[Cosmic][PATCH 0/1] fsnotify: Fix fsnotify_mark_connector race

Joseph Salisbury joseph.salisbury at canonical.com
Thu May 24 15:08:10 UTC 2018

On 05/24/2018 11:03 AM, Seth Forshee wrote:
> On Thu, May 24, 2018 at 10:39:05AM -0400, Joseph Salisbury wrote:
>> On 05/24/2018 07:59 AM, Thadeu Lima de Souza Cascardo wrote:
>>> On Thu, May 24, 2018 at 07:32:53AM -0400, Joseph Salisbury wrote:
>>>> On 05/23/2018 02:27 PM, Joseph Salisbury wrote:
>>>>> BugLink: http://bugs.launchpad.net/bugs/1771075
>>>>> == Cosmic Justification ==
>>>>> A regression was introduced by commit 08991e83b728.  This regression
>>>>> causes problems on hosts that make heavy use of inotify and causes panics / lockups 
>>>>> in the kernel in inotify-related code.
>>>>> This regression is fixed buy commit d90a10e2444b as of v4.17-rc3.  Commit d90a10e2444b
>>>>> was cc'd to stable and is already in Artful and Bionic but not Cosmic.
>>>>> The commit was added to A and B by bug 1765564.
>>>>> == Fix ==
>>>>> d90a10e2444b ("fsnotify: Fix fsnotify_mark_connector race")
>>>>> == Regression Potential ==
>>>>> Low.  Fixes a current regression. The patch was cc'd to stable, so it has had
>>>>> additional upstream review.  This commit is already applied to Artful
>>>>> and Bionic.
>>>>> == Test Case ==
>>>>> A test kernel was built with this patch and tested by the original bug reporter.
>>>>> The bug reporter states the test kernel resolved the bug.
>>>>> Robert Kolchmeyer (1):
>>>>>   fsnotify: Fix fsnotify_mark_connector race
>>>>>  include/linux/fsnotify_backend.h | 4 +---
>>>>>  1 file changed, 1 insertion(+), 3 deletions(-)
>>>> This patch is also being requested in linux-gcp.  Should I send a v2, so
>>>> cosmic and linux-gcp are both listed in the subject?
>>> Is this really targeting cosmic? Or bionic? Though cosmic is still carrying the
>>> bionic kernel, it will soon have a 4.17 based kernel, which already has that
>>> commit.
>>> Cascardo.
>> It is targeting cosmic(And linux-gce).  Bionic is getting the commit via
>> bug 1765564, but that bug does not have cosmic as a bug task(And cosmic
>> was not requested in the SRU for that bug).  Cosmic will get this commit
>> when it is rebased, but I was not sure when that would happen.  I
>> figured I'd send a request to have the commit in cosmic in the mean time. 
> I think the point is that there is no cosmic kernel distinct from the
> bionic kernel yet, so it's ambiguous when you say "cosmic" whether you
> mean the bionic kernel that gets copied forward to cosmic or the
> unstable kernel which will eventually become the cosmic kernel.

Ahh, Ok.  So if Bionic fixes get carried forward, then this can probably
be NAK'd.

More information about the kernel-team mailing list