El Wed, Apr 01, 2009 at 10:30:08AM +0100 Catalin Marinas ha dit: > On Mon, 2009-03-30 at 14:24 +0200, Matthias Kaehlcke wrote: > > linux-next 20090330 causes the following page fault on an edb9302 > > (ARM) like board: > > > > # Unable to handle kernel paging request at virtual address 00535447 > > pgd = c56b0000 > > [00535447] *pgd=00000000 > > Internal error: Oops: 3 [#1] PREEMPT > > Modules linked in: > > CPU: 0 Not tainted (2.6.29-next-20090330 #6) > > PC is at kmemleak_scan+0x158/0x3ac > > LR is at find_and_get_object+0xe0/0x104 > > pc : [] lr : [] psr: 00000013 > > sp : c55bff78 ip : c55bff28 fp : c55bffa4 > > r10: c0322718 r9 : c035fa6c r8 : c035f048 > > r7 : c035f044 r6 : 00005800 r5 : 00004981 r4 : 00093020 > > r3 : 00535443 r2 : 414f4c5f r1 : 00000001 r0 : c0400020 > > Flags: nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel > > Control: 0000717f Table: 0567c000 DAC: 00000017 > > Process kmemleak (pid: 343, stack limit = 0xc55be268) > [...] > > the fault is reproducible and happens some seconds after having > > finished the boot process. please tell me if you need more information > > (like the .config, ...) in order to track this down > > Yes, the .config would be useful. I'm mainly interested in which slab > allocator you are using. I'm testing kmemleak mainly on ARM and haven't > seen any issues. attached is the .config for linux-next 20090401, which should be roughly the same than the 20090330 one. some hours ago i booted the board with 20090401 and the page fault didn't happen -- Matthias Kaehlcke Embedded Linux Engineer Barcelona La libertad es como la maņana. Hay quienes esperan dormidos a que llegue, pero hay quienes desvelan y caminan la noche para alcanzarla (Subcomandante Marcos) .''`. using free software / Debian GNU/Linux | http://debian.org : :' : `. `'` gpg --keyserver pgp.mit.edu --recv-keys 47D8E5D4 `-