[GEDI] [PATCH v4 04/31] error: auto propagated local_err

[Markus Armbruster]

Vladimir Sementsov-Ogievskiy <vsementsov at virtuozzo.com> writes:

> 09.10.2019 11:04, Markus Armbruster wrote:
>> Vladimir Sementsov-Ogievskiy <vsementsov at virtuozzo.com> writes:
>> 
>>> Here is introduced ERRP_AUTO_PROPAGATE macro, to be used at start of
>>> functions with errp OUT parameter.
>>>
>>> It has three goals:
>>>
>>> 1. Fix issue with error_fatal & error_prepend/error_append_hint: user
>>> can't see this additional information, because exit() happens in
>>> error_setg earlier than information is added. [Reported by Greg Kurz]
>> 
>> Done in PATCH 07-31.
>> 
>>> 2. Fix issue with error_abort & error_propagate: when we wrap
>>> error_abort by local_err+error_propagate, resulting coredump will
>>> refer to error_propagate and not to the place where error happened.
>>> (the macro itself doesn't fix the issue, but it allows to [3.] drop all
>>> local_err+error_propagate pattern, which will definitely fix the issue)
>>> [Reported by Kevin Wolf]
>>>
>>> 3. Drop local_err+error_propagate pattern, which is used to workaround
>>> void functions with errp parameter, when caller wants to know resulting
>>> status. (Note: actually these functions could be merely updated to
>>> return int error code).
>> 
>> Not done.  Can you prototype this part?  A few manually done examples
>> would give us an idea how the complete solution would look like.  A
>> (semi-)automated complete conversion of a subsystem would additionally
>> give us an idea how to actually do the conversion.
>> 
>> We can discuss applying ERRP_AUTO_PROPAGATE() as a bug fix for 1., and
>> leave 2. and 3. for later.  Feels like a half-done job to me.  We've got
>> too many of those in the tree already.  Dunno.
>> 
>
> If update everything it's about 90-140 patches.
>
> The whole thing was done in "[RFC v2 0/9] error: auto propagated local_err"
> <20190923161231.22028-1-vsementsov at virtuozzo.com>
> https://lists.gnu.org/archive/html/qemu-devel/2019-09/msg05157.html
>
> - it was so huge, that I decided to postpone 2 and 3...

Suggest to post unsplit patches as RFC.  This lets us see how the macro
gets used, and can also help us decide on a sensible split.  The
automatic split isn't quite right in places, and feels excessively
fine-grained to me.