linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Lukas Wunner <lukas@wunner.de>
To: Stuart Hayes <stuart.w.hayes@gmail.com>
Cc: Bjorn Helgaas <bhelgaas@google.com>,
	Austin Bolen <austin_bolen@dell.com>,
	Keith Busch <kbusch@kernel.org>,
	Alexandru Gagniuc <mr.nuke.me@gmail.com>,
	"Rafael J . Wysocki" <rafael.j.wysocki@intel.com>,
	Mika Westerberg <mika.westerberg@linux.intel.com>,
	Andy Shevchenko <andy.shevchenko@gmail.com>,
	"Gustavo A . R . Silva" <gustavo@embeddedor.com>,
	Sinan Kaya <okaya@kernel.org>,
	Oza Pawandeep <poza@codeaurora.org>,
	linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org,
	narendra_k@dell.com
Subject: Re: [PATCH v3] PCI: pciehp: Make sure pciehp_isr clears interrupt events
Date: Sun, 9 Feb 2020 19:07:22 +0100	[thread overview]
Message-ID: <20200209180722.ikuyjignnd7ddfp5@wunner.de> (raw)
In-Reply-To: <20200209150328.2x2zumhqbs6fihmc@wunner.de>

On Sun, Feb 09, 2020 at 04:03:28PM +0100, Lukas Wunner wrote:
> Using a for (;;) or do/while loop that you jump out of if
> (!status || !pci_dev_msi_enabled(pdev)) might be more readable
> than a goto, but I'm not sure.

Actually, scratch that.  After thinking about this problem for a day
I've come up with a much simpler and more elegant solution.  Could you
test if the below works for you?

This solution has the added benefit that the IRQ thread is started up
once the first event bits have been collected.  If more event bits are
found in the additional loop iterations, they're added to the collected
event bits and the IRQ thread will pick them up asynchronously.  Once no
more bits are found, the hardirq handler exits with IRQ_NONE.  This means
that the genirq code won't wake the IRQ thread but that doesn't matter
because the ISR has already done that itself.  Should also work correctly
in poll mode and the behavior in INTx mode should be as before.

-- >8 --
diff --git a/drivers/pci/hotplug/pciehp_hpc.c b/drivers/pci/hotplug/pciehp_hpc.c
index c3e3f53..707324d 100644
--- a/drivers/pci/hotplug/pciehp_hpc.c
+++ b/drivers/pci/hotplug/pciehp_hpc.c
@@ -553,6 +553,7 @@ static irqreturn_t pciehp_isr(int irq, void *dev_id)
 		}
 	}
 
+read_status:
 	pcie_capability_read_word(pdev, PCI_EXP_SLTSTA, &status);
 	if (status == (u16) ~0) {
 		ctrl_info(ctrl, "%s: no response from device\n", __func__);
@@ -609,6 +610,17 @@ static irqreturn_t pciehp_isr(int irq, void *dev_id)
 
 	/* Save pending events for consumption by IRQ thread. */
 	atomic_or(events, &ctrl->pending_events);
+
+	/*
+	 * In MSI mode, all event bits must be zero before the port will send
+	 * a new interrupt (PCIe Base Spec r5.0 sec 6.7.3.4).  So re-read the
+	 * Slot Status register in case a bit was set between read and write.
+	 */
+	if (pci_dev_msi_enabled(pdev) && !pciehp_poll_mode) {
+		irq_wake_thread(irq, ctrl);
+		goto read_status;
+	}
+
 	return IRQ_WAKE_THREAD;
 }
 

  reply	other threads:[~2020-02-09 18:07 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-02-07 19:54 [PATCH v3] PCI: pciehp: Make sure pciehp_isr clears interrupt events Stuart Hayes
2020-02-09 15:03 ` Lukas Wunner
2020-02-09 18:07   ` Lukas Wunner [this message]
2020-02-09 20:25     ` Lukas Wunner
2020-02-10 21:40       ` Stuart Hayes
2020-02-13 20:19         ` Stuart Hayes
2020-02-19 15:57       ` Lukas Wunner
2020-02-19 14:31 ` [PATCH v4] PCI: pciehp: Fix MSI interrupt race Lukas Wunner
2020-03-06 19:59   ` Stuart Hayes
2020-03-20 15:16   ` Joerg Roedel
2020-03-28 21:27   ` Bjorn Helgaas

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=20200209180722.ikuyjignnd7ddfp5@wunner.de \
    --to=lukas@wunner.de \
    --cc=andy.shevchenko@gmail.com \
    --cc=austin_bolen@dell.com \
    --cc=bhelgaas@google.com \
    --cc=gustavo@embeddedor.com \
    --cc=kbusch@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=mika.westerberg@linux.intel.com \
    --cc=mr.nuke.me@gmail.com \
    --cc=narendra_k@dell.com \
    --cc=okaya@kernel.org \
    --cc=poza@codeaurora.org \
    --cc=rafael.j.wysocki@intel.com \
    --cc=stuart.w.hayes@gmail.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).