[U-Boot] [PATCH] designware_i2c: Enhance DesignWare I2C driver address support

Enhance the DesignWare I2C driver to support address length more than 1 byte. This enhancement is required as some I2C slave device such as EEPROM chip might have 16 bit address byte.
Signed-off-by: Chin Liang See clsee@altera.com Cc: Alexey Brodkin Alexey.Brodkin@synopsys.com Cc: Tom Rini trini@ti.com cc: Armando Visconti armando.visconti@st.com Cc: Stefan Roese sr@denx.de Cc: Albert ARIBAUD albert.u.boot@aribaud.net Cc: Heiko Schocher hs@denx.de Cc: Vipin KUMAR vipin.kumar@st.com Cc: Tom Rix Tom.Rix@windriver.com Cc: Mischa Jonker mjonker@synopsys.com --- drivers/i2c/designware_i2c.c | 24 +++++++++--------------- 1 file changed, 9 insertions(+), 15 deletions(-)
diff --git a/drivers/i2c/designware_i2c.c b/drivers/i2c/designware_i2c.c index cb2ac04..c0ac5f7 100644 --- a/drivers/i2c/designware_i2c.c +++ b/drivers/i2c/designware_i2c.c @@ -205,27 +205,21 @@ static int check_params(uint addr, int alen, uchar *buffer, int len) return 1; }
- if (alen > 1) { - printf("addr len %d not supported\n", alen); - return 1; - } - - if (addr + len > 256) { - printf("address out of range\n"); - return 1; - } - return 0; }
-static int i2c_xfer_init(uchar chip, uint addr) +static int i2c_xfer_init(uchar chip, uint addr, int alen) { if (i2c_wait_for_bb()) return 1;
i2c_setaddress(chip); - writel(addr, &i2c_regs_p->ic_cmd_data); - + while (alen) { + alen--; + /* high byte address going out first */ + writel((addr >> (alen * 8)) & 0xff, + &i2c_regs_p->ic_cmd_data); + } return 0; }
@@ -269,7 +263,7 @@ int i2c_read(uchar chip, uint addr, int alen, uchar *buffer, int len) if (check_params(addr, alen, buffer, len)) return 1;
- if (i2c_xfer_init(chip, addr)) + if (i2c_xfer_init(chip, addr, alen)) return 1;
start_time_rx = get_timer(0); @@ -310,7 +304,7 @@ int i2c_write(uchar chip, uint addr, int alen, uchar *buffer, int len) if (check_params(addr, alen, buffer, len)) return 1;
- if (i2c_xfer_init(chip, addr)) + if (i2c_xfer_init(chip, addr, alen)) return 1;
start_time_tx = get_timer(0);

On Wed, 2014-01-15 at 08:58 -0600, Chin Liang See wrote:
Enhance the DesignWare I2C driver to support address length more than 1 byte. This enhancement is required as some I2C slave device such as EEPROM chip might have 16 bit address byte. diff --git a/drivers/i2c/designware_i2c.c b/drivers/i2c/designware_i2c.c index cb2ac04..c0ac5f7 100644 --- a/drivers/i2c/designware_i2c.c +++ b/drivers/i2c/designware_i2c.c @@ -205,27 +205,21 @@ static int check_params(uint addr, int alen, uchar *buffer, int len) return 1; }
- if (alen > 1) {
printf("addr len %d not supported\n", alen);
return 1;
- }
- if (addr + len > 256) {
printf("address out of range\n");
return 1;
- }
- return 0;
}
Hi Chin,
if you strip down functionality of "check_params()" to one single check I would recommend you to remove "check_params()" at all and do in-place check for "buffer" existence.
Moreover you may just use "assert" for this check because this buffer is passed by u-boot (no need to check every parameter passed to any function in run-time) so in production/release build it won't exist at all.
Regards, Alexey

Hi Alexey,
On Wed, 2014-01-15 at 15:30 +0000, Alexey Brodkin wrote:
On Wed, 2014-01-15 at 08:58 -0600, Chin Liang See wrote:
Enhance the DesignWare I2C driver to support address length more than 1 byte. This enhancement is required as some I2C slave device such as EEPROM chip might have 16 bit address byte. diff --git a/drivers/i2c/designware_i2c.c b/drivers/i2c/designware_i2c.c index cb2ac04..c0ac5f7 100644 --- a/drivers/i2c/designware_i2c.c +++ b/drivers/i2c/designware_i2c.c @@ -205,27 +205,21 @@ static int check_params(uint addr, int alen, uchar *buffer, int len) return 1; }
- if (alen > 1) {
printf("addr len %d not supported\n", alen);
return 1;
- }
- if (addr + len > 256) {
printf("address out of range\n");
return 1;
- }
- return 0;
}
Hi Chin,
if you strip down functionality of "check_params()" to one single check I would recommend you to remove "check_params()" at all and do in-place check for "buffer" existence.
Moreover you may just use "assert" for this check because this buffer is passed by u-boot (no need to check every parameter passed to any function in run-time) so in production/release build it won't exist at all.
Good suggestion. I agreed that we can strip off the check_params. Let me send the v2 patch. Thanks and have a nice day!
Chin Liang
Regards, Alexey
participants (2)
-
Alexey Brodkin
-
Chin Liang See