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]
Date:   Mon, 2 May 2022 10:49:19 -0700
From:   Matthias Kaehlcke <mka@...omium.org>
To:     Greg KH <greg@...ah.com>
Cc:     Stephen Rothwell <sfr@...b.auug.org.au>,
        Ravi Chandra Sadineni <ravisadineni@...omium.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Linux Next Mailing List <linux-next@...r.kernel.org>,
        Douglas Anderson <dianders@...omium.org>
Subject: Re: linux-next: build failure after merge of the usb tree

On Mon, May 02, 2022 at 07:32:54PM +0200, Greg KH wrote:
> On Mon, May 02, 2022 at 10:09:25AM -0700, Matthias Kaehlcke wrote:
> > On Mon, May 02, 2022 at 01:40:45PM +0200, Greg KH wrote:
> > > On Mon, May 02, 2022 at 09:07:28PM +1000, Stephen Rothwell wrote:
> > > > Hi all,
> > > > 
> > > > After merging the usb tree, today's linux-next build (x86_64
> > > > modules_install) failed like this:
> > > > 
> > > > depmod: ERROR: Cycle detected: usbcore -> onboard_usb_hub -> usbcore
> > > > depmod: ERROR: Found 2 modules in dependency cycles!
> > > > 
> > > > Caused by commit
> > > > 
> > > >   c40b62216c1a ("usb: core: hcd: Create platform devices for onboard hubs in probe()")
> > > > 
> > > > I have reverted that commit for today.
> > > 
> > > argh, I thought the build issues were fixed :(
> > 
> > Sorry about that :(
> > 
> > The *build* issues were actually fixed, however I did not try
> > modules_install ...
> > 
> > > I'll go revert that series from my tree later today, thanks for the
> > > report.
> > 
> > I think the dependency situation can only be resolved by linking
> > onboard_hub_create/destroy_pdevs() into the USB core module. My initial
> > idea was to build them into the kernel binary, however that doesn't
> > work because onboard_hub_create_pdevs() calls usb_of_get_device_node(),
> > which is part of the core module when CONFIG_USB=m. The two function
> > are relatively lightweight and don't depend on internals of the
> > onboard_usb_hub driver (besides the device id table) so linking them
> > into the core module doesn't seem too ugly.
> 
> I'll try to look at this next week, it shouldn't be that complex.  If it
> is, something feels wrong...

Thanks!

The root of all these dependency issues is that the core/HCD code
needs to call APIs of the onboard hub driver to create/destroy
platform devices based on the device tree. That's fine when CONFIG_USB=y
and CONFIG_USB_ONBOARD_HUB=y, but gets complicated when modules
come into play, since the core/HCD code depends on the aforementioned
driver functions and the driver depends on symbols from the USB core.

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ