Hi Fabio,
Thanks for testing.
On 11.01.2016 13:10, Fabio Estevam wrote:
> On Mon, Jan 11, 2016 at 10:04 AM, Fabio Estevam <festevam@xxxxxxxxx> wrote:
>
>> This patch causes the following issue in linux-next:
>>
>> [ 2.526984] ------------[ cut here ]------------
>> [ 2.531632] WARNING: CPU: 1 PID: 1 at kernel/locking/lockdep.c:2755
>> lockdep_trace_alloc+0xf4/0x124()
>> [ 2.540771] DEBUG_LOCKS_WARN_ON(irqs_disabled_flags(flags))
>> [ 2.546175] Modules linked in:
>> [ 2.549447] CPU: 1 PID: 1 Comm: swapper/0 Not tainted
>> 4.4.0-rc8-next-20160111 #204
>> [ 2.557021] Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree)
>> [ 2.563553] Backtrace:
>> [ 2.566040] [<c00136d8>] (dump_backtrace) from [<c0013874>]
>> (show_stack+0x18/0x1c)
>> [ 2.573615] r6:00000ac3 r5:00000000 r4:00000000 r3:00000000
>> [ 2.579362] [<c001385c>] (show_stack) from [<c02de89c>]
>> (dump_stack+0x88/0xa4)
>> [ 2.586607] [<c02de814>] (dump_stack) from [<c002bcac>]
>> (warn_slowpath_common+0x80/0xbc)
>> [ 2.594702] r5:c0071ed0 r4:ef055b90
>> [ 2.598326] [<c002bc2c>] (warn_slowpath_common) from [<c002bd8c>]
>> (warn_slowpath_fmt+0x38/0x40)
>> [ 2.607028] r8:00000004 r7:00000004 r6:024080c0 r5:024080c0 r4:60000093
>> [ 2.613829] [<c002bd58>] (warn_slowpath_fmt) from [<c0071ed0>]
>> (lockdep_trace_alloc+0xf4/0x124)
>> [ 2.622532] r3:c09a1634 r2:c099dc0c
>> [ 2.626161] [<c0071ddc>] (lockdep_trace_alloc) from [<c010bc0c>]
>> (kmem_cache_alloc+0x30/0x174)
>> [ 2.634778] r4:ef001f00 r3:c0b02a88
>> [ 2.638407] [<c010bbdc>] (kmem_cache_alloc) from [<c03edac8>]
>> (regcache_rbtree_write+0x150/0x724)
>> [ 2.647283] r10:00000000 r9:00000010 r8:00000004 r7:00000004
>> r6:0000002c r5:00000000
>> [ 2.655203] r4:00000000
>> [ 2.657767] [<c03ed978>] (regcache_rbtree_write) from [<c03ec7d8>]
>> (regcache_write+0x5c/0x64)
>
> This fixes the warning:
>
> --- a/sound/soc/fsl/fsl_ssi.c
> +++ b/sound/soc/fsl/fsl_ssi.c
> @@ -180,7 +180,6 @@ static const struct regmap_config fsl_ssi_regconfig = {
> .volatile_reg = fsl_ssi_volatile_reg,
> .precious_reg = fsl_ssi_precious_reg,
> .writeable_reg = fsl_ssi_writeable_reg,
> - .cache_type = REGCACHE_RBTREE,
> };
>
> Is this the correct fix?
>
This will disable register cache so it isn't right.
Could you try REGCACHE_FLAT instead, please?
Looks like the problem here is rbtree cache does some non-atomic allocations in
read / write path when not supplied with default register values.
Best regards,
Maciej Szmigiero
_______________________________________________
Alsa-devel mailing list
Alsa-devel@xxxxxxxxxxxxxxxx
http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
