number of bytes to save here is really small, and it has been
>> suggested to maybe try to support segments? I don't know anything about
>> PCI segments; I vaguely recall that it allows for disjoint bus
>> intervals, with each interval having at most 256 elements. Maybe we
>> could accommodate that with a uint32_t element type?

> While I dont' really care about the type,
> Pmultiple pci segments correspond to multiple *host bridges*,
> as opposed to one host bridge with multiple root bridges.

Noted. UINT8 is fine then.

Thanks!
Laszlo

      reply	other threads:[~2015-06-04 16:27 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <556DC60B.20402@redhat.com>
     [not found] ` <556DD8F7.7020502@redhat.com>
     [not found]   ` <556EB7B7.5030004@redhat.com>
     [not found]     ` <556ED4D0.1030105@redhat.com>
     [not found]       ` <556F64C1.7010406@redhat.com>
2015-06-03 23:11         ` [Qemu-devel] [RFC] edk2 support for a new QEMU device - PXB (PCI Expander Device) Laszlo Ersek
2015-06-04  9:42           ` Marcel Apfelbaum
2015-06-04 13:04             ` Laszlo Ersek
2015-06-04 14:21               ` Michael S. Tsirkin
2015-06-04 15:18               ` Marcel Apfelbaum
2015-06-04 16:27                 ` Laszlo Ersek [this message]

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=55707C5D.9060000@redhat.com \
    --to=lersek@redhat.com \
    --cc=edk2-devel@lists.sourceforge.net \
    --cc=gsomlo@gmail.com \
    --cc=marcel@redhat.com \
    --cc=mst@redhat.com \
    --cc=qemu-devel@nongnu.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).