diff mbox

platforms/astbmc: Move prd_init calls to astbmc_early_init()

Message ID 1442197954-8361-1-git-send-email-jk@ozlabs.org
State Accepted
Headers show

Commit Message

Jeremy Kerr Sept. 14, 2015, 2:32 a.m. UTC
Currently, most astbmc platforms do their own call to prd_init(), but
garrison is out-of-sync.

This change moves the prd_init call to astbmc_early_init, so we don't
need to enable it on every platform.

Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
---
 platforms/astbmc/common.c    | 2 ++
 platforms/astbmc/firestone.c | 1 -
 platforms/astbmc/habanero.c  | 2 --
 platforms/astbmc/palmetto.c  | 2 --
 4 files changed, 2 insertions(+), 5 deletions(-)

Comments

Joel Stanley Sept. 14, 2015, 3:10 a.m. UTC | #1
On Mon, Sep 14, 2015 at 12:02 PM, Jeremy Kerr <jk@ozlabs.org> wrote:
>
> Currently, most astbmc platforms do their own call to prd_init(), but
> garrison is out-of-sync.
>
> This change moves the prd_init call to astbmc_early_init, so we don't
> need to enable it on every platform.

Good idea.

> Signed-off-by: Jeremy Kerr <jk@ozlabs.org>

Reviewed-by: Joel Stanley <joel@jms.id.au>

> ---
>  platforms/astbmc/common.c    | 2 ++
>  platforms/astbmc/firestone.c | 1 -
>  platforms/astbmc/habanero.c  | 2 --
>  platforms/astbmc/palmetto.c  | 2 --
>  4 files changed, 2 insertions(+), 5 deletions(-)
>
> diff --git a/platforms/astbmc/common.c b/platforms/astbmc/common.c
> index 8d37785..40a9fc8 100644
> --- a/platforms/astbmc/common.c
> +++ b/platforms/astbmc/common.c
> @@ -348,4 +348,6 @@ void astbmc_early_init(void)
>
>         /* Setup UART and use it as console with interrupts */
>         uart_init(true);
> +
> +       prd_init();

Should it always be last? Do we need a comment?

>  }
> diff --git a/platforms/astbmc/firestone.c b/platforms/astbmc/firestone.c
> index 4eed429..aaed13e 100644
> --- a/platforms/astbmc/firestone.c
> +++ b/platforms/astbmc/firestone.c
> @@ -138,7 +138,6 @@ static bool firestone_probe(void)
>
>         /* Lot of common early inits here */
>         astbmc_early_init();
> -       prd_init();
>         slot_table_init(firestone_phb_table);
>
>         return true;
> diff --git a/platforms/astbmc/habanero.c b/platforms/astbmc/habanero.c
> index 4ab2cdc..738aa63 100644
> --- a/platforms/astbmc/habanero.c
> +++ b/platforms/astbmc/habanero.c
> @@ -134,8 +134,6 @@ static bool habanero_probe(void)
>         /* Lot of common early inits here */
>         astbmc_early_init();
>
> -       prd_init();
> -
>         slot_table_init(habanero_phb_table);
>
>         return true;
> diff --git a/platforms/astbmc/palmetto.c b/platforms/astbmc/palmetto.c
> index ca91908..033f103 100644
> --- a/platforms/astbmc/palmetto.c
> +++ b/platforms/astbmc/palmetto.c
> @@ -39,8 +39,6 @@ static bool palmetto_probe(void)
>         /* Lot of common early inits here */
>         astbmc_early_init();
>
> -       prd_init();
> -
>         return true;
>  }
>
> --
> 2.1.4
>
> _______________________________________________
> Skiboot mailing list
> Skiboot@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/skiboot
Jeremy Kerr Sept. 14, 2015, 7:13 a.m. UTC | #2
Hi Joel,

>> --- a/platforms/astbmc/common.c
>> +++ b/platforms/astbmc/common.c
>> @@ -348,4 +348,6 @@ void astbmc_early_init(void)
>>
>>         /* Setup UART and use it as console with interrupts */
>>         uart_init(true);
>> +
>> +       prd_init();
> 
> Should it always be last? Do we need a comment?

There's no need for it to be early. I've added it last as each of the
individual platform definitions called prd_init after
astbmc_early_init(), so we have the smallest functional change here.

Not sure we'd need a comment - it "does what it says on the tin".

Cheers,


Jeremy
Stewart Smith Sept. 15, 2015, 6:39 a.m. UTC | #3
Jeremy Kerr <jk@ozlabs.org> writes:

> Hi Joel,
>
>>> --- a/platforms/astbmc/common.c
>>> +++ b/platforms/astbmc/common.c
>>> @@ -348,4 +348,6 @@ void astbmc_early_init(void)
>>>
>>>         /* Setup UART and use it as console with interrupts */
>>>         uart_init(true);
>>> +
>>> +       prd_init();
>> 
>> Should it always be last? Do we need a comment?
>
> There's no need for it to be early. I've added it last as each of the
> individual platform definitions called prd_init after
> astbmc_early_init(), so we have the smallest functional change here.
>
> Not sure we'd need a comment - it "does what it says on the tin".

Solidly agreed - I casually detest the pattern of "comment above
obviously named function" and umm and ahh over the day I'll go through
skiboot and rip out all of them.

Also, merged as of ef186c6
diff mbox

Patch

diff --git a/platforms/astbmc/common.c b/platforms/astbmc/common.c
index 8d37785..40a9fc8 100644
--- a/platforms/astbmc/common.c
+++ b/platforms/astbmc/common.c
@@ -348,4 +348,6 @@  void astbmc_early_init(void)
 
 	/* Setup UART and use it as console with interrupts */
 	uart_init(true);
+
+	prd_init();
 }
diff --git a/platforms/astbmc/firestone.c b/platforms/astbmc/firestone.c
index 4eed429..aaed13e 100644
--- a/platforms/astbmc/firestone.c
+++ b/platforms/astbmc/firestone.c
@@ -138,7 +138,6 @@  static bool firestone_probe(void)
 
 	/* Lot of common early inits here */
 	astbmc_early_init();
-	prd_init();
 	slot_table_init(firestone_phb_table);
 
 	return true;
diff --git a/platforms/astbmc/habanero.c b/platforms/astbmc/habanero.c
index 4ab2cdc..738aa63 100644
--- a/platforms/astbmc/habanero.c
+++ b/platforms/astbmc/habanero.c
@@ -134,8 +134,6 @@  static bool habanero_probe(void)
 	/* Lot of common early inits here */
 	astbmc_early_init();
 
-	prd_init();
-
 	slot_table_init(habanero_phb_table);
 
 	return true;
diff --git a/platforms/astbmc/palmetto.c b/platforms/astbmc/palmetto.c
index ca91908..033f103 100644
--- a/platforms/astbmc/palmetto.c
+++ b/platforms/astbmc/palmetto.c
@@ -39,8 +39,6 @@  static bool palmetto_probe(void)
 	/* Lot of common early inits here */
 	astbmc_early_init();
 
-	prd_init();
-
 	return true;
 }