Russell,
Shall I put these two patches in the patch tracker for 3.11?
On 06/14/13 13:32, Stephen Boyd wrote:
> Failure to add the mapping created in debug_ll_io_init() can lead
> to the BUG_ON() triggering in lib/ioremap.c:27 if the static
> virtual address decided for the debug_ll mapping overlaps with
> another mapping that is created later. This happens because the
> generic ioremap code has no idea there is a mapping there and it
> tries to place a mapping in the same location and blows up when
> it sees that there is a pte already present.
>
> kernel BUG at lib/ioremap.c:27!
> Internal error: Oops - BUG: 0 [#1] PREEMPT SMP ARM
> Modules linked in:
> CPU: 0 PID: 1 Comm: swapper/0 Not tainted 3.10.0-rc2-00042-g2af0c67-dirty #316
> task: ef088000 ti: ef082000 task.ti: ef082000
> PC is at ioremap_page_range+0x16c/0x198
> LR is at ioremap_page_range+0xf0/0x198
> pc : []lr : []psr: 2113
> sp : ef083e78 ip : af14 fp : ef083ebc
> r10: ef7fc100 r9 : ef7fc104 r8 : 000af174
> r7 : 0647 r6 : beff r5 : f004c000 r4 : f004
> r3 : af173417 r2 : 16440653 r1 : af173e07 r0 : ef7fc8fc
> Flags: nzCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel
> Control: 10c5787d Table: 8020406a DAC: 0015
> Process swapper/0 (pid: 1, stack limit = 0xef082238)
> Stack: (0xef083e78 to 0xef084000)
> 3e60: 0004 ef083eec
> 3e80: bf134000 f004bfff c0207c00 f004c000 c02fc120 f000c000 c15e7800 0004
> 3ea0: ef083eec 0647 c098ba9c c0953544 ef083edc ef083ec0 c021b82c c04cb714
> 3ec0: c09cdc50 0040 ef0f1e00 ef1003c0 ef083f14 ef083ee0 c09535bc c021b7bc
> 3ee0: c0953544 c04d0c6c c094e2cc c1600be4 c07440c4 c09a6888 0002 c0a15f00
> 3f00: ef082000 ef083f54 ef083f18 c0208728 c0953550 0002 c1600bfc
> 3f20: c08e3fac c0839918 ef083f54 c1600b80 c09a6888 c0a15f00 008b c094e2cc
> 3f40: c098ba9c c098bab8 ef083f94 ef083f58 c094ea0c c020865c 0002 0002
> 3f60: c094e2cc c025b674 c06ff860
> 3f80: ef083fac ef083f98 c06ff878 c094e910
> 3fa0: ef083fb0 c020efe8 c06ff86c
> 3fc0:
> 3fe0: 0013 c0595108
> [] (ioremap_page_range+0x16c/0x198) from []
> (__alloc_remap_buffer.isra.18+0x7c/0xc4)
> [] (__alloc_remap_buffer.isra.18+0x7c/0xc4) from []
> (atomic_pool_init+0x78/0x128)
> [] (atomic_pool_init+0x78/0x128) from []
> (do_one_initcall+0xd8/0x198)
> [] (do_one_initcall+0xd8/0x198) from []
> (kernel_init_freeable+0x108/0x1d0)
> [] (kernel_init_freeable+0x108/0x1d0) from []
> (kernel_init+0x18/0xf4)
> [] (kernel_init+0x18/0xf4) from []
> (ret_from_fork+0x14/0x20)
> Code: e50b0040 ebf54b2f e51b0040 eaee (e7f001f2)
>
> Fix it by telling generic layers about the static mapping via
> iotable_init(). This also has the nice side effect of letting
> you see the mapping in procfs' vmallocinfo file.
>
> Cc: Rob Herring
> Cc: Stephen Warren
> Signed-off-by: Stephen Boyd
> ---
>
> Changes since v1:
> * Use iotable_init() instead of open coding it
>
> arch/arm/mm/mmu.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c
> index e0d8565..18b8aeb 100644
> --- a/arch/arm/mm/mmu.c
> +++ b/arch/arm/mm/mmu.c
> @@ -947,7 +947,7 @@ void __init debug_ll_io_init(void)
> map.virtual &= PAGE_MASK;
> map.length = PAGE_SIZE;
> map.type = MT_DEVICE;
> - create_mapping(&map);
> + iotable_init(&map, 1);
> }
> #endif
>
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
hosted by The Linux Foundation
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
