Skip to contentSkip to conversation

drm/omap: dsi: Fix deferred probe warnings

PATCH In review Tony Lindgren 4 participants 11 messages Apr 12, 2023 – Sep 18, 2023

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 drm/omap: dsi: Fix deferred probe warnings

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

Let's fix the warnings by checking for a valid dsi->dsidev.

---
 drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -1
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
 {
 	struct dsi_data *dsi = host_to_omap(host);
 
-	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
 		return -EINVAL;
 
 	cancel_delayed_work_sync(&dsi->dsi_disable_work);
Signature · 1 line
2.40.0
#1 lore raw

Hi Tony,

Thank you for the patch.

Quoted 6 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

How can this happen ? I assume .detach() can't be called without a priori successful call to .attach(), that that sets dsi->dsidev.

Quoted 19 lines

Let's fix the warnings by checking for a valid dsi->dsidev.

---
 drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -1
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
 {
 	struct dsi_data *dsi = host_to_omap(host);
 
-	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
 		return -EINVAL;
 
 	cancel_delayed_work_sync(&dsi->dsi_disable_work);
Signature · 3 lines
Regards,

Laurent Pinchart
maintainer
#2 lore raw
Quoted 9 lines · Laurent Pinchart

Hi Tony,

Thank you for the patch.

Quoted 4 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

Quoted 5 lines
Quoted 2 lines

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

How can this happen ? I assume .detach() can't be called without a priori successful call to .attach(), that that sets dsi->dsidev.

I had a quick look, and the driver calls mipi_dsi_host_register() in probe, and mipi_dsi_host_unregister() in remove.

mipi_dsi_host_unregister() always calls mipi_dsi_detach(), but I don't think mipi_dsi_host_register() always calls attach, which happens later when the peripheral probes.

Tomi

Quoted 20 lines
Quoted 19 lines

Let's fix the warnings by checking for a valid dsi->dsidev.

---
  drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -1
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
  {
  	struct dsi_data *dsi = host_to_omap(host);
 
-	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
  		return -EINVAL;
 
  	cancel_delayed_work_sync(&dsi->dsi_disable_work);
#3 lore raw
Quoted 16 lines · Tomi Valkeinen
Quoted 9 lines · Laurent Pinchart

Hi Tony,

Thank you for the patch.

Quoted 4 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

Quoted 5 lines
Quoted 2 lines

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

How can this happen ? I assume .detach() can't be called without a priori successful call to .attach(), that that sets dsi->dsidev.

Quoted 6 lines

I had a quick look, and the driver calls mipi_dsi_host_register() in probe, and mipi_dsi_host_unregister() in remove.

mipi_dsi_host_unregister() always calls mipi_dsi_detach(), but I don't think mipi_dsi_host_register() always calls attach, which happens later when the peripheral probes.

Is this something that should be addressed in the DRM MIPI DSI helpers, to only detach after an attach ?

Quoted 19 lines
Quoted 19 lines
Quoted 19 lines

Let's fix the warnings by checking for a valid dsi->dsidev.

---
  drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -1
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
  {
  	struct dsi_data *dsi = host_to_omap(host);
 
-	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
  		return -EINVAL;
 
  	cancel_delayed_work_sync(&dsi->dsi_disable_work);
Signature · 3 lines
Regards,

Laurent Pinchart
#4 lore raw
Quoted 20 lines · Laurent Pinchart
Quoted 19 lines · Tomi Valkeinen
Quoted 9 lines · Laurent Pinchart

Hi Tony,

Thank you for the patch.

Quoted 4 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

Quoted 5 lines
Quoted 2 lines

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

How can this happen ? I assume .detach() can't be called without a priori successful call to .attach(), that that sets dsi->dsidev.

I had a quick look, and the driver calls mipi_dsi_host_register() in probe, and mipi_dsi_host_unregister() in remove.

Quoted 6 lines
Quoted 3 lines

mipi_dsi_host_unregister() always calls mipi_dsi_detach(), but I don't think mipi_dsi_host_register() always calls attach, which happens later when the peripheral probes.

Is this something that should be addressed in the DRM MIPI DSI helpers, to only detach after an attach ?

Tomi, any comments on detach after attach?

Regards,

Tony

Quoted 25 lines
Quoted 19 lines
Quoted 19 lines
Quoted 19 lines

Let's fix the warnings by checking for a valid dsi->dsidev.

---
  drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -1
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
  {
  	struct dsi_data *dsi = host_to_omap(host);
 
-	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
  		return -EINVAL;
 
  	cancel_delayed_work_sync(&dsi->dsi_disable_work);
Signature · 3 lines
Regards,

Laurent Pinchart
maintainer
#5 lore raw
Quoted 21 lines · Tony Lindgren
Quoted 20 lines · Laurent Pinchart
Quoted 19 lines · Tomi Valkeinen
Quoted 9 lines · Laurent Pinchart

Hi Tony,

Thank you for the patch.

Quoted 4 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

Quoted 5 lines
Quoted 2 lines

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

How can this happen ? I assume .detach() can't be called without a priori successful call to .attach(), that that sets dsi->dsidev.

I had a quick look, and the driver calls mipi_dsi_host_register() in probe, and mipi_dsi_host_unregister() in remove.

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

mipi_dsi_host_unregister() always calls mipi_dsi_detach(), but I don't think mipi_dsi_host_register() always calls attach, which happens later when the peripheral probes.

Is this something that should be addressed in the DRM MIPI DSI helpers, to only detach after an attach ?

Tomi, any comments on detach after attach?

I sent a comment to the patch.

As Laurent said, I think this really should be fixed in the framework side. Calling detach in drivers without attach feels just plain wrong.

However, I don't see any harm in the patch (but perhaps some changes needed as per my comments), and it will fix the issue for omapdrm, until someone has the time and energy to look at a real fix.

Tomi

#6 lore raw
In reply to Laurent Pinchart
Quoted 12 lines

Am 12.04.2023 um 10:50 schrieb Laurent Pinchart <[email protected]>:

Hi Tony,

Thank you for the patch.

Quoted 4 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

Quoted 5 lines
Quoted 2 lines

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

How can this happen ? I assume .detach() can't be called without a priori successful call to .attach(), that that sets dsi->dsidev.

I have a similar patch (not submitted because it looks like a workaround) in our LetuxOS kernel:

https://git.goldelico.com/?p=letux-kernel.git;a=commitdiff;h=0656acb534439d4546b33826de62694e1df1b8ad

There I had commented:

dsi_init_output() called by dsi_probe() may fail. In that case mipi_dsi_host_unregister() is called which may call omap_dsi_host_detach() with uninitialized dsi->dsidev because omap_dsi_host_attach() was never called before.

This happens if the panel driver asks for an EPROBE_DEFER.

So let's suppress the WARN() in this special case.

[    7.416759] WARNING: CPU: 0 PID: 32 at drivers/gpu/drm/omapdrm/dss/dsi.c:4419 omap_dsi_host_detach+0x3c/0xbc [omapdrm]
[    7.436053] Modules linked in: ina2xx_adc snd_soc_ts3a227e bq2429x_charger bq27xxx_battery_i2c(+) bq27xxx_battery ina2xx tca8418_keypad as5013(+) omapdrm hci_uart cec palmas_pwrbutton btbcm bmp280_spi palmas_gpadc bluetooth usb3503 ecdh_generic bmc150_accel_i2c bmg160_i2c ecc bmc150_accel_core bmg160_core bmc150_magn_i2c bmp280_i2c bmc150_magn bno055 industrialio_triggered_buffer bmp280 kfifo_buf snd_soc_omap_aess display_connector drm_kms_helper syscopyarea snd_soc_omap_mcbsp snd_soc_ti_sdma sysfillrect ti_tpd12s015 sysimgblt fb_sys_fops wwan_on_off snd_soc_gtm601 generic_adc_battery drm snd_soc_w2cbw003_bt industrialio drm_panel_orientation_quirks pwm_bl pwm_omap_dmtimer ip_tables x_tables ipv6 autofs4
[    7.507068] CPU: 0 PID: 32 Comm: kworker/u4:2 Tainted: G        W          6.1.0-rc3-letux-lpae+ #11107
[    7.516964] Hardware name: Generic OMAP5 (Flattened Device Tree)
[    7.523284] Workqueue: events_unbound deferred_probe_work_func
[    7.529456]  unwind_backtrace from show_stack+0x10/0x14
[    7.534972]  show_stack from dump_stack_lvl+0x40/0x4c
[    7.540315]  dump_stack_lvl from __warn+0xb0/0x164
[    7.545379]  __warn from warn_slowpath_fmt+0x70/0x9c
[    7.550625]  warn_slowpath_fmt from omap_dsi_host_detach+0x3c/0xbc [omapdrm]
[    7.558137]  omap_dsi_host_detach [omapdrm] from mipi_dsi_remove_device_fn+0x10/0x20
[    7.566376]  mipi_dsi_remove_device_fn from device_for_each_child+0x60/0x94
[    7.573729]  device_for_each_child from mipi_dsi_host_unregister+0x20/0x54
[    7.580992]  mipi_dsi_host_unregister from dsi_probe+0x5d8/0x744 [omapdrm]
[    7.588315]  dsi_probe [omapdrm] from platform_probe+0x58/0xa8
[    7.594542]  platform_probe from really_probe+0x144/0x2ac
[    7.600249]  really_probe from __driver_probe_device+0xc4/0xd8
[    7.606411]  __driver_probe_device from driver_probe_device+0x3c/0xb8
[    7.613216]  driver_probe_device from __device_attach_driver+0x58/0xbc
[    7.620115]  __device_attach_driver from bus_for_each_drv+0xa0/0xb4
[    7.626737]  bus_for_each_drv from __device_attach+0xdc/0x150
[    7.632808]  __device_attach from bus_probe_device+0x28/0x80
[    7.638792]  bus_probe_device from deferred_probe_work_func+0x84/0xa0
[    7.645595]  deferred_probe_work_func from process_one_work+0x1a4/0x2d8
[    7.652587]  process_one_work from worker_thread+0x214/0x2b8
[    7.658567]  worker_thread from kthread+0xe4/0xf0
[    7.663542]  kthread from ret_from_fork+0x14/0x1c
[    7.668515] Exception stack(0xf01b5fb0 to 0xf01b5ff8)
[    7.673827] 5fa0:                                     00000000 00000000 00000000 00000000
[    7.682435] 5fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[    7.691038] 5fe0: 00000000 00000000 00000000 00000000 00000013 00000000

Hope this helps to find the real cause.

BR,
Nikolaus

maintainer
#7 lore raw
In reply to Tony Lindgren
Quoted 18 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

Let's fix the warnings by checking for a valid dsi->dsidev.

---
  drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

drivers/gpu/drm/omapdrm/dss/dsi.c static int omap_dsi_host_detach(struct mipi_dsi_host *host, · line 4417

Quoted 8 lines
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -1
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
  {
  	struct dsi_data *dsi = host_to_omap(host);
 
-	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
  		return -EINVAL;
 
  	cancel_delayed_work_sync(&dsi->dsi_disable_work);

Shouldn't this rather be

if (!dsi->dsidev)

	return 0;

before the if (WARN_ON(dsi->dsidev != client)) line?

If dsi->dsidev is NULL, then attach hasn't been called, and we shouldn't do anything in the detach callback either.

With your change we'll end up doing all the work in the detach callback, without ever doing their counterpart in the attach side.

Tomi

#8 lore raw
Quoted 31 lines · Tomi Valkeinen
Quoted 18 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.

The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

Let's fix the warnings by checking for a valid dsi->dsidev.

---
  drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

drivers/gpu/drm/omapdrm/dss/dsi.c static int omap_dsi_host_detach(struct mipi_dsi_host *host, · line 4415

Quoted 6 lines
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -1
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
  {
  	struct dsi_data *dsi = host_to_omap(host);
-	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
  		return -EINVAL;
  	cancel_delayed_work_sync(&dsi->dsi_disable_work);

Shouldn't this rather be

if (!dsi->dsidev)

	return 0;
Quoted 7 lines

before the if (WARN_ON(dsi->dsidev != client)) line?

If dsi->dsidev is NULL, then attach hasn't been called, and we shouldn't do anything in the detach callback either.

With your change we'll end up doing all the work in the detach callback, without ever doing their counterpart in the attach side.

Oops, I'll check that thanks.

Tony

#9 lore raw
In reply to Tomi Valkeinen

Hi Tomi and Tony,

Quoted 26 lines

Am 13.09.2023 um 13:59 schrieb Tomi Valkeinen <[email protected]>:

Quoted 14 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.
The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

Let's fix the warnings by checking for a valid dsi->dsidev.

---
 drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

drivers/gpu/drm/omapdrm/dss/dsi.c static int omap_dsi_host_detach(struct mipi_dsi_host *host, · line 4416

Quoted 6 lines
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -0
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
 {
 	struct dsi_data *dsi = host_to_omap(host);
 -	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
 		return -EINVAL;
   	cancel_delayed_work_sync(&dsi->dsi_disable_work);

Shouldn't this rather be

Quoted 4 lines

if (!dsi->dsidev)

	return 0;

before the if (WARN_ON(dsi->dsidev != client)) line?

Yes you are right. We have a different variant in our Pyra kernel:

What we currently have in our Pyra tree is: https://git.goldelico.com/?p=letux-kernel.git;a=commitdiff;h=5bf7bd64eec1eb924e794e8d6600919f0dae8c5a;hp=27a0cd6263194d1465e9c53293d35f8c8c988f9d

        struct dsi_data *dsi = host_to_omap(host);

- if (WARN_ON(dsi->dsidev != client))
+printk("%s\n", __func__);
+
+ if (!dsi->dsidev || WARN_ON(dsi->dsidev != client))

                return -EINVAL;
 
        cancel_delayed_work_sync(&dsi->dsi_disable_work);
Quoted 4 lines

If dsi->dsidev is NULL, then attach hasn't been called, and we shouldn't do anything in the detach callback either.

With your change we'll end up doing all the work in the detach callback, without ever doing their counterpart in the attach side.

If useful, I can post above mentioned patch (without printk).

BR and thanks,
Nikolaus

#10 lore raw
Quoted 48 lines · H. Nikolaus Schaller

Hi Tomi and Tony,

Quoted 26 lines

Am 13.09.2023 um 13:59 schrieb Tomi Valkeinen <[email protected]>:

Quoted 14 lines · Tony Lindgren

We may not have dsi->dsidev initialized during probe, and that can lead into various dsi related warnings as omap_dsi_host_detach() gets called with dsi->dsidev set to NULL.
The warnings can be "Fixed dependency cycle(s)" followed by a

WARNING: CPU: 0 PID: 787 at drivers/gpu/drm/omapdrm/dss/dsi.c:4414.

Let's fix the warnings by checking for a valid dsi->dsidev.

---
 drivers/gpu/drm/omapdrm/dss/dsi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

drivers/gpu/drm/omapdrm/dss/dsi.c static int omap_dsi_host_detach(struct mipi_dsi_host *host, · line 4416

Quoted 6 lines
drivers/gpu/drm/omapdrm/dss/dsi.c +1 -0
@@ -4411,7 +4411,7 @@ static int omap_dsi_host_detach(struct mipi_dsi_host *host,
 {
 	struct dsi_data *dsi = host_to_omap(host);
 -	if (WARN_ON(dsi->dsidev != client))
+	if (dsi->dsidev && WARN_ON(dsi->dsidev != client))
 		return -EINVAL;
   	cancel_delayed_work_sync(&dsi->dsi_disable_work);

Shouldn't this rather be

Quoted 4 lines

if (!dsi->dsidev)

	return 0;

before the if (WARN_ON(dsi->dsidev != client)) line?

Yes you are right. We have a different variant in our Pyra kernel:

What we currently have in our Pyra tree is: https://git.goldelico.com/?p=letux-kernel.git;a=commitdiff;h=5bf7bd64eec1eb924e794e8d6600919f0dae8c5a;hp=27a0cd6263194d1465e9c53293d35f8c8c988f9d

        struct dsi_data *dsi = host_to_omap(host);
drivers/gpu/drm/omapdrm/dss/dsi.c +3 -1
-       if (WARN_ON(dsi->dsidev != client))
+printk("%s\n", __func__);
+
+       if (!dsi->dsidev || WARN_ON(dsi->dsidev != client))
                return -EINVAL;
 
        cancel_delayed_work_sync(&dsi->dsi_disable_work);
Quoted 2 lines
Quoted 5 lines
Quoted 3 lines

If dsi->dsidev is NULL, then attach hasn't been called, and we shouldn't do anything in the detach callback either.

With your change we'll end up doing all the work in the detach callback, without ever doing their counterpart in the attach side.

If useful, I can post above mentioned patch (without printk).

Sounds good to me.

Thanks,

Tony