On 6/6/19 10:16 AM, Lukasz Majewski wrote:
On Thu, 6 Jun 2019 10:08:04 +0200 Marek Vasut marex@denx.de wrote:
On 6/6/19 10:00 AM, Lukasz Majewski wrote:
On Thu, 6 Jun 2019 09:26:04 +0200 Marek Vasut marex@denx.de wrote:
On 6/6/19 9:09 AM, Lukasz Majewski wrote:
On Wed, 5 Jun 2019 18:24:11 +0200 Lukasz Majewski lukma@denx.de wrote:
Hi Marek,
> On Sun, 12 May 2019 23:34:52 +0200 > Marek Vasut marex@denx.de wrote: > >> Use CONFIG_IS_ENABLED(WDT) to permit use of WDT in SPL without >> DM, while the full U-Boot can use rich DM/DT WDT driver. >> >> Signed-off-by: Marek Vasut marex@denx.de >> Cc: Peng Fan Peng.Fan@freescale.com >> Cc: Stefano Babic sbabic@denx.de >> --- >> common/board_r.c | 2 +- >> common/spl/spl.c | 2 +- >> drivers/watchdog/Makefile | 2 +- >> include/asm-generic/global_data.h | 2 +- >> include/wdt.h | 2 +- >> 5 files changed, 5 insertions(+), 5 deletions(-) >> >> diff --git a/common/board_r.c b/common/board_r.c >> index 150e8cd424..988e40abb2 100644 >> --- a/common/board_r.c >> +++ b/common/board_r.c >> @@ -678,7 +678,7 @@ static init_fnc_t init_sequence_r[] = { >> #ifdef CONFIG_DM >> initr_dm, >> #endif >> -#if defined(CONFIG_WDT) >> +#if CONFIG_IS_ENABLED(WDT) >> initr_watchdog, >> #endif >> #if defined(CONFIG_ARM) || defined(CONFIG_NDS32) || >> defined(CONFIG_RISCV) || \ diff --git a/common/spl/spl.c >> b/common/spl/spl.c index 0a6a47c202..f22f854718 100644 >> --- a/common/spl/spl.c >> +++ b/common/spl/spl.c >> @@ -601,7 +601,7 @@ void board_init_r(gd_t *dummy1, ulong >> dummy2) spl_board_init(); >> #endif >> >> -#if defined(CONFIG_SPL_WATCHDOG_SUPPORT) && >> defined(CONFIG_WDT) +#if defined(CONFIG_SPL_WATCHDOG_SUPPORT) >> && CONFIG_IS_ENABLED(WDT) initr_watchdog(); >> #endif >> >> diff --git a/drivers/watchdog/Makefile >> b/drivers/watchdog/Makefile index 40b2f4bc66..4b94ae988c 100644 >> --- a/drivers/watchdog/Makefile >> +++ b/drivers/watchdog/Makefile >> @@ -16,7 +16,7 @@ obj-$(CONFIG_OMAP_WATCHDOG) += omap_wdt.o >> obj-$(CONFIG_DESIGNWARE_WATCHDOG) += designware_wdt.o >> obj-$(CONFIG_TANGIER_WATCHDOG) += tangier_wdt.o >> obj-$(CONFIG_ULP_WATCHDOG) += ulp_wdog.o >> -obj-$(CONFIG_WDT) += wdt-uclass.o >> +obj-$(CONFIG_$(SPL_TPL_)WDT) += wdt-uclass.o >> obj-$(CONFIG_WDT_SANDBOX) += sandbox_wdt.o >> obj-$(CONFIG_WDT_ARMADA_37XX) += armada-37xx-wdt.o >> obj-$(CONFIG_WDT_ASPEED) += ast_wdt.o >> diff --git a/include/asm-generic/global_data.h >> b/include/asm-generic/global_data.h index >> 02a3ed6838..7c2220643b 100644 --- >> a/include/asm-generic/global_data.h +++ >> b/include/asm-generic/global_data.h @@ -137,7 +137,7 @@ >> typedef struct global_data { #if >> defined(CONFIG_TRANSLATION_OFFSET) fdt_addr_t >> translation_offset; /* optional translation offset */ >> #endif -#if defined(CONFIG_WDT) >> +#if CONFIG_IS_ENABLED(WDT) >> struct udevice *watchdog_dev; >> #endif >> } gd_t; >> diff --git a/include/wdt.h b/include/wdt.h >> index aa77d3e9b4..5bcff24ab3 100644 >> --- a/include/wdt.h >> +++ b/include/wdt.h >> @@ -106,7 +106,7 @@ struct wdt_ops { >> int (*expire_now)(struct udevice *dev, ulong flags); >> }; >> >> -#if defined(CONFIG_WDT) >> +#if CONFIG_IS_ENABLED(WDT) >> #ifndef CONFIG_WATCHDOG_TIMEOUT_MSECS >> #define CONFIG_WATCHDOG_TIMEOUT_MSECS (60 * 1000) >> #endif > > Tested-by: Lukasz Majewski lukma@denx.de > > Test HW: display5 i.MX6Q device >
Unfortunately this series causes build break when run on Travis-CI for some Atmel/Microchip:
arm: + picosam9g45 +lib/built-in.o: In function `udelay': +lib/time.c:167: undefined reference to `watchdog_reset' +drivers/built-in.o: In function `atmel_serial_getc': +drivers/serial/atmel_usart.c:103: undefined reference to `watchdog_reset' +make[2]: *** [spl/u-boot-spl] Error 1 +make[1]: *** [spl/u-boot-spl] Error 2 +make: *** [sub-make] Error 2 14 38 1 /53 axm boards.cfg is up to date. Nothing to do. Summary of current source for 53 boards (2 threads, 1 job per thread)
u-boot master SHA1: e1a2ed7180adeefb6164239a18249dca5701319d
The fix is under Travis-CI testing: https://travis-ci.org/lmajewski/u-boot-dfu/builds/542118403
I will post patches when it finishes.
What's the fix ?
https://github.com/lmajewski/u-boot-dfu/commit/98c95790fee02296023a290681109...
I've explicitly enabled SPL_WDT=y for affected boards.
I don't think that's the right approach, it should be possible to do without explicit board config modification. I'm building another take, since some boards still failed to build.
Just to be clear - if it works with imply - I'm for this solution.
However, if it breaks, then we may need to find another solution (as for example shown here: https://github.com/lmajewski/u-boot-dfu/commit/98c95790fee02296023a290681109...
- the number of affected boards is small - just 3 of them).
I would very much prefer to make it work without defconfig modifications. This has to be possible, since it worked before the split.