[U-Boot] [PATCH 2/4] x86: ivybridge: Use reset_cpu()

[Simon Glass]

Hi Bin,

On 26 April 2015 at 22:58, Bin Meng <bmeng.cn at gmail.com> wrote:
> Hi Simon,
>
> On Sat, Apr 25, 2015 at 11:04 PM, Simon Glass <sjg at chromium.org> wrote:
>> Now that reset_cpu() functions correctly, use it instead of directly
>> accessing the port.
>>
>> Signed-off-by: Simon Glass <sjg at chromium.org>
>> ---
>>
>>  arch/x86/cpu/ivybridge/cpu.c      | 5 ++---
>>  arch/x86/cpu/ivybridge/early_me.c | 7 +++----
>>  arch/x86/cpu/ivybridge/sdram.c    | 3 +--
>>  3 files changed, 6 insertions(+), 9 deletions(-)
>>
>> diff --git a/arch/x86/cpu/ivybridge/cpu.c b/arch/x86/cpu/ivybridge/cpu.c
>> index 37f3731..a62b30e 100644
>> --- a/arch/x86/cpu/ivybridge/cpu.c
>> +++ b/arch/x86/cpu/ivybridge/cpu.c
>> @@ -92,7 +92,7 @@ static int set_flex_ratio_to_tdp_nominal(void)
>>
>>         /* Issue warm reset, will be "CPU only" due to soft reset data */
>>         outb(0x0, PORT_RESET);
>> -       outb(0x6, PORT_RESET);
>> +       outb(SYS_RST | RST_CPU, PORT_RESET);
>>         cpu_hlt();
>>
>>         /* Not reached */
>> @@ -286,8 +286,7 @@ int print_cpuinfo(void)
>>
>>                 /* System is not happy after keyboard reset... */
>>                 debug("Issuing CF9 warm reset\n");
>> -               outb(0x6, 0xcf9);
>> -               cpu_hlt();
>> +               reset_cpu();
>>         }
>>
>>         /* Early chipset init required before RAM init can work */
>> diff --git a/arch/x86/cpu/ivybridge/early_me.c b/arch/x86/cpu/ivybridge/early_me.c
>> index 356bbb4..0347197 100644
>> --- a/arch/x86/cpu/ivybridge/early_me.c
>> +++ b/arch/x86/cpu/ivybridge/early_me.c
>> @@ -183,9 +183,8 @@ int intel_early_me_init_done(u8 status)
>>         }
>>
>>         /* Perform the requested reset */
>> -       if (reset) {
>> -               outb(reset, 0xcf9);
>> -               cpu_hlt();
>> -       }
>> +       if (reset)
>> +               reset_cpu();
>> +
>
> I think the following codes before this code block need to be updated
> to use the reset bit enum to assign to the reset variable.

Ah yes, I didn't look very hard. I'll fix this.

>
>         reset = 0;
>         switch (hfs.ack_data) {
>         case ME_HFS_ACK_CONTINUE:
>                 /* Continue to boot */
>                 return 0;
>         case ME_HFS_ACK_RESET:
>                 /* Non-power cycle reset */
>                 set_global_reset(0);
>                 reset = 0x06;
>                 break;
>         case ME_HFS_ACK_PWR_CYCLE:
>                 /* Power cycle reset */
>                 set_global_reset(0);
>                 reset = 0x0e;
>                 break;
>         case ME_HFS_ACK_GBL_RESET:
>                 /* Global reset */
>                 set_global_reset(1);
>                 reset = 0x0e;
>                 break;
>
>>         return -1;
>>  }
>> diff --git a/arch/x86/cpu/ivybridge/sdram.c b/arch/x86/cpu/ivybridge/sdram.c
>> index 9a6da37..42e1e60 100644
>> --- a/arch/x86/cpu/ivybridge/sdram.c
>> +++ b/arch/x86/cpu/ivybridge/sdram.c
>> @@ -393,8 +393,7 @@ int sdram_initialise(struct pei_data *pei_data)
>>         /* If MRC data is not found we cannot continue S3 resume. */
>>         if (pei_data->boot_mode == PEI_BOOT_RESUME && !pei_data->mrc_input) {
>>                 debug("Giving up in sdram_initialize: No MRC data\n");
>> -               outb(0x6, PORT_RESET);
>> -               cpu_hlt();
>> +               reset_cpu();
>>         }
>>
>>         /* Pass console handler in pei_data */
>> --

Regards,
Simon