Re: [dpdk-dev] [PATCH v2 4/4] add ABI checks

[Ferruh Yigit <ferruh.yigit@intel.com>]

On 2/3/2020 11:50 AM, Neil Horman wrote:
> On Mon, Feb 03, 2020 at 09:30:06AM +0000, Ferruh Yigit wrote:
>> On 2/2/2020 2:41 PM, Ananyev, Konstantin wrote:
>>>
>>>> 31/01/2020 15:16, Trahe, Fiona:
>>>>> On 1/30/2020 8:18 PM, Thomas Monjalon wrote:
>>>>>> 30/01/2020 17:09, Ferruh Yigit:
>>>>>>> On 1/29/2020 8:13 PM, Akhil Goyal wrote:
>>>>>>>>
>>>>>>>> I believe these enums will be used only in case of ASYM case which is experimental.
>>>>>>>
>>>>>>> Independent from being experiment and not, this shouldn't be a problem, I think
>>>>>>> this is a false positive.
>>>>>>>
>>>>>>> The ABI break can happen when a struct has been shared between the application
>>>>>>> and the library (DPDK) and the layout of that memory know differently by
>>>>>>> application and the library.
>>>>>>>
>>>>>>> Here in all cases, there is no layout/size change.
>>>>>>>
>>>>>>> As to the value changes of the enums, since application compiled with old DPDK,
>>>>>>> it will know only up to '6', 7 and more means invalid to the application. So it
>>>>>>> won't send these values also it should ignore these values from library. Only
>>>>>>> consequence is old application won't able to use new features those new enums
>>>>>>> provide but that is expected/normal.
>>>>>>
>>>>>> If library give higher value than expected by the application,
>>>>>> if the application uses this value as array index,
>>>>>> there can be an access out of bounds.
>>>>>
>>>>> [Fiona] All asymmetric APIs are experimental so above shouldn't be a problem.
>>>>> But for the same issue with sym crypto below, I believe Ferruh's explanation makes
>>>>> sense and I don't see how there can be an API breakage.
>>>>> So if an application hasn't compiled against the new lib it will be still using the old value
>>>>> which will be within bounds. If it's picking up the higher new value from the lib it must
>>>>> have been compiled against the lib so shouldn't have problems.
>>>>
>>>> You say there is no ABI issue because the application will be re-compiled
>>>> for the updated library. Indeed, compilation fixes compatibility issues.
>>>> But this is not relevant for ABI compatibility.
>>>> ABI compatibility means we can upgrade the library without recompiling
>>>> the application and it must work.
>>>> You think it is a false positive because you assume the application
>>>> "picks" the new value. I think you miss the case where the new value
>>>> is returned by a function in the upgraded library.
>>>>
>>>>> There are also no structs on the API which contain arrays using this
>>>>> for sizing, so I don't see an opportunity for an appl to have a
>>>>> mismatch in memory addresses.
>>>>
>>>> Let me demonstrate where the API may "use" the new value
>>>> RTE_CRYPTO_AEAD_CHACHA20_POLY1305 and how it impacts the application.
>>>>
>>>> Once upon a time a DPDK application counting the number of devices
>>>> supporting each AEAD algo (in order to find the best supported algo).
>>>> It is done in an array indexed by algo id:
>>>> int aead_dev_count[RTE_CRYPTO_AEAD_LIST_END];
>>>> The application is compiled with DPDK 19.11,
>>>> where RTE_CRYPTO_AEAD_LIST_END = 3.
>>>> So the size of the application array aead_dev_count is 3.
>>>> This binary is run with DPDK 20.02,
>>>> where RTE_CRYPTO_AEAD_CHACHA20_POLY1305 = 3.
>>>> When calling rte_cryptodev_info_get() on a device QAT_GEN3,
>>>> rte_cryptodev_info.capabilities.sym.aead.algo is set to
>>>> RTE_CRYPTO_AEAD_CHACHA20_POLY1305 (= 3).
>>>> The application uses this value:
>>>> ++ aead_dev_count[info.capabilities.sym.aead.algo];
>>>> The application is crashing because of out of bound access.
>>>
>>> I'd say this is an example of bad written app.
>>> It probably should check that returned by library value doesn't
>>> exceed its internal array size.
>>
>> +1
>>
>> Application should ignore values >= MAX.
>>
> The example is still somewhat valid in it general principle though.  While
> extending an ennumeration may be flagged by libabigail as an ABI breakage, its
> not necessecarily a false positive.  By extending the ennumeration, all the
> previous entries in an array defined by said ennumeration remain constant in
> their offsets, so you can 'get away with such a change' in terms of preserving
> backwards compatibility in the above example, but you cannot, for example,
> shuffle the values in the ennumeration, as doing so would cause a functional
> breakage (i.e. requesting an instance of RTE_CRYPTO_AEAD_CHACHA20_POLY1305 might
> instead give you an instance of RTE_CRYPTO_AEAD_AES_GCM.  

+1 the change/shuffle of the existing values are problematic, but we don't have
it in this case.

> 
> These sorts of changes are the type that we could collectively waive in terms of
> ABI checking, as they should be ok, but the errors from libabigail should be
> taken as an indicator that this API could be rewritten (for example by removing
> the abi entirely, and adding an API call that returns an array of instance name
> and ids), so that changes of the above sort arent required.

We can spend more time on it, but I can't see for now how to escape returning
enumaration as indication of type, and this looks legitimate sage as long as
other side verifies the received value is valid in the type range.

> 
> 
>> Do you suggest we don't extend any enum or define between ABI breakage releases
>> to be sure bad written applications not affected?
>>
> As noted above, we could waive such corner cases, and probably be fine, but the
> error from the ABI check still serves a valid purpose in that its an indicator
> that your library API is ABI sensitive to code changes that re-architecture may
> address
> 

The concern is when there are cases we can waive, we can't directly rely on the
tool and automate it. These indicators good for improving the code, but not good
to use it as build time checker.
Is there any way to reduce the failure only to definite ABI breakages?