qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Paolo Bonzini <pbonzini@redhat.com>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: Alexey Kardashevskiy <aik@ozlabs.ru>,
	Yongji Xie <xyjxie@linux.vnet.ibm.com>,
	QEMU Developers <qemu-devel@nongnu.org>,
	Alex Williamson <alex.williamson@redhat.com>,
	zhong@linux.vnet.ibm.com,
	David Gibson <david@gibson.dropbear.id.au>,
	Paul Mackerras <paulus@samba.org>
Subject: Re: [Qemu-devel] [PATCH] memory: make ram device read/write endian sensitive
Date: Thu, 23 Feb 2017 13:53:50 +0100	[thread overview]
Message-ID: <41630a89-e645-7d7e-b7c2-356fd6dcadee@redhat.com> (raw)
In-Reply-To: <CAFEAcA9gcArSqhbRi6-YdR=5eE9dtq66C7N0qTbCuBiY_Zr+tA@mail.gmail.com>



On 23/02/2017 13:26, Peter Maydell wrote:
> On 23 February 2017 at 11:43, Paolo Bonzini <pbonzini@redhat.com> wrote:
>> On 23/02/2017 12:34, Peter Maydell wrote:
>>> We should probably update the doc comment to note that the
>>> pointer is to host-endianness memory (and that this is not
>>> like normal RAM which is target-endian)...
>>
>> I wouldn't call it host-endianness memory, and I disagree that normal
>> RAM is target-endian---in both cases it's just a bunch of bytes.
>>
>> However, the access done by the MemoryRegionOps callbacks needs to match
>> the endianness declared by the MemoryRegionOps themselves.
> 
> Well, if the guest stores a bunch of integers to the memory, which
> way round do you see them when you look at the bunch of bytes?

You see them in whatever endianness the guest used.  So let's say a
little-endian guest is writing 0x12345678 to a register.  With
DEVICE_HOST_ENDIAN, some examples are:

- little-endian host, little-endian device.  The guest will write
0x12345678, QEMU will write 0x12345678 which becomes 0x78 0x56 0x34
0x12.  The little-endian device is happy because it sees it as 0x12345678.

- little-endian host, big-endian device.  The guest will write
0x78563412 to account for the different endianness of the host.  QEMU
will also write 0x78563412 which becomes 0x12 0x34 0x56 0x78.  The
big-endian device is happy because it sees it as 0x12345678.

- big-endian host, little-endian device.  The guest will write
0x12345678, memory.c will flip it and QEMU will write 0x78563412.  In
bytes this becomes 0x78 0x56 0x34 0x12.  The little-endian device is
happy because it sees it as 0x12345678.

- big-endian host, big-endian device.  The guest will write 0x78563412
to account for the different endianness of the host.  Memory.c will flip
it and QEMU will write 0x12345678, which becomes 0x12 0x34 0x56 0x78.
The big-endian device is happy because it sees it as 0x12345678.

With DEVICE_LITTLE_ENDIAN you add two flips on big-endian hosts; with
DEVICE_BIG_ENDIAN you add two flips on little-endian hosts.

Paolo

> The answer is different for the two things, so something is different
> about their endianness...

  reply	other threads:[~2017-02-23 12:54 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-02-21  6:46 [Qemu-devel] [PATCH] memory: make ram device read/write endian sensitive Yongji Xie
2017-02-21 16:21 ` Alex Williamson
2017-02-21 16:34   ` Paolo Bonzini
2017-02-21 18:09     ` Peter Maydell
2017-02-21 18:44       ` Alex Williamson
2017-02-22  7:54         ` Yongji Xie
2017-02-22 10:53         ` Paolo Bonzini
2017-02-21 18:53       ` Paolo Bonzini
2017-02-21 19:40         ` Peter Maydell
2017-02-23  4:20 ` Alexey Kardashevskiy
2017-02-23  8:35   ` Paolo Bonzini
2017-02-23 10:02     ` Peter Maydell
2017-02-23 10:10       ` Paolo Bonzini
2017-02-23 10:23         ` Peter Maydell
2017-02-23 10:33           ` Paolo Bonzini
2017-02-23 11:34             ` Peter Maydell
2017-02-23 11:43               ` Paolo Bonzini
2017-02-23 12:26                 ` Peter Maydell
2017-02-23 12:53                   ` Paolo Bonzini [this message]
2017-02-23 14:35                     ` Peter Maydell
2017-02-23 15:21                       ` Paolo Bonzini
2017-02-23 15:29                         ` Peter Maydell
2017-02-23 15:58                           ` Paolo Bonzini
2017-02-23 16:08                             ` Peter Maydell
2017-02-23 16:15                               ` Paolo Bonzini
2017-02-23 17:14                                 ` Yongji Xie
2017-02-24  3:28                                   ` David Gibson
2017-02-23 23:36                           ` Paul Mackerras
2017-02-23 15:39                         ` Alex Williamson
2017-02-23 15:47                           ` Paolo Bonzini
2017-02-23 16:08                             ` Alex Williamson
2017-02-24  3:26                 ` David Gibson
2017-02-23 11:04     ` Alexey Kardashevskiy
2017-02-27  2:25   ` Michael Roth
2017-02-27  3:25     ` Alexey Kardashevskiy
2017-02-27  4:28       ` Yongji Xie

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=41630a89-e645-7d7e-b7c2-356fd6dcadee@redhat.com \
    --to=pbonzini@redhat.com \
    --cc=aik@ozlabs.ru \
    --cc=alex.williamson@redhat.com \
    --cc=david@gibson.dropbear.id.au \
    --cc=paulus@samba.org \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=xyjxie@linux.vnet.ibm.com \
    --cc=zhong@linux.vnet.ibm.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 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).