Messages in this thread | | | From | Sven Joachim <> | Subject | Re: [stable] [PATCH] Fix 'flush_old_exec()/setup_new_exec()' split | Date | Fri, 05 Feb 2010 11:23:26 +0100 |
| |
On 2010-02-04 23:25 +0100, Ben Hutchings wrote:
> In a Debian initramfs, /init is a script interpreted by /bin/sh.
Yes, I tested without an initramfs though. The crash happens actually before init is run.
> I've now tested i386 userland on an x86_64 kernel with all the relevant > patches on top of 2.6.32.7: > > fdpic-respect-pt_gnu_stack-exec-protection-markings-when-creating-nommu-stack.patch > split-flush_old_exec-into-two-functions.patch > sparc-tif_abi_pending-bit-removal.patch > x86-get-rid-of-the-insane-tif_abi_pending-bit.patch > fix-flush_old_exec-setup_new_exec-split.patch > powerpc-tif_abi_pending-bit-removal.patch > > This works perfectly, so far as I can see.
It seems so. The faulty patch is actually the cherry-pick of commit 28f6aeea (net: restore ip source validation), reverting that leads to a kernel that runs fine. And this has nothing to do with 64-bit vs 32-bit.
Sven
| |