Hello,
On 04/21/2014 04:25 PM, Akshay Saraswat wrote:
Used a script to recheck/verify pin mapping and existing mapping appears to be fine, returning correct bank and pin values. Did ./MAKEALL -a arm and found all Exynos/S5P related boards compiled successfully. Couldn't test booting over all of them. Tested U-Boot bootup over SMDK5420, SMDK5250, Snow. Requesting maintaners to please test over affected SoCs/boards.
Changes in V2:
- Enabled CMD_GPIO as suggested by Simon Glass and supported same for EXYNOS5.
Changes in V3:
- New patch added to rename S5P GPIO definitions to S5P_GPIO.
- GPIO Table added to calculate the base address of input gpio bank.
Changes in V4:
- To have consistent 0..n-1 GPIO numbering the banks are divided into different parts where ever they have holes in them.
- Function and table to support gpio command moved to s5p-gpio driver.
- Rebased on latest u-boot-samsung tree.
Changes in V5:
- Rebased on latest u-boot-samsung tree.
- Removed Exynos5 specific code in gpio driver api to get bank.
- Added #define HAVE_GENERIC_GPIO in config file to remove conditinal CPU check in gpio driver.
Changes in V6:
- Isolated config changes in a new patch.
- Updated patches with corresponding changes for Exynos 5420.
Changes in V7:
- Added changes for other SoCs like Exynos 4412, 4210 etc.
Changes in V8:
- Changed Subject of patch 2/2 to reflect affected SoCs/boards.
- Fixed arndale board compile time errors introduced due to patch-set v7.
Changes in V9:
- Fixed checkpatch errors.
- Fixed naming error in exynosxxxx_gpio_data arrays which could be the possible reason behind data abort witnessed over Exynos4 boards.
Akshay Saraswat (2): EXYNOS: Add GPIO pin numbering and rename definitions S5P: Exynos: Config: Enable Generic GPIO and CMD configs
arch/arm/cpu/armv7/exynos/pinmux.c | 403 +++---- arch/arm/include/asm/arch-exynos/cpu.h | 17 +- arch/arm/include/asm/arch-exynos/gpio.h | 1786 +++++++++++++++++++++++++----- arch/arm/include/asm/arch-s5pc1xx/gpio.h | 941 +++++++++++++--- board/samsung/arndale/arndale.c | 11 +- board/samsung/goni/goni.c | 26 +- board/samsung/smdk5250/exynos5-dt.c | 20 +- board/samsung/smdk5250/smdk5250.c | 19 +- board/samsung/smdk5420/smdk5420.c | 15 +- board/samsung/smdkc100/smdkc100.c | 5 +- board/samsung/smdkv310/smdkv310.c | 17 +- board/samsung/trats/trats.c | 39 +- board/samsung/trats2/trats2.c | 74 +- board/samsung/universal_c210/universal.c | 51 +- drivers/gpio/s5p_gpio.c | 195 +++- include/configs/arndale.h | 1 + include/configs/exynos5-dt.h | 3 + include/configs/origen.h | 1 + include/configs/s5p_goni.h | 5 +- include/configs/s5pc210_universal.h | 17 +- include/configs/smdkc100.h | 1 + include/configs/smdkv310.h | 2 + include/configs/trats.h | 9 +- include/configs/trats2.h | 5 +- 24 files changed, 2806 insertions(+), 857 deletions(-)
I think that you missed my last few comments. You are still using exynos_gpio_get() - what for? It returns some number and next it's interpreted as gpio continuous number - but it isn't continuous. It's because each samsung board has now defined "HAVE_GENERIC_GPIO", so gpio numbers are interpreted as linear. We don't need additional macro since you introduce linear numbering.
Please read my last comments again.
I tried test this on trats2 device but data abort occurs at PMIC init, it's probably because exynos_gpio_get(). Please use just gpio numbering as you defined in this patchset.
Thank you