From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?utf-8?Q?Beno=C3=AEt_Th=C3=A9baudeau?= Date: Sat, 11 Aug 2012 16:23:35 +0200 (CEST) Subject: [U-Boot] [PATCH] mxc_spi: Round up clock divider In-Reply-To: <50263605.1000601@denx.de> Message-ID: <1126440883.2300301.1344695015769.JavaMail.root@advansee.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Hi Stefano, On 08/11/2012 12:37, Stefano Babic wrote: > On 10/08/2012 20:51, Beno?t Th?baudeau wrote: > > Since the input frequency of the API is a maximum that should not > > be exceeded in > > order for the devices to operate properly, the SPI clock divider > > should be > > rounded up, not truncated. > > > > Hi Beno?t, > > > Signed-off-by: Beno?t Th?baudeau > > Cc: Wolfgang Denk > > Cc: Stefano Babic > > --- > > .../drivers/spi/mxc_spi.c | 4 ++-- > > 1 file changed, 2 insertions(+), 2 deletions(-) > > > > diff --git u-boot-4d3c95f.orig/drivers/spi/mxc_spi.c > > u-boot-4d3c95f/drivers/spi/mxc_spi.c > > index 2e15318..cf1462f 100644 > > --- u-boot-4d3c95f.orig/drivers/spi/mxc_spi.c > > +++ u-boot-4d3c95f/drivers/spi/mxc_spi.c > > @@ -96,7 +96,7 @@ static s32 spi_cfg_mxc(struct mxc_spi_slave > > *mxcs, unsigned int cs, > > > > clk_src = mxc_get_clock(MXC_CSPI_CLK); > > > > - div = clk_src / max_hz; > > + div = DIV_ROUND_UP(clk_src, max_hz); > > div = get_cspi_div(div); > > > > debug("clk %d Hz, div %d, real clk %d Hz\n", > > @@ -147,7 +147,7 @@ static s32 spi_cfg_mxc(struct mxc_spi_slave > > *mxcs, unsigned int cs, > > * The following computation is taken directly from Freescale's > > code. > > */ > > if (clk_src > max_hz) { > > - pre_div = clk_src / max_hz; > > + pre_div = DIV_ROUND_UP(clk_src, max_hz); > > if (pre_div > 16) { > > post_div = pre_div / 16; > > pre_div = 15; > > > > Agree - do you also get a case where the divider was too low and the > SPI > does not work or it was only examining the code ? It was only examining the code. I had many clock issues on i.MX51, so I made a global review of clock usage that led to that. Most i.MX5 clock functions are broken. I have a series fixing these that I will post next week. > Acked-by: Stefano Babic > > Best regards, > Stefano Babic Best regards, Beno?t