Omap fixes for v2.6.32-rc6

Hi all,

Here are still few more fixes for review for the current -rc series.

Regards,

Tony

---

Grazvydas Ignotas (2):
omap: update Pandora defconfig
omap3: pandora: Fix keypad keymap

Janusz Krzysztofik (1):
omap1: Amstrad Delta defconfig fixes

Jarkko Nikula (1):
omap3: beagle: Fix USB host port power control

Mike Turquette (1):
omap3: Decrease cpufreq transition latency

Sergio Aguirre (1):
omap3: 3430sdp: Enable Linux Regulator framework

Tao Hu (1):
omap: Fix race condition in omap dma driver

Tobias Klauser (1):
omap: Use resource_size

 arch/arm/configs/ams_delta_defconfig     |   24 +++++++--
 arch/arm/configs/omap3_pandora_defconfig |   49 +++++++++++++++++--
 arch/arm/configs/omap_3430sdp_defconfig  |    9 +++
 arch/arm/mach-omap2/board-omap3beagle.c  |    2 -
 arch/arm/mach-omap2/board-omap3pandora.c |   78 +++++++++++++++---------------
 arch/arm/mach-omap2/gpmc.c               |    2 -
 arch/arm/mach-omap2/mailbox.c            |    2 -
 arch/arm/plat-omap/cpu-omap.c            |    2 -
 arch/arm/plat-omap/dma.c                 |    6 ++
 9 files changed, 118 insertions(+), 56 deletions(-)

Signature

From: Tao Hu <[email protected]>

The bug could cause irq enable bit of one DMA channel is cleared/set unexpectedly when 2 (or more) drivers are calling omap_request_dma()/omap_free_dma() simultaneously

---
 arch/arm/plat-omap/dma.c |    6 ++++++
 1 files changed, 6 insertions(+), 0 deletions(-)

index b53125f..02ed945 100644
@@ -691,13 +691,16 @@ static inline void disable_lnk(int lch)
 static inline void omap2_enable_irq_lch(int lch)
 {
 	u32 val;
+	unsigned long flags;
 
 	if (!cpu_class_is_omap2())
 		return;
 
+	spin_lock_irqsave(&dma_chan_lock, flags);
 	val = dma_read(IRQENABLE_L0);
 	val |= 1 << lch;
 	dma_write(val, IRQENABLE_L0);
+	spin_unlock_irqrestore(&dma_chan_lock, flags);
 }
 
 int omap_request_dma(int dev_id, const char *dev_name,

@@ -799,10 +802,13 @@ void omap_free_dma(int lch)
 
 	if (cpu_class_is_omap2()) {
 		u32 val;
+
+		spin_lock_irqsave(&dma_chan_lock, flags);
 		/* Disable interrupts */
 		val = dma_read(IRQENABLE_L0);
 		val &= ~(1 << lch);
 		dma_write(val, IRQENABLE_L0);
+		spin_unlock_irqrestore(&dma_chan_lock, flags);
 
 		/* Clear the CSR register and IRQ status register */
 		dma_write(OMAP2_DMA_CSR_CLEAR_MASK, CSR(lch));

From: Tobias Klauser <[email protected]>

Use the resource_size function instead of manually calculating the resource size. This reduces the chance of introducing off-by-one errors and actually fixes one in mailbox.c.

---
 arch/arm/mach-omap2/gpmc.c    |    2 +-
 arch/arm/mach-omap2/mailbox.c |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

index 1587682..f865756 100644
@@ -366,7 +366,7 @@ int gpmc_cs_request(int cs, unsigned long size, unsigned long *base)
 	if (r < 0)
 		goto out;
 
-	gpmc_cs_enable_mem(cs, res->start, res->end - res->start + 1);
+	gpmc_cs_enable_mem(cs, res->start, resource_size(res));
 	*base = res->start;
 	gpmc_cs_set_reserved(cs, 1);
 out:

index c035ad3..ef57b38 100644
@@ -300,7 +300,7 @@ static int __devinit omap2_mbox_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "invalid mem resource\n");
 		return -ENODEV;
 	}
-	mbox_base = ioremap(res->start, res->end - res->start);
+	mbox_base = ioremap(res->start, resource_size(res));
 	if (!mbox_base)
 		return -ENOMEM;
 

Here

One more fix that I forgot about.

Tony

From: Janusz Krzysztofik <[email protected]>

The patch provides the following fixes:

- keep kernel small enough to boot with standard tools,
- ensure compatibility with both new and legacy distros,
- turn on support for recently added or fixed hardware features.

Created and tested against linux-2.6.32-rc5.

---
 arch/arm/configs/ams_delta_defconfig |   24 +++++++++++++++++-------
 1 files changed, 17 insertions(+), 7 deletions(-)

index 7647325..e8f7380 100644
@@ -55,10 +55,10 @@ CONFIG_BLK_DEV_INITRD=y
 CONFIG_INITRAMFS_SOURCE=""
 CONFIG_CC_OPTIMIZE_FOR_SIZE=y
 CONFIG_SYSCTL=y
-# CONFIG_EMBEDDED is not set
+CONFIG_EMBEDDED=y
 CONFIG_UID16=y
 CONFIG_SYSCTL_SYSCALL=y
-CONFIG_KALLSYMS=y
+# CONFIG_KALLSYMS is not set
 # CONFIG_KALLSYMS_ALL is not set
 # CONFIG_KALLSYMS_EXTRA_PASS is not set
 CONFIG_HOTPLUG=y

@@ -224,7 +224,7 @@ CONFIG_CPU_CP15_MMU=y
 #
 # Processor Features
 #
-# CONFIG_ARM_THUMB is not set
+CONFIG_ARM_THUMB=y
 # CONFIG_CPU_ICACHE_DISABLE is not set
 # CONFIG_CPU_DCACHE_DISABLE is not set
 CONFIG_CPU_DCACHE_WRITETHROUGH=y

@@ -248,7 +248,7 @@ CONFIG_CPU_DCACHE_WRITETHROUGH=y
 # CONFIG_HIGH_RES_TIMERS is not set
 CONFIG_PREEMPT=y
 CONFIG_HZ=100
-# CONFIG_AEABI is not set
+CONFIG_AEABI=y
 # CONFIG_ARCH_DISCONTIGMEM_ENABLE is not set
 CONFIG_SELECT_MEMORY_MODEL=y
 CONFIG_FLATMEM_MANUAL=y

@@ -299,7 +299,9 @@ CONFIG_BINFMT_ELF=y
 #
 # Power management options
 #
-# CONFIG_PM is not set
+CONFIG_PM=y
+# CONFIG_SUSPEND is not set
+CONFIG_PM_RUNTIME=y
 
 #
 # Networking

@@ -670,7 +672,7 @@ CONFIG_INPUT_MOUSEDEV_SCREEN_X=1024
 CONFIG_INPUT_MOUSEDEV_SCREEN_Y=768
 # CONFIG_INPUT_JOYDEV is not set
 # CONFIG_INPUT_TSDEV is not set
-# CONFIG_INPUT_EVDEV is not set
+CONFIG_INPUT_EVDEV=y
 # CONFIG_INPUT_EVBUG is not set
 
 #

@@ -784,6 +786,7 @@ CONFIG_I2C_OMAP=y
 #
 # CONFIG_SPI is not set
 # CONFIG_SPI_MASTER is not set
+CONFIG_GPIO_SYSFS=y
 
 #
 # Dallas's 1-wire bus

@@ -820,6 +823,7 @@ CONFIG_LEDS_AMS_DELTA=y
 CONFIG_LEDS_TRIGGERS=y
 CONFIG_LEDS_TRIGGER_TIMER=y
 CONFIG_LEDS_TRIGGER_HEARTBEAT=y
+CONFIG_LEDS_TRIGGER_DEFAULT_ON=y
 
 #
 # Multimedia devices

@@ -896,7 +900,13 @@ CONFIG_LOGO_LINUX_CLUT224=y
 #
 # Sound
 #
-# CONFIG_SOUND is not set
+CONFIG_SOUND=y
+CONFIG_SND=y
+CONFIG_SND_MIXER_OSS=y
+CONFIG_SND_PCM_OSS=y
+CONFIG_SND_SOC=y
+CONFIG_SND_OMAP_SOC=y
+CONFIG_SND_OMAP_SOC_AMS_DELTA=y
 
 #
 # HID Devices

Wednesday 11 November 2009 04:00:57 Tony Lindgren napisa?(a):

Tony,

Please do not apply this patch yet, or revert it if already applied. It would exhibit the LCD DMA not working bug when CONFIG_PM=y, as I tried to state while submitting v2 of my fix that corrected the issue
(http://www.spinics.net/lists/linux-omap/msg19963.html).

Sorry for not stating that again with this submission (or not waiting with the submission after the bug is fixed first). That happened after I misinterpreted your comment "Sounds good to me."
(http://www.spinics.net/lists/linux-omap/msg19965.html) as your acceptance for the patch v2 as posted, with my self proposed changes, that resulted in v3 submission directly (http://www.spinics.net/lists/linux-omap/msg20059.html),
that has been objected by you next.

Thanks,
Janusz

OK, sorry looks like I forgot that issue..

I suggest we keep this patch, and also add your v3 patch before this one.

I'll post your v3 patch to this thread with an added comment about moving the OMAP_LCDC_ defines during the next merge cycle.

Regards,

Tony

From: Grazvydas Ignotas <[email protected]>

The original TWL4030 keypad driver from linux-omap used KEY() macro defined as (col, row), but while it was merged upstream it was changed to use matrix keypad infrastructure, which uses (row, col) format. Update the keymap in board file to match layout of mainline driver.

---
 arch/arm/mach-omap2/board-omap3pandora.c |   78 +++++++++++++++---------------
 1 files changed, 39 insertions(+), 39 deletions(-)

index 5326e0d..7519edb 100644
@@ -134,50 +134,50 @@ static void __init pandora_keys_gpio_init(void)
 }
 
 static int board_keymap[] = {
-	/* col, row, code */
+	/* row, col, code */
 	KEY(0, 0, KEY_9),
-	KEY(0, 1, KEY_0),
-	KEY(0, 2, KEY_BACKSPACE),
-	KEY(0, 3, KEY_O),
-	KEY(0, 4, KEY_P),
-	KEY(0, 5, KEY_K),
-	KEY(0, 6, KEY_L),
-	KEY(0, 7, KEY_ENTER),
-	KEY(1, 0, KEY_8),
+	KEY(0, 1, KEY_8),
+	KEY(0, 2, KEY_I),
+	KEY(0, 3, KEY_J),
+	KEY(0, 4, KEY_N),
+	KEY(0, 5, KEY_M),
+	KEY(1, 0, KEY_0),
 	KEY(1, 1, KEY_7),
-	KEY(1, 2, KEY_6),
-	KEY(1, 3, KEY_5),
-	KEY(1, 4, KEY_4),
-	KEY(1, 5, KEY_3),
-	KEY(1, 6, KEY_2),
-	KEY(1, 7, KEY_1),
-	KEY(2, 0, KEY_I),
-	KEY(2, 1, KEY_U),
+	KEY(1, 2, KEY_U),
+	KEY(1, 3, KEY_H),
+	KEY(1, 4, KEY_B),
+	KEY(1, 5, KEY_SPACE),
+	KEY(2, 0, KEY_BACKSPACE),
+	KEY(2, 1, KEY_6),
 	KEY(2, 2, KEY_Y),
-	KEY(2, 3, KEY_T),
-	KEY(2, 4, KEY_R),
-	KEY(2, 5, KEY_E),
-	KEY(2, 6, KEY_W),
-	KEY(2, 7, KEY_Q),
-	KEY(3, 0, KEY_J),
-	KEY(3, 1, KEY_H),
-	KEY(3, 2, KEY_G),
+	KEY(2, 3, KEY_G),
+	KEY(2, 4, KEY_V),
+	KEY(2, 5, KEY_FN),
+	KEY(3, 0, KEY_O),
+	KEY(3, 1, KEY_5),
+	KEY(3, 2, KEY_T),
 	KEY(3, 3, KEY_F),
-	KEY(3, 4, KEY_D),
-	KEY(3, 5, KEY_S),
-	KEY(3, 6, KEY_A),
-	KEY(3, 7, KEY_LEFTSHIFT),
-	KEY(4, 0, KEY_N),
-	KEY(4, 1, KEY_B),
-	KEY(4, 2, KEY_V),
-	KEY(4, 3, KEY_C),
+	KEY(3, 4, KEY_C),
+	KEY(4, 0, KEY_P),
+	KEY(4, 1, KEY_4),
+	KEY(4, 2, KEY_R),
+	KEY(4, 3, KEY_D),
 	KEY(4, 4, KEY_X),
-	KEY(4, 5, KEY_Z),
-	KEY(4, 6, KEY_DOT),
-	KEY(4, 7, KEY_COMMA),
-	KEY(5, 0, KEY_M),
-	KEY(5, 1, KEY_SPACE),
-	KEY(5, 2, KEY_FN),
+	KEY(5, 0, KEY_K),
+	KEY(5, 1, KEY_3),
+	KEY(5, 2, KEY_E),
+	KEY(5, 3, KEY_S),
+	KEY(5, 4, KEY_Z),
+	KEY(6, 0, KEY_L),
+	KEY(6, 1, KEY_2),
+	KEY(6, 2, KEY_W),
+	KEY(6, 3, KEY_A),
+	KEY(6, 4, KEY_DOT),
+	KEY(7, 0, KEY_ENTER),
+	KEY(7, 1, KEY_1),
+	KEY(7, 2, KEY_Q),
+	KEY(7, 3, KEY_LEFTSHIFT),
+	KEY(7, 4, KEY_COMMA),
 };
 
 static struct matrix_keymap_data board_map_data = {

From: Jarkko Nikula <[email protected]>

The host port power is enabled by driving the nEN_USB_PWR low as stated in the comment. This fix is originally from Steve Sakoman <[email protected]>.

---
 arch/arm/mach-omap2/board-omap3beagle.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

index 70df6b4..08b0816 100644
@@ -162,7 +162,7 @@ static int beagle_twl_gpio_setup(struct device *dev,
 
 	/* TWL4030_GPIO_MAX + 0 == ledA, EHCI nEN_USB_PWR (out, active low) */
 	gpio_request(gpio + TWL4030_GPIO_MAX, "nEN_USB_PWR");
-	gpio_direction_output(gpio + TWL4030_GPIO_MAX, 1);
+	gpio_direction_output(gpio + TWL4030_GPIO_MAX, 0);
 
 	/* TWL4030_GPIO_MAX + 1 == ledB, PMU_STAT (out, active low LED) */
 	gpio_leds[2].gpio = gpio + TWL4030_GPIO_MAX + 1;

From: Sergio Aguirre <[email protected]>

Some drivers have dependencies on this, and therefore should be enabled.

---
 arch/arm/configs/omap_3430sdp_defconfig |    9 ++++++++-
 1 files changed, 8 insertions(+), 1 deletions(-)

index 8a4a7e2..5a305f0 100644
@@ -1703,7 +1703,14 @@ CONFIG_RTC_DRV_TWL4030=y
 # on-CPU RTC drivers
 #
 # CONFIG_DMADEVICES is not set
-# CONFIG_REGULATOR is not set
+CONFIG_REGULATOR=y
+# CONFIG_REGULATOR_DEBUG is not set
+# CONFIG_REGULATOR_FIXED_VOLTAGE is not set
+# CONFIG_REGULATOR_VIRTUAL_CONSUMER is not set
+# CONFIG_REGULATOR_USERSPACE_CONSUMER is not set
+# CONFIG_REGULATOR_BQ24022 is not set
+# CONFIG_REGULATOR_MAX1586 is not set
+CONFIG_REGULATOR_TWL4030=y
 # CONFIG_UIO is not set
 # CONFIG_STAGING is not set
 

From: Grazvydas Ignotas <[email protected]>

This patch updates defconfig to enable options needed to properly boot OMAP3 pandora board. It also enables MMC, OTG, GPIO LEDs, TWL4030 GPIO and sound drivers.

---
 arch/arm/configs/omap3_pandora_defconfig |   49 +++++++++++++++++++++++++++---
 1 files changed, 44 insertions(+), 5 deletions(-)

index b54ad2e..150deaf 100644
@@ -611,7 +611,7 @@ CONFIG_INPUT_KEYBOARD=y
 # CONFIG_KEYBOARD_XTKBD is not set
 # CONFIG_KEYBOARD_NEWTON is not set
 # CONFIG_KEYBOARD_STOWAWAY is not set
-# CONFIG_KEYBOARD_GPIO is not set
+CONFIG_KEYBOARD_GPIO=y
 CONFIG_INPUT_MOUSE=y
 # CONFIG_MOUSE_PS2 is not set
 # CONFIG_MOUSE_SERIAL is not set

@@ -634,7 +634,8 @@ CONFIG_TOUCHSCREEN_ADS7846=y
 # CONFIG_TOUCHSCREEN_TOUCHWIN is not set
 # CONFIG_TOUCHSCREEN_USB_COMPOSITE is not set
 # CONFIG_TOUCHSCREEN_TOUCHIT213 is not set
-# CONFIG_INPUT_MISC is not set
+CONFIG_INPUT_MISC=y
+CONFIG_INPUT_TWL4030_PWRBUTTON=y
 
 #
 # Hardware I/O ports

@@ -834,7 +835,29 @@ CONFIG_DAB=y
 #
 # CONFIG_VGA_CONSOLE is not set
 CONFIG_DUMMY_CONSOLE=y
-# CONFIG_SOUND is not set
+CONFIG_SOUND=y
+CONFIG_SOUND_OSS_CORE=y
+CONFIG_SOUND_OSS_CORE_PRECLAIM=y
+CONFIG_SND=y
+CONFIG_SND_TIMER=y
+CONFIG_SND_PCM=y
+CONFIG_SND_JACK=y
+CONFIG_SND_OSSEMUL=y
+CONFIG_SND_MIXER_OSS=y
+CONFIG_SND_PCM_OSS=y
+CONFIG_SND_PCM_OSS_PLUGINS=y
+CONFIG_SND_SUPPORT_OLD_API=y
+CONFIG_SND_VERBOSE_PROCFS=y
+CONFIG_SND_VERBOSE_PRINTK=y
+CONFIG_SND_DRIVERS=y
+CONFIG_SND_USB=y
+CONFIG_SND_SOC=y
+CONFIG_SND_OMAP_SOC=y
+CONFIG_SND_OMAP_SOC_MCBSP=y
+CONFIG_SND_OMAP_SOC_OMAP3_PANDORA=y
+CONFIG_SND_SOC_I2C_AND_SPI=y
+CONFIG_SND_SOC_TWL4030=y
+
 CONFIG_HID_SUPPORT=y
 CONFIG_HID=y
 # CONFIG_HID_DEBUG is not set

@@ -1020,7 +1043,13 @@ CONFIG_MMC_BLOCK_BOUNCE=y
 # CONFIG_MMC_SPI is not set
 # CONFIG_MEMSTICK is not set
 # CONFIG_ACCESSIBILITY is not set
-# CONFIG_NEW_LEDS is not set
+CONFIG_NEW_LEDS=y
+CONFIG_LEDS_CLASS=y
+CONFIG_LEDS_GPIO=y
+CONFIG_LEDS_GPIO_PLATFORM=y
+CONFIG_LEDS_TRIGGERS=y
+CONFIG_LEDS_TRIGGER_DEFAULT_ON=y
+
 CONFIG_RTC_LIB=y
 CONFIG_RTC_CLASS=y
 CONFIG_RTC_HCTOSYS=y

@@ -1084,9 +1113,12 @@ CONFIG_RTC_DRV_TWL4030=y
 # on-CPU RTC drivers
 #
 # CONFIG_DMADEVICES is not set
-# CONFIG_REGULATOR is not set
 # CONFIG_UIO is not set
 
+CONFIG_REGULATOR=y
+CONFIG_REGULATOR_FIXED_VOLTAGE=y
+CONFIG_REGULATOR_TWL4030=y
+
 #
 # File systems
 #

@@ -1407,3 +1439,10 @@ CONFIG_PLIST=y
 CONFIG_HAS_IOMEM=y
 CONFIG_HAS_IOPORT=y
 CONFIG_HAS_DMA=y
+
+# added by hand for now
+CONFIG_KEYBOARD_TWL4030=y
+CONFIG_USB_OTG_UTILS=y
+CONFIG_TWL4030_USB=y
+CONFIG_MMC_OMAP_HS=y
+

From: Mike Turquette <[email protected]>

Adjust OMAP3 frequency transition latency from 10,000,000uS to a more reasonable 300,000uS. This causes ondemand and conservative governors to sample CPU load more often resulting in more responsive behavior.

Tested on Android 2.6.29; using this value and conservative governor, CORE power consumption on Zoom2 was comparable to the old and unresponsive 10,000,000uS value while UI responsiveness was greatly improved.

---
 arch/arm/plat-omap/cpu-omap.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

index 1868c0d..341235c 100644
@@ -127,7 +127,7 @@ static int __init omap_cpu_init(struct cpufreq_policy *policy)
 	}
 
 	/* FIXME: what's the actual transition time? */
-	policy->cpuinfo.transition_latency = 10 * 1000 * 1000;
+	policy->cpuinfo.transition_latency = 300 * 1000;
 
 	return 0;
 }

I still haven't been able to find the reason why the Ralink drivers haven't been offered in make config/menuconfig/xconfig - cross compiling on Fedora 12 x86_64 as per my previous post. I downloaded the latest linux-omap snapshot Tuesday, 6 hours after it was up. Regards
Sid.

Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
Specialist, Cricket Coach
Microsoft Windows Free Zone - Linux used for all Computing Tasks

Sounds like that should be patched via drivers/net.

Regards,

Tony

CONFIG_CFG80211=m and CONFIG_MAC80211=m fixed that problem, though I'm sure I tried the same some days ago and it didn't work - your snapshot of today built and is working.

One remaining problem is getting FB working, the LCD status light lights up blue, but the display is blank.
For that I shall start another thread.

Regards
Sid.

Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
Specialist, Cricket Coach
Microsoft Windows Free Zone - Linux used for all Computing Tasks