Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 2 Jul 2019 14:21:17 +0200
From:   Lukas Wunner <lukas@wunner.de>
To:     linmiaohe <linmiaohe@huawei.com>
Cc:     "bhelgaas@google.com" <bhelgaas@google.com>,
        "andy.shevchenko@gmail.com" <andy.shevchenko@gmail.com>,
        "rafael.j.wysocki@intel.com" <rafael.j.wysocki@intel.com>,
        "linux-pci@vger.kernel.org" <linux-pci@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Mingfangsen <mingfangsen@huawei.com>
Subject: Re: A question about shpcp.
Message-ID: <20190702122117.vl2opo55zpsykjf6@wunner.de>
References: <d87dfa281d0d4e7da3f6bf714a7c2e5a@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <d87dfa281d0d4e7da3f6bf714a7c2e5a@huawei.com>
User-Agent: NeoMutt/20170113 (1.7.2)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Jun 27, 2019 at 12:16:18PM +0000, linmiaohe wrote:
> In qemu+shpcp+pcie scene, hotplug a network card(virtio) would take more
> than 5 seconds. It's because 5 seconds delayed_work in func
> handle_button_press_event with case STATIC_STATE. And this will break some
> protocols with timeout within 5 seconds.
> It's very nice of you if you could tell me why is 5*HZ there and if this
> delay can be reduced?

According to Table 2-4 of the PCI Standard Hot-Plug Controller and
Subsystem Specification, "System software is waiting 5 seconds to
provide the user with an opportunity to cancel the hot-plug operation."
(http://drydkim.com/MyDocuments/PCI%20Spec/specifications/shpc1_0.pdf)

So the reason the delay can't be reduced or removed is because the spec
mandates it.

Thanks,

Lukas