Re: [PATCH] fix UBSAN splat in test_scanf

[Christoph von Recklinghausen]

On 8/10/21 9:31 AM, Andy Shevchenko wrote:
> On Mon, Aug 09, 2021 at 06:20:13AM -0400, Chris von Recklinghausen wrote:
>
> Thanks for the report and fix!
>
> First of all, have you seen this:
> https://lore.kernel.org/lkml/20210727150132.28920-1-andriy.shevchenko@xxxxxxxxxxxxxxx/T/#u
> ?


No I hadn't seen it. I'll defer to this fix then.

Thanks,

Chris


> > On a system with CONFIG_UBSAN_SHIFT=y, the following splat was observed:
> > [  104.676826] test_scanf: loaded.
> > [  104.680862] ================================================================================
> > [  104.690304] UBSAN: shift-out-of-bounds in lib/test_scanf.c:274:51
> > [  104.697144] shift exponent 32 is too large for 32-bit type 'unsigned int'
> As you may see in the above the rule of thumb is to include only meaningful
> parts of the oops.