On Tue, 14 Sep 2021 11:01:31 -0700 Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > All of these are just plain wrong: I'll revert these, and expect Masami to address the issues you present. Note, they do fix a real bug as shown by: https://lore.kernel.org/all/e5c7ce5b-93f0-b305-de32-c99d0390eb28@xxxxxxx/ > > > +static void *init_xbc_data_copy __initdata; > > +static phys_addr_t init_xbc_data_size __initdata; > > + init_xbc_data_copy = copy; > > + init_xbc_data_size = size + 1; > > + memblock_free(__pa(init_xbc_data_copy), init_xbc_data_size); > > because the xbc code already saves these as xbc_data/xbc_data_size and > that final free should just be done in xbc_destroy_all(). > > So this fix is pointlessly ugly to begin with. > > But what I _really_ ended up reacting to was that > > > + memblock_free(__pa(copy), size + 1); Should we hold off until mm has fixed this issue? -- Steve