flush_icache_range on AMCC 44x targets

flush_icache_range on AMCC 44x targets

[Mike]

[-- Attachment #1: Type: text/plain, Size: 1525 bytes --]

Hi,

i was reading trough arch/powerpc/kernel/misc32.S looking at the icbi and
iccci instructions, from whats on print in
http://s.eeweb.com/members/kvks_kumar/answers/1356585717-PPC440_UM2013.pdf(page
272) iccci should be used once in the power-on / reset routine, and
as far as flush_icache_range goes presumably before icbi is called?

So should not flush_icache_range go
#ifdef CONFIG_44x
iccci    0, r0
#endif
icbi    0,r6

arch/powerpc/kernel/misc32.S:
/*
 * Write any modified data cache blocks out to memory
 * and invalidate the corresponding instruction cache blocks.
 * This is a no-op on the 601.
 *
 * flush_icache_range(unsigned long start, unsigned long stop)
 */
_KPROBE(__flush_icache_range)
BEGIN_FTR_SECTION
    blr                /* for 601, do nothing */
END_FTR_SECTION_IFSET(CPU_FTR_COHERENT_ICACHE)
    li    r5,L1_CACHE_BYTES-1
    andc    r3,r3,r5
    subf    r4,r3,r4
    add    r4,r4,r5
    srwi.    r4,r4,L1_CACHE_SHIFT
    beqlr
    mtctr    r4
    mr    r6,r3
1:    dcbst    0,r3
    addi    r3,r3,L1_CACHE_BYTES
    bdnz    1b
    sync                /* wait for dcbst's to get to ram */
#ifndef CONFIG_44x
    mtctr    r4
2:    icbi    0,r6
    addi    r6,r6,L1_CACHE_BYTES
    bdnz    2b
#else
    /* Flash invalidate on 44x because we are passed kmapped addresses and
       this doesn't work for userspace pages due to the virtually tagged
       icache.  Sigh. */
    iccci    0, r0
#endif
    sync                /* additional sync needed on g4 */
    isync
    blr


Best regards
-Mike

[-- Attachment #2: Type: text/html, Size: 1935 bytes --]

Re: flush_icache_range on AMCC 44x targets

[Josh Boyer]

On Tue, Apr 30, 2013 at 02:17:59PM +0200, Mike wrote:
>Hi,
>
>i was reading trough arch/powerpc/kernel/misc32.S looking at the icbi and
>iccci instructions, from whats on print in
>http://s.eeweb.com/members/kvks_kumar/answers/1356585717-PPC440_UM2013.pdf(page
>272) iccci should be used once in the power-on / reset routine, and
>as far as flush_icache_range goes presumably before icbi is called?

I'm not understanding your question.

>So should not flush_icache_range go
>#ifdef CONFIG_44x
>iccci    0, r0
>#endif
>icbi    0,r6

The icbi isn't ever executed on 44x at all.

>arch/powerpc/kernel/misc32.S:
>/*
> * Write any modified data cache blocks out to memory
> * and invalidate the corresponding instruction cache blocks.
> * This is a no-op on the 601.
> *
> * flush_icache_range(unsigned long start, unsigned long stop)
> */
>_KPROBE(__flush_icache_range)
>BEGIN_FTR_SECTION
>    blr                /* for 601, do nothing */
>END_FTR_SECTION_IFSET(CPU_FTR_COHERENT_ICACHE)
>    li    r5,L1_CACHE_BYTES-1
>    andc    r3,r3,r5
>    subf    r4,r3,r4
>    add    r4,r4,r5
>    srwi.    r4,r4,L1_CACHE_SHIFT
>    beqlr
>    mtctr    r4
>    mr    r6,r3
>1:    dcbst    0,r3
>    addi    r3,r3,L1_CACHE_BYTES
>    bdnz    1b
>    sync                /* wait for dcbst's to get to ram */
>#ifndef CONFIG_44x

This part above is "if not defined CONFIG_44X", which means execute
everything below here until you hit #else if you are running on a
processor that isn't 4xx.

>    mtctr    r4
>2:    icbi    0,r6
>    addi    r6,r6,L1_CACHE_BYTES
>    bdnz    2b
>#else

Otherwise, use iccci.  Which is what you're sort of suggesting be done.

>    /* Flash invalidate on 44x because we are passed kmapped addresses and
>       this doesn't work for userspace pages due to the virtually tagged
>       icache.  Sigh. */
>    iccci    0, r0
>#endif

So.  I think the code is already doing what you think it should?

josh