[0/4] ASoC: simple-card / audio-graph re-cleanup

Hi Mark, Guillaume

KernelCI had reported that my previous cleanup patches for simple-card / audio-graph
broke kontron-sl28-var3-ads2 sound card probing.
It reported that it is using same name for dailink->name.

At first I thought that the issue was fsl,vf610-sai doesn't have .name on driver.
But real issue was that cpu->dai_name removed on simple_parse_node(),
and dailink->name was based on it.
We need to set/get dailink->name first, and call simple_parse_node().
This patches are do it.
audio-graph has same issue. [4/4] patch is for it.

I hope these patches has no issues on kontron-sl28-var3-ads2.

Link: https://lore.kernel.org/r/87h7k0i437.wl-kuninori.morimoto.gx@renesas.com
Link: https://lore.kernel.org/r/20210423175318.13990-1-broonie@kernel.org
Link: https://lore.kernel.org/r/3ca62063-41b4-c25b-a7bc-8a8160e7b684@collabora.com

Kuninori Morimoto (4):
  ASoC: simple-card: add simple_parse_node()
  ASoC: simple-card: add simple_link_init()
  ASoC: audio-graph: tidyup graph_dai_link_of_dpcm()
  ASoC: audio-graph: tidyup dai_name seting timing

 sound/soc/generic/audio-graph-card.c |  48 +++----
 sound/soc/generic/simple-card.c      | 187 ++++++++++++++-------------
 2 files changed, 122 insertions(+), 113 deletions(-)

Hi,

Am 2021-05-10 03:51, schrieb Kuninori Morimoto:
> Hi Mark, Guillaume
> 
> KernelCI had reported that my previous cleanup patches for simple-card
> / audio-graph
> broke kontron-sl28-var3-ads2 sound card probing.
> It reported that it is using same name for dailink->name.
> 
> At first I thought that the issue was fsl,vf610-sai doesn't have .name
> on driver.
> But real issue was that cpu->dai_name removed on simple_parse_node(),
> and dailink->name was based on it.
> We need to set/get dailink->name first, and call simple_parse_node().
> This patches are do it.
> audio-graph has same issue. [4/4] patch is for it.
> 
> I hope these patches has no issues on kontron-sl28-var3-ads2.

I've just tested this on kontron-sl28-var3-ads2 (based on the latest
next). Now I'm getting even less useful names ;)

[    6.769932] sysfs: cannot create duplicate filename 
'/devices/platform/sound/(null)-(null)'
[    6.778529] CPU: 1 PID: 65 Comm: kworker/u4:1 Not tainted 
5.12.0-next-20210506+ #527
[    6.786397] Hardware name: Kontron SMARC-sAL28 (Single PHY) on SMARC 
Eval 2.0 carrier (DT)
[    6.794789] Workqueue: events_unbound deferred_probe_work_func
[    6.800727] Call trace:
[    6.803209]  dump_backtrace+0x0/0x1b8
[    6.806933]  show_stack+0x20/0x30
[    6.810301]  dump_stack+0x100/0x170
[    6.813846]  sysfs_warn_dup+0x6c/0x88
[    6.817565]  sysfs_create_dir_ns+0xe8/0x100
[    6.821813]  kobject_add_internal+0x9c/0x290
[    6.826150]  kobject_add+0xa0/0x108
[    6.829693]  device_add+0xfc/0x840
[    6.833147]  device_register+0x28/0x38
[    6.836954]  snd_soc_add_pcm_runtime+0x274/0x760
[    6.841647]  snd_soc_bind_card+0x330/0x9b8
[    6.842884] hid-generic 0003:064F:2AF9.0001: device has no listeners, 
quitting
[    6.845811]  snd_soc_register_card+0x10c/0x128
[    6.845823]  devm_snd_soc_register_card+0x4c/0xa8
[    6.845830]  asoc_simple_probe+0x1e8/0x3c8
[    6.845838]  platform_probe+0x70/0xe0
[    6.845846]  really_probe+0xec/0x3c0
[    6.845853]  driver_probe_device+0x6c/0xd0
[    6.845861]  __device_attach_driver+0x98/0xe0
[    6.845869]  bus_for_each_drv+0x84/0xd8
[    6.886424]  __device_attach+0xf0/0x150
[    6.890322]  device_initial_probe+0x1c/0x28
[    6.894572]  bus_probe_device+0xa4/0xb0
[    6.898468]  deferred_probe_work_func+0x90/0xd0
[    6.903069]  process_one_work+0x2b8/0x720
[    6.907141]  worker_thread+0x4c/0x488
[    6.910860]  kthread+0x164/0x168
[    6.914141]  ret_from_fork+0x10/0x30
[    6.917924] kobject_add_internal failed for (null)-(null) with 
-EEXIST, don't try to register things with the same name in the same 
directory.
[    6.931455] asoc-simple-card: probe of sound failed with error -12

-michael

On Mon, May 10, 2021 at 01:27:17PM +0100, Guillaume Tucker wrote:

> Could you please add this trailer, or maybe the maintainers can
> add it when applying the patches?

>   Reported-by: "kernelci.org bot" <bot@kernelci.org>

If you're going to send stuff like this please send any tags you're
trying to get picked up in the normal format without any indentation at
the start of the line, that way tooling will pick it up.  Though in this
case Morimoto-san had already put this on patch 4 anyway.