Skip to contentSkip to conversation

patch thinkpad-acpi-fix-incorrect-use-of-tpacpi_brght_mode_ecnvram.patch added to 2.6.30-stable tree

DISCUSSION Stefan Lippers-Hollmann 3 participants 6 messages Aug 6, 2009 – Aug 14, 2009 1 message not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw

Hi

Quoted 8 lines · On Friday 07 August 2009, [email protected] wrote:

This is a note to let you know that we have just queued up the patch titled

Subject: thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM

to the 2.6.30-stable tree. Its filename is

thinkpad-acpi-fix-incorrect-use-of-tpacpi_brght_mode_ecnvram.patch

Quoted 8 lines

A git repo of this tree can be found at
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

From 59fe4fe34d7afdf63208124f313be9056feaa2f4 Mon Sep 17 00:00:00 2001
From: Henrique de Moraes Holschuh <[email protected]>
Date: Sat, 1 Aug 2009 12:04:20 -0300
Subject: thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM

This patch fails to compile if applied to 2.6.30 + current stable queue.

drivers/platform/x86/thinkpad_acpi.c:5769: error: array type has incomplete element type drivers/platform/x86/thinkpad_acpi.c:5771: error: implicit declaration of function 'TPACPI_Q_IBM' drivers/platform/x86/thinkpad_acpi.c: In function 'brightness_init':
drivers/platform/x86/thinkpad_acpi.c:5799: error: implicit declaration of function 'tpacpi_check_quirks' drivers/platform/x86/thinkpad_acpi.c:5800: warning: type defaults to 'int' in declaration of 'type name' drivers/platform/x86/thinkpad_acpi.c:5800: warning: type defaults to 'int' in declaration of 'type name' drivers/platform/x86/thinkpad_acpi.c:5800: error: size of array 'type name' is negative make[6]: *** [drivers/platform/x86/thinkpad_acpi.o] Error 1
make[5]: *** [drivers/platform/x86] Error 2
make[4]: *** [drivers/platform] Error 2
make[4]: *** Waiting for unfinished jobs....

It seems to depend on the new quirk handling, introduced in:

commit 7d95a3d564901e88ed42810f054e579874151999

Author: Henrique de Moraes Holschuh <[email protected]>
Date:   Sat May 30 13:25:06 2009 -0300

thinkpad-acpi: add quirklist engine

Add a quirklist engine suitable for matching ThinkPad firmware, and change the code to use it.

Regards
Stefan Lippers-Hollmann

Quoted 125 lines

From: Henrique de Moraes Holschuh <[email protected]>

commit 59fe4fe34d7afdf63208124f313be9056feaa2f4 upstream.

HBRV-based default selection of backlight control strategy didn't work well, at least the X41 defines it but doesn't use it and I don't think it will stop there.

Switch to a white/blacklist. All models that have HBRV defined have been included in the list, and initially all ATI GPUs will get ECNVRAM, and the Intel GPUs will get UCMS_STEP.

Symptoms of incorrect backlight mode selection are:

1. Non-working backlight control through sysfs;

2. Backlight gets reset to the lowest level at every shutdown, reboot and when thinkpad-acpi gets unloaded;

This fixes a regression in 2.6.30, bugzilla #13826

---
 drivers/platform/x86/thinkpad_acpi.c |   61 ++++++++++++++++++++++++++---------
 1 file changed, 47 insertions(+), 14 deletions(-)
drivers/platform/x86/thinkpad_acpi.c +47 -14
@@ -5757,14 +5757,48 @@ static struct backlight_ops ibm_backligh
 
 /* --------------------------------------------------------------------- */
 
+/*
+ * These are only useful for models that have only one possibility
+ * of GPU.  If the BIOS model handles both ATI and Intel, don't use
+ * these quirks.
+ */
+#define TPACPI_BRGHT_Q_NOEC	0x0001	/* Must NOT use EC HBRV */
+#define TPACPI_BRGHT_Q_EC	0x0002  /* Should or must use EC HBRV */
+#define TPACPI_BRGHT_Q_ASK	0x8000	/* Ask for user report */
+
+static const struct tpacpi_quirk brightness_quirk_table[] __initconst = {
+	/* Models with ATI GPUs known to require ECNVRAM mode */
+	TPACPI_Q_IBM('1', 'Y', TPACPI_BRGHT_Q_EC),	/* T43/p ATI */
+
+	/* Models with ATI GPUs (waiting confirmation) */
+	TPACPI_Q_IBM('1', 'R', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+	TPACPI_Q_IBM('1', 'Q', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+	TPACPI_Q_IBM('7', '6', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+	TPACPI_Q_IBM('7', '8', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_EC),
+
+	/* Models with Intel Extreme Graphics 2 (waiting confirmation) */
+	TPACPI_Q_IBM('1', 'V', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_NOEC),
+	TPACPI_Q_IBM('1', 'W', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_NOEC),
+	TPACPI_Q_IBM('1', 'U', TPACPI_BRGHT_Q_ASK|TPACPI_BRGHT_Q_NOEC),
+
+	/* Models with Intel GMA900 */
+	TPACPI_Q_IBM('7', '0', TPACPI_BRGHT_Q_NOEC),	/* T43, R52 */
+	TPACPI_Q_IBM('7', '4', TPACPI_BRGHT_Q_NOEC),	/* X41 */
+	TPACPI_Q_IBM('7', '5', TPACPI_BRGHT_Q_NOEC),	/* X41 Tablet */
+};
+
 static int __init brightness_init(struct ibm_init_struct *iibm)
 {
 	int b;
+	unsigned long quirks;
 
 	vdbg_printk(TPACPI_DBG_INIT, "initializing brightness subdriver\n");
 
 	mutex_init(&brightness_mutex);
 
+	quirks = tpacpi_check_quirks(brightness_quirk_table,
+				ARRAY_SIZE(brightness_quirk_table));
+
 	/*
 	 * We always attempt to detect acpi support, so as to switch
 	 * Lenovo Vista BIOS to ACPI brightness mode even if we are not
@@ -5821,23 +5855,13 @@ static int __init brightness_init(struct
 	/* TPACPI_BRGHT_MODE_AUTO not implemented yet, just use default */
 	if (brightness_mode == TPACPI_BRGHT_MODE_AUTO ||
 	    brightness_mode == TPACPI_BRGHT_MODE_MAX) {
-		if (thinkpad_id.vendor == PCI_VENDOR_ID_IBM) {
-			/*
-			 * IBM models that define HBRV probably have
-			 * EC-based backlight level control
-			 */
-			if (acpi_evalf(ec_handle, NULL, "HBRV", "qd"))
-				/* T40-T43, R50-R52, R50e, R51e, X31-X41 */
-				brightness_mode = TPACPI_BRGHT_MODE_ECNVRAM;
-			else
-				/* all other IBM ThinkPads */
-				brightness_mode = TPACPI_BRGHT_MODE_UCMS_STEP;
-		} else
-			/* All Lenovo ThinkPads */
+		if (quirks & TPACPI_BRGHT_Q_EC)
+			brightness_mode = TPACPI_BRGHT_MODE_ECNVRAM;
+		else
 			brightness_mode = TPACPI_BRGHT_MODE_UCMS_STEP;
 
 		dbg_printk(TPACPI_DBG_BRGHT,
-			   "selected brightness_mode=%d\n",
+			   "driver auto-selected brightness_mode=%d\n",
 			   brightness_mode);
 	}
 
@@ -5858,6 +5882,15 @@ static int __init brightness_init(struct
 	vdbg_printk(TPACPI_DBG_INIT | TPACPI_DBG_BRGHT,
 			"brightness is supported\n");
 
+	if (quirks & TPACPI_BRGHT_Q_ASK) {
+		printk(TPACPI_NOTICE
+			"brightness: will use unverified default: "
+			"brightness_mode=%d\n", brightness_mode);
+		printk(TPACPI_NOTICE
+			"brightness: please report to %s whether it works well "
+			"or not on your ThinkPad\n", TPACPI_MAIL);
+	}
+
 	ibm_backlight_device->props.max_brightness =
 				(tp_features.bright_16levels)? 15 : 7;
 	ibm_backlight_device->props.brightness = b & TP_EC_BACKLIGHT_LVLMSK;
#2 lore raw
Quoted 1 line · Stefan Lippers-Hollmann

This patch fails to compile if applied to 2.6.30 + current stable queue.

[...]

Quoted 2 lines

It seems to depend on the new quirk handling, introduced in:
commit 7d95a3d564901e88ed42810f054e579874151999

Good catch. And I _will_ need that quirk handling stuff to keep 2.6.30 in shape :-(

The dependencies for the "thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM" patch are:

commit 050df107c408a3df048524b3783a5fc6d4dccfdb

Author: Henrique de Moraes Holschuh <[email protected]>
Date:   Sat May 30 13:25:05 2009 -0300

thinkpad-acpi: store fw version with strict checking

Extend the thinkpad model and firmware identification data with the release serial number for the BIOS and firmware (when available), as that is easier to parse and compare than the version strings.

We're going to greatly extend the use of the ThinkPad DMI data through quirk lists, so it is best to be quite strict and make sure what we get from DMI is exactly what we expect, otherwise quirk matching may result in quite insane things.

IBM (and Lenovo, at least for the ThinkPad line) uses this schema for firmware versioning and model:

Firmware model: Two digits, [0-9A-Z]

Firmware version: AABBCCDD, where
AA = firmware model, see above
BB = "ET" for BIOS, "HT" for EC
CC = release version, two digits, [0-9A-Z],
"00" < "09" < "0A" < "10" < "A0" < "ZZ"
DD = "WW"

And, as identified by Stefan,

commit 7d95a3d564901e88ed42810f054e579874151999

Author: Henrique de Moraes Holschuh <[email protected]>
Date:   Sat May 30 13:25:06 2009 -0300

thinkpad-acpi: add quirklist engine

Add a quirklist engine suitable for matching ThinkPad firmware, and change the code to use it.

I cherry picked the commits on the order below, on top of v2.6.30.4 and it compiled properly (and visual inspection of the result shows it is fine):

050df107c408a3df048524b3783a5fc6d4dccfdb
7d95a3d564901e88ed42810f054e579874151999
59fe4fe34d7afdf63208124f313be9056feaa2f4

Now, commit 7d95a3d564901e88ed42810f054e579874151999 could be reduced a bit for -stable if you guys prefer (only the first hunk is really necessary), but that would just make things more complicated later on if I find myself in need to patch anything in the fan control subdriver for 2.6.30.y (unlikely, but still...)

Signature · 4 lines
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh
#3 lore raw
Quoted 8 lines · Henrique de Moraes Holschuh
Quoted 1 line · Stefan Lippers-Hollmann

This patch fails to compile if applied to 2.6.30 + current stable queue.

[...]

Quoted 2 lines

It seems to depend on the new quirk handling, introduced in:
commit 7d95a3d564901e88ed42810f054e579874151999

Good catch. And I _will_ need that quirk handling stuff to keep 2.6.30 in shape :-(

Oops, sorry for missing this.

Quoted 2 lines

The dependencies for the "thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM" patch are:

Due to the need to add a new feature just to get these quirks to work, I'll just drop this single patch, which should keep everything building properly, right?

thanks,

greg k-h

#4 lore raw

Hi

Quoted 10 lines · Greg KH
Quoted 8 lines · Henrique de Moraes Holschuh
Quoted 1 line · Stefan Lippers-Hollmann

This patch fails to compile if applied to 2.6.30 + current stable queue.

[...]

Quoted 2 lines

It seems to depend on the new quirk handling, introduced in:
commit 7d95a3d564901e88ed42810f054e579874151999

Good catch. And I _will_ need that quirk handling stuff to keep 2.6.30 in shape :-(

Quoted 8 lines

Oops, sorry for missing this.

Quoted 2 lines

The dependencies for the "thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM" patch are:

Due to the need to add a new feature just to get these quirks to work, I'll just drop this single patch, which should keep everything building properly, right?

Yes, dropping this patch thinkpad-acpi-fix-incorrect-use-of-tpacpi_brght_mode_ecnvram.patch results in a buildable tree again.

Quoted 3 lines

thanks,

greg k-h

Regards
Stefan Lippers-Hollmann

#5 lore raw
Quoted 18 lines · Stefan Lippers-Hollmann

Hi

Quoted 15 lines · Greg KH
Quoted 8 lines · Henrique de Moraes Holschuh
Quoted 1 line · Stefan Lippers-Hollmann

This patch fails to compile if applied to 2.6.30 + current stable queue.

[...]

Quoted 2 lines

It seems to depend on the new quirk handling, introduced in:
commit 7d95a3d564901e88ed42810f054e579874151999

Good catch. And I _will_ need that quirk handling stuff to keep 2.6.30 in shape :-(

Oops, sorry for missing this.

Quoted 2 lines

The dependencies for the "thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM" patch are:

Quoted 7 lines
Quoted 3 lines

Due to the need to add a new feature just to get these quirks to work, I'll just drop this single patch, which should keep everything building properly, right?

Yes, dropping this patch thinkpad-acpi-fix-incorrect-use-of-tpacpi_brght_mode_ecnvram.patch results in a buildable tree again.

Thanks again so much for verifying this, I really appreciate it.

greg k-h

#6 lore raw
In reply to Greg KH
Quoted 10 lines · Greg KH
Quoted 8 lines · Henrique de Moraes Holschuh
Quoted 1 line · Stefan Lippers-Hollmann

This patch fails to compile if applied to 2.6.30 + current stable queue.

[...]

Quoted 2 lines

It seems to depend on the new quirk handling, introduced in:
commit 7d95a3d564901e88ed42810f054e579874151999

Good catch. And I _will_ need that quirk handling stuff to keep 2.6.30 in shape :-(

Quoted 8 lines

Oops, sorry for missing this.

Quoted 2 lines

The dependencies for the "thinkpad-acpi: fix incorrect use of TPACPI_BRGHT_MODE_ECNVRAM" patch are:

Due to the need to add a new feature just to get these quirks to work, I'll just drop this single patch, which should keep everything building properly, right?

Right. I will hack something and submit it to the next stable.

Signature · 4 lines
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh