qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v8 3/3] docs: qcow2: introduce compression type feature


From: Vladimir Sementsov-Ogievskiy
Subject: Re: [PATCH v8 3/3] docs: qcow2: introduce compression type feature
Date: Fri, 18 Oct 2019 14:38:34 +0000

18.10.2019 17:04, Eric Blake wrote:
> On 10/18/19 4:47 AM, Vladimir Sementsov-Ogievskiy wrote:
>> The patch add new additional field to qcow2 header: compression_type,
>> which specifies compression type. If field is absent or zero, default
>> compression type is set: ZLIB, which corresponds to current behavior.
>>
>> New compression type (ZSTD) is to be added in further commit.
>>
>> Suggested-by: Denis Plotnikov <address@hidden>
>> Signed-off-by: Vladimir Sementsov-Ogievskiy <address@hidden>
>> ---
>>   docs/interop/qcow2.txt | 19 ++++++++++++++++++-
>>   1 file changed, 18 insertions(+), 1 deletion(-)
>>
>> diff --git a/docs/interop/qcow2.txt b/docs/interop/qcow2.txt
>> index b971e59b1a..4eabd81363 100644
>> --- a/docs/interop/qcow2.txt
>> +++ b/docs/interop/qcow2.txt
>> @@ -109,6 +109,12 @@ least next five fields, up to the @header_length field.
>>                                   An External Data File Name header 
>> extension may
>>                                   be present if this bit is set.
>> +                    Bit 3:      Compression type bit.  If this bit is set,
>> +                                non-default compression is used for 
>> compressed
>> +                                clusters. In this case, @header_length must
>> +                                be at least 105 and @compression_type field
>> +                                must be non-zero.
> 
> I want to insist on 8-byte alignment, so this should be at least 112.
> 
> Also, as pointed out in v7, use of the decoration '@header_length' and 
> '@compression_type' is not consistent with the rest of the document. Drop the 
> @.

Missed this, sorry(

> 
>> +
>>                       Bits 3-63:  Reserved (set to 0)
>>            80 -  87:  compatible_features
>> @@ -183,7 +189,18 @@ It's allowed for the header end to cut some field in 
>> the middle (in this case
>>   the field is considered as absent), but in this case the part of the field
>>   which is covered by @header_length must be zeroed.
>> -        < ... No additional fields in the header currently ... >
>> +              104:  compression_type
>> +                    Defines the compression method used for compressed 
>> clusters.
>> +                    A single compression type is applied to all compressed 
>> image
>> +                    clusters.
>> +                    If incompatible compression type bit is set: the field 
>> must
>> +                    exist (i.e. @header_length >= 105) and must be non-zero 
>> (
>> +                    which means non-zlib compression type)
>> +                    If incompatible compression type bit is unset: the field
>> +                    may not exist (if @header_length < 105) or it must be 
>> zero
>> +                    (which means zlib).
>> +                    Available compression type values:
>> +                        0: zlib <https://www.zlib.net/>
> 
> One byte for the field is fine, but I'd still explicitly document 7 bytes of 
> unused padding, since I want to insist on an 8-byte multiple.
> 

OK

-- 
Best regards,
Vladimir

reply via email to

[Prev in Thread] Current Thread [Next in Thread]