mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Du Huanpeng <u74147@gmail.com>
To: Ahmad Fatoum <a.fatoum@pengutronix.de>
Cc: barebox@lists.infradead.org
Subject: Re: [PATCH] MIPS: DEBUG_LL_UART_DIVISOR is 0, use a0 instead
Date: Thu, 17 Sep 2020 21:56:08 +0800	[thread overview]
Message-ID: <20200917135606.GA12263@beer> (raw)
In-Reply-To: <c56b316a-1fba-3e2e-530f-da5d9c810d1a@pengutronix.de>

Hi, this patch should not intoduce any regression to existing boards,
because: 

 1. pass divisor by DEBUG_LL_UART_DIVISOR

#define DEBUG_LL_UART_DIVISOR NUMBER

which the ``NUMBER'' is not ``0'' (zero). In this case, this patch
has no effect and this line of code will be removed by preprocessor,
before compile :)
        move    t1, a0

 2. Pass divisor by a0

#define DEBUG_LL_UART_DIVISOR 0

recalculate the divisor is needed when the pll/div setting is changed
outside of barebox. e.g. ejtag. and load barebox and run in ram.

Because 0 maybe an invalid value for a divisor, so I take this value for
a hint to pass a0 as divisor. this will not conflict with other boards
which pass divisor by defining DEBUG_LL_UART_DIVISOR with a valied value.

> you should migrate all users of this function all at once to supply a valid divisor
> (probably `move a0, DEBUG_LL_UART_DIVISOR`).

MIPS use ``li'' load immediate, and move to copy inter registers.
so, maybe change my patch to:

#if DEBUG_LL_UART_DIVISOR
	li	t1, DEBUG_LL_UART_DIVISOR
#else
	move	t1, a0
#endif

- - - 
or even do a bit tricky:

--- a/arch/mips/include/asm/debug_ll_ns16550.h
+++ b/arch/mips/include/asm/debug_ll_ns16550.h
@@ -65,7 +65,7 @@ static inline void PUTC_LL(char ch)
 	li	t1, UART_LCR_DLAB		/* DLAB on */
 	sb	t1, UART_LCR(t0)		/* Write it out */
 
-	li	t1, DEBUG_LL_UART_DIVISOR
+	add	t1, zero, DEBUG_LL_UART_DIVISOR
 	sb	t1, UART_DLL(t0)		/* write low order byte */
 	srl	t1, t1, 8
 	sb	t1, UART_DLM(t0)		/* write high order byte */
-- 
2.7.4

this will allow to pass divisor by a number or a register.
add is a pseudo insertion can be compile to addi or add depends
on the DEBUG_LL_UART_DIVISOR is a number or a register.


On Fri, Sep 11, 2020 at 04:07:43PM +0200, Ahmad Fatoum wrote:
> On 9/11/20 2:33 PM, Du Huanpeng wrote:
> > this make it possiable pass a0 as divisor to calculate uart baudrate
> > in run time on boot.
> 
> This is not Ok. Your last commit has your board pass the divisor in a0 now, but other
> boards, like the netgear-wg102, don't. Your change might break their UART because the
> code there doesn't expect that a0 needs to have a valid value. If you want to do this,
> you should migrate all users of this function all at once to supply a valid divisor
> (probably `move a0, DEBUG_LL_UART_DIVISOR`).
> 
> It's same thing in C. You don't add a new parameter to a function without checking
> that all callsites handle this correctly.
> 
> > 
> > Signed-off-by: Du Huanpeng <u74147@gmail.com>
> > ---
> >  arch/mips/include/asm/debug_ll_ns16550.h | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/arch/mips/include/asm/debug_ll_ns16550.h b/arch/mips/include/asm/debug_ll_ns16550.h
> > index df58c4c..a33587a 100644
> > --- a/arch/mips/include/asm/debug_ll_ns16550.h
> > +++ b/arch/mips/include/asm/debug_ll_ns16550.h
> > @@ -66,6 +66,9 @@ static inline void PUTC_LL(char ch)
> >  	sb	t1, UART_LCR(t0)		/* Write it out */
> >  
> >  	li	t1, DEBUG_LL_UART_DIVISOR
> > +#if DEBUG_LL_UART_DIVISOR == 0
> > +	move	t1, a0
> > +#endif
> >  	sb	t1, UART_DLL(t0)		/* write low order byte */
> >  	srl	t1, t1, 8
> >  	sb	t1, UART_DLM(t0)		/* write high order byte */
> > 
> 
> -- 
> Pengutronix e.K.                           |                             |
> Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
> 31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
> Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |


  Regards,
  duhuanpeng


_______________________________________________
barebox mailing list
barebox@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/barebox

  reply	other threads:[~2020-09-17 13:56 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-11 12:33 [PATCH] MIPS: loongson1: enable to run from spi flash Du Huanpeng
2020-09-11 12:33 ` [PATCH] MIPS: DEBUG_LL_UART_DIVISOR is 0, use a0 instead Du Huanpeng
2020-09-11 14:07   ` Ahmad Fatoum
2020-09-17 13:56     ` Du Huanpeng [this message]
2020-09-17 17:42       ` Ahmad Fatoum
2020-09-25 16:42         ` Du Huanpeng
2020-09-25 17:54           ` Ahmad Fatoum

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200917135606.GA12263@beer \
    --to=u74147@gmail.com \
    --cc=a.fatoum@pengutronix.de \
    --cc=barebox@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox