On Fri, Jul 19, 2013 at 08:50:20AM +0200, Martin Schwidefsky wrote: > On Thu, 18 Jul 2013 09:35:41 -0400 > Vivek Goyal <vgoyal at redhat.com> wrote: > > > On Thu, Jul 18, 2013 at 12:47:54PM +0200, Martin Schwidefsky wrote: > > > On Thu, 18 Jul 2013 12:40:04 +0200 > > > Michael Holzheu <holzheu at linux.vnet.ibm.com> wrote: > > > > > > > On Wed, 17 Jul 2013 17:42:07 -0400 > > > > Vivek Goyal <vgoyal at redhat.com> wrote: > > > > > On Wed, Jul 17, 2013 at 06:00:49PM +0200, Michael Holzheu wrote: > > > > > > > > [snip] > > > > > > > > > > But this is all additional effort now and would not be necessary if we > > > > > > integrate this patch series in 3.11. > > > > > > > > > > > > Perhaps we should let Andrew decide here. > > > > > > > > > > Hi Michael, > > > > > > > > > > Given the fact that andrew too prefers a fix to get s390 working at this > > > > > stage can we modify s390 copy_from_oldmem() to be able to copy to > > > > > vmalloc() memory area. > > > > > > > > > > For mmap(), let us disable it on s390. And rest of the cleanups w.r.t > > > > > ELF header swap etc, let us now target that for 3.12. > > > > > > > > > > Sounds reasonable? > > > > > > > > Hi Vivek, > > > > > > > > Ok this is not our preferred solution but we can't expect that life is > > > > always easy ;-) > > > > > > > > Our s390 kernel maintainer Martin Schwidefsky agreed to send the following > > > > two patches upstream for 3.11: > > > > > > > > * s390/kdump: Disable mmap for s390 > > > > * s390/kdump: Allow copy_oldmem_page() copy to virtual memory > > > > > > Patches have been added to > > > git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux.git features > > > > > > They will go upstream with my next pull request. > > > > > > > If everything related to crash dump is going through Andrew, wouldn't > > it make sense that even these fixes go through him? > > > > https://git.kernel.org/cgit/linux/kernel/git/s390/linux.git/commit/?h=features&id=5a74953ff56aa870d6913ef4d81934f5c620c59d > > > > https://git.kernel.org/cgit/linux/kernel/git/s390/linux.git/commit/?h=features&id=191a2fa0a8d2bbb64c98f9b1976fcb37ee5eae6b > > Most of the architecture specific kdump patches for s390 have gone through me, > and these two fall into that category. If you insist we can route them over > Andrew, it just seems easier to me via the s390 tree. > In this case one patch is modifying proc/vmcore.c and that's not arch specific change only. Secondly, in this specific instance, these changes are in the context of mmap() related changes which just were pushed. I find it easier if fixes flow through same channel where previous big set of patches flowed through. Hence, I was expecting that Michael will route these fixes through Andrew. But I am not too particular about it, so go ahead and push it to Linus. Thanks Vivek
