memblock_is_region_reserved() returns true in the case of a partial overlap, meaning that the current code fails to reserve the non-overlapping portion.
I observed this causing a Midway system with a buggy fdt (the header declares itself to be larger than it really is) failing to boot because the over-inflated size of the fdt was causing it to seem to run into the swapper_pg_dir region, meaning the DT wasn't reserved. The symptoms were failing to find an disks or network and failing to boot. I think it is work printing a warning in this case. However given the ambiguity of whether things like the initrd are covered by /memreserve/ and similar I think it is best to also register the region rather than just ignoring it. Since memblock_reserve() handles overlaps just fine lets just warn and carry on. Signed-off-by: Ian Campbell <[email protected]> Cc: Grant Likely <[email protected]> Cc: Rob Herring <[email protected]> Cc: [email protected] --- drivers/of/fdt.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/of/fdt.c b/drivers/of/fdt.c index 30e97bc..4f0f24c 100644 --- a/drivers/of/fdt.c +++ b/drivers/of/fdt.c @@ -965,7 +965,8 @@ int __init __weak early_init_dt_reserve_memory_arch(phys_addr_t base, phys_addr_t size, bool nomap) { if (memblock_is_region_reserved(base, size)) - return -EBUSY; + pr_warning("DT reserved region at 0x%pa is already reserved\n", + &base); if (nomap) return memblock_remove(base, size); return memblock_reserve(base, size); -- 1.7.10.4 -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
