-----Original Message----- From: york sun Sent: Thursday, October 20, 2016 10:46 AM To: Prabhakar Kushwaha prabhakar.kushwaha@nxp.com; u- boot@lists.denx.de Cc: Priyanka Jain priyanka.jain@nxp.com Subject: Re: [PATCH 1/5] armv8: lsch3: Use SVR based timer base address detection
On 10/19/2016 10:28 PM, Prabhakar Kushwaha wrote:
Hi York
-----Original Message----- From: Priyanka Jain [mailto:priyanka.jain@nxp.com] Sent: Wednesday, October 19, 2016 3:07 PM To: u-boot@lists.denx.de Cc: Priyanka Jain priyanka.jain@nxp.com; Prabhakar Kushwaha prabhakar.kushwaha@nxp.com Subject: [PATCH 1/5] armv8: lsch3: Use SVR based timer base address detection
Timer base address has been changed from LS2080A SoC to new SoCs like LS2088A, LS1088A.
Use SVR based timer base address detection to avoid compile time #ifdef.
Signed-off-by: Priyanka Jain priyanka.jain@nxp.com Signed-off-by: Prabhakar Kushwaha prabhakar.kushwaha@nxp.com
arch/arm/cpu/armv8/fsl-layerscape/cpu.c | 14 +++++++++++++- .../include/asm/arch-fsl-layerscape/immap_lsch3.h | 3 ++- 2 files changed, 15 insertions(+), 2 deletions(-)
diff --git a/arch/arm/cpu/armv8/fsl-layerscape/cpu.c b/arch/arm/cpu/armv8/fsl-layerscape/cpu.c index b7a2e0c..ce04e48 100644 --- a/arch/arm/cpu/armv8/fsl-layerscape/cpu.c +++ b/arch/arm/cpu/armv8/fsl-layerscape/cpu.c @@ -424,8 +424,10 @@ int arch_early_init_r(void)
int timer_init(void) {
- u32 __iomem *cntcr = (u32 *)CONFIG_SYS_FSL_TIMER_ADDR;
- u32 __iomem *cntcr;
#ifdef CONFIG_FSL_LSCH3
- struct ccsr_gur __iomem *gur = (void
*)(CONFIG_SYS_FSL_GUTS_ADDR);
- u32 svr, ver; u32 __iomem *cltbenr = (u32 *)CONFIG_SYS_FSL_PMU_CLTBENR;
#endif
#ifdef CONFIG_LS2080A @@ -439,6 +441,16 @@ int timer_init(void) #endif
#ifdef CONFIG_FSL_LSCH3
- svr = gur_in32(&gur->svr);
- ver = SVR_SOC_VER(svr);
- if ((ver == SVR_LS2080A) || (ver == SVR_LS2040A) ||
(ver == SVR_LS2085A) || (ver == SVR_LS2045A))cntcr = (u32 *)LS2080A_LS2085A_TIMER_ADDR;- else
+#endif
cntcr = (u32 *)CONFIG_SYS_FSL_TIMER_ADDR;+#ifdef CONFIG_FSL_LSCH3 /* Enable timebase for all clusters. * It is safe to do so even some clusters are not enabled. */ diff --git a/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch3.h b/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch3.h index 7acba27..e6cdfcb 100644 --- a/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch3.h +++ b/arch/arm/include/asm/arch-fsl-layerscape/immap_lsch3.h @@ -23,7 +23,8 @@ #define CONFIG_SYS_IFC_ADDR (CONFIG_SYS_IMMR
0x01240000) #define CONFIG_SYS_NS16550_COM1
(CONFIG_SYS_IMMR +
0x011C0500) #define CONFIG_SYS_NS16550_COM2
(CONFIG_SYS_IMMR +
0x011C0600) -#define CONFIG_SYS_FSL_TIMER_ADDR 0x023d0000 +#define LS2080A_LS2085A_TIMER_ADDR 0x023d0000
Is this ok to define new constant without CONFIG_? Only problem, it is not consistent with existing defines.
Prabhakar,
I opinion on this is
- Let's check if a Kconfig option is better for it If yes, let's convert it to
Kconfig. If not, go to 2)
- Let's use a different name space for the macros.
I suggest to use SYS_FSL_* for fixed value for our hardware.
York
LS2080A_LS2085A_TIMER_ADDR is defining the timer controller (register) offset. Kconfig option does not look to be very helpful here. I am thinking of renaming this to "SYS_FSL_LS2080A_LS2085A_TIMER_ADDR"
Please confirm if this is fine.
Priyanka