Skip to contentSkip to conversation

Fix Rock Pi 4 serial port

PATCH In review Joshua Watt 3 participants 5 messages Dec 11, 2020 – Dec 17, 2020

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 Fix Rock Pi 4 serial port

Fixes the serial port output stopping mid way through the boot process by reverting the kernel commit that caused it.

---
 ...-resolve-supply-after-creating-regul.patch | 53 +++++++++++++++++++
 recipes-kernel/linux/linux-yocto_5.8.bbappend |  4 ++
 2 files changed, 57 insertions(+)
 create mode 100644 recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch
 create mode 100644 recipes-kernel/linux/linux-yocto_5.8.bbappend
recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch new file +53 -0
new file mode 100644
index 0000000..3dd336b
@@ -0,0 +1,53 @@
+From a414d39b848002e15531f2538d2b6427ce51d07d Mon Sep 17 00:00:00 2001
+From: Joshua Watt <[email protected]>
+Date: Thu, 10 Dec 2020 15:59:47 -0600
+Subject: [PATCH] Revert "regulator: resolve supply after creating regulator"
+
+This commit prevents the serial console from working on the Rock Pi 4
+for some reason. It *appears* to possibly be fixed by some other commit
+upstream, but after a lot of head scratching and bisecting, I was unable
+to find which one, so just revert it for now and we can deal with it
+later.
+
+This reverts commit 96c6b5d5775637b3095ef934f871044811fd4db7.
+
+---
+ drivers/regulator/core.c | 21 ++++++++-------------
+ 1 file changed, 8 insertions(+), 13 deletions(-)
+
+diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
+index 25e601bf9383..be8c709a7488 100644
+--- a/drivers/regulator/core.c
++++ b/drivers/regulator/core.c
+@@ -5187,20 +5187,15 @@ regulator_register(const struct regulator_desc *regulator_desc,
+ 	else if (regulator_desc->supply_name)
+ 		rdev->supply_name = regulator_desc->supply_name;
+ 
++	/*
++	 * Attempt to resolve the regulator supply, if specified,
++	 * but don't return an error if we fail because we will try
++	 * to resolve it again later as more regulators are added.
++	 */
++	if (regulator_resolve_supply(rdev))
++		rdev_dbg(rdev, "unable to resolve supply\n");
++
+ 	ret = set_machine_constraints(rdev, constraints);
+-	if (ret == -EPROBE_DEFER) {
+-		/* Regulator might be in bypass mode and so needs its supply
+-		 * to set the constraints */
+-		/* FIXME: this currently triggers a chicken-and-egg problem
+-		 * when creating -SUPPLY symlink in sysfs to a regulator
+-		 * that is just being created */
+-		ret = regulator_resolve_supply(rdev);
+-		if (!ret)
+-			ret = set_machine_constraints(rdev, constraints);
+-		else
+-			rdev_dbg(rdev, "unable to resolve supply early: %pe\n",
+-				 ERR_PTR(ret));
+-	}
+ 	if (ret < 0)
+ 		goto wash;
+ 
+-- 
+2.29.2
+
recipes-kernel/linux/linux-yocto_5.8.bbappend new file +4 -0
new file mode 100644
index 0000000..5a31842
@@ -0,0 +1,4 @@
+FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
+
+SRC_URI_append_rock-pi-4 = " file://0001-Revert-regulator-resolve-supply-after-creating-regul.patch"
+
Signature · 1 line
2.29.2
#1 lore raw
Quoted 3 lines · Joshua Watt

Fixes the serial port output stopping mid way through the boot process by reverting the kernel commit that caused it.

If you want, I can also pick this up and merge it directly into linux-yocto and do a SRCREV bump.

Have you tested against 5.10 ? I'm working through issues with it now, and it would be nice to fix this before it pops up again on the next reference kernel bump.

Cheers,

Bruce

Quoted 79 lines
---
 ...-resolve-supply-after-creating-regul.patch | 53 +++++++++++++++++++
 recipes-kernel/linux/linux-yocto_5.8.bbappend |  4 ++
 2 files changed, 57 insertions(+)
 create mode 100644 recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch
 create mode 100644 recipes-kernel/linux/linux-yocto_5.8.bbappend
recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch new file +53 -0
new file mode 100644
index 0000000..3dd336b
@@ -0,0 +1,53 @@
+From a414d39b848002e15531f2538d2b6427ce51d07d Mon Sep 17 00:00:00 2001
+From: Joshua Watt <[email protected]>
+Date: Thu, 10 Dec 2020 15:59:47 -0600
+Subject: [PATCH] Revert "regulator: resolve supply after creating regulator"
+
+This commit prevents the serial console from working on the Rock Pi 4
+for some reason. It *appears* to possibly be fixed by some other commit
+upstream, but after a lot of head scratching and bisecting, I was unable
+to find which one, so just revert it for now and we can deal with it
+later.
+
+This reverts commit 96c6b5d5775637b3095ef934f871044811fd4db7.
+
+---
+ drivers/regulator/core.c | 21 ++++++++-------------
+ 1 file changed, 8 insertions(+), 13 deletions(-)
+
+diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
+index 25e601bf9383..be8c709a7488 100644
+--- a/drivers/regulator/core.c
++++ b/drivers/regulator/core.c
+@@ -5187,20 +5187,15 @@ regulator_register(const struct regulator_desc *regulator_desc,
+       else if (regulator_desc->supply_name)
+               rdev->supply_name = regulator_desc->supply_name;
+
++      /*
++       * Attempt to resolve the regulator supply, if specified,
++       * but don't return an error if we fail because we will try
++       * to resolve it again later as more regulators are added.
++       */
++      if (regulator_resolve_supply(rdev))
++              rdev_dbg(rdev, "unable to resolve supply\n");
++
+       ret = set_machine_constraints(rdev, constraints);
+-      if (ret == -EPROBE_DEFER) {
+-              /* Regulator might be in bypass mode and so needs its supply
+-               * to set the constraints */
+-              /* FIXME: this currently triggers a chicken-and-egg problem
+-               * when creating -SUPPLY symlink in sysfs to a regulator
+-               * that is just being created */
+-              ret = regulator_resolve_supply(rdev);
+-              if (!ret)
+-                      ret = set_machine_constraints(rdev, constraints);
+-              else
+-                      rdev_dbg(rdev, "unable to resolve supply early: %pe\n",
+-                               ERR_PTR(ret));
+-      }
+       if (ret < 0)
+               goto wash;
+
+--
+2.29.2
+
recipes-kernel/linux/linux-yocto_5.8.bbappend new file +4 -1
new file mode 100644
index 0000000..5a31842
@@ -0,0 +1,4 @@
+FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
+
+SRC_URI_append_rock-pi-4 = " file://0001-Revert-regulator-resolve-supply-after-creating-regul.patch"
+
--
Quoted 5 lines

2.29.2

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

#2 lore raw
Quoted 9 lines · Bruce Ashfield
Quoted 2 lines · Joshua Watt

Fixes the serial port output stopping mid way through the boot process by reverting the kernel commit that caused it.

If you want, I can also pick this up and merge it directly into linux-yocto and do a SRCREV bump.

Have you tested against 5.10 ? I'm working through issues with it now, and it would be nice to fix this before it pops up again on the next reference kernel bump.

When I use linux-yocto-dev as the kernel provider, it works just fine (in fact, I can't seem to find the commit that *fixes* the problem there....), so I think 5.10 will be OK? Is there something else I should try instead?

I'm not sure if we want to backport it to linux-yocto or not... I presume upstream backported it because it fixed something, so I'd be a little worried about a regression somewhere else.

Quoted 93 lines

Cheers,

Bruce

Quoted 78 lines
---
  ...-resolve-supply-after-creating-regul.patch | 53 +++++++++++++++++++
  recipes-kernel/linux/linux-yocto_5.8.bbappend |  4 ++
  2 files changed, 57 insertions(+)

create mode 100644 recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch create mode 100644 recipes-kernel/linux/linux-yocto_5.8.bbappend

recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch new file +53 -0
new file mode 100644
index 0000000..3dd336b
@@ -0,0 +1,53 @@
+From a414d39b848002e15531f2538d2b6427ce51d07d Mon Sep 17 00:00:00 2001
+From: Joshua Watt <[email protected]>
+Date: Thu, 10 Dec 2020 15:59:47 -0600
+Subject: [PATCH] Revert "regulator: resolve supply after creating regulator"
+
+This commit prevents the serial console from working on the Rock Pi 4
+for some reason. It *appears* to possibly be fixed by some other commit
+upstream, but after a lot of head scratching and bisecting, I was unable
+to find which one, so just revert it for now and we can deal with it
+later.
+
+This reverts commit 96c6b5d5775637b3095ef934f871044811fd4db7.
+
+---
+ drivers/regulator/core.c | 21 ++++++++-------------
+ 1 file changed, 8 insertions(+), 13 deletions(-)
+
+diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
+index 25e601bf9383..be8c709a7488 100644
+--- a/drivers/regulator/core.c
++++ b/drivers/regulator/core.c
+@@ -5187,20 +5187,15 @@ regulator_register(const struct regulator_desc *regulator_desc,
+       else if (regulator_desc->supply_name)
+               rdev->supply_name = regulator_desc->supply_name;
+
++      /*
++       * Attempt to resolve the regulator supply, if specified,
++       * but don't return an error if we fail because we will try
++       * to resolve it again later as more regulators are added.
++       */
++      if (regulator_resolve_supply(rdev))
++              rdev_dbg(rdev, "unable to resolve supply\n");
++
+       ret = set_machine_constraints(rdev, constraints);
+-      if (ret == -EPROBE_DEFER) {
+-              /* Regulator might be in bypass mode and so needs its supply
+-               * to set the constraints */
+-              /* FIXME: this currently triggers a chicken-and-egg problem
+-               * when creating -SUPPLY symlink in sysfs to a regulator
+-               * that is just being created */
+-              ret = regulator_resolve_supply(rdev);
+-              if (!ret)
+-                      ret = set_machine_constraints(rdev, constraints);
+-              else
+-                      rdev_dbg(rdev, "unable to resolve supply early: %pe\n",
+-                               ERR_PTR(ret));
+-      }
+       if (ret < 0)
+               goto wash;
+
+--
+2.29.2
+
recipes-kernel/linux/linux-yocto_5.8.bbappend new file +4 -1
new file mode 100644
index 0000000..5a31842
@@ -0,0 +1,4 @@
+FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
+
+SRC_URI_append_rock-pi-4 = " file://0001-Revert-regulator-resolve-supply-after-creating-regul.patch"
+
--
Quoted 5 lines

2.29.2

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

#3 lore raw
Quoted 9 lines · Joshua Watt
Quoted 6 lines · Bruce Ashfield
Quoted 2 lines · Joshua Watt

Fixes the serial port output stopping mid way through the boot process by reverting the kernel commit that caused it.

If you want, I can also pick this up and merge it directly into linux-yocto and do a SRCREV bump.

Quoted 9 lines
Quoted 3 lines

Have you tested against 5.10 ? I'm working through issues with it now, and it would be nice to fix this before it pops up again on the next reference kernel bump.

When I use linux-yocto-dev as the kernel provider, it works just fine (in fact, I can't seem to find the commit that *fixes* the problem there....), so I think 5.10 will be OK? Is there something else I should try instead?

Nope, that should be enough to show that it has already been addressed upstream. By default, you'll get a v5.10-rc with -dev in master, so that confirms the fix.

Quoted 5 lines

I'm not sure if we want to backport it to linux-yocto or not... I presume upstream backported it because it fixed something, so I'd be a little worried about a regression somewhere else.

True that. And the time spent to prove that, or identify the fix isn't worth it (since 5.8 is going away). If this was a longer term version, I'd create a BSP branch and we could isolate it that way, but again, just not worth the wrangling.

Bruce

Quoted 93 lines
Quoted 93 lines

Cheers,

Bruce

Quoted 78 lines
---
  ...-resolve-supply-after-creating-regul.patch | 53 +++++++++++++++++++
  recipes-kernel/linux/linux-yocto_5.8.bbappend |  4 ++
  2 files changed, 57 insertions(+)

create mode 100644 recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch create mode 100644 recipes-kernel/linux/linux-yocto_5.8.bbappend

recipes-kernel/linux/linux-yocto/0001-Revert-regulator-resolve-supply-after-creating-regul.patch new file +53 -0
new file mode 100644
index 0000000..3dd336b
@@ -0,0 +1,53 @@
+From a414d39b848002e15531f2538d2b6427ce51d07d Mon Sep 17 00:00:00 2001
+From: Joshua Watt <[email protected]>
+Date: Thu, 10 Dec 2020 15:59:47 -0600
+Subject: [PATCH] Revert "regulator: resolve supply after creating regulator"
+
+This commit prevents the serial console from working on the Rock Pi 4
+for some reason. It *appears* to possibly be fixed by some other commit
+upstream, but after a lot of head scratching and bisecting, I was unable
+to find which one, so just revert it for now and we can deal with it
+later.
+
+This reverts commit 96c6b5d5775637b3095ef934f871044811fd4db7.
+
+---
+ drivers/regulator/core.c | 21 ++++++++-------------
+ 1 file changed, 8 insertions(+), 13 deletions(-)
+
+diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
+index 25e601bf9383..be8c709a7488 100644
+--- a/drivers/regulator/core.c
++++ b/drivers/regulator/core.c
+@@ -5187,20 +5187,15 @@ regulator_register(const struct regulator_desc *regulator_desc,
+       else if (regulator_desc->supply_name)
+               rdev->supply_name = regulator_desc->supply_name;
+
++      /*
++       * Attempt to resolve the regulator supply, if specified,
++       * but don't return an error if we fail because we will try
++       * to resolve it again later as more regulators are added.
++       */
++      if (regulator_resolve_supply(rdev))
++              rdev_dbg(rdev, "unable to resolve supply\n");
++
+       ret = set_machine_constraints(rdev, constraints);
+-      if (ret == -EPROBE_DEFER) {
+-              /* Regulator might be in bypass mode and so needs its supply
+-               * to set the constraints */
+-              /* FIXME: this currently triggers a chicken-and-egg problem
+-               * when creating -SUPPLY symlink in sysfs to a regulator
+-               * that is just being created */
+-              ret = regulator_resolve_supply(rdev);
+-              if (!ret)
+-                      ret = set_machine_constraints(rdev, constraints);
+-              else
+-                      rdev_dbg(rdev, "unable to resolve supply early: %pe\n",
+-                               ERR_PTR(ret));
+-      }
+       if (ret < 0)
+               goto wash;
+
+--
+2.29.2
+
recipes-kernel/linux/linux-yocto_5.8.bbappend new file +4 -1
new file mode 100644
index 0000000..5a31842
@@ -0,0 +1,4 @@
+FILESEXTRAPATHS_prepend := "${THISDIR}/${PN}:"
+
+SRC_URI_append_rock-pi-4 = " file://0001-Revert-regulator-resolve-supply-after-creating-regul.patch"
+
--
Quoted 5 lines

2.29.2

--
- Thou shalt not follow the NULL pointer, for chaos and madness await thee at its end
- "Use the force Harry" - Gandalf, Star Trek II

Signature · 3 lines
- Thou shalt not follow the NULL pointer, for chaos and madness await
thee at its end
- "Use the force Harry" - Gandalf, Star Trek II
#4 lore raw
In reply to Joshua Watt
Quoted 2 lines · Joshua Watt

Fixes the serial port output stopping mid way through the boot process by reverting the kernel commit that caused it.

Interesting.

As you know (but for reference for other people) the baud rate of the rock pi 4's serial console is 1,500,000bps and not the usual 115,200bps that everything else uses.

I have *always* seen instability on the console. The best way to reproduce this instability is with the following test:

	root@rock-pi-4:~# cd /
	root@rock-pi-4:/# ls -lR

This never lasts more than a second or two before freezing up.

The ironic thing is, before applying this patch I thought I'd give the latest of everything a test (I don't test every change every day on my board).

With all the latest software but before applying your patch, I just went through 5 cycles of:
- reboot
- ls -lR

I didn't see a single freeze-up on the console! I've never seen the console survive the ls -lR thing, nevermind 5 in a row.

Weird.

I did, however, recently change the device I'm using to supply my USB-C power. I had noticed my last power supply was flaky, my local computer shop had some new fandangled ones on sale, so I upgraded.

In any case I'm happy to apply the patch, if it solves a problem for you then others are likely to bump into it as well. But interesting nonetheless.

Best regards,
Trevor