On 08/08/2017 02:34 PM, Paolo Bonzini wrote:
> ----- Original Message -----
>> From: "John Snow" <js...@redhat.com>
>> To: qemu-bl...@nongnu.org
>> Cc: kw...@redhat.com, email@example.com, dgilb...@redhat.com,
>> stefa...@redhat.com, pbonz...@redhat.com,
>> p...@redhat.com, "John Snow" <js...@redhat.com>
>> Sent: Tuesday, August 8, 2017 7:57:10 PM
>> Subject: [PATCH 3/4] block-backend: shift in-flight counter to BB from BDS
>> From: Kevin Wolf <kw...@redhat.com>
>> This allows us to detect errors in cache flushing (ENOMEDIUM)
>> without choking on a null dereference because we assume that
>> blk_bs(bb) is always defined.
>> Signed-off-by: Kevin Wolf <kw...@redhat.com>
>> Signed-off-by: John Snow <js...@redhat.com>
> This is not enough, as discussed in the thread.
Sure, I amended Kevin's later fix and rolled it into one patch and split
the tests out. The cover letter states that this is busted, but I wanted
it on the list instead of buried in a now-unrelated thread.
So now it's here as a patch, can we keep discussion here instead of on
the other thread?