From mboxrd@z Thu Jan  1 00:00:00 1970
From: Tom Rix <tom@bumblecow.com>
Date: Sun, 02 May 2010 18:33:54 -0500
Subject: [U-Boot] [PATCH v2 03/17] SPEAr : SMI erase and write timeouts
 increased
In-Reply-To: <1272250610-15439-4-git-send-email-vipin.kumar@st.com>
References: <1272250610-15439-1-git-send-email-vipin.kumar@st.com>	<1272250610-15439-2-git-send-email-vipin.kumar@st.com>	<1272250610-15439-3-git-send-email-vipin.kumar@st.com>
	<1272250610-15439-4-git-send-email-vipin.kumar@st.com>
Message-ID: <4BDE0BE2.2070904@bumblecow.com>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

Vipin KUMAR wrote:
> SMI driver fails because of low timeout values. Increasing the erase and write
> timeouts to 3 seconds
> 
> Signed-off-by: Vipin Kumar <vipin.kumar@st.com>
> ---
>  arch/arm/include/asm/arch-spear/spr_smi.h |    6 +++---
>  drivers/mtd/spr_smi.c                     |    8 ++++----
>  2 files changed, 7 insertions(+), 7 deletions(-)
> 
> diff --git a/arch/arm/include/asm/arch-spear/spr_smi.h b/arch/arm/include/asm/arch-spear/spr_smi.h
> index 06df745..e2e5e8f 100644
> --- a/arch/arm/include/asm/arch-spear/spr_smi.h
> +++ b/arch/arm/include/asm/arch-spear/spr_smi.h
> @@ -108,8 +108,8 @@ struct flash_dev {
>  	ushort sector_count;
>  };
>  
> -#define SFLASH_PAGE_SIZE	0x100	/* flash page size */
> -#define XFER_FINISH_TOUT	2	/* xfer finish timeout */
> -#define WMODE_TOUT		2	/* write enable timeout */
> +#define SFLASH_PAGE_SIZE	0x100		/* flash page size */
> +#define XFER_FINISH_TOUT	(3 * CONFIG_SYS_HZ)
> +#define WMODE_TOUT		(3 * CONFIG_SYS_HZ)
>  

Restore the comments for xfer and write timeouts

>  #endif
> diff --git a/drivers/mtd/spr_smi.c b/drivers/mtd/spr_smi.c
> index 9a70a19..189ee6d 100644
> --- a/drivers/mtd/spr_smi.c
> +++ b/drivers/mtd/spr_smi.c
> @@ -60,11 +60,11 @@ static struct flash_dev flash_ids[] = {
>   */
>  static void smi_wait_xfer_finish(int timeout)
>  {
> -	while (timeout--) {
> +	do {
>  		if (readl(&smicntl->smi_sr) & TFF)
>  			break;
>  		udelay(1000);
> -	}
> +	} while (timeout--);
>  }
>  

Earlier (Jan 11), I recommended that if there is timeout, that this is a failure
The status of the failure must be passed up.

smi_write_enable checks the timeout
smi_read_id,  smi_read_sr does not

Why was the logic for the loops changed?
If you want to change the logic, you should also change
smi_wait_till_ready to be consistent.

Tom

>  /*
> @@ -215,11 +215,11 @@ static int smi_write_enable(int bank)
>  	/* Restore the CTRL REG1 state */
>  	writel(ctrlreg1, &smicntl->smi_cr1);
>  
> -	while (timeout--) {
> +	do {
>  		if (smi_read_sr(bank) & (1 << (bank + WM_SHIFT)))
>  			break;
>  		udelay(1000);
> -	}
> +	} while (timeout--);
>  
>  	if (timeout)
>  		return 0;