linux-mips
[Top] [All Lists]

Re: [PATCH V15 02/12] MIPS: Loongson: Add basic Loongson-3 CPU support

To: Huacai Chen <chenhc@lemote.com>
Subject: Re: [PATCH V15 02/12] MIPS: Loongson: Add basic Loongson-3 CPU support
From: Aurelien Jarno <aurelien@aurel32.net>
Date: Mon, 30 Dec 2013 22:33:54 +0100
Cc: Ralf Baechle <ralf@linux-mips.org>, John Crispin <john@phrozen.org>, "Steven J. Hill" <Steven.Hill@imgtec.com>, linux-mips@linux-mips.org, Fuxin Zhang <zhangfx@lemote.com>, Zhangjin Wu <wuzhangjin@gmail.com>, Hongliang Tao <taohl@lemote.com>, Hua Yan <yanh@lemote.com>
In-reply-to: <1387109676-540-3-git-send-email-chenhc@lemote.com>
List-archive: <http://www.linux-mips.org/archives/linux-mips/>
List-help: <mailto:ecartis@linux-mips.org?Subject=help>
List-id: linux-mips <linux-mips.eddie.linux-mips.org>
List-owner: <mailto:ralf@linux-mips.org>
List-post: <mailto:linux-mips@linux-mips.org>
List-software: Ecartis version 1.0.0
List-subscribe: <mailto:ecartis@linux-mips.org?subject=subscribe%20linux-mips>
List-unsubscribe: <mailto:ecartis@linux-mips.org?subject=unsubscribe%20linux-mips>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <1387109676-540-1-git-send-email-chenhc@lemote.com> <1387109676-540-3-git-send-email-chenhc@lemote.com>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mutt/1.5.21 (2010-09-15)
On Sun, Dec 15, 2013 at 08:14:26PM +0800, Huacai Chen wrote:
> Basic Loongson-3 CPU support include CPU probing and TLB/cache
> initializing.
> 
> Signed-off-by: Huacai Chen <chenhc@lemote.com>
> Signed-off-by: Hongliang Tao <taohl@lemote.com>
> Signed-off-by: Hua Yan <yanh@lemote.com>
> ---
>  arch/mips/include/asm/cpu-type.h |    4 +++
>  arch/mips/kernel/cpu-probe.c     |   14 +++++++---
>  arch/mips/mm/c-r4k.c             |   56 
> ++++++++++++++++++++++++++++++++++++++
>  arch/mips/mm/tlb-r4k.c           |    3 +-
>  arch/mips/mm/tlbex.c             |    1 +
>  5 files changed, 73 insertions(+), 5 deletions(-)
> 
> diff --git a/arch/mips/include/asm/cpu-type.h 
> b/arch/mips/include/asm/cpu-type.h
> index 4a402cc..a591e63 100644
> --- a/arch/mips/include/asm/cpu-type.h
> +++ b/arch/mips/include/asm/cpu-type.h
> @@ -20,6 +20,10 @@ static inline int __pure __get_cpu_type(const int cpu_type)
>       case CPU_LOONGSON2:
>  #endif
>  
> +#ifdef CONFIG_SYS_HAS_CPU_LOONGSON3
> +     case CPU_LOONGSON3:
> +#endif
> +
>  #ifdef CONFIG_SYS_HAS_CPU_LOONGSON1B
>       case CPU_LOONGSON1:
>  #endif
> diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c
> index c814287..4bc81b2 100644
> --- a/arch/mips/kernel/cpu-probe.c
> +++ b/arch/mips/kernel/cpu-probe.c
> @@ -639,17 +639,23 @@ static inline void cpu_probe_legacy(struct cpuinfo_mips 
> *c, unsigned int cpu)
>                            MIPS_CPU_LLSC;
>               c->tlbsize = 64;
>               break;
> -     case PRID_IMP_LOONGSON2:
> -             c->cputype = CPU_LOONGSON2;
> -             __cpu_name[cpu] = "ICT Loongson-2";
> -
> +     case PRID_IMP_LOONGSON2: /* Loongson-2/3 have the same PRID_IMP field */

In short it means there is one PRID_IMP for 32-bit Loongson CPUs, and
one for 64-bit ones. Maybe the name should be changed?

>               switch (c->processor_id & PRID_REV_MASK) {
>               case PRID_REV_LOONGSON2E:
> +                     c->cputype = CPU_LOONGSON2;
> +                     __cpu_name[cpu] = "ICT Loongson-2E";
>                       set_elf_platform(cpu, "loongson2e");
>                       break;
>               case PRID_REV_LOONGSON2F:
> +                     c->cputype = CPU_LOONGSON2;
> +                     __cpu_name[cpu] = "ICT Loongson-2F";
>                       set_elf_platform(cpu, "loongson2f");
>                       break;
> +             case PRID_REV_LOONGSON3A:
> +                     c->cputype = CPU_LOONGSON3;
> +                     __cpu_name[cpu] = "ICT Loongson-3A";
> +                     set_elf_platform(cpu, "loongson3a");
> +                     break;
>               }

I have mixed feelings about the Loongson-2 name change. On one side it's
clearly better to have 2E and 2F instead of 2 V0.2 and 2 V0.3, and it
should have been like that since the beginning. That said changing that
now is kind of breaking the userland. I know that it would break debian
installer support for example, though that should not be a real problem
as we ship the installer with a given kernel version. I don't know if
there are other usages that can cause a problem. Any opinion from
others?

Also it's totally out of scope of this patch, but having a different
value of cputype for Loongson 2E and 2F might be a good idea to remove
a few ifdef.

>               set_isa(c, MIPS_CPU_ISA_III);
> diff --git a/arch/mips/mm/c-r4k.c b/arch/mips/mm/c-r4k.c
> index 73f02da..69dd882 100644
> --- a/arch/mips/mm/c-r4k.c
> +++ b/arch/mips/mm/c-r4k.c
> @@ -348,6 +348,7 @@ static inline void local_r4k___flush_cache_all(void * 
> args)
>  {
>       switch (current_cpu_type()) {
>       case CPU_LOONGSON2:
> +     case CPU_LOONGSON3:
>       case CPU_R4000SC:
>       case CPU_R4000MC:
>       case CPU_R4400SC:
> @@ -1003,6 +1004,31 @@ static void probe_pcache(void)
>               c->dcache.waybit = 0;
>               break;
>  
> +     case CPU_LOONGSON3:
> +             config1 = read_c0_config1();
> +             if ((lsize = ((config1 >> 19) & 7)))
> +                     c->icache.linesz = 2 << lsize;
> +             else
> +                     c->icache.linesz = lsize;
                
Here you can use directly:
                c->icache.linesz = 0;

> +             c->icache.sets = 64 << ((config1 >> 22) & 7);
> +             c->icache.ways = 1 + ((config1 >> 16) & 7);
> +             icache_size = c->icache.sets *
> +                                       c->icache.ways *
> +                                       c->icache.linesz;
> +             c->icache.waybit = 0;
> +
> +             if ((lsize = ((config1 >> 10) & 7)))
> +                     c->dcache.linesz = 2 << lsize;
> +             else
> +                     c->dcache.linesz = lsize;

Same as above, you can set the value as 0 directly.

> +             c->dcache.sets = 64 << ((config1 >> 13) & 7);
> +             c->dcache.ways = 1 + ((config1 >> 7) & 7);
> +             dcache_size = c->dcache.sets *
> +                                       c->dcache.ways *
> +                                       c->dcache.linesz;
> +             c->dcache.waybit = 0;
> +             break;
> +
>       default:
>               if (!(config & MIPS_CONF_M))
>                       panic("Don't know how to probe P-caches on this cpu.");
> @@ -1222,6 +1248,32 @@ static void __init loongson2_sc_init(void)
>       c->options |= MIPS_CPU_INCLUSIVE_CACHES;
>  }
>  
> +static void __init loongson3_sc_init(void)
> +{
> +     struct cpuinfo_mips *c = &current_cpu_data;
> +     unsigned int config2, lsize;
> +
> +     config2 = read_c0_config2();
> +     if ((lsize = ((config2 >> 4) & 15)))
> +             c->scache.linesz = 2 << lsize;
> +     else
> +             c->scache.linesz = lsize;

same as above, set to 0;

> +     c->scache.sets = 64 << ((config2 >> 8) & 15);
> +     c->scache.ways = 1 + (config2 & 15);
> +
> +     scache_size = c->scache.sets *
> +                               c->scache.ways *
> +                               c->scache.linesz;
> +     /* Loongson-3 has 4 cores, 1MB scache for each. scaches are shared */
> +     scache_size *= 4;
> +     c->scache.waybit = 0;
> +     pr_info("Unified secondary cache %ldkB %s, linesize %d bytes.\n",
> +            scache_size >> 10, way_string[c->scache.ways], c->scache.linesz);
> +     if (scache_size)
> +             c->options |= MIPS_CPU_INCLUSIVE_CACHES;
> +     return;
> +}
> +
>  extern int r5k_sc_init(void);
>  extern int rm7k_sc_init(void);
>  extern int mips_sc_init(void);
> @@ -1274,6 +1326,10 @@ static void setup_scache(void)
>               loongson2_sc_init();
>               return;
>  
> +     case CPU_LOONGSON3:
> +             loongson3_sc_init();
> +             return;
> +
>       case CPU_XLP:
>               /* don't need to worry about L2, fully coherent */
>               return;
> diff --git a/arch/mips/mm/tlb-r4k.c b/arch/mips/mm/tlb-r4k.c
> index da3b0b9..10ab29e 100644
> --- a/arch/mips/mm/tlb-r4k.c
> +++ b/arch/mips/mm/tlb-r4k.c
> @@ -53,13 +53,14 @@ extern void build_tlb_refill_handler(void);
>  #endif /* CONFIG_MIPS_MT_SMTC */
>  
>  /*
> - * LOONGSON2 has a 4 entry itlb which is a subset of dtlb,
> + * LOONGSON2/3 has a 4 entry itlb which is a subset of dtlb,
>   * unfortrunately, itlb is not totally transparent to software.

unfortunately

>   */
>  static inline void flush_itlb(void)
>  {
>       switch (current_cpu_type()) {
>       case CPU_LOONGSON2:
> +     case CPU_LOONGSON3:
>               write_c0_diag(4);
>               break;
>       default:
> diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c
> index 183f2b5..f89124c 100644
> --- a/arch/mips/mm/tlbex.c
> +++ b/arch/mips/mm/tlbex.c
> @@ -579,6 +579,7 @@ static void build_tlb_write_entry(u32 **p, struct 
> uasm_label **l,
>       case CPU_BMIPS4380:
>       case CPU_BMIPS5000:
>       case CPU_LOONGSON2:
> +     case CPU_LOONGSON3:
>       case CPU_R5500:
>               if (m4kc_tlbp_war())
>                       uasm_i_nop(p);

Beside the few comments above, this patch globally looks fine to me.


-- 
Aurelien Jarno                          GPG: 1024D/F1BCDB73
aurelien@aurel32.net                 http://www.aurel32.net

<Prev in Thread] Current Thread [Next in Thread>