On Fri, 18 Feb 2022 at 19:06, Jann Horn <jannh@xxxxxxxxxx> wrote:
>
> When the "block" flag is false, the old code would sometimes still call
> check_var_size(), which wrongly tells ->query_variable_store() that it can
> block.
>
> As far as I can tell, this can't really materialize as a bug at the moment,
> because ->query_variable_store only does something on X86 with generic EFI,
> and in that configuration we always take the efivar_entry_set_nonblocking()
> path. So I'm not marking this for stable backporting.
>
> Fixes: ca0e30dcaa53 ("efi: Add nonblocking option to efi_query_variable_store()")
> Signed-off-by: Jann Horn <jannh@xxxxxxxxxx>
Thanks Jann. I'll queue this up.
> ---
>
> drivers/firmware/efi/vars.c | 5 ++++-
> 1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/firmware/efi/vars.c b/drivers/firmware/efi/vars.c
> index abdc8a6a3963..cae590bd08f2 100644
> --- a/drivers/firmware/efi/vars.c
> +++ b/drivers/firmware/efi/vars.c
> @@ -742,6 +742,7 @@ int efivar_entry_set_safe(efi_char16_t *name, efi_guid_t vendor, u32 attributes,
> {
> const struct efivar_operations *ops;
> efi_status_t status;
> + unsigned long varsize;
>
> if (!__efivars)
> return -EINVAL;
> @@ -764,15 +765,17 @@ int efivar_entry_set_safe(efi_char16_t *name, efi_guid_t vendor, u32 attributes,
> return efivar_entry_set_nonblocking(name, vendor, attributes,
> size, data);
>
> + varsize = size + ucs2_strsize(name, 1024);
> if (!block) {
> if (down_trylock(&efivars_lock))
> return -EBUSY;
> + status = check_var_size_nonblocking(attributes, varsize);
> } else {
> if (down_interruptible(&efivars_lock))
> return -EINTR;
> + status = check_var_size(attributes, varsize);
> }
>
> - status = check_var_size(attributes, size + ucs2_strsize(name, 1024));
> if (status != EFI_SUCCESS) {
> up(&efivars_lock);
> return -ENOSPC;
>
> base-commit: 83e396641110663d3c7bb25b9bc0c6a750359ecf
> --
> 2.35.1.473.g83b2b277ed-goog
>
