linux-kernel - Re: [PATCH 2.6.26] PCI: refuse to re-add a device to a bus upon pci_scan_child

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

Message-ID: <488C5570.1050305@extricom.com>
Date:	Sun, 27 Jul 2008 14:01:04 +0300
From:	Eran Liberty <liberty@...ricom.com>
To:	Eran Liberty <liberty@...ricom.com>
CC:	Matthew Wilcox <matthew@....cx>,
	eran liberty <eran.liberty@...il.com>,
	linux-kernel@...r.kernel.org, linux-pci@...r.kernel.org,
	Jesse Barnes <jbarnes@...tuousgeek.org>
Subject: Re: [PATCH 2.6.26] PCI: refuse to re-add a device to a bus upon pci_scan_child_bus()

Eran Liberty wrote:

> Soooo ...
>   
> If I really wanted to make it, working my way around the current code, I would have done something like this.
>   
> bus = null;
>   
> while ((bus = pci_find_next_bus(bus)) != NULL) {
>   
>     for (devfn = 0; devfn < 0x100; devfn += 8) {
>   
>         for (func = 0; func < 8; func++) {
>   
>             struct pci_dev *dev =  <http://liberty/lxr/ident?v=e500-linux-2.6.26-rc4;i=pci_get_slot>pci_get_slot(struct pci_bus *bus, unsigned int devfn);
>   
>             if (dev) 
>                 continue;
>   
>             dev = pci_scan_single_device(bus, devfn);
>   
>             if (!dev)
>   
>                 continue;
>   
>             pci_device_add(dev, bus);
>   
>         }
>   
>     }
>   
>        pci_bus_assign_resources(bus);
>   
>        pci_bus_add_devices(bus);
>   
> }
>   
Found some time to test my own idea... it seems to work. :)

Its final version looks like this:

Before reloading the programmable unit, remove all the device implemented by it

	struct pci_dev *dev;
	while ((dev = pci_get_device(PCI_VENDOR_ID_MYCOMP,PCI_DEVICE_ID_MYCOMP_MYDEV,NULL)) != NULL) {
		pci_remove_bus_device(dev);
		pci_dev_put(dev);
	}

After the programmable unit is loaded, scan and add only newly added device!

 	struct pci_bus *bus = NULL;
	while ((bus = pci_find_next_bus(bus)) != NULL) {
		int devfn;
		for (devfn = 0; devfn < 0x100; devfn++) {
			struct pci_dev *dev=pci_get_slot(bus,devfn);
			if (dev) 
				continue;  // <---- do not scan devices already present on the bus. Missing magic line :)
			(void)pci_scan_single_device(bus, devfn);
		}
		pci_bus_assign_resources(bus);
		pci_bus_add_devices(bus);
	}

So... no need for patch after all.
Unless... Hey whats this... behind it... on the side.... naaa, no need of patch... :(

Liberty

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/