Re: [PATCH] of/fdt: Make sure no-map does not remove already reserved regions

[Nicolas Boichat]

On Wed, Jul 17, 2019 at 7:17 AM Florian Fainelli <f.fainelli@xxxxxxxxx> wrote:
>
> On 7/16/19 4:12 PM, Rob Herring wrote:
> > On Tue, Jul 16, 2019 at 4:46 PM Florian Fainelli <f.fainelli@xxxxxxxxx> wrote:
> >>
> >> On 7/2/19 10:08 PM, Nicolas Boichat wrote:
> >>> If the device tree is incorrectly configured, and attempts to
> >>> define a "no-map" reserved memory that overlaps with the kernel
> >>> data/code, the kernel would crash quickly after boot, with no
> >>> obvious clue about the nature of the issue.
> >>>
> >>> For example, this would happen if we have the kernel mapped at
> >>> these addresses (from /proc/iomem):
> >>> 40000000-41ffffff : System RAM
> >>>   40080000-40dfffff : Kernel code
> >>>   40e00000-411fffff : reserved
> >>>   41200000-413e0fff : Kernel data
> >>>
> >>> And we declare a no-map shared-dma-pool region at a fixed address
> >>> within that range:
> >>> mem_reserved: mem_region {
> >>>       compatible = "shared-dma-pool";
> >>>       reg = <0 0x40000000 0 0x01A00000>;
> >>>       no-map;
> >>> };
> >>>
> >>> To fix this, when removing memory regions at early boot (which is
> >>> what "no-map" regions do), we need to make sure that the memory
> >>> is not already reserved. If we do, __reserved_mem_reserve_reg
> >>> will throw an error:
> >>> [    0.000000] OF: fdt: Reserved memory: failed to reserve memory
> >>>    for node 'mem_region': base 0x0000000040000000, size 26 MiB
> >>> and the code that will try to use the region should also fail,
> >>> later on.
> >>>
> >>> We do not do anything for non-"no-map" regions, as memblock
> >>> explicitly allows reserved regions to overlap, and the commit
> >>> that this fixes removed the check for that precise reason.
> >>>
> >>> Fixes: 094cb98179f19b7 ("of/fdt: memblock_reserve /memreserve/ regions in the case of partial overlap")
> >>> Signed-off-by: Nicolas Boichat <drinkcat@xxxxxxxxxxxx>
> >>> ---
> >>>  drivers/of/fdt.c | 10 +++++++++-
> >>>  1 file changed, 9 insertions(+), 1 deletion(-)
> >>>
> >>> diff --git a/drivers/of/fdt.c b/drivers/of/fdt.c
> >>> index cd17dc62a71980a..a1ded43fc332d0c 100644
> >>> --- a/drivers/of/fdt.c
> >>> +++ b/drivers/of/fdt.c
> >>> @@ -1138,8 +1138,16 @@ int __init __weak early_init_dt_mark_hotplug_memory_arch(u64 base, u64 size)
> >>>  int __init __weak early_init_dt_reserve_memory_arch(phys_addr_t base,
> >>>                                       phys_addr_t size, bool nomap)
> >>>  {
> >>> -     if (nomap)
> >>> +     if (nomap) {
> >>> +             /*
> >>> +              * If the memory is already reserved (by another region), we
> >>> +              * should not allow it to be removed altogether.
> >>> +              */
> >>> +             if (memblock_is_region_reserved(base, size))
> >>> +                     return -EBUSY;
> >>> +
> >>>               return memblock_remove(base, size);
> >>
> >> While you are it, the nomap argument (introduced with
> >> e8d9d1f5485b52ec3c4d7af839e6914438f6c285) predates the introduction of
> >> memblock_is_nomap() (bf3d3cc580f9960883ebf9ea05868f336d9491c2), so
> >> should just remove memblock_remove() and use memblock_mark_nomap()
> >> instead here.
> >
> > Perhaps like this patch[1]? Though the reasoning is different and the
> > commit message here is more thorough, so can I get a combined patch.
>
> From a quick reading it does look like memblock_isolate_range(), as
> called by memblock_setclr_flag() should be able to detect this region
> was already reserved, though I have not tried it.

I quickly tested it, and just using memblock_mark_nomap does not seem
be be enough (the call does not fail, and the nomap memory is still
allocated).

> > However, I don't under how handling a misconfigured DT and aligned
> > with EFI are the same patch. What's considered valid for EFI is not
> > for DT regions?
>
> That I don't know how to answer.
> --
> Florian