Re: [U-Boot] [PATCH] wait_bit: add big endian version of wait_for_bit function

On Tue, Jan 02, 2018 at 10:37:31AM +0100, Álvaro Fernández Rojas wrote:

Hello Tom,

El 01/01/2018 a las 14:41, Tom Rini escribió:

...

On Sat, Dec 30, 2017 at 11:12:36AM +0100, Álvaro Fernández Rojas wrote:

...

The only difference with the existing wait_for_bit function is the fact that wait_for_bit_be expects the register size to be read.

Signed-off-by: Álvaro Fernández Rojas noltari@gmail.com

include/wait_bit.h | 69 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 69 insertions(+)

diff --git a/include/wait_bit.h b/include/wait_bit.h index 06ad43a122..828fdcda02 100644 --- a/include/wait_bit.h +++ b/include/wait_bit.h @@ -69,5 +69,74 @@ static inline int wait_for_bit(const char *prefix, const u32 *reg, return -ETIMEDOUT; } +/**

• • wait_for_bit_be() waits for bit set/cleared in BE register
• • Function polls register waiting for specific bit(s) change
• • (either 0->1 or 1->0). It can fail under two conditions:
• • • Timeout
• • • User interaction (CTRL-C)
• • Function succeeds only if all bits of masked register are set/cleared
• • (depending on set option).
• • @param prefix Prefix added to timeout messagge (message visible only

• • 	with debug enabled)
    

• • @param reg Register that will be read (using read_be())
• • @param size Size of the register that will be read
• • @param mask Bit(s) of register that must be active
• • @param set Selects wait condition (bit set or clear)
• • @param timeout_ms Timeout (in miliseconds)
• • @param breakable Enables CTRL-C interruption
• • @return 0 on success, -EINVAL, -ETIMEDOUT or -EINTR on failure
• */

+static inline int wait_for_bit_be(const char *prefix,

• 		  const u32 *reg, const u8 size,
  

• 		  const u32 mask, const bool set,
  

• 		  const unsigned int timeout_ms,
  

• 		  const bool breakable)
  

+{

• u32 val;
• unsigned long start = get_timer(0);
• while (1) {

• switch(size)
  

• {
  

• 	case 8:
  

• 		val = readb_be(reg);
  

• 		break;
  

• 	case 16:
  

• 		val = readw_be(reg);
  

• 		break;
  

• 	case 32:
  

• 		val = readl_be(reg);
  

• 		break;
  

• 	default:
  

• 		debug("%s: invalid size %u\n", prefix, size);
  

• 		return -EINVAL;
  

• }
  

• if (!set)
  

• 	val = ~val;
  

• if ((val & mask) == mask)
  

• 	return 0;
  

• if (get_timer(start) > timeout_ms)
  

• 	break;
  

• if (breakable && ctrlc()) {
  

• 	puts("Abort\n");
  

• 	return -EINTR;
  

• }
  

• udelay(1);
  

• WATCHDOG_RESET();
  

• }
• debug("%s: Timeout (reg=%p mask=%08x wait_set=%i)\n", prefix, reg, mask,

•      set);
  

• return -ETIMEDOUT;

+}

Where is this used? Thanks!

I was planning on using this for both brcm63xx SPI drivers, as stated here: https://lists.denx.de/pipermail/u-boot/2017-December/315263.html https://lists.denx.de/pipermail/u-boot/2017-December/315264.html

Maybe it would be better to define a macro and use it for both LE/BE, with functions like: wait_for_bit_8 wait_for_bit_le16/wait_for_bit_be16 wait_for_bit_le32/wait_for_bit_be32

What do you think?

Whatever feels cleaner with the rest of your patches is fine. But please post it as part of the series. Thanks!