On Thu, 3 Apr 2014 21:58:01 -0500 Corey Wright <undefined@pobox.com> wrote: > applying patch-3.10.33-vs2.3.6.8.diff to linux-3.10.36 fails. > > the change of atomically setting ext4 inode flags in fs/ext4/inode.c [1] > causes two linux-vserver patch rejects. > > Hunk #1 FAILED at 38. > Hunk #2 FAILED at 4045. > > [1] > http://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/?id=0a0ae7b3fb0fb301da83f7c7da38807c76b2b869 > > the first failure is only because of including a new header (and > invalidating the context). > > to address the second failure, i modified the linux-vserver code (which > includes both i_flags and i_vflags) to update inode flags atomically (ie call > set_mask_bits), just like 3.10.36 introduced. the patch also references > ext4_get_inode_flags(), but that's only because the linux-vserver changes > within that function are in the same context as the ext4_set_inode_flags > conflicts, not because that function changed in 3.10.36. > > the attached patch fixes the two failures (and Makefile's EXTRAVERSION, too). > > instructions: > 1. acquire linux-3.10.36 > 2. apply patch-3.10.33-vs2.3.6.8.diff > 3. apply the attached patch-3.10.33-36-vs2.3.6.8.diff > > i've successfully built the resulting kernel, but i'm currently unable to > test it (against testme.sh and testfs.sh in virtualbox) until this weekend. > if somebody tests the resulting kernel before then, then please reply with > your results. i've tested the resulting kernel: * against testme.sh and testfs.sh in virtualbox * for 10 hours on a 30-guest production host corey -- undefined@pobox.com > corey > -- > undefined@pobox.com