linux-kernel - Re: [PATCH 1/3] efi/libstub: fix efi_parse_options() ignoring the default command line

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <5364ea9f-9042-3fba-5cb1-c42ae94b0d4b@marek.ca>
Date: Sat, 12 Oct 2024 07:55:21 -0400
From: Jonathan Marek <jonathan@...ek.ca>
To: Ard Biesheuvel <ardb@...nel.org>
Cc: linux-efi@...r.kernel.org,
 Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@...ux.intel.com>,
 Ingo Molnar <mingo@...nel.org>, open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 1/3] efi/libstub: fix efi_parse_options() ignoring the
 default command line

On 10/12/24 3:46 AM, Ard Biesheuvel wrote:
> Hi Jonathan,
> 
> Please use a cover letter when sending more than a single patch.
> 
> On Sat, 12 Oct 2024 at 00:51, Jonathan Marek <jonathan@...ek.ca> wrote:
>>
>> efi_convert_cmdline() always returns a size of at least 1 because it counts
>> the NUL terminator, so the "cmdline_size == 0" condition is not possible.
>>
>> Change it to compare against 1 to get the intended behavior: to use
>> CONFIG_CMDLINE when load_options_size is 0.
>>
>> Fixes: 60f38de7a8d4 ("efi/libstub: Unify command line param parsing")
>> Signed-off-by: Jonathan Marek <jonathan@...ek.ca>
>> ---
>>   drivers/firmware/efi/libstub/efi-stub.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/firmware/efi/libstub/efi-stub.c b/drivers/firmware/efi/libstub/efi-stub.c
>> index 958a680e0660d..709ae2d41a632 100644
>> --- a/drivers/firmware/efi/libstub/efi-stub.c
>> +++ b/drivers/firmware/efi/libstub/efi-stub.c
>> @@ -129,7 +129,7 @@ efi_status_t efi_handle_cmdline(efi_loaded_image_t *image, char **cmdline_ptr)
>>
>>          if (IS_ENABLED(CONFIG_CMDLINE_EXTEND) ||
>>              IS_ENABLED(CONFIG_CMDLINE_FORCE) ||
>> -           cmdline_size == 0) {
>> +           cmdline_size == 1) {
> 
> I'd prefer it if we could keep the weirdness local to
> efi_convert_cmdline(). Would the below fix things too?
> 
> --- a/drivers/firmware/efi/libstub/efi-stub-helper.c
> +++ b/drivers/firmware/efi/libstub/efi-stub-helper.c
> @@ -395,9 +395,7 @@
>                  }
>          }
> 
> -       options_bytes++;        /* NUL termination */
> -
> -       status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, options_bytes,
> +       status = efi_bs_call(allocate_pool, EFI_LOADER_DATA, options_bytes + 1,
>                               (void **)&cmdline_addr);
>          if (status != EFI_SUCCESS)
>                  return NULL;
> 
> Note that the only other caller of efi_convert_cmdline() in x86-stub.c
> ignores this value entirely.
> 

Just changing this would just make things more broken, the following 
snprintf would remove the last character of the command line because it 
uses options_bytes.

Since this patch has a Fixes: tag, I wanted to make the fix as simple as 
possible. If you think comparing the size to 1 is "weird", the fix could 
instead check if cmdline[0] is non-NUL (or just strlen(cmdline)==0 if 
you don't like that either).

And then my followup cleanup patch can just remove the cmd_line_len 
argument from efi_convert_cmdline().