linux-kernel - Re: [PATCH V5 3/3] PCI: Mask and unmask hotplug interrupts during reset

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20180729123251.GB4364@wunner.de>
Date:   Sun, 29 Jul 2018 14:32:51 +0200
From:   Lukas Wunner <lukas@...ner.de>
To:     poza@...eaurora.org
Cc:     okaya@...eaurora.org, linux-pci@...r.kernel.org,
        linux-arm-msm@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        Keith Busch <keith.busch@...el.com>,
        open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH V5 3/3] PCI: Mask and unmask hotplug interrupts during
 reset

On Tue, Jul 03, 2018 at 06:41:33PM +0530, poza@...eaurora.org wrote:
> pciehp_unconfigure_device doing little more than enumeration to quiescence
> the bus.
> 
>		/*
> 		 * Ensure that no new Requests will be generated from
> 		 * the device.
> 		 */
> 		if (presence) {
> 			pci_read_config_word(dev, PCI_COMMAND, &command);
> 			command &= ~(PCI_COMMAND_MASTER | PCI_COMMAND_SERR);
> 			command |= PCI_COMMAND_INTX_DISABLE;
> 			pci_write_config_word(dev, PCI_COMMAND, command);
> 		}

That piece of code is supposed to be executed on safe removal via sysfs
or an Attention Button press:  The card remains in the slot, even though
the slot is brought down.  So the card is quiesced.

However IIUC, on fatal error the link goes down and for pciehp, that's
essentially a surprise removal.  In that case, the above code is not
intended to be executed, rather the devices below the hotplug bridge
are marked disconnected.  See this patch I posted yesterday:
https://www.spinics.net/lists/linux-pci/msg74763.html

Thanks,

Lukas