linux-pci.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Gerd Hoffmann <kraxel@redhat.com>
Cc: linux-pci@vger.kernel.org, Bjorn Helgaas <bhelgaas@google.com>,
	open list <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] pciehp: fast unplug for virtual machines
Date: Thu, 11 Nov 2021 12:00:45 -0500	[thread overview]
Message-ID: <20211111115931-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <20211111090225.946381-1-kraxel@redhat.com>

On Thu, Nov 11, 2021 at 10:02:24AM +0100, Gerd Hoffmann wrote:
> The PCIe specification asks the OS to wait five seconds after the
> attention button has been pressed before actually un-plugging the
> device.  This gives the operator the chance to cancel the operation
> by pressing the attention button again within those five seconds.
> 
> For physical hardware this makes sense.  Picking the wrong button
> by accident can easily happen and it can be corrected that way.
> 
> For virtual hardware the benefits are questionable.  Typically
> users find the five second delay annoying.
> 
> This patch adds the fast_virtual_unplug module parameter to the
> pciehp driver.  When enabled (which is the default) the linux
> kernel will simply skip the delay for virtual pcie ports, which
> reduces the total time for the unplug operation from 6-7 seconds
> to 1-2 seconds.

BTW how come it's still taking seconds, not milliseconds?

> Virtual pcie ports are identified by PCI ID.  For now the qemu
> pcie root ports are detected, other hardware can be added easily.
> 
> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
> ---
>  drivers/pci/hotplug/pciehp.h      |  3 +++
>  drivers/pci/hotplug/pciehp_core.c |  5 +++++
>  drivers/pci/hotplug/pciehp_ctrl.c | 11 ++++++++++-
>  3 files changed, 18 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/hotplug/pciehp.h b/drivers/pci/hotplug/pciehp.h
> index 69fd401691be..131ffec2e947 100644
> --- a/drivers/pci/hotplug/pciehp.h
> +++ b/drivers/pci/hotplug/pciehp.h
> @@ -79,6 +79,7 @@ extern int pciehp_poll_time;
>   * @request_result: result of last user request submitted to the IRQ thread
>   * @requester: wait queue to wake up on completion of user request,
>   *	used for synchronous slot enable/disable request via sysfs
> + * @is_virtual: virtual machine pcie port.
>   *
>   * PCIe hotplug has a 1:1 relationship between controller and slot, hence
>   * unlike other drivers, the two aren't represented by separate structures.
> @@ -109,6 +110,8 @@ struct controller {
>  	unsigned int ist_running;
>  	int request_result;
>  	wait_queue_head_t requester;
> +
> +	bool is_virtual;
>  };
>  
>  /**
> diff --git a/drivers/pci/hotplug/pciehp_core.c b/drivers/pci/hotplug/pciehp_core.c
> index ad3393930ecb..28867ec33f5b 100644
> --- a/drivers/pci/hotplug/pciehp_core.c
> +++ b/drivers/pci/hotplug/pciehp_core.c
> @@ -227,6 +227,11 @@ static int pciehp_probe(struct pcie_device *dev)
>  		goto err_out_shutdown_notification;
>  	}
>  
> +	if (dev->port->vendor == PCI_VENDOR_ID_REDHAT &&
> +	    dev->port->device == 0x000c)
> +		/* qemu pcie root port */
> +		ctrl->is_virtual = true;
> +
>  	pciehp_check_presence(ctrl);
>  
>  	return 0;
> diff --git a/drivers/pci/hotplug/pciehp_ctrl.c b/drivers/pci/hotplug/pciehp_ctrl.c
> index 529c34808440..119bb11f87d6 100644
> --- a/drivers/pci/hotplug/pciehp_ctrl.c
> +++ b/drivers/pci/hotplug/pciehp_ctrl.c
> @@ -15,12 +15,17 @@
>  
>  #define dev_fmt(fmt) "pciehp: " fmt
>  
> +#include <linux/moduleparam.h>
>  #include <linux/kernel.h>
>  #include <linux/types.h>
>  #include <linux/pm_runtime.h>
>  #include <linux/pci.h>
>  #include "pciehp.h"
>  
> +static bool fast_virtual_unplug = true;
> +module_param(fast_virtual_unplug, bool, 0644);
> +MODULE_PARM_DESC(fast_virtual_unplug, "Fast unplug (don't wait 5 seconds) for virtual machines.");
> +
>  /* The following routines constitute the bulk of the
>     hotplug controller logic
>   */
> @@ -176,7 +181,11 @@ void pciehp_handle_button_press(struct controller *ctrl)
>  		/* blink power indicator and turn off attention */
>  		pciehp_set_indicators(ctrl, PCI_EXP_SLTCTL_PWR_IND_BLINK,
>  				      PCI_EXP_SLTCTL_ATTN_IND_OFF);
> -		schedule_delayed_work(&ctrl->button_work, 5 * HZ);
> +		if (ctrl->is_virtual && fast_virtual_unplug) {
> +			schedule_delayed_work(&ctrl->button_work, 1);
> +		} else {
> +			schedule_delayed_work(&ctrl->button_work, 5 * HZ);
> +		}
>  		break;
>  	case BLINKINGOFF_STATE:
>  	case BLINKINGON_STATE:
> -- 
> 2.33.1


  reply	other threads:[~2021-11-11 17:00 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-11  9:02 [PATCH] pciehp: fast unplug for virtual machines Gerd Hoffmann
2021-11-11 17:00 ` Michael S. Tsirkin [this message]
2021-11-12  9:56   ` Gerd Hoffmann
2021-11-12 10:09     ` Michael S. Tsirkin
2021-11-11 21:50 ` Bjorn Helgaas
2021-11-12 11:28   ` Gerd Hoffmann
2021-11-12  0:00 ` Krzysztof Wilczyński
2021-11-14 16:39 ` Lukas Wunner
2021-11-14 17:24   ` Michael S. Tsirkin
2021-11-14 18:06     ` Lukas Wunner
2021-11-14 22:37       ` Michael S. Tsirkin
2021-11-15  9:59       ` Gerd Hoffmann
2021-11-15  7:45     ` Lukas Wunner
2021-11-15 10:09   ` Gerd Hoffmann

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=20211111115931-mutt-send-email-mst@kernel.org \
    --to=mst@redhat.com \
    --cc=bhelgaas@google.com \
    --cc=kraxel@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.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).