Skip to contentSkip to conversation

Fix LPSS clock divider for XPS 9530

PATCH Applied Aleksandrs Vinarskis 3 participants 8 messages Dec 21, 2023 – Jan 26, 2024 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

v4 cover Fix LPSS clock divider for XPS 9530

Dell XPS 9530 (2023) uses spi-pxa2xx with clock-divider enabled from intel-lpss with the ratio of 1:32767 (Dell's firmware bug). This caps SPI controller's speed at very low value of 3051Hz, which makes the interface practically unusable. Since either spi-pxa2xx or intel-lpss should have clock divider enabled, not both, and SPI controller can have higher speed than requested by the device, it is preffered to disable intel-lpss clock-divider, and let SPI controller handle it.

Fixing this issue directly in Dell firmware by setting lpss divider to 1:1 would've been the ideal solution, but is unlikely to ever happen.

Particular driver already implements customized solution for handling buggy ACPI tables for select Microsoft devices. This patch series converts it to a more generic quirk table, and adds a new quirk QUIRK_CLOCK_DIVIDER_UNITY which forces clock divider to be set to 1:1. In the future, devices with similar bug (if any) can be easily added to the same pci id lookup table.

Changes since v3:
- Altered "{ PCI.." style of quirk table
- Added CC to Hans de Goede

Changes since v2:
- Added missing periods, moved variable declaration
- Altered "}, {" style of quirk table
- Confirmed to compile for 32-bit without warnings

Changes since v1:
- Applied suggestions by Andy Shevchenko

Aleksandrs Vinarskis (2):
mfd: intel-lpss: Switch to generalized quirk table
mfd: intel-lpss: Introduce QUIRK_CLOCK_DIVIDER_UNITY for XPS 9530

 drivers/mfd/intel-lpss-pci.c | 28 ++++++++++++++++++++--------
 drivers/mfd/intel-lpss.c     |  9 ++++++++-
 drivers/mfd/intel-lpss.h     | 14 +++++++++++++-
 3 files changed, 41 insertions(+), 10 deletions(-)
Signature · 1 line
2.40.1
#1 lore raw

PATCH v4 1/2 mfd: intel-lpss: Switch to generalized quirk table

Introduce generic quirk table, and port existing walkaround for select Microsoft devices to it. This is a preparation for QUIRK_CLOCK_DIVIDER_UNITY.

---
 drivers/mfd/intel-lpss-pci.c | 23 +++++++++++++++--------
 drivers/mfd/intel-lpss.c     |  2 +-
 drivers/mfd/intel-lpss.h     |  9 ++++++++-
 3 files changed, 24 insertions(+), 10 deletions(-)
drivers/mfd/intel-lpss-pci.c +15 -8
index 4621d3950b8f..07713a2f694f 100644
@@ -23,12 +23,17 @@
 
 #include "intel-lpss.h"
 
-/* Some DSDTs have an unused GEXP ACPI device conflicting with I2C4 resources */
-static const struct pci_device_id ignore_resource_conflicts_ids[] = {
-	/* Microsoft Surface Go (version 1) I2C4 */
-	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_INTEL, 0x9d64, 0x152d, 0x1182), },
-	/* Microsoft Surface Go 2 I2C4 */
-	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_INTEL, 0x9d64, 0x152d, 0x1237), },
+static const struct pci_device_id quirk_ids[] = {
+	{
+		/* Microsoft Surface Go (version 1) I2C4 */
+		PCI_DEVICE_SUB(PCI_VENDOR_ID_INTEL, 0x9d64, 0x152d, 0x1182),
+		.driver_data = QUIRK_IGNORE_RESOURCE_CONFLICTS,
+	},
+	{
+		/* Microsoft Surface Go 2 I2C4 */
+		PCI_DEVICE_SUB(PCI_VENDOR_ID_INTEL, 0x9d64, 0x152d, 0x1237),
+		.driver_data = QUIRK_IGNORE_RESOURCE_CONFLICTS,
+	},
 	{ }
 };
 
@@ -36,6 +41,7 @@ static int intel_lpss_pci_probe(struct pci_dev *pdev,
 				const struct pci_device_id *id)
 {
 	const struct intel_lpss_platform_info *data = (void *)id->driver_data;
+	const struct pci_device_id *quirk_pci_info;
 	struct intel_lpss_platform_info *info;
 	int ret;
 
@@ -55,8 +61,9 @@ static int intel_lpss_pci_probe(struct pci_dev *pdev,
 	info->mem = pci_resource_n(pdev, 0);
 	info->irq = pci_irq_vector(pdev, 0);
 
-	if (pci_match_id(ignore_resource_conflicts_ids, pdev))
-		info->ignore_resource_conflicts = true;
+	quirk_pci_info = pci_match_id(quirk_ids, pdev);
+	if (quirk_pci_info)
+		info->quirks = quirk_pci_info->driver_data;
 
 	pdev->d3cold_delay = 0;
 
drivers/mfd/intel-lpss.c +1 -1
index eff423f7dd28..aafa0da5f8db 100644
@@ -412,7 +412,7 @@ int intel_lpss_probe(struct device *dev,
 		return ret;
 
 	lpss->cell->swnode = info->swnode;
-	lpss->cell->ignore_resource_conflicts = info->ignore_resource_conflicts;
+	lpss->cell->ignore_resource_conflicts = info->quirks & QUIRK_IGNORE_RESOURCE_CONFLICTS;
 
 	intel_lpss_init_dev(lpss);
 
drivers/mfd/intel-lpss.h +8 -1
index c1d72b117ed5..2fa9ef916258 100644
@@ -11,16 +11,23 @@
 #ifndef __MFD_INTEL_LPSS_H
 #define __MFD_INTEL_LPSS_H
 
+#include <linux/bits.h>
 #include <linux/pm.h>
 
+/*
+ * Some DSDTs have an unused GEXP ACPI device conflicting with I2C4 resources.
+ * Set to ignore resource conflicts with ACPI declared SystemMemory regions.
+ */
+#define QUIRK_IGNORE_RESOURCE_CONFLICTS BIT(0)
+
 struct device;
 struct resource;
 struct software_node;
 
 struct intel_lpss_platform_info {
 	struct resource *mem;
-	bool ignore_resource_conflicts;
 	int irq;
+	unsigned int quirks;
 	unsigned long clk_rate;
 	const char *clk_con_id;
 	const struct software_node *swnode;
Signature · 1 line
2.40.1
maintainer
#2 lore raw

Re: [PATCH v4 1/2] mfd: intel-lpss: Switch to generalized quirk table

Quoted 10 lines · Aleksandrs Vinarskis

Introduce generic quirk table, and port existing walkaround for select Microsoft devices to it. This is a preparation for QUIRK_CLOCK_DIVIDER_UNITY.

---
 drivers/mfd/intel-lpss-pci.c | 23 +++++++++++++++--------
 drivers/mfd/intel-lpss.c     |  2 +-
 drivers/mfd/intel-lpss.h     |  9 ++++++++-
 3 files changed, 24 insertions(+), 10 deletions(-)

An RB from Andy would make for an easy apply.

Signature · 1 line
Lee Jones [李琼斯]
maintainer
#3 lore raw

Re: [PATCH v4 1/2] mfd: intel-lpss: Switch to generalized quirk table

Quoted 6 lines · Lee Jones
Quoted 4 lines · Aleksandrs Vinarskis

Introduce generic quirk table, and port existing walkaround for select Microsoft devices to it. This is a preparation for QUIRK_CLOCK_DIVIDER_UNITY.

Quoted 8 lines
Quoted 6 lines
---
 drivers/mfd/intel-lpss-pci.c | 23 +++++++++++++++--------
 drivers/mfd/intel-lpss.c     |  2 +-
 drivers/mfd/intel-lpss.h     |  9 ++++++++-
 3 files changed, 24 insertions(+), 10 deletions(-)

An RB from Andy would make for an easy apply.

Nevermind, I found it! Old eyes!!

Signature · 1 line
Lee Jones [李琼斯]
#4 lore raw

PATCH v4 2/2 mfd: intel-lpss: Introduce QUIRK_CLOCK_DIVIDER_UNITY for XPS 9530

In reply to Aleksandrs Vinarskis

Some devices (eg. Dell XPS 9530, 2023) due to a firmware bug have a misconfigured clock divider, which should've been 1:1. This introduces quirk which conditionally re-configures the clock divider to 1:1.

---
 drivers/mfd/intel-lpss-pci.c | 5 +++++
 drivers/mfd/intel-lpss.c     | 7 +++++++
 drivers/mfd/intel-lpss.h     | 5 +++++
 3 files changed, 17 insertions(+)
drivers/mfd/intel-lpss-pci.c +5 -0
index 07713a2f694f..8c00e0c695c5 100644
@@ -34,6 +34,11 @@ static const struct pci_device_id quirk_ids[] = {
 		PCI_DEVICE_SUB(PCI_VENDOR_ID_INTEL, 0x9d64, 0x152d, 0x1237),
 		.driver_data = QUIRK_IGNORE_RESOURCE_CONFLICTS,
 	},
+	{
+		/* Dell XPS 9530 (2023) */
+		PCI_DEVICE_SUB(PCI_VENDOR_ID_INTEL, 0x51fb, 0x1028, 0x0beb),
+		.driver_data = QUIRK_CLOCK_DIVIDER_UNITY,
+	},
 	{ }
 };
 
drivers/mfd/intel-lpss.c +7 -0
index aafa0da5f8db..2a9018112dfc 100644
@@ -300,6 +300,7 @@ static int intel_lpss_register_clock_divider(struct intel_lpss *lpss,
 {
 	char name[32];
 	struct clk *tmp = *clk;
+	int ret;
 
 	snprintf(name, sizeof(name), "%s-enable", devname);
 	tmp = clk_register_gate(NULL, name, __clk_get_name(tmp), 0,
@@ -316,6 +317,12 @@ static int intel_lpss_register_clock_divider(struct intel_lpss *lpss,
 		return PTR_ERR(tmp);
 	*clk = tmp;
 
+	if (lpss->info->quirks & QUIRK_CLOCK_DIVIDER_UNITY) {
+		ret = clk_set_rate(tmp, lpss->info->clk_rate);
+		if (ret)
+			return ret;
+	}
+
 	snprintf(name, sizeof(name), "%s-update", devname);
 	tmp = clk_register_gate(NULL, name, __clk_get_name(tmp),
 				CLK_SET_RATE_PARENT, lpss->priv, 31, 0, NULL);
drivers/mfd/intel-lpss.h +5 -0
index 2fa9ef916258..6f8f668f4c6f 100644
@@ -19,6 +19,11 @@
  * Set to ignore resource conflicts with ACPI declared SystemMemory regions.
  */
 #define QUIRK_IGNORE_RESOURCE_CONFLICTS BIT(0)
+/*
+ * Some devices have misconfigured clock divider due to a firmware bug.
+ * Set this to force the clock divider to 1:1 ratio.
+ */
+#define QUIRK_CLOCK_DIVIDER_UNITY		BIT(1)
 
 struct device;
 struct resource;
Signature · 1 line
2.40.1
reviewer Reviewed-by on the v4 series
#5 lore raw
In reply to Aleksandrs Vinarskis
Quoted 8 lines · Aleksandrs Vinarskis

Dell XPS 9530 (2023) uses spi-pxa2xx with clock-divider enabled from intel-lpss with the ratio of 1:32767 (Dell's firmware bug). This caps SPI controller's speed at very low value of 3051Hz, which makes the interface practically unusable. Since either spi-pxa2xx or intel-lpss should have clock divider enabled, not both, and SPI controller can have higher speed than requested by the device, it is preffered to disable intel-lpss clock-divider, and let SPI controller handle it.

Quoted 8 lines

Fixing this issue directly in Dell firmware by setting lpss divider to 1:1 would've been the ideal solution, but is unlikely to ever happen.

Particular driver already implements customized solution for handling buggy ACPI tables for select Microsoft devices. This patch series converts it to a more generic quirk table, and adds a new quirk QUIRK_CLOCK_DIVIDER_UNITY which forces clock divider to be set to 1:1. In the future, devices with similar bug (if any) can be easily added to the same pci id lookup table.

LGTM now,

Signature · 2 lines
With Best Regards,
Andy Shevchenko
maintainer Applied
#6 lore raw
In reply to Aleksandrs Vinarskis
Quoted 9 lines · Aleksandrs Vinarskis

Dell XPS 9530 (2023) uses spi-pxa2xx with clock-divider enabled from intel-lpss with the ratio of 1:32767 (Dell's firmware bug). This caps SPI controller's speed at very low value of 3051Hz, which makes the interface practically unusable. Since either spi-pxa2xx or intel-lpss should have clock divider enabled, not both, and SPI controller can have higher speed than requested by the device, it is preffered to disable intel-lpss clock-divider, and let SPI controller handle it.

[...]

Applied, thanks!

[1/2] mfd: intel-lpss: Switch to generalized quirk table

      commit: d43b5230c38ed6001f996eb9262b457713b31ef8
[2/2] mfd: intel-lpss: Introduce QUIRK_CLOCK_DIVIDER_UNITY for XPS 9530
      commit: 106362167f49a0e83f4e6c26f9653f3061575229

--
Lee Jones [李琼斯]

A message that is not in the archive (sent off-list or not yet received): CAMcHhXrZvw3QGuoHk+fBaetpPfpTqWa3hrfkENtFuggOZ7UHSQ@mail.gmail.com

maintainer
#8 lore raw

Please don't drop the list. Other people may benefit from your questions.

I'm redacting some of the text which may cause conflict and adding back the original recipient's list.

Quoted 19 lines · alex vinarskis
Quoted 9 lines · Aleksandrs Vinarskis
Quoted 8 lines

Dell XPS 9530 (2023) uses spi-pxa2xx with clock-divider enabled from intel-lpss with the ratio of 1:32767 (Dell's firmware bug). This caps SPI controller's speed at very low value of 3051Hz, which makes the interface practically unusable. Since either spi-pxa2xx or intel-lpss should have clock divider enabled, not both, and SPI controller can have higher speed than requested by the device, it is preffered to disable intel-lpss clock-divider, and let SPI controller handle it.

Quoted 3 lines
Quoted 1 line

[...]

Applied, thanks!

Hi,

I've got a quick question regarding the merging procedure, apologies if this was trivial,
I am still rather new to this:

Quoted 5 lines

I noticed that this series was staged for for-mfd-next-next, while only for-mfd-next
was merged to 6.8-rc1. Does that imply that following series is not planned to be included
in 6.8, but rather in 6.9?

Yes, that's correct. The for-mfd-next-next branch comes into play when for-mfd-next gets locked down for inclusion and gets merged into for-mfd-next once the merge window has closed.

Quoted 2 lines

If yes, could I ask why? It was under the impression that it was submitted in time (at the late stage of 6.7 testing) to make it to 6.8

It was not. Most maintainers stop merging new patches in the mid to late -rcs for inclusion into the next merge-window. I like to have all new additions soak in Linux Next (-next) for at least 2 to 3 weeks before sending to Linus.

Quoted 1 line

(sadly, not in time for LTS > 6.7).

I think you're confusing 2 different things. I assume you mean the -rcs. LTS (Long Term Stable) is for fixes. Commits end up there *after* they land in Mainline so long as they conform to the rules set out in:

Documentation/process/stable-kernel-rules.rst

Quoted 8 lines

Reason I am asking is
because this is the last series required to fix audio on XPS 9530
laptops: [1] was merged in
6.7, [2] was merged in 6.8-rc1 and backported to 6.7.1-rc1. Thus me and other XPS users
are extremely eager to see this land in the next kernel release, as there was quite some
community effort to get this done.

Adding support for new devices isn't usually a good enough reason to submit to the -rcs. The -rcs are used to fix issues that broke during the merge-window of the same release and for urgent bug fixes.

This set will be applied to v6.8-rc1. From there you can apply to have it backported (or better still, backport and submit it yourself) to Stable (a.k.a. LTS). Since it's the LTS' that you will usually find running on your distro of choice, your audio will likely be fixed with an upcoming apt upgrade or equivalent.

Quoted 16 lines

Once again, apologies if this question is obvious, Thanks,
Alex

[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/[email protected]/

Quoted 9 lines

[1/2] mfd: intel-lpss: Switch to generalized quirk table

      commit: d43b5230c38ed6001f996eb9262b457713b31ef8
[2/2] mfd: intel-lpss: Introduce QUIRK_CLOCK_DIVIDER_UNITY for XPS 9530
      commit: 106362167f49a0e83f4e6c26f9653f3061575229

--
Lee Jones [李琼斯]

Signature · 1 line
Lee Jones [李琼斯]