- 17 Sep, 2009 3 commits
-
-
Ben Dooks authored
-
Ben Dooks authored
A number of machines that need s3c_device_nand present do not select this in their Kconfig entries. Add the necessary selection of the configuration S3C_DEV_NAND so that we avoid the following error: arch/arm/mach-s3c2410/built-in.o: In function `bast_map_io': arch/arm/mach-s3c2410/mach-bast.c:634: undefined reference to `s3c_device_nand' arch/arm/mach-s3c2412/built-in.o: In function `s3c2412_init_uarts': arch/arm/mach-s3c2412/s3c2412.c:109: undefined reference to `s3c_device_nand' arch/arm/mach-s3c2412/built-in.o: In function `jive_machine_init': arch/arm/mach-s3c2412/mach-jive.c:673: undefined reference to `s3c_device_nand' Signed-off-by:Ben Dooks <ben-linux@fluff.org>
-
Mark Brown authored
Set up some IRQ space to allocation to off-SoC interrupt controllers. Default this to 16 IRQs. If individual boards require more than this then they will need to modify this file so allocating a small number helps reduce the number of modifications required. Signed-off-by:
Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by:
-
- 15 Sep, 2009 6 commits
-
-
Jassi authored
s3c2410_dma_enqueue makes call to kzalloc and dma_pool_alloc with GFP_KERNEL flag set, this can be an issue for drivers, like I2S, which call s3c2410_dma_enqueue from dma-bufferdone callback. Change the flag GFP_KERNEL to GFP_ATOMIC to avoid any problems. Signed-Off-by:
Jassi <jassi.brar@samsung.com> [ben-linux@fluff.org: Minor description edit and re-wrap] Signed-off-by:
-
Jassi authored
Provide actual minimum(struct pl080s_lli) size of block to dma_pool_create call, instead of hardcoded 32 bytes. Signed-Off-by:
-
Jassi authored
Correct the lli structure in struct s3c64xx_dma_buff which should have been 'struct pl080s_lli' (samsung specific) instead of the generic version 'struct pl080_lli' Signed-Off-by:
-
Thomas Abraham authored
In s3c64xx_roundrate_clksrc function, the calculation is wrong. This patch fixes this calculation. Signed-off-by:
Thomas Abraham <thomas.ab@samsung.com> Signed-off-by:
-
Thomas Abraham authored
This ensures the clock hierarchy data structures are updated when we change the clock source in the actual hardware registers. Signed-off-by:
-
Thomas Abraham authored
In s3c64xx_setrate_clksrc() we used sclk->shift, but actually need to use sclk->divider_shift to correctly calculate the value for the divider register. Signed-off-by:
-
- 16 Aug, 2009 10 commits
-
-
Ben Dooks authored
The show_fiq_list() call in arch/arm/kernel/irq.c currently depends on CONFIG_ARCH_ACORN, but this is not the only architecture that supports the usage of FIQ. Change to calling this if CONFIG_FIQ is set (which is what arch/arm/kernel/fiq.c is built by). Signed-off-by:
Ben Dooks <ben@simtec.co.uk> Signed-off-by:
-
Ben Dooks authored
Currently the S5PC100 does not define S3C_PA_NAND, leaving the NAND device definitions in arch/arm/plat-s3c/dev-nand.c unbuildable. Add a KConfig entry to select whether this is built. As backwards compatibility, both the S3C24XX and S3C64XX define the new configuration in their main Kconfig files until better support for basing this selection on a per-machine basis can be sorted out. Signed-off-by: -
Byungho Min authored
SMDKC100 board support. The board can be obtained from meritech (http://www.meritech.co.kr) Signed-off-by:
Byungho Min <bhmin@samsung.com> [ben-linux@fluff.org: fixup subject and description] Signed-off-by:
-
Byungho Min authored
S5PC100 is a new SoC with ARM coretex-A8 and numerous peripherals. This SoC is successor of S3C64XX. S5PC100 has peripherals which are still similar to S3C families so some drivers in "arch/arm/plat-s3c" can be shared. S5PC100 specific drivers will be added in "arch/arm/plat-s5pcxx" or "arch/arm/mach-s5pc100" Signed-off-by:
-
Byungho Min authored
S5PC100 has 4 PLLs (APLL,MPLL,EPLL,HPLL) and 3 clock domains. Clock scheme is implemented here. Signed-off-by:
-
Byungho Min authored
S5PC100 has 3 VICs(Vectored Interrupt Controller). The VICs come from S3C64xx series, so the driver source code can be shared with S3C families. The S5PC100 has 3 VICs while S3C64xx has only 2. Signed-off-by:
-
Byungho Min authored
S5PC100 has more GPIO group then previous one. It has 34 groups of GPIO, while S3C6410 has 17 groups. For now, only header files are written. Signed-off-by:
-
Byungho Min authored
Signed-off-by:
-
Byungho Min authored
Serial driver of S5PC100 is the same as S3C6400, so S5PC100 shares the serial driver with S3C6400. Uart driver is copied from plat-s3c64xx to plat-s5pc1xx, as I do not use plat-s3c64xx directory. Signed-off-by:
-
Byungho Min authored
S5PC100's the physical IO space starts at 0xe000"0000. To maximize space for vmalloc, the virtual IO space starts at 0xf400"0000 as same as other samsung CPUs(s3c24xx and s3c64xx) do. Signed-off-by:
-
- 14 Aug, 2009 10 commits
-
-
Ben Dooks authored
-
Ben Dooks authored
-
Ben Dooks authored
-
Ben Dooks authored
-
Ben Dooks authored
-
Ben Dooks authored
-
Ben Dooks authored
Add configuration callback for SPI bus 1 on GPD[8..10] and ensure the correct GPIO configuration register definitions in regs-gpio.h Signed-off-by:
-
Mark Brown authored
Move the definition of the "generic" IRQ in the process. Signed-off-by:
-
Ben Dooks authored
Add support for routing an IRQ from the normal ARM IRQ mechanism to the FIQ input of the processor. Note, also fix a bug where the init_FIQ() function has not been called when CONFIG_FIQ is enabled. Signed-off-by; Ben Dooks <ben@simtec.co.uk> Signed-off-by: -
Mark Brown authored
Allowing us to make the Kconfig a little bit saner. Signed-off-by:
-
- 13 Aug, 2009 4 commits
-
-
Mark Brown authored
Signed-off-by:
-
Mark Brown authored
Signed-off-by:
-
Matt Hsu authored
Add proper uartcfg for UART port 2,3 and tidy it up on SMDK6410. Signed-off-by:
Matt Hsu <matt@0xlab.org> Signed-off-by:
-
Marek Szyprowski authored
Make ncp_iodesc struct static to clean a public namespace a bit and move it to __initdata section to save memory a bit. Reviewed-by:
Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by:
Marek Szyprowski <m.szyprowski@samsung.com> Signed-off-by:
-
- 30 Jul, 2009 7 commits
-
-
Peter Korsgaard authored
The s3c64xx devices use the same hardware core as s3c24xx, so move the timer/pwm handling to plat-s3c so it can be used by both. Signed-off-by:
Peter Korsgaard <jacmet@sunsite.dk> Signed-off-by:
-
Peter Korsgaard authored
Move the s3c_device_nand platform device from plat-s3c24xx to plat-s3c, now that the nand driver also support the s3c64xx devices. Signed-off-by:
-
Peter Korsgaard authored
Add support for the Airgoo HMT (home media terminal) device. The HMT is a tablet device with a s3c6410, a 7" LCD and a number of peripheral connections. For more details of the hardware specs, see: http://article.gmane.org/gmane.linux.debian.devel.embedded/4307Signed-off-by:
-
Ben Dooks authored
Add debugfs support for the cpufreq driver to allow information about the system state to be exported to the user. Signed-off-by:
-
Ben Dooks authored
Add documentation for the S3C24XX style CPUFREQ driver. Signed-off-by:
-
Ben Dooks authored
Add armclk for use with the cpufreq support and anything else that may want it. This clock is just a direct descendant of fclk. Signed-off-by:
-
Ben Dooks authored
Add board support for CPUFREQ with the Simtec BAST board registering the necessary information with the core. Signed-off-by:
-
