All of lore.kernel.org
 help / color / mirror / Atom feed
From: Oleksandr Andrushchenko <Oleksandr_Andrushchenko@epam.com>
To: "Roger Pau Monné" <roger.pau@citrix.com>
Cc: "xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>,
	"julien@xen.org" <julien@xen.org>,
	"sstabellini@kernel.org" <sstabellini@kernel.org>,
	Oleksandr Tyshchenko <Oleksandr_Tyshchenko@epam.com>,
	Volodymyr Babchuk <Volodymyr_Babchuk@epam.com>,
	Artem Mygaiev <Artem_Mygaiev@epam.com>,
	"jbeulich@suse.com" <jbeulich@suse.com>,
	"andrew.cooper3@citrix.com" <andrew.cooper3@citrix.com>,
	"george.dunlap@citrix.com" <george.dunlap@citrix.com>,
	"paul@xen.org" <paul@xen.org>,
	Bertrand Marquis <bertrand.marquis@arm.com>,
	Rahul Singh <rahul.singh@arm.com>,
	Oleksandr Andrushchenko <Oleksandr_Andrushchenko@epam.com>
Subject: Re: [PATCH v5 05/14] vpci: add hooks for PCI device assign/de-assign
Date: Tue, 1 Feb 2022 08:56:49 +0000	[thread overview]
Message-ID: <ef8feba1-695f-5cd6-bdf6-8ef6c55e62e1@epam.com> (raw)
In-Reply-To: <f4ea5ece-551e-8a91-90c6-d9f806c97d5b@epam.com>

Hi, Roger!

On 31.01.22 10:45, Oleksandr Andrushchenko wrote:
> Hi, Roger!
>
> On 13.01.22 13:40, Roger Pau Monné wrote:
>> On Thu, Nov 25, 2021 at 01:02:42PM +0200, Oleksandr Andrushchenko wrote:
>>> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
>>> +#ifdef CONFIG_HAS_VPCI_GUEST_SUPPORT
>>> +/* Notify vPCI that device is assigned to guest. */
>>> +int vpci_assign_device(struct domain *d, struct pci_dev *pdev)
>>> +{
>>> +    int rc;
>>> +
>>> +    /* It only makes sense to assign for hwdom or guest domain. */
>>> +    if ( is_system_domain(d) || !has_vpci(d) )
>>> +        return 0;
>>> +
>>> +    spin_lock(&pdev->vpci_lock);
>>> +    rc = run_vpci_init(pdev);
>> Following my comment below, this will likely need to call
>> vpci_add_handlers in order to allocate the pdev->vpci field.
>>
>> It's not OK to carry the contents of pdev->vpci across domain
>> assignations, as the device should be reset, and thus the content of
>> pdev->vpci would be stale.
>>
>>> +    spin_unlock(&pdev->vpci_lock);
>>> +    if ( rc )
>>> +        vpci_deassign_device(d, pdev);
>>> +
>>> +    return rc;
>>> +}
>>> +
>>> +/* Notify vPCI that device is de-assigned from guest. */
>>> +int vpci_deassign_device(struct domain *d, struct pci_dev *pdev)
>>> +{
>>> +    /* It only makes sense to de-assign from hwdom or guest domain. */
>>> +    if ( is_system_domain(d) || !has_vpci(d) )
>>> +        return 0;
>>> +
>>> +    spin_lock(&pdev->vpci_lock);
>>> +    vpci_remove_device_handlers_locked(pdev);
>> You need to free the pdev->vpci structure on deassign. I would expect
>> the device to be reset on deassign, so keeping the pdev->vpci contents
>> would be wrong.
> Sure, I will re-allocate pdev->vpci then
After thinking a bit more on this I have realized that we cannot free
pdev->vpci on de-assign. The reason for that is the fact that vpci
structure contains vital data which is collected and managed at different
stages: for example, BAR types are collected while we run for the
hardware domain and in init_bars we collect the types of the BARS etc.
This is then used while assigning device to construct guest's representation
of the device. Freeing vpci will lead to that data is lost and the required
data is not populated into vpci.
So, it is no possible to free vpci structure and I am about to leave the
approach as it is.
>> Thanks, Roger.
> Thank you,
> Oleksandr
Thank you,
Oleksandr

  reply	other threads:[~2022-02-01  8:57 UTC|newest]

Thread overview: 130+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-25 11:02 [PATCH v5 00/14] PCI devices passthrough on Arm, part 3 Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 01/14] rangeset: add RANGESETF_no_print flag Oleksandr Andrushchenko
2021-11-25 11:06   ` Jan Beulich
2021-11-25 11:08     ` Oleksandr Andrushchenko
2021-12-15  3:20   ` Volodymyr Babchuk
2021-12-15  5:53     ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 02/14] vpci: fix function attributes for vpci_process_pending Oleksandr Andrushchenko
2021-12-10 17:55   ` Julien Grall
2021-12-11  8:20     ` Roger Pau Monné
2021-12-11  8:57       ` Oleksandr Andrushchenko
2022-01-26  8:31         ` Oleksandr Andrushchenko
2022-01-26 10:54           ` Jan Beulich
2021-11-25 11:02 ` [PATCH v5 03/14] vpci: move lock outside of struct vpci Oleksandr Andrushchenko
2022-01-11 15:17   ` Roger Pau Monné
2022-01-12 14:42     ` Jan Beulich
2022-01-26  8:40       ` Oleksandr Andrushchenko
2022-01-26 11:13         ` Roger Pau Monné
2022-01-31  7:41           ` Oleksandr Andrushchenko
2022-01-12 14:57   ` Jan Beulich
2022-01-12 15:42     ` Roger Pau Monné
2022-01-12 15:52       ` Jan Beulich
2022-01-13  8:58         ` Roger Pau Monné
2022-01-28 14:15           ` Oleksandr Andrushchenko
2022-01-31  8:56             ` Roger Pau Monné
2022-01-31  9:00               ` Oleksandr Andrushchenko
2022-01-28 14:12     ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 04/14] vpci: cancel pending map/unmap on vpci removal Oleksandr Andrushchenko
2022-01-11 16:57   ` Roger Pau Monné
2022-01-12 15:27   ` Jan Beulich
2022-01-28 12:21     ` Oleksandr Andrushchenko
2022-01-31  7:53   ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 05/14] vpci: add hooks for PCI device assign/de-assign Oleksandr Andrushchenko
2022-01-12 12:12   ` Roger Pau Monné
2022-01-31  8:43     ` Oleksandr Andrushchenko
2022-01-13 11:40   ` Roger Pau Monné
2022-01-31  8:45     ` Oleksandr Andrushchenko
2022-02-01  8:56       ` Oleksandr Andrushchenko [this message]
2022-02-01 10:23         ` Roger Pau Monné
2021-11-25 11:02 ` [PATCH v5 06/14] vpci/header: implement guest BAR register handlers Oleksandr Andrushchenko
2021-11-25 16:28   ` Bertrand Marquis
2021-11-26 12:19     ` Oleksandr Andrushchenko
2022-02-03 12:36       ` Oleksandr Andrushchenko
2022-02-03 12:44         ` Jan Beulich
2022-02-03 12:48           ` Oleksandr Andrushchenko
2022-02-03 12:50             ` Jan Beulich
2022-02-03 12:53               ` Oleksandr Andrushchenko
2022-01-12 12:35   ` Roger Pau Monné
2022-01-31  9:47     ` Oleksandr Andrushchenko
2022-01-31 10:40       ` Oleksandr Andrushchenko
2022-01-31 10:54         ` Jan Beulich
2022-01-31 11:04           ` Oleksandr Andrushchenko
2022-01-31 11:27             ` Roger Pau Monné
2022-01-31 11:30               ` Oleksandr Andrushchenko
2022-01-31 11:10         ` Roger Pau Monné
2022-01-31 11:23           ` Oleksandr Andrushchenko
2022-01-31 11:31             ` Roger Pau Monné
2022-01-31 11:39             ` Jan Beulich
2022-01-31 13:30               ` Oleksandr Andrushchenko
2022-01-31 13:36                 ` Jan Beulich
2022-01-31 13:41                   ` Oleksandr Andrushchenko
2022-01-31 13:51                     ` Jan Beulich
2022-01-31 13:58                       ` Oleksandr Andrushchenko
2022-01-31 11:04       ` Roger Pau Monné
2022-01-31 14:51         ` Oleksandr Andrushchenko
2022-01-31 15:06     ` Oleksandr Andrushchenko
2022-01-31 15:50       ` Jan Beulich
2022-02-01  7:31         ` Oleksandr Andrushchenko
2022-02-01 10:10           ` Roger Pau Monné
2022-02-01 10:41             ` Oleksandr Andrushchenko
2022-01-12 17:34   ` Roger Pau Monné
2022-01-31  9:53     ` Oleksandr Andrushchenko
2022-01-31 10:56       ` Roger Pau Monné
2022-02-03 12:45       ` Oleksandr Andrushchenko
2022-02-03 12:54         ` Jan Beulich
2022-02-03 13:30           ` Oleksandr Andrushchenko
2022-02-03 14:04             ` Jan Beulich
2022-02-03 14:19               ` Oleksandr Andrushchenko
2022-02-03 14:05             ` Roger Pau Monné
2022-02-03 14:26               ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 07/14] vpci/header: handle p2m range sets per BAR Oleksandr Andrushchenko
2022-01-12 15:15   ` Roger Pau Monné
2022-01-12 15:18     ` Jan Beulich
2022-02-02  6:44     ` Oleksandr Andrushchenko
2022-02-02  9:56       ` Roger Pau Monné
2022-02-02 10:02         ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 08/14] vpci/header: program p2m with guest BAR view Oleksandr Andrushchenko
2022-01-13 10:22   ` Roger Pau Monné
2022-02-02  8:23     ` Oleksandr Andrushchenko
2022-02-02  9:46       ` Oleksandr Andrushchenko
2022-02-02 10:34         ` Roger Pau Monné
2022-02-02 10:44           ` Oleksandr Andrushchenko
2022-02-02 11:11             ` Jan Beulich
2022-02-02 11:14               ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 09/14] vpci/header: emulate PCI_COMMAND register for guests Oleksandr Andrushchenko
2022-01-13 10:50   ` Roger Pau Monné
2022-02-02 12:49     ` Oleksandr Andrushchenko
2022-02-02 13:32       ` Jan Beulich
2022-02-02 13:47         ` Oleksandr Andrushchenko
2022-02-02 14:18           ` Jan Beulich
2022-02-02 14:26             ` Oleksandr Andrushchenko
2022-02-02 14:31               ` Jan Beulich
2022-02-02 15:04                 ` Oleksandr Andrushchenko
2022-02-02 15:08                   ` Jan Beulich
2022-02-02 15:12                     ` Oleksandr Andrushchenko
2022-02-02 15:31                       ` Jan Beulich
2021-11-25 11:02 ` [PATCH v5 10/14] vpci/header: reset the command register when adding devices Oleksandr Andrushchenko
2022-01-13 11:07   ` Roger Pau Monné
2022-02-02 12:58     ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 11/14] vpci: add initial support for virtual PCI bus topology Oleksandr Andrushchenko
2022-01-12 15:39   ` Jan Beulich
2022-02-02 13:15     ` Oleksandr Andrushchenko
2022-01-13 11:35   ` Roger Pau Monné
2022-02-02 13:17     ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 12/14] xen/arm: translate virtual PCI bus topology for guests Oleksandr Andrushchenko
2022-01-13 12:18   ` Roger Pau Monné
2022-02-02 13:58     ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 13/14] xen/arm: account IO handlers for emulated PCI MSI-X Oleksandr Andrushchenko
2022-01-13 13:23   ` Roger Pau Monné
2022-02-02 14:08     ` Oleksandr Andrushchenko
2021-11-25 11:02 ` [PATCH v5 14/14] vpci: add TODO for the registers not explicitly handled Oleksandr Andrushchenko
2021-11-25 11:17   ` Jan Beulich
2021-11-25 11:20     ` Oleksandr Andrushchenko
2022-01-13 13:27     ` Roger Pau Monné
2022-01-13 13:38       ` Jan Beulich
2022-01-28 13:03         ` Oleksandr Andrushchenko
2021-12-15 11:56 ` [PATCH v5 00/14] PCI devices passthrough on Arm, part 3 Oleksandr Andrushchenko
2021-12-15 12:07   ` Jan Beulich
2021-12-15 12:22     ` Oleksandr Andrushchenko
2021-12-15 14:51       ` Roger Pau Monné
2021-12-15 15:02         ` Oleksandr Andrushchenko

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=ef8feba1-695f-5cd6-bdf6-8ef6c55e62e1@epam.com \
    --to=oleksandr_andrushchenko@epam.com \
    --cc=Artem_Mygaiev@epam.com \
    --cc=Oleksandr_Tyshchenko@epam.com \
    --cc=Volodymyr_Babchuk@epam.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=bertrand.marquis@arm.com \
    --cc=george.dunlap@citrix.com \
    --cc=jbeulich@suse.com \
    --cc=julien@xen.org \
    --cc=paul@xen.org \
    --cc=rahul.singh@arm.com \
    --cc=roger.pau@citrix.com \
    --cc=sstabellini@kernel.org \
    --cc=xen-devel@lists.xenproject.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 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.