All of lore.kernel.org
 help / color / mirror / Atom feed
From: dmukhin@xen.org
To: Mykola Kvach <xakep.amatop@gmail.com>
Cc: xen-devel@lists.xenproject.org, andrew.cooper3@citrix.com,
	anthony.perard@vates.tech, jbeulich@suse.com, julien@xen.org,
	michal.orzel@amd.com, roger.pau@citrix.com,
	sstabellini@kernel.org, dmukhin@ford.com
Subject: Re: [PATCH v7 01/16] emul/vuart: introduce framework for UART emulators
Date: Sat, 13 Sep 2025 11:09:35 -0700	[thread overview]
Message-ID: <aMWzXzL3bhSkKNEp@kraken> (raw)
In-Reply-To: <CAGeoDV92gvzfF4fEo2KBPhvYba2ULK5yW2LGBBQ2e8z2FU2yyQ@mail.gmail.com>

On Wed, Sep 10, 2025 at 10:57:14AM +0300, Mykola Kvach wrote:
> > +int vuart_init(struct domain *d, const struct vuart_info *info)
> > +{
> > +    const struct vuart_emulator *emulator;
> > +    struct vuart *vuart;
> > +    int rc;
> > +
> > +    if ( d->console.vuart )
> > +        return -EBUSY;
> > +
> > +    emulator = vuart_match_by_compatible(d, info->compatible);
> > +    if ( !emulator )
> > +        return -ENODEV;
> > +
> > +    vuart = xzalloc(typeof(*vuart));
> > +    if ( !vuart )
> > +        return -ENOMEM;
> > +
> > +    vuart->info = xvzalloc(typeof(*vuart->info));
> > +    if ( !vuart->info )
> > +    {
> > +        rc = -ENOMEM;
> > +        goto err_out1;
> > +    }
> > +    memcpy(vuart->info, info, sizeof(*info));
> > +
> > +    vuart->vdev = emulator->alloc(d, vuart->info);
> > +    if ( IS_ERR(vuart->vdev) )
> > +    {
> > +        rc = PTR_ERR(vuart->vdev);
> > +        goto err_out2;
> > +    }
> > +
> > +    vuart->emulator = emulator;
> > +    vuart->owner = d;
> > +    vuart->flags |= VUART_CONSOLE_INPUT;
> > +
> > +    d->console.input_allowed = true;
> 
> I'm not a specialist in the area of consoles, but I'm wondering:
> Does the input_allowed flag serve the same purpose as
> VUART_CONSOLE_INPUT? If so, do we need both, or
> could one be removed to simplify the code?

At this point these two flags must be in sync.

`console.input_allowed` is a permission for a domain to take the physical
console input.

`VUART_CONSOLE_INPUT` in `vuart->flags` is a permission for vUART to take
the physical console input.

pvshim does not have vUART, but can have console focus.

And not every vUART can be configured to have a console input (e.g. Arm's
"hwdom vuart").

> > +/*
> > + * Put character to the first emulated UART's FIFO with the physical console
> > + * forwarding enabled.
> > + *
> > + * Must be called under rcu_lock_domain().
> > + */
> > +int vuart_put_rx(struct domain *d, char c)
> > +{
> > +    const struct vuart *vuart = vuart_find_by_flags(d, VUART_CONSOLE_INPUT);
> 
> The call to vuart_find_by_flags() with VUART_CONSOLE_INPUT in
> vuart_put_rx() appears unnecessary. Every vUART console is always
> initialized with VUART_CONSOLE_INPUT, so even if multiple consoles
> exist, the search will always return the first console. It would be
> simpler and clearer to use d->console.vuart directly.

There's no certain order in which multiple vUARTs are initialized, so there
should be something which scans the vUART list and selects the vUART with
console input permission. Follow on Arm's change will add multiple vUARTs
and I decided to generalize logic in this patch, so that there will be 
minimal update in vuart_find_by_flags() only.

> 
> Consider updating the function to remove the flag-based search and add a
> short comment explaining why checking the flag isn’t needed. This will
> help avoid confusion for future maintainers. Alternatively, we could
> pass flags to the init functions instead of hardcoding
> VUART_CONSOLE_INPUT for every console.

That will work, thanks for suggestion.


  reply	other threads:[~2025-09-13 18:09 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-08 21:11 [PATCH v7 00/16] x86: introduce NS16550-compatible UART emulator dmukhin
2025-09-08 21:11 ` [PATCH v7 01/16] emul/vuart: introduce framework for UART emulators dmukhin
2025-09-10  7:57   ` Mykola Kvach
2025-09-13 18:09     ` dmukhin [this message]
2025-09-08 21:11 ` [PATCH v7 02/16] xen/8250-uart: update definitions dmukhin
2025-09-09 10:05   ` Jan Beulich
2025-09-09 19:42     ` dmukhin
2025-09-10  8:39     ` Mykola Kvach
2025-09-13 17:50       ` dmukhin
2025-09-08 21:11 ` [PATCH v7 03/16] emul/ns16x50: implement emulator stub dmukhin
2025-09-10 10:05   ` Mykola Kvach
2025-09-13 17:29     ` dmukhin
2025-11-14  5:19     ` dmukhin
2025-09-15 10:16   ` Mykola Kvach
2025-11-14  5:28     ` dmukhin
2025-09-08 21:11 ` [PATCH v7 04/16] emul/ns16x50: implement DLL/DLM registers dmukhin
2025-09-10 10:16   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 05/16] emul/ns16x50: implement SCR register dmukhin
2025-09-12  7:23   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 06/16] emul/ns16x50: implement IER/IIR registers dmukhin
2025-09-15  6:00   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 07/16] emul/ns16x50: implement LCR/LSR registers dmukhin
2025-09-15  6:00   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 08/16] emul/ns16x50: implement MCR/MSR registers dmukhin
2025-09-15  6:00   ` Mykola Kvach
2025-09-15 14:49     ` Jan Beulich
2025-09-16  8:00       ` Mykola Kvach
2025-09-16 14:13         ` Jan Beulich
2025-09-08 21:11 ` [PATCH v7 09/16] emul/ns16x50: implement RBR register dmukhin
2025-11-18  6:00   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 10/16] emul/ns16x50: implement THR register dmukhin
2025-11-18  6:00   ` Mykola Kvach
2026-05-14 23:23     ` dmukhin
2025-09-08 21:11 ` [PATCH v7 11/16] emul/ns16x50: implement FCR register (write-only) dmukhin
2025-11-18  6:00   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 12/16] emul/ns16550: implement dump_state() hook dmukhin
2025-11-18  6:00   ` Mykola Kvach
2026-05-14 23:35     ` dmukhin
2025-09-08 21:11 ` [PATCH v7 13/16] emul/ns16x50: add Kconfig options dmukhin
2025-11-18  6:00   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 14/16] x86/domain: enable per-domain I/O port bitmaps dmukhin
2025-11-18  6:00   ` Mykola Kvach
2026-05-14 23:52     ` dmukhin
2025-09-08 21:11 ` [PATCH v7 15/16] xen/domain: allocate d->irq_caps before arch-specific initialization dmukhin
2025-11-18  6:00   ` Mykola Kvach
2025-09-08 21:11 ` [PATCH v7 16/16] emul/ns16x50: implement IRQ emulation via vIOAPIC dmukhin

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=aMWzXzL3bhSkKNEp@kraken \
    --to=dmukhin@xen.org \
    --cc=andrew.cooper3@citrix.com \
    --cc=anthony.perard@vates.tech \
    --cc=dmukhin@ford.com \
    --cc=jbeulich@suse.com \
    --cc=julien@xen.org \
    --cc=michal.orzel@amd.com \
    --cc=roger.pau@citrix.com \
    --cc=sstabellini@kernel.org \
    --cc=xakep.amatop@gmail.com \
    --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.