[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-block] [PATCH v2 0/9] block: Delay poll when ending drained se
From: |
Max Reitz |
Subject: |
Re: [Qemu-block] [PATCH v2 0/9] block: Delay poll when ending drained sections |
Date: |
Tue, 16 Jul 2019 18:24:11 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.7.2 |
On 16.07.19 16:40, Kevin Wolf wrote:
> Am 19.06.2019 um 17:25 hat Max Reitz geschrieben:
>> Hi,
>>
>> This is v2 to “block: Keep track of parent quiescing”.
>>
>> Please read this cover letter, because I’m very unsure about the design
>> in this series and I’d appreciate some comments.
>>
>> As Kevin wrote in his reply to that series, the actual problem is that
>> bdrv_drain_invoke() polls on every node whenever ending a drain. This
>> may cause graph changes, which is actually forbidden.
>>
>> To solve that problem, this series makes the drain code construct a list
>> of undrain operations that have been initiated, and then polls all of
>> them on the root level once graph changes are acceptable.
>>
>> Note that I don’t like this list concept very much, so I’m open to
>> alternatives.
>
> So drain_end is different from drain_begin in that it wants to wait only
> for all bdrv_drain_invoke() calls to complete, but not for other
> requests that are in flight. Makes sense.
>
> Though instead of managing a whole list, wouldn't a counter suffice?
>
>> Furthermore, all BdrvChildRoles with BDS parents have a broken
>> .drained_end() implementation. The documentation clearly states that
>> this function is not allowed to poll, but it does. So this series
>> changes it to a variant (using the new code) that does not poll.
>>
>> There is a catch, which may actually be a problem, I don’t know: The new
>> variant of that .drained_end() does not poll at all, never. As
>> described above, now every bdrv_drain_invoke() returns an object that
>> describes when it will be done and which can thus be polled for. These
>> objects are just discarded when using BdrvChildRole.drained_end(). That
>> does not feel quite right. It would probably be more correct to let
>> BdrvChildRole.drained_end() return these objects so the top level
>> bdrv_drained_end() can poll for their completion.
>>
>> I decided not to do this, for two reasons:
>> (1) Doing so would spill the “list of objects to poll for” design to
>> places outside of block/io.c. I don’t like the design very much as
>> it is, but I can live with it as long as it’s constrained to the
>> core drain code in block/io.c.
>> This is made worse by the fact that currently, those objects are of
>> type BdrvCoDrainData. But it shouldn’t be a problem to add a new
>> type that is externally visible (we only need the AioContext and
>> whether bdrv_drain_invoke_entry() is done).
>>
>> (2) It seems to work as it is.
>>
>> The alternative would be to add the same GSList ** parameter to
>> BdrvChildRole.drained_end() that I added in the core drain code in patch
>> 2, and then let the .drained_end() implementation fill that with objects
>> to poll for. (Which would be accomplished by adding a frontend to
>> bdrv_do_drained_end() that lets bdrv_child_cb_drained_poll() pass the
>> parameter through.)
>>
>> Opinions?
>
> I think I would add an int* to BdrvChildRole.drained_end() so that we
> can just increase the counter whereever we need to.
So you mean just polling the @bs for which a caller gave poll=true until
the counter reaches 0? I’ll try, sounds good (if I can get it to work).
>> And then we have bdrv_replace_child_noperm(), which actually wants a
>> polling BdrvChildRole.drained_end(). So this series adds
>> BdrvChildRole.drained_end_unquiesce(), which takes that role (if there
>> is any polling to do).
>>
>> Note that if I implemented the alternative described above
>> (.drained_end() gets said GSList ** parameter), a
>> .drained_end_unquiesce() wouldn’t be necessary.
>> bdrv_parent_drained_end_single() could just poll the list returned by
>> .drained_end() by itself.
>
> The split between .drained_end/.drained_end_unquiesce feels wrong. It
> shouldn't be the job of the BdrvChildRole to worry about this. Polling
> should be handled inside bdrv_parent_drained_end_single(), like we do in
> bdrv_parent_drained_begin_single(), so that the BdrvChildRole never has
> to poll.
If it’s just an int pointer, there’s no point in having two functions, I
suppose.
Thanks for you suggestion!
>> Finally, patches 1, 8, and 9 are unmodified from v1.
>> [...]
>>
>> include/block/block.h | 22 +++++-
>> include/block/block_int.h | 23 ++++++
>> block.c | 24 +++---
>> block/io.c | 155 ++++++++++++++++++++++++++++++-------
>> python/qemu/__init__.py | 5 +-
>> tests/qemu-iotests/040 | 40 +++++++++-
>> tests/qemu-iotests/040.out | 4 +-
>> tests/qemu-iotests/255 | 2 +-
>> 8 files changed, 231 insertions(+), 44 deletions(-)
>
> I feel this series should add something to tests/test-bdrv-drain.c, too.
> qemu-iotests can only test high-level block job commands that happen to
> trigger the bug today, but that code may change in the future. Unit
> tests allow us to test the problematic cases more directly.
Well, I’m glad if test-bdrv-drain just keeps working. :-)
I’ll try.
Max
signature.asc
Description: OpenPGP digital signature