linux-pci.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Lukas Wunner <lukas@wunner.de>
To: "Pali Rohár" <pali@kernel.org>
Cc: linux-pci@vger.kernel.org
Subject: Re: PCI service interrupt handlers & access to PCI config space
Date: Sat, 10 Apr 2021 16:25:24 +0200	[thread overview]
Message-ID: <20210410142524.GA31187@wunner.de> (raw)
In-Reply-To: <20210410122845.nhenihbygmcjlegn@pali>

On Sat, Apr 10, 2021 at 02:28:45PM +0200, Pali Rohár wrote:
> I see that more PCI service drivers in their interrupt handlers are
> accessing PCI config space.
> 
> E.g. in PCIe Hot Plug interrupt handler pciehp_isr handler is called
> pcie_capability_read_word() function.
> 
> It is correct? Because these capability functions are protected by
> pci_lock_config() / pci_unlock_config() calls.

Looks fine to me.  That's a raw_spin_lock, which is allowed to be taken
in interrupt context.

> And what would happen when during execution of reading or writing to
> PCIe config space (e.g. via pcie_capability_read_word()) is triggered
> PCIe HP interrupt? Would not enter interrupt handler function in
> deadlock (waiting for unlocking config space)?

The interrupt handler would spin until the lock is released.
raw_spin_locks are not supposed to be held for a prolonged
period of time.

The interrupt is masked when it triggers and until it's been handled,
so the interrupt handler never runs multiple times concurrently.
See e.g. handle_level_irq() in kernel/irq/chip.c.

Thanks,

Lukas

  reply	other threads:[~2021-04-10 14:25 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-10 12:28 PCI service interrupt handlers & access to PCI config space Pali Rohár
2021-04-10 14:25 ` Lukas Wunner [this message]
2021-04-10 15:17   ` Pali Rohár
2021-04-10 16:26     ` Lukas Wunner
2021-04-12  2:25       ` Keith Busch
2021-04-12 14:04         ` Pali Rohár
2021-04-12 15:09           ` Keith Busch
2021-04-13  9:33             ` Pali Rohár
2021-04-13  9:35       ` Pali Rohár

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=20210410142524.GA31187@wunner.de \
    --to=lukas@wunner.de \
    --cc=linux-pci@vger.kernel.org \
    --cc=pali@kernel.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).