List of kernel bugs tagged suspend, resume, or hibernate

Steve Conklin steve.conklin at
Wed Jan 14 21:10:43 GMT 2009

On 01/14/2009 09:45 AM, Andy Whitcroft wrote:
> On Tue, Jan 13, 2009 at 09:09:46AM -0800, Leann Ogasawara wrote:
>> On Mon, 2009-01-12 at 22:34 -0600, Steve Conklin wrote:
>>> On 01/12/2009 04:35 PM, Leann Ogasawara wrote:
>>>> I assume the bugs that will get reported via the test script (apport)
>>>> will also be tagged something like "apport-bug".  Steve, can you comment
>>>> if the tagging I'm expecting to happen here is a valid assumption?  This
>>>> will make it easy to then distinguish those bugs reported via the script
>>>> vs manually reported from a user.  I could then also easily showcase
>>>> this in the bug list display.
>>> As it stands now, all suspend/resume bugs files by the apport automatic
>>> detection will be tagged with all of the following: "suspend", "resume",
>>> "hiberate", and "apport-kerneloops" (not sure of the exact spelling of
>>> the last one). This can all be changed if desired, the tags are
>>> hard-coded in the script.
>> I'd prefer the use of "apport-bug" vs. "apport-kerneloops".
>> "apport-kerneloops" implies an actual Oops was detected and is being
>> reported.
> I think we might be stuck with whatever Steve put in his patches to
> apport for the start of alpha-3, we are not likely to get apport changed
> again before the release.
> -apw

I couldn't get 'bug' to work when reporting against the kernel. I talked
with Martin Pitt about this, and there are a few options open to us,
just not in time for the alpha. The options we discussed are:

1. Tag it as a bug (apport-bug) against the pm-utils package

2. Create a new class of things to be reported by apport, just for
suspend/resume bugs.

Another problem with using kerneloops is that apport presents the user
with a message that says something like "you have experiences a kernel
crash", which is not really accurate.

Implementing (1) above is relatively easy, implementing (2) would allow
us to customize apport so that the user message is specific to

Martin is on holiday this week, he should be involved in any decisions
about this. This alpha release was a prototype, and we can improve it.


More information about the kernel-team mailing list