Re: [PATCH v6] error: rename errp to errp_in where it is IN-argument

[Markus Armbruster]

Vladimir Sementsov-Ogievskiy <address@hidden> writes:

> 29.11.2019 17:35, Markus Armbruster wrote:
>> Vladimir Sementsov-Ogievskiy <address@hidden> writes:
>> 
>>> 28.11.2019 23:29, Markus Armbruster wrote:
>>>> Vladimir Sementsov-Ogievskiy <address@hidden> writes:
>>>>
>>>>> 28.11.2019 17:23, Markus Armbruster wrote:
>>>>>> Vladimir Sementsov-Ogievskiy <address@hidden> writes:
>>>>>>
>>>>>>> Error **errp is almost always OUT-argument: it's assumed to be NULL, or
>>>>>>> pointer to NULL-initialized pointer, or pointer to error_abort or
>>>>>>> error_fatal, for callee to report error.
>>>>>>>
>>>>>>> But very few functions instead get Error **errp as IN-argument:
>>>>>>> it's assumed to be set (or, maybe, NULL), and callee should clean it,
>>>>>>> or add some information.
>>>>>>>
>>>>>>> In such cases, rename errp to errp_in.
>>>>>>
>>>>>> Missing: why is the rename useful?
>>>>>
>>>>> The main reason is to prepare for coccinelle part.
>>>>
>>>> It's not a prerequisite for applying the patches Coccinelle produces,
>>>> only a prerequisite for running Coccinelle.
>>>>
>>>>>> It's useful if it helps readers recognize unusual Error ** parameters,
>>>>>> and recognizing unusual Error ** parameters is actually a problem.  I'm
>>>>>> not sure it is, but my familiarity with the Error interface may blind
>>>>>> me.
>>>>>>
>>>>>> How many functions have unusual Error **parameters?  How are they used?
>>>>>> Any calls that could easily be mistaken as the usual case?  See [*]
>>>>>> below.
>> [...]
>>>>>> [*] The information I asked for above is buried in these patches.  I'll
>>>>>> try to dig it up as I go reviewing them.
>>>>>>
>>>>>> Second, it risks some of these "further patches" overtake this one, and
>>>>>> then its commit message will be misleading.  Moreover, the other commits
>>>>>> will lack context.
>> [...]
>>>>>>> diff --git a/util/error.c b/util/error.c
>>>>>>> index d4532ce318..275586faa8 100644
>>>>>>> --- a/util/error.c
>>>>>>> +++ b/util/error.c
>> [...]
>>>>>>> @@ -271,11 +271,11 @@ void error_free(Error *err)
>>>>>>>         }
>>>>>>>     }
>>>>>>>     
>>>>>>> -void error_free_or_abort(Error **errp)
>>>>>>> +void error_free_or_abort(Error **errp_in)
>>>>>>>     {
>>>>>>> -    assert(errp && *errp);
>>>>>>> -    error_free(*errp);
>>>>>>> -    *errp = NULL;
>>>>>>> +    assert(errp_in && *errp_in);
>>>>>>> +    error_free(*errp_in);
>>>>>>> +    *errp_in = NULL;
>>>>
>>>> This one is actually in/out.
>>>>
>>>> To make the compiler check errp_in is truly an in-argument, we can
>>>> declare it as Error *const *errp_in.
>>>>
>>>> But we can save ourselves the trouble of renaming it; the const should
>>>> suffice to tell both human readers and Coccinelle that this is not your
>>>> common out-argument.  I think I like this better than relying on a
>>>> naming convention.  What about you?
>>>
>>> I think it's good idea... But what to do with error_free_or_abort, and 
>>> other functions
>>> which get filled errp, and want to set it to NULL? Patch 21 adds three such 
>>> functions..
>>>
>>> Maybe, add assert(errp) at start of such functions, and catch it by 
>>> coccinelle (not sure
>>> that it is possible)?
>> 
>> I went through the git branch you provided.
>> 
>> These get rid of unusual Error ** parameters:
>> 
>>      01e10667d1 hmp: drop Error pointer indirection in hmp_handle_error
>>      606bfb7520 vnc: drop Error pointer indirection in vnc_client_io_error
>> 
>>      Get rid of them by peeling off an indirection.
>> 
>> These try to make unusual Error ** parameters stand out:
>> 
>>      51e73b3305 error: rename errp to errp_in where it is IN-argument
>> 
>>      Four renames.
>> 
>>      Three functions modify the error, name @errp_in is okay, my const
>>      proposal works.
>> 
>>      error_free_or_abort() clears *errp_in, name @errp_in is misleading,
>>      const doesn't work.
>> 
>>      f7bdfd42da qdev-monitor: well form error hint helpers
>> 
>>      Two renames.  Both functions modify the error, rename is okay, const
>>      works.
>> 
>>      9d4aac7299 nbd: well form nbd_iter_channel_error errp handler
>> 
>>      One rename, from @local_err.  nbd_iter_channel_error() clears *errp_in,
>>      name @errp_in is misleading, const doesn't work.
>> 
>>      fb1bd83c40 ppc: well form kvmppc_hint_smt_possible error hint helper
>> 
>>      One rename.  The function modify the error, rename is okay, const works.
>> 
>>      e024e89b10 9pfs: well form error hint helpers
>> 
>>      Two renames.  Both functions modify the error, rename is okay, const
>>      works.
>> 
>> These make usual Error ** parameters look, well, more usual:
>> 
>>      c01649d999 hw/core/qdev: cleanup Error ** variables
>>      a5f6424163 block/snapshot: rename Error ** parameter to more common errp
>>      ae200ca21e hw/i386/amd_iommu: rename Error ** parameter to more common 
>> errp
>>      561f73e681 qga: rename Error ** parameter to more common errp
>>      12589a96cd monitor/qmp-cmds: rename Error ** parameter to more common 
>> errp
>>      f608fc5999 hw/s390x: rename Error ** parameter to more common errp
>>      747a90d044 hw/tpm: rename Error ** parameter to more common errp
>>      3d19e66610 hw/usb: rename Error ** parameter to more common errp
>>      07366648ef include/qom/object.h: rename Error ** parameter to more 
>> common errp
>> 
>> These don't touch Error ** parameter declarations:
>> 
>>      f6e4fffc16 hw/core/loader-fit: fix freeing errp in fit_load_fdt
>>      b5bba880ae net/net: Clean up variable shadowing in net_client_init()
>>      4a4462ce4c hw/sd: drop extra whitespace in sdhci_sysbus_realize() header
>>      d52d44e822 backends/cryptodev: drop local_err from 
>> cryptodev_backend_complete()
>>      7e95d30766 hw/vfio/ap: drop local_err from vfio_ap_realize
>> 
>> Unusual Error ** parameters that can be made Error *const *errp should
>> not also need a rename, neither to avoid confusion about @errp's role,
>> nor to help Coccinelle.
>> 
>> Unusual Error ** parameters that can't be made Error *const *errp:
>> 
>>      nbd_iter_channel_error(): parameter is called @local_err.  Confusion
>>      seems as unlikely without the rename as with it.  Coccinelle should
>>      not need the rename.  No need to touch.  I'm willing to accept your
>>      assertion change.
>
> Hmm but you reverted it.. Will you keep an assertion?

You decide whether to change the assertion in v7.

>>      error_free_or_abort(): parameter is called @errp.  Confusion seems
>>      as unlikely without the rename as with it.  Coccinelle should not
>>      need the rename.  I'm willing to accept a rename to @local_err
>>      regardless.
>
> So, either we rename it to local_err, or coccinelle should distinguish it
> by assertion at the top.

Yes.

I'm also willing to consider a rename to a name other than @local_err,
if you can come up with a good, non-misleading one.

>> I pushed my fixups to git://repo.or.cz/qemu/armbru.git branch error-prep
>> for your convenience.  The commit messages of the fixed up commits need
>> rephrasing, of course.
>> 
>
> Looked through fixups, looks OK for me, thanks! What next?

Let me finish my fixing incorrect dereferences of errp, and then we
figure out what to include in v7.