On Tue, 19 Jan 2021 13:55:29 -0600
Timur Tabi <ti...@kernel.org> wrote:
> > case DUMP_PREFIX_ADDRESS:
> > printk("%s%s%p: %s\n",
> > - level, prefix_str, ptr + i, linebuf);
> > + level, prefix_str, addr + i, linebuf);
>
> Well, this is better than nothing, but not by much. Again, as long as
> %px exists for printk(), I just cannot understand any resistance to
> allowing it in print_hex_dump().
>
> Frankly, I think this patch and my patch should both be added. During
> debugging, it's very difficult if not impossible to work with hashed
> addresses. I use print_hex_dump() with an unhashed address all the
> time, either by applying my patch to my own kernel or just replacing the
> %p with %px.
I'm curious, what is the result if you replaced %p with %pS?
That way you get a kallsyms offset version of the output, which could still
be very useful depending on what you are dumping.
-- Steve
