Re: [PATCH 1/2] dev-core: fix build break when DEBUG is enabled

[Greg KH <gregkh@linuxfoundation.org>]

On Thu, Aug 15, 2013 at 07:53:13PM +0300, Dmitry Kasatkin wrote:
> On 15/08/13 19:37, Greg KH wrote:
> > On Thu, Aug 15, 2013 at 07:04:54PM +0300, Dmitry Kasatkin wrote:
> >> When DEBUG is defined, dev_dbg_ratelimited uses dynamic debug data
> >> structures even when CONFIG_DYNAMIC_DEBUG is not defined.
> >> It leads to build break.
> >> For example, when I try to use dev_dbg_ratelimited in USB code and
> >> CONFIG_USB_DEBUG is enabled, but CONFIG_DYNAMIC_DEBUG is not, I get:
> >>
> >>   CC [M]  drivers/usb/host/xhci-ring.o
> >>   drivers/usb/host/xhci-ring.c: In function ‘xhci_queue_intr_tx’:
> >>   drivers/usb/host/xhci-ring.c:3059:3: error: implicit declaration of function ‘DEFINE_DYNAMIC_DEBUG_METADATA’ [-Werror=implicit-function-declaration]
> >>   drivers/usb/host/xhci-ring.c:3059:3: error: ‘descriptor’ undeclared (first use in this function)
> >>   drivers/usb/host/xhci-ring.c:3059:3: note: each undeclared identifier is reported only once for each function it appears in
> >>   drivers/usb/host/xhci-ring.c:3059:3: error: implicit declaration of function ‘__dynamic_pr_debug’ [-Werror=implicit-function-declaration]
> >>   drivers/usb/host/xhci-ring.c: In function ‘xhci_queue_isoc_tx_prepare’:
> >>   drivers/usb/host/xhci-ring.c:3847:3: error: ‘descriptor’ undeclared (first use in this function)
> >>   cc1: some warnings being treated as errors
> >>   make[2]: *** [drivers/usb/host/xhci-ring.o] Error 1
> >>   make[1]: *** [drivers/usb/host] Error 2
> >>   make: *** [drivers/usb/] Error 2
> >>
> >> This patch separates definition for CONFIG_DYNAMIC_DEBUG and DEBUG cases.
> >>
> >> Signed-off-by: Dmitry Kasatkin <d.kasatkin@samsung.com>
> >> Cc: stable@vger.kernel.org
> > How is this a stable issue?  I don't see how the rules listed in
> > Documentation/stable_kernel_rules.txt apply here, what am I missing?
> >
> > Not to say your patch isn't correct, just that it's not stable material,
> > right?
> >
> > thanks,
> >
> > greg k-h
> >
> 
> There are few drivers which uses dev_dbg_ratelimited().
> In the case someone tries to build kernel with those drivers with DEBUG
> enabled and without CONFIG_DYNAMIC_DEBUG,
> will get the build break. It will happen also on stable kernel.

Do you have specific examples of this happening?  Given that this code
has been this way for a very long time now, and Randy's excellent
'random .config bot' normally catches this type of thing, I think the
applicability to a stable release is pretty low.

Please read that Documentation file again about "actual" issues, not
just theoretical ones.

thanks,

greg k-h