- Oct 17, 2011
-
-
Mark Brown authored
If we don't know the SoC type then we won't add any chips which is rather unfortunate as neither GPIO nor pinmux APIs will work, breaking lots of different subsystems. Logging at least provides a hint to the user as to what's gone wrong. Signed-off-by:
Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by:
Kukjin Kim <kgene.kim@samsung.com>
-
Mark Brown authored
The IS_SAMSUNG_CPU() macro works by comparing the CPU ID mask exactly with the CPU ID. This was failing for S3C64xx SoCs as in order to support identification of the exact device the mask covers both variants of the chip, meaning that the test would always fail on S3C6410 devices. This in turn caused the core GPIO subsystem to fail to identify the CPU and not support any GPIOs, crippling the system. As a minimally invasive fix change the test for the class to be done by checking each implementation and oring them together. Signed-off-by:
-
- Oct 14, 2011
-
-
Axel Lin authored
Since commit af337f3e "ARM: S3C2443: Move parts of the clock code to common clock file", the init_clocks array is moved to arch/arm/plat-s3c24xx/s3c2443-clock.c. Now we call s3c_register_clocks for init_clocks in s3c2443_common_init_clocks. Thus we can remove the empty init_clocks array here and remove the redundant s3c_register_clocks call for init_clocks in s3c2443_init_clocks. Signed-off-by:
Axel Lin <axel.lin@gmail.com> Signed-off-by:
-
Heiko Stuebner authored
S3C2443 uses hsmmc1 as its only hsmmc device and for S3C2416/S3C2450 it's the second hsmmc channel with the same PCLKCON bit. The hsmmc-if clocks on both systems already got a devname, as did the hsmmc pclk for hsmmc0 on the S3C2416. So to make it possible to identify the hsmmc1 pclk on S3C2416 add the correct devname for it. The sclk name on S3C2443 also is s3c-sdhci.1. Signed-off-by:
Heiko Stuebner <heiko@sntech.de> Signed-off-by:
-
Heiko Stuebner authored
Previously the fclk rate was calculated by dividing the pll through the divider value of the armdiv. With a real armdiv clk in place it's possible to simply read its value, which does essentially the same. This change makes the whole fdiv_fn function pointers supplied to s3c2443_common_init_clocks and s3c2443_common_setup_clocks obsolete, so remove it too. Signed-off-by:
-
Heiko Stuebner authored
Cpufreq uses frequencies in kHz and not Hz, so set_rate and round_rate would be called with a frequency of 266666000 instead of 266666666 but the clock functions check for rates smaller or equal to the targetrate. As the armdiv does not support steps this small we can accommodate this by simply also setting the last 3 digits of the calculated rate to zero. Signed-off-by:
-
Heiko Stuebner authored
The armdiv array may contain unset divider values. Check the relevant value to prevent division by zero errors. Also check for set nr_armdiv and armdivmask before meddling with clkdiv0. Signed-off-by:
-
Heiko Stuebner authored
Signed-off-by:
-
Heiko St?bner authored
Signed-off-by:
-
Heiko St?bner authored
The system-layout of the armdiv and armclk is common to S3C2443/S3C2416/S3C2450 and only differs in the array of possible dividers. Therefore it is possible to reuse the clock definitions for all of these SoCs. Signed-off-by:
-
Heiko Stuebner authored
This is needed for making the armdiv clock common to S3C2443 and S3C2416/2450. Signed-off-by:
-
Heiko Stuebner authored
The S3C2416/2450 has only 3 bits for the armdiv setting instead of the 4 bits of the S3C2443. Signed-off-by:
-
Sylwester Nawrocki authored
Add voltage regulator and platform data definition for M-5MOLS sensor and MIPI-CSI receiver drivers. Add CAM power domain dependencies for FIMC device and set up camera port A GPIO. Configure I2C0 bus timings. Signed-off-by:
Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by:
HeungJun Kim <riverful.kim@samsung.com> Signed-off-by:
Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by:
-
Sachin Kamat authored
This patch enables multi-format codec (MFC) support on ORIGEN board. Signed-off-by:
Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by:
-
Heiko Stuebner authored
The ADC of the S3C2416/2450 SoC is 10 or 12 bit wide, has its source selection in the register base+0x18 and its width selection in bit 03 of the ADCCON register. Signed-off-by:
-
Heiko Stuebner authored
The S3C2443-adc is 10 bit wide and has its mux-select in an extra register at base+0x18 Signed-off-by:
-
- Oct 12, 2011
-
-
Heiko Stuebner authored
The adc blocks of S3C2443 and S3C2416 contain quirks not present in the stock S3C24xx adc. Therefore allow them to alter the device name via s3c_adc_setname. Signed-off-by:
-
Heiko Stuebner authored
The mux bits in the adccon register should be cleared only if muxing is really done in ADCCON and not another register. This patch introduces a conditional for this. Signed-off-by:
-
Heiko Stuebner authored
The adc blocks of the S3C2443 and S3C2416 define some additional registers and bits. Signed-off-by:
-
Tushar Behera authored
In drivers/gpio/gpio-samsung.c, there are certain structures and functions which are not getting used if the particular CPU is not selected. These code segments are moved under CPU specific macros to remove compilation warnings. Signed-off-by:
Tushar Behera <tushar.behera@linaro.org> Acked-by:
Grant Likely <grant.likely@secretlab.ca> Signed-off-by:
-
Thomas Abraham authored
Some of the gpio chips of exynos4 are assigned a default gpio config without the exynos4 specific pull up/down callbacks which resulted in incorrect setting of pull up/down configuration. Fix this by adding two new exynos4 specific entries in the array of default configs with set_pull and get_pull callbacks set to exynos4 specific callbacks The new default gpio configs can then be used for exynos4 gpio chips. Signed-off-by:
Thomas Abraham <thomas.abraham@linaro.org> Signed-off-by:
-
Mark Brown authored
The S3C64xx CPUs have TCMs so enable the kernel support for it on these systems. Signed-off-by:
Linus Walleij <linus.walleij@linaro.org> Signed-off-by:
-
Peter Korsgaard authored
Only register gpio banks provided by SoC instead of the maximum possible to lessen confusion, get rid of a warning from gpiolib and stop it from eating into the extra gpios for configs with S3C24XX_GPIO_EXTRA != 0. Signed-off-by:
Peter Korsgaard <jacmet@sunsite.dk> Signed-off-by:
-
Peter Korsgaard authored
S3C_GPIO_SPACE cannont be NON-zero, not zero. Signed-off-by:
-
Kukjin Kim authored
The 'nuri_cm_devices' is defined but not used. Cc: Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by: -
Sachin Kamat authored
This patch registers all the available power domains on ORIGEN board. Signed-off-by:
-
Sachin Kamat authored
Since the DC5V line is connected directly to the HDMI connector, "hdmi-en" regulator would become a dummy regulator on origen board (by defining REGULATOR_DUMMY in the kernel config file). Signed-off-by:
-
- Oct 04, 2011
-
-
Abhilash Kesavan authored
Add suspend-to-ram support for SMDK6440/50 Signed-off-by:
Abhilash Kesavan <a.kesavan@samsung.com> Signed-off-by:
-
Abhilash Kesavan authored
The sleep code for S5PV210 and EXYNOS4 are identical; moreover S5p64X0 and S5PC100 for which support will be added soon can use the same procedure. Create a common sleep code in the plat-s5p directory so that it can be re-used. Signed-off-by:
-
Abhilash Kesavan authored
Pave the way for adding PM support on S5P64X0, which is more similar to the S3C64XX series than the S5P series. Hence, the common pm code (containing dummy functions) should not be used for S5P64X0. Signed-off-by:
-
Kukjin Kim authored
-
Kukjin Kim authored
Conflicts: arch/arm/plat-s5p/include/plat/pll.h
-
Kukjin Kim authored
Conflicts: arch/arm/mach-exynos4/clock.c arch/arm/mach-s3c2412/gpio.c arch/arm/mach-s5p64x0/dma.c arch/arm/mach-s5p64x0/gpiolib.c
-
Kukjin Kim authored
-
Kukjin Kim authored
-
Kukjin Kim authored
This patch moves header files from plat-s5p to plat-samsung to remove plat-s5p directory to make one plat-samsung directory for Samsung SoCs. Signed-off-by: -
Kukjin Kim authored
This patch moves header files from plat-s3c24xx to plat-samsung to remove plat-s3c24xx directory to make one plat-samsung directory for Samsung SoCs. And this patch includes fixing coding style, too. Cc: Ben Dooks <ben-linux@fluff.org> Signed-off-by: -
Kukjin Kim authored
This patch moves SoC header files for supporting each SoCs to plat-samsung directory. This is required to make one plat- directory for Samsung SoCs. Cc: Ben Dooks <ben-linux@fluff.org> Signed-off-by: -
Kukjin Kim authored
Removed - arch/arm/plat-s3c24xx/include/plat/pll.h - arch/arm/mach-s3c64xx/include/mach/pll.h - arch/arm/plat-s5p/include/plat/pll.h - arch/arm/plat-samsung/include/plat/pll6553x.h And created - arch/arm/plat-samsung/include/plat/pll.h Cc: Ben Dooks <ben-linux@fluff.org> [kgene.kim@samsung.com: changed title] [kgene.kim@samsung.com: fixed conflicts in plat-s5p/include/pll.h] Signed-off-by: -
Kukjin Kim authored
Removed - arch/arm/plat-s3c24xx/include/mach/pwm-clock.h - arch/arm/mach-s3c64xx/include/mach/pwm-clock.h - arch/arm/mach-s5p64x0/include/mach/pwm-clock.h - arch/arm/mach-s5pc100/include/mach/pwm-clock.h - arch/arm/mach-s5pv210/include/mach/pwm-clock.h - arch/arm/mach-exynos4/include/mach/pwm-clock.h And created - arch/arm/plat-samsung/include/plat/pwm-clock.h Cc: Ben Dooks <ben-linux@fluff.org> [kgene.kim@samsung.com: changed title] Signed-off-by:
-
