Thanks, I corrected this in the referenced repository fork.
Will be included in v4.

> On 27 Jan 2023, at 20:22, Pedro Falcato <pedro.falc...@gmail.com> wrote:
> 
> On Fri, Jan 27, 2023 at 9:29 AM Savva Mitrofanov <savva...@gmail.com> wrote:
>> 
>> Missing check for wrong s_log_block_size exponent leads to shift out of
>> bounds. Limit block size to 2 MiB
>> 
>> Cc: Marvin Häuser <mhaeu...@posteo.de>
>> Cc: Pedro Falcato <pedro.falc...@gmail.com>
>> Cc: Vitaly Cheptsov <vit9...@protonmail.com>
>> Fixes: d9ceedca6c8f ("Ext4Pkg: Add Ext4Dxe driver.")
>> Signed-off-by: Savva Mitrofanov <savva...@gmail.com>
>> ---
>> Features/Ext4Pkg/Ext4Dxe/Ext4Dxe.h    | 14 ++++++++++++++
>> Features/Ext4Pkg/Ext4Dxe/Superblock.c |  5 +++++
>> 2 files changed, 19 insertions(+)
>> 
>> diff --git a/Features/Ext4Pkg/Ext4Dxe/Ext4Dxe.h 
>> b/Features/Ext4Pkg/Ext4Dxe/Ext4Dxe.h
>> index 2e489ce4dd86..a23323319a59 100644
>> --- a/Features/Ext4Pkg/Ext4Dxe/Ext4Dxe.h
>> +++ b/Features/Ext4Pkg/Ext4Dxe/Ext4Dxe.h
>> @@ -40,6 +40,20 @@
>> #define EXT4_EFI_PATH_MAX    4096
>> #define EXT4_DRIVER_VERSION  0x0000
>> 
>> +//
>> +// The EXT4 Specification doesn't strictly limit block size and this value 
>> could be up to 2^31,
>> +// but in practice it is limited by PAGE_SIZE due to performance 
>> significant impact.
>> +// Many EXT4 implementations have size of block limited to PAGE_SIZE. In 
>> many cases it's limited
>> +// to 4096, which is a commonly supported page size on most MMU-capable 
>> hardware, and up to 65536.
>> +// So, to take a balance between compatibility and security measures, it is 
>> decided to use the
>> +// value of 2MiB as the limit, which is equal to page size on new hardware.
> Nit: s/page size/large page size/
> I can change this for you when pushing, no need for a v4 on this one.
>> +// As for supporting big block sizes, EXT4 has a RO_COMPAT_FEATURE called 
>> BIGALLOC, which changes
>> +// EXT4 to use clustered allocation, so that each bit in the ext4 block 
>> allocation bitmap addresses
>> +// a power of two number of blocks. So it would be wiser to implement and 
>> use this feature
>> +// if there is such a need instead of big block size.
>> +//
>> +#define EXT4_LOG_BLOCK_SIZE_MAX  11
>> +
>> /**
>>    Opens an ext4 partition and installs the Simple File System protocol.
>> 
>> diff --git a/Features/Ext4Pkg/Ext4Dxe/Superblock.c 
>> b/Features/Ext4Pkg/Ext4Dxe/Superblock.c
>> index be3527e4d618..3f56de93c105 100644
>> --- a/Features/Ext4Pkg/Ext4Dxe/Superblock.c
>> +++ b/Features/Ext4Pkg/Ext4Dxe/Superblock.c
>> @@ -248,6 +248,11 @@ Ext4OpenSuperblock (
>>     return EFI_VOLUME_CORRUPTED;
>>   }
>> 
>> +  if (Sb->s_log_block_size > EXT4_LOG_BLOCK_SIZE_MAX) {
>> +    DEBUG ((DEBUG_ERROR, "[ext4] SuperBlock s_log_block_size %lu is too 
>> big\n", Sb->s_log_block_size));
>> +    return EFI_UNSUPPORTED;
>> +  }
>> +
>>   Partition->BlockSize = (UINT32)LShiftU64 (1024, Sb->s_log_block_size);
>> 
>>   // The size of a block group can also be calculated as 8 * 
>> Partition->BlockSize
>> --
>> 2.39.0
>> 
> 
> Reviewed-by: Pedro Falcato <pedro.falc...@gmail.com>
> 
> -- 
> Pedro



-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#99180): https://edk2.groups.io/g/devel/message/99180
Mute This Topic: https://groups.io/mt/96562695/21656
Group Owner: devel+ow...@edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-


Reply via email to