Re: [PATCH rdma-next 1/4] gcov: Open-code kmemdup() to work correctly with kernel and user space pointers

[Leon Romanovsky]

On Wed, Sep 02, 2020 at 11:27:14AM -0700, Linus Torvalds wrote:
> On Wed, Sep 2, 2020 at 10:46 AM Leon Romanovsky <leon@xxxxxxxxxx> wrote:
> >
> > Definitely my explanation is wrong, but it was my interpretation of
> > "BUG: KASAN: global-out-of-bounds in kmemdup+0x43/0x70" line. I saw
> > that the failure was in memcpy() inside of kmemdup(), so I changed
> > from memcpy to be copy_from_user() and it solved the KASAN warning.
>
> But the actual patch attached to that explanation *doesn't* use
> copy_from_user().
>
> So your "changed from memcpy to be copy_from_user() solved the KASAN
> warning" explanation makes even less sense. Because that's not at all
> what the patch does.

I already don't remember why, but copy_from_user() caused to second flow
of gcov_info_dup() through gcov_event() to generate another set of warnings.

As a summary, I have a workaround, but I don't know why it works and not
proud about it.

Thanks

>
>              Linus