On Tue Feb 3, 2026 at 7:43 PM IST, Bryan Brattlof wrote: > On February 3, 2026 thus sayeth Anshul Dalal: >> We currently provide default board names for each board in their >> respective evm.c file. However for custom boards, this behaviour >> overwrites the default DT as set in the defconfig >> (CONFIG_DEFAULT_FDT_FILE or CONFIG_DEFAULT_DEVICE_TREE). > > Are people using the TI_I2C_BOARD_DETECT logic on their custom boards?
Users are expected to not use TI_I2C_BOARD_DETECT for custom boards but the correct behaviour should be to fallback to what's configured in the .config regardless. > >> >> This patch changes the default name to be NULL which prevents this >> overwrite and allows ti_set_fdt_env to instead fallback to the correct >> DT as set in Kconfig. >> >> Signed-off-by: Anshul Dalal <[email protected]> >> --- >> board/ti/am64x/evm.c | 2 +- >> board/ti/j721e/evm.c | 2 +- >> board/ti/j721s2/evm.c | 2 +- >> 3 files changed, 3 insertions(+), 3 deletions(-) >> >> diff --git a/board/ti/am64x/evm.c b/board/ti/am64x/evm.c >> index 25076a8a588..c6ddc44d14c 100644 >> --- a/board/ti/am64x/evm.c >> +++ b/board/ti/am64x/evm.c >> @@ -138,7 +138,7 @@ static struct ti_fdt_map ti_am64_evm_fdt_map[] = { >> >> static void setup_board_eeprom_env(void) >> { >> - char *name = "am64x_gpevm"; >> + char *name = NULL; > > Would it make sense to add this default to the else case? > I assume you're suggesting something like the following: diff --git a/board/ti/am64x/evm.c b/board/ti/am64x/evm.c index 25076a8a588..7d59e5c376c 100644 --- a/board/ti/am64x/evm.c +++ b/board/ti/am64x/evm.c @@ -147,9 +147,11 @@ static void setup_board_eeprom_env(void) name = "am64x_gpevm"; else if (board_is_am64x_skevm()) name = "am64x_skevm"; - else + else { + name = NULL; printf("Unidentified board claims %s in eeprom header\n", board_ti_get_name()); + } invalid_eeprom: set_board_info_env_am6(name); We could do so but it would be functionally identical except for the cases where do_board_detect fails and we still fallback to presuming a GP board. > If I understand the problem correctly we still get the benefit of the > default for boards we don't have a match for in eeprom but if > do_board_detect() fails we pass the NULL to ti_set_fdt_env()? > IMO passing NULL to set_board_info_env_am6 and especially ti_set_fdt_env is the correct behavior here as my patch originally intends to. For set_board_info_env_am6, it ensures the "board_name" in env is set to "unknown" instead of the default GP board which might not be correct. And for ti_set_fdt_env it allows the user to specify the DT from .config instead of having to modify ti_fdt_map. > ~Bryan
