Re: [PATCH] x86/ioremap: Fix off-by-one in e820 check in memremap_should_map_decrypted()

[Tom Lendacky <thomas.lendacky@amd.com>]

On 4/18/25 08:59, Dmytro Maluka wrote:
> The end address in e820__get_entry_type() is exclusive, not inclusive.
> 
> Note: untested, bug found by code inspection.
> 
> Signed-off-by: Dmytro Maluka <dmaluka@chromium.org>
> ---
>  arch/x86/mm/ioremap.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c
> index 331e101bf801..a44800a6196e 100644
> --- a/arch/x86/mm/ioremap.c
> +++ b/arch/x86/mm/ioremap.c
> @@ -578,7 +578,7 @@ static bool memremap_should_map_decrypted(resource_size_t phys_addr,
>  	}
>  
>  	/* Check if the address is outside kernel usable area */
> -	switch (e820__get_entry_type(phys_addr, phys_addr + size - 1)) {
> +	switch (e820__get_entry_type(phys_addr, phys_addr + size)) {

I don't think removing the " - 1" is correct. For example, if you are
mapping a page-aligned value for page-size (4K), then not subtracting 1
will place you on the next page, which is incorrect, because you are not
mapping that page.

Thanks,
Tom

>  	case E820_TYPE_RESERVED:
>  	case E820_TYPE_ACPI:
>  	case E820_TYPE_NVS: