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: <20190604091443.GA458@kroah.com>
Date:   Tue, 4 Jun 2019 11:14:43 +0200
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     Chunfeng Yun <chunfeng.yun@...iatek.com>
Cc:     Felipe Balbi <felipe.balbi@...ux.intel.com>,
        Matthias Brugger <matthias.bgg@...il.com>,
        linux-usb@...r.kernel.org, devicetree@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        linux-mediatek@...ts.infradead.org
Subject: Re: [PATCH v4] usb: create usb_debug_root for gadget only

On Tue, Jun 04, 2019 at 04:59:14PM +0800, Chunfeng Yun wrote:
> On Tue, 2019-06-04 at 09:37 +0200, Greg Kroah-Hartman wrote:
> > On Tue, Jun 04, 2019 at 03:34:07PM +0800, Chunfeng Yun wrote:
> > > When CONFIG_USB is not set, and CONFIG_USB_GADGET is set,
> > > there is an issue, e.g.:
> > > 
> > > drivers/usb/mtu3/mtu3_debugfs.o: in function 'ssusb_debugfs_create_root':
> > > mtu3_debugfs.c:(.text+0xba3): undefined reference to 'usb_debug_root'
> > > 
> > > usb_debug_root is currently only built when host is supported
> > > (CONFIG_USB is set), for convenience, we also want it created when
> > > gadget only is enabled, this patch try to support it.
> > > 
> > > Reported-by: Randy Dunlap <rdunlap@...radead.org>
> > > Signed-off-by: Chunfeng Yun <chunfeng.yun@...iatek.com>
> > > ---
> > > v4:
> > >   move common API into common.c suggested by Felipe
> > > 
> > > v3:
> > >   1. still create usb_debug_root for gadget only
> > >   2. abandon mtu3's change
> > >   3. drop acked-by Randy
> > > 
> > > v2(resend): add acked-by Randy
> > > 
> > > v1: fix mtu3's build error, replace usb_debug_root by NULL;
> > > ---
> > >  drivers/usb/common/common.c   | 26 ++++++++++++++++++++++++++
> > >  drivers/usb/core/usb.c        | 16 ++++------------
> > >  drivers/usb/gadget/udc/core.c |  3 +++
> > >  include/linux/usb.h           |  2 ++
> > >  4 files changed, 35 insertions(+), 12 deletions(-)
> > > 
> > > diff --git a/drivers/usb/common/common.c b/drivers/usb/common/common.c
> > > index 18f5dcf58b0d..c52e9c9f58ec 100644
> > > --- a/drivers/usb/common/common.c
> > > +++ b/drivers/usb/common/common.c
> > > @@ -15,6 +15,7 @@
> > >  #include <linux/usb/of.h>
> > >  #include <linux/usb/otg.h>
> > >  #include <linux/of_platform.h>
> > > +#include <linux/debugfs.h>
> > >  
> > >  static const char *const ep_type_names[] = {
> > >  	[USB_ENDPOINT_XFER_CONTROL] = "ctrl",
> > > @@ -139,6 +140,31 @@ enum usb_dr_mode usb_get_dr_mode(struct device *dev)
> > >  }
> > >  EXPORT_SYMBOL_GPL(usb_get_dr_mode);
> > >  
> > > +struct dentry *usb_debug_root;
> > > +EXPORT_SYMBOL_GPL(usb_debug_root);
> > > +
> > > +static atomic_t usb_debug_root_refcnt = ATOMIC_INIT(0);
> > 
> > Ick, no.
> > 
> > > +
> > > +struct dentry *usb_debugfs_init(void)
> > > +{
> > > +	if (!usb_debug_root)
> > > +		usb_debug_root = debugfs_create_dir("usb", NULL);
> > > +
> > > +	atomic_inc(&usb_debug_root_refcnt);
> > > +
> > > +	return usb_debug_root;
> > > +}
> > > +EXPORT_SYMBOL_GPL(usb_debugfs_init);
> > > +
> > > +void usb_debugfs_cleanup(void)
> > > +{
> > > +	if (atomic_dec_and_test(&usb_debug_root_refcnt)) {
> > > +		debugfs_remove_recursive(usb_debug_root);
> > > +		usb_debug_root = NULL;
> > > +	}
> > > +}
> > > +EXPORT_SYMBOL_GPL(usb_debugfs_cleanup);
> > 
> > Only remove the debugfs subdir if the usbcore module is removed. 
> Both usbcore module and gadget module will use this "usb" subdir now.
> Gadget module may still use it when remove usbcore module.

Did you try to remove the usb-common module with the udc gadget module
remaining loaded with this change in place?  I think you will find that
it is impossible given that the udc gadget module now depends on a
symbol in the usb-common code.

So again, just use the module reference counting logic to drive this
functionality, don't create an atomic variable that duplicates the
logic we already have in place today.

thanks,

greg k-h

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ