Messages in this thread | | | Date | Sun, 5 Oct 2008 11:35:23 +0200 | From | Ingo Molnar <> | Subject | Re: [PATCH] dumpstack: x86: various small unification steps |
| |
* Alexander van Heukelum <heukelum@fastmail.fm> wrote:
> Hi Ingo, > > Here is the first round of unification of dumpstack_32.c > and dumpstack_64.c. The first patch can also be seen as > a clean-up of the traps.c-unification as I forgot to move > one function at the time. Anyhow, the series depends on > the traps unification.
great - i've created tip/x86/dumpstack for this and applied your patches there. (that branch embedds tip/x86/core which already embedds tip/x86/traps)
> B.T.W., I could reproduce the spontaneous reboot with the > traps unification with glibc. The change GATE_INTERRUPT -> > GATE_TRAP fixed the crash there. Usually I test only with > a small klibc-based userspace, and there the reboot does > not happen. I guess the int 0x80 interface is less picky. > (Just to say that I do test the changes a bit ;) )
it also passed -tip testing still then so the changes are fine.
Generally if you see your commits show up and stay in tip/master it means they get tested with a newly built random kernel about once every two minutes or so.
regarding klibc, that's interesting: is that the in-kernel klibc from hpa? Which tree are you using to pull that into tip/master? (unless i misunderstood what you are doing)
Ingo
| |