svn commit: r195843 - in head: lib/libkvm sys/kern sys/sys

[Thomas Backman]

On Sep 8, 2009, at 9:43 PM, Brooks Davis wrote:

> On Sun, Sep 06, 2009 at 04:47:00AM +0100, Bruce Cran wrote:
>> On Fri, 4 Sep 2009 17:14:32 -0500
>> Brooks Davis <brooks at FreeBSD.org> wrote:
>>
>>> On Sat, Aug 29, 2009 at 08:12:28PM +0100, Bruce Cran wrote:
>>>> On Fri, 24 Jul 2009 15:03:10 +0000 (UTC)
>>>> Brooks Davis <brooks at FreeBSD.org> wrote:
>>>>
>>>>> Author: brooks
>>>>> Date: Fri Jul 24 15:03:10 2009
>>>>> New Revision: 195843
>>>>> URL: http://svn.freebsd.org/changeset/base/195843
>>>>>
>>>>> Log:
>>>>>  Revert the changes to struct kinfo_proc in r194498.  Instead,
>>>>> fill in up to 16 (KI_NGROUPS) values and steal a bit from
>>>>> ki_cr_flags (all bits currently unused) to indicate overflow with
>>>>> the new flag KI_CRF_GRP_OVERFLOW.
>>>>>
>>>>>  This fixes procstat -s.
>>>>>
>>>>>  Approved by: re (kib)
>>>>
>>>> Hi Brooks,
>>>>
>>>> This checkin appears to have broken crash dump support: the bcopy  
>>>> in
>>>> kvm_proc.c crashes when running "ps -ax -M /var/crash/vmcore.x
>>>> -N /boot/kernel/kernel", apparently because ucred.cr_groups isn't
>>>> valid. Does it need converted using KVM_READ first?
>>>
>>> Sorry for the delay, I believe you are correct we need to use  
>>> kvm_read
>>> here instead of the bcopy.  Do you still have a core handy you can
>>> test a patch against?  The following should do it.
>>
>> The patch has fixed the crash, and data is now returned too.
>
> Thanks for testing!  I've committed the patch and will MFC in the next
> week or so.
MFC coming up? The patch fixes the segfault for me as well, and this  
should *really* make it into 8.0.

Regards,
Thomas