[U-Boot] Broken: U-Boot SPL 2016.01
In 2016.01 SPL is broken on an omap board with DM3730:
U-Boot SPL 2016.01-00076-g72128b2 (Jan 14 2016 - 22:07:33) wrong hwadapnr: 1075903588
The error message comes from drivers/i2c/omap24xx_i2c.c:
static struct i2c *omap24_get_base(struct i2c_adapter *adap) { switch (adap->hwadapnr) { case 0: return (struct i2c *)I2C_BASE1; break; case 1: return (struct i2c *)I2C_BASE2; break; #if (I2C_BUS_MAX > 2) case 2: return (struct i2c *)I2C_BASE3; break; #if (I2C_BUS_MAX > 3) case 3: return (struct i2c *)I2C_BASE4; break; #if (I2C_BUS_MAX > 4) case 4: return (struct i2c *)I2C_BASE5; break; #endif #endif #endif default: printf("wrong hwadapnr: %d\n", adap->hwadapnr); break; } return NULL; }
Has anything changed in the I2C configuration for SPL?
Peter
On Thu, Jan 14, 2016 at 10:29:44PM +0100, Peter Kümmel wrote:
In 2016.01 SPL is broken on an omap board with DM3730:
U-Boot SPL 2016.01-00076-g72128b2 (Jan 14 2016 - 22:07:33) wrong hwadapnr: 1075903588
Seems like gd isn't pointing in the right place or looking at garbage or something.
Am 14.01.2016 um 22:49 schrieb Tom Rini:
On Thu, Jan 14, 2016 at 10:29:44PM +0100, Peter Kümmel wrote:
In 2016.01 SPL is broken on an omap board with DM3730:
U-Boot SPL 2016.01-00076-g72128b2 (Jan 14 2016 - 22:07:33) wrong hwadapnr: 1075903588
Seems like gd isn't pointing in the right place or looking at garbage or something.
After disabling CONFIG_EXPERT (and implicitly CONFIG_SYS_MALLOC_CLEAR_ON_INIT) it works again.
Seems with for a GCC update from 5.2 to 5.3 this flag is very critical.
Many thanks, Peter
On Fri, Jan 15, 2016 at 12:32:04AM +0100, Peter Kümmel wrote:
Am 14.01.2016 um 22:49 schrieb Tom Rini:
On Thu, Jan 14, 2016 at 10:29:44PM +0100, Peter Kümmel wrote:
In 2016.01 SPL is broken on an omap board with DM3730:
U-Boot SPL 2016.01-00076-g72128b2 (Jan 14 2016 - 22:07:33) wrong hwadapnr: 1075903588
Seems like gd isn't pointing in the right place or looking at garbage or something.
After disabling CONFIG_EXPERT (and implicitly CONFIG_SYS_MALLOC_CLEAR_ON_INIT) it works again.
Seems with for a GCC update from 5.2 to 5.3 this flag is very critical.
To be clear, disabling or enabling CONFIG_SYS_MALLOC_CLEAR_ON_INIT ?
participants (2)
-
Peter Kümmel -
Tom Rini