2023-09-07 16:22:15

by Robin Murphy

[permalink] [raw]
Subject: Re: [PATCH] Revert "dma-contiguous: check for memory region overlap"

On 2023-09-07 09:03, Zhenhua Huang wrote:
> This reverts commit 3fa6456ebe13adab3ba1817c8e515a5b88f95dce.
>
> The Commit broke the CMA region creation through DT on arm64,
> as showed below logs with "memblock=debug":
> [ 0.000000] memblock_phys_alloc_range: 41943040 bytes align=0x200000
> from=0x0000000000000000 max_addr=0x00000000ffffffff
> early_init_dt_alloc_reserved_memory_arch+0x34/0xa0
> [ 0.000000] memblock_reserve: [0x00000000fd600000-0x00000000ffdfffff]
> memblock_alloc_range_nid+0xc0/0x19c
> [ 0.000000] Reserved memory: overlap with other memblock reserved region
>
> From call flow, region we defined in DT was always reserved before entering
> into rmem_cma_setup. Also, rmem_cma_setup has one routine cma_init_reserved_mem
> to ensure the region was reserved. Checking the region not reserved here seems
> not correct.
>
> early_init_fdt_scan_reserved_mem:
> fdt_scan_reserved_mem
> __reserved_mem_reserve_reg
> early_init_dt_reserve_memory
> memblock_reserve(using “reg” prop case)
> fdt_init_reserved_mem
> __reserved_mem_alloc_size
> *early_init_dt_alloc_reserved_memory_arch*
> memblock_reserve(dynamic alloc case)
> __reserved_mem_init_node
> rmem_cma_setup(region overlap check here should always fail)
>
> Example DT can be used to reproduce issue:
>
> dump_mem: mem_dump_region {
> compatible = "shared-dma-pool";
> alloc-ranges = <0x0 0x00000000 0x0 0xffffffff>;
> reusable;
> size = <0 0x2800000>;
> };
>
> Signed-off-by: Zhenhua Huang <[email protected]>
> ---
> kernel/dma/contiguous.c | 5 -----
> 1 file changed, 5 deletions(-)
>
> diff --git a/kernel/dma/contiguous.c b/kernel/dma/contiguous.c
> index 88c595e..f005c66 100644
> --- a/kernel/dma/contiguous.c
> +++ b/kernel/dma/contiguous.c
> @@ -473,11 +473,6 @@ static int __init rmem_cma_setup(struct reserved_mem *rmem)
> return -EBUSY;
> }
>
> - if (memblock_is_region_reserved(rmem->base, rmem->size)) {
> - pr_info("Reserved memory: overlap with other memblock reserved region\n");
> - return -EBUSY;
> - }

Um, indeed... failing to process a reserved memory region if it is
reserved is clearly nonsensical. I mean, a few steps later we reach
cma_init_reserved_mem() which already relies on the exact opposite check:

if (!size || !memblock_is_region_reserved(base, size))
return -EINVAL;

There's no way 3fa6456ebe13 could have worked or been properly tested :/

Thanks,
Robin.

> -
> if (!of_get_flat_dt_prop(node, "reusable", NULL) ||
> of_get_flat_dt_prop(node, "no-map", NULL))
> return -EINVAL;