On 4/12/19 10:55 AM, Eric Blake wrote: > On 4/12/19 7:21 AM, Kevin Wolf wrote: >> Am 10.04.2019 um 20:27 hat Eric Blake geschrieben: >>> On 4/10/19 1:03 PM, Cole Robinson wrote: >>>> This gives us the consistent 'Error:' prefix added in 66363e9a43f, >>>> which helps users like libvirt who still need to scrape hmp error >>>> messages to detect failure. >>>> >>>> Signed-off-by: Cole Robinson <crobi...@redhat.com> >>>> --- >>>> hmp.c | 7 ++++--- >>>> 1 file changed, 4 insertions(+), 3 deletions(-) >>> >>> Not enough to drive -rc4 on its own, but worth adding to our wishlist of >>> potential easy patches if we do have a release blocker surface. >> >> As we are going to have an -rc4, I had a look at this. >> >> Commit 66363e9a43f was in February and explicitly says "Note: Some >> places don't use hmp_handle_error". So this doesn't seem to be a >> regression and even if it's fixed, it's likely not the last place that >> doesn't use the "Error:" prefix. This would suggest that this isn't for >> -rc4. >> >> Am I misunderstanding the situation? > > No, I think your read is accurate, and delaying this to 4.1 is okay. >
Yup, this isn't really fixing any specific thing in libvirt, just a bit of future proofing > >>>> + error_prepend(&err, >>>> + "Error while deleting snapshot on device '%s': ", >>> >>> Do we want to reword the message (maybe s/Error while //) to avoid the >>> word "Error" twice in the same line? > > especially since we still would want this resolved via a v2, rather than > taking this patch as-is. > > I'll send a v2 after 4.0 is out Thanks, Cole
