Re: [PATCH 6.1 000/219] 6.1.77-rc1 review

[Guenter Roeck]

On 2/3/24 07:42, Greg Kroah-Hartman wrote:
> On Sat, Feb 03, 2024 at 12:23:33AM -0600, Daniel Díaz wrote:
> > Hello!
> >
> > On 02/02/24 10:02 p. m., Greg Kroah-Hartman wrote:
> > > This is the start of the stable review cycle for the 6.1.77 release.
> > > There are 219 patches in this series, all will be posted as a response
> > > to this one.  If anyone has any issues with these being applied, please
> > > let me know.
> > >
> > > Responses should be made by Mon, 05 Feb 2024 03:51:47 +0000.
> > > Anything received after that time might be too late.
> > >
> > > The whole patch series can be found in one patch at:
> > > 	https://www.kernel.org/pub/linux/kernel/v6.x/stable-review/patch-6.1.77-rc1.gz
> > > or in the git tree and branch at:
> > > 	git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-6.1.y
> > > and the diffstat can be found below.
> > >
> > > thanks,
> > >
> > > greg k-h
> >
> > We're seeing lots of build problems and warnings with Clang 17 and Clang nightly:
> >
> > -----8<-----
> >    /builds/linux/drivers/usb/host/xhci.c:1684:37: error: variable 'slot_id' is uninitialized when used here [-Werror,-Wuninitialized]
> >     1684 |                         ret = xhci_check_maxpacket(xhci, slot_id,
> >          |                                                          ^~~~~~~
> >    /builds/linux/drivers/usb/host/xhci.c:1652:22: note: initialize the variable 'slot_id' to silence this warning
> >     1652 |         unsigned int slot_id, ep_index;
> >          |                             ^
> >          |                              = 0
> >    1 error generated.
> >    make[5]: *** [/builds/linux/scripts/Makefile.build:250: drivers/usb/host/xhci.o] Error 1
> > ----->8-----
> >
> > Bisection points to:
> >
> >    commit 37ef029fe9a5639f12250f75f5d1594c6a11e181
> >    Author: Mathias Nyman <mathias.nyman@xxxxxxxxxxxxxxx>
> >    Date:   Fri Dec 1 17:06:47 2023 +0200
> >
> >        xhci: fix possible null pointer deref during xhci urb enqueue
> >        [ Upstream commit e2e2aacf042f52854c92775b7800ba668e0bdfe4 ]
> >
> > Reverting the patch makes the build pass.
>
> Thanks, will drop.  Odd that gcc doesn't catch this as well :(

I have never been able to track down the reason, but gcc on many if not
most architectures doesn't catch problems like this. I often see a warning
pointing to a real problem on, say, parisc builds but not elsewhere, even
though the compiler version is the same.

The above problem is seen because commit e34900f46cd6 ("xhci: Reconfigure
endpoint 0 max packet size only during endpoint reset") removed the call
to xhci_check_maxpacket(), but that commit is not in stable releases.

Guenter