linux-kernel - Re: [PATCH v3] uio:uio_pci_generic:Don't clear master bit when the process does not exit

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]

Date:   Thu, 9 Mar 2023 12:05:07 +0100
From:   Greg KH <gregkh@...uxfoundation.org>
To:     Su Weifeng <suweifeng1@...wei.com>
Cc:     mst@...hat.com, linux-kernel@...r.kernel.org, shikemeng@...wei.com,
        liuzhiqiang26@...wei.com, linfeilong@...wei.com,
        zhanghongtao22@...wei.com
Subject: Re: [PATCH v3] uio:uio_pci_generic:Don't clear master bit when the
 process does not exit

On Sat, Mar 04, 2023 at 03:43:16PM +0800, Su Weifeng wrote:
> From: Weifeng Su <suweifeng1@...wei.com>
> 
> The /dev/uioX device has concurrent operations in a few scenarios.

And you just broke that :(

> For example, when a process is operating the uio0 device, someone executes
> like "cat /dev/uio0" command. In this case, the bus master bit is cleared
> unconditionally.

It is cleared when the close happens, not when "cat" runs.

So prevent userspace from doing that with permissions, why must the
kernel enforce this policy you are making?

> As a result, the running program cannot work commands
> or I/Os, which is usually unaware of. This happens after
> 865a11f("uio/uio_pci_generic: Disable bus-mastering on release");

Please always reference commits in the documented way.

> The restriction on the process that uses the PCI device is added. The new
> process can be used only after the process that uses the PCI device exits.
> Otherwise, the system returns a message indicating that the device is busy.

Again, you are changing the functionality of the kernel, are you sure
you did not just now break something else?

thanks,

greg k-h