Magnus Damm wrote: > On 12/22/06, Jay Lan <[EMAIL PROTECTED]> wrote: >> Magnus Damm wrote: >> > On 12/19/06, Jay Lan <[EMAIL PROTECTED]> wrote: >> > I'd like to spend time on fixing up ia64 support properly, but it has >> > been sort of a moving target because of heavy development. So to be >> >> That was what i thought. Is the old kexec tree no longer maintained? > > I would not call the old tree maintained. Other people may have > different opinions though. > > Simon (Horms) and I felt that it was painful to keep track of all > patches floating around for the old tree - they were never picked up - > so to solve this problem Simon created the kexec-tools-testing tree > and he has been picking up patches ever since.
You and Simon's work are greatly appreciated! > >> I assume you have included all ia64 patches from Nan-hai? There is >> certainly work needed to be done to make ia64 support solid in >> kexec-tools-testing. > > I know that Simon has been trying hard to locate all patches and > include them in the tree, but he may have missed a few. It is also > possible that some patches are in a deadlock state because of review > and different opinions on how things should be implemented. But the > most likely problem IMO is that the kernel code and the kexec-tools > code are out of sync because of all recent development. > > Good thing is that all these things are solvable - it's just a matter > of spending time. And now when kexec for ia64 is included in > 2.6.20-rc1 we should have some "firm ground" to work with which is > nice. Agreed. > >> > Are you still experiencing this problem? >> >> Yes. > > Ack. =) > >> > >> > You mention the RHEL5 kexec plus your patch, do you have a pointer to >> > a SRPM of the version you are using? >> > >> > What kernel are you using? Any patches? Have you tried 2.6.20-rc1? >> >> I have not tried 2.6.20-rc1 yet, but i believe it is a very good target. >> It contains all kernel patches IA64 SN needs, i think. I recently >> focused on kernel and kexec-tools of rhel5. > > I see. I've mainly worked with vanilla kernels and kexec-tools-testing > so we need to find some common ground. Maybe it is a good idea that > you test the RHEL5 kernel with kexec-tools-testing and old > kexec-tools, while I do the same with 2.6.20-rc... Or we could both > focus on 2.6.20-rc and then backport whatever needed for the kernel. Sound good. I will test 2.6.20-rc on SN with both old kexec-tools and the kexec-tools-testing and go from there. Thanks, - jay > >> > Do you have time for more testing? >> > >> > I will go on vacation over the christmas and new years, but if you >> > have time after that I'd be more than happy to work with you to >> > resolve this. I suspect it will be a matter of bisecting and testing >> > until we find what is broken. I have a few ia64 boxes here that I can >> > use too, but no altix. =) >> >> I would be very glad to work with you to make kexec-tools-testing >> solid for IA64 SN. I can not speak for other IA64 boxes since i do >> not have access to them (except one zx workstation which i had some >> problem with.) Any time after new year would be perfect. > > Great, let's exercise our machine park and minds after new year then! =) > > Cheers, > > / magnus _______________________________________________ fastboot mailing list [email protected] https://lists.osdl.org/mailman/listinfo/fastboot
