[U-Boot] [PATCH RFT 1/2] TI ARMv7: Don't use GD before crt0.S has set it
Prior to this change we set the gd pointer early so that we can store data in it. This becomes problematic for DM changes as well as being odd in general. Re-work the code paths so that we don't need to set the gd pointer so early and instead can rely upon the normal setting of it.
In order to do this we do need to move certain calls from s_init into spl_board_init(), mainly preloader_console_init and save_omap_boot_params.
Tested on: Beaglebone Black, AM43xx GP EVM, Beagleboard, Beagleboard xM, OMAP5 uEVM, DRA7xx EVM Signed-off-by: Tom Rini trini@ti.com --- arch/arm/cpu/armv7/am33xx/board.c | 11 ----------- arch/arm/cpu/armv7/omap-common/boot-common.c | 10 ++++++++++ arch/arm/cpu/armv7/omap-common/hwinit-common.c | 11 ----------- arch/arm/cpu/armv7/omap3/board.c | 9 +-------- 4 files changed, 11 insertions(+), 30 deletions(-)
diff --git a/arch/arm/cpu/armv7/am33xx/board.c b/arch/arm/cpu/armv7/am33xx/board.c index eaf09d1..81477aa 100644 --- a/arch/arm/cpu/armv7/am33xx/board.c +++ b/arch/arm/cpu/armv7/am33xx/board.c @@ -285,14 +285,6 @@ void s_init(void) #ifdef CONFIG_NOR_BOOT enable_norboot_pin_mux(); #endif - /* - * Save the boot parameters passed from romcode. - * We cannot delay the saving further than this, - * to prevent overwrites. - */ -#ifdef CONFIG_SPL_BUILD - save_omap_boot_params(); -#endif watchdog_disable(); set_uart_mux_conf(); setup_clocks_for_console(); @@ -301,9 +293,6 @@ void s_init(void) gd->baudrate = CONFIG_BAUDRATE; serial_init(); gd->have_console = 1; -#elif defined(CONFIG_SPL_BUILD) - gd = &gdata; - preloader_console_init(); #endif #if defined(CONFIG_SPL_AM33XX_ENABLE_RTC32K_OSC) /* Enable RTC32K clock */ diff --git a/arch/arm/cpu/armv7/omap-common/boot-common.c b/arch/arm/cpu/armv7/omap-common/boot-common.c index cb18908..3ba62db 100644 --- a/arch/arm/cpu/armv7/omap-common/boot-common.c +++ b/arch/arm/cpu/armv7/omap-common/boot-common.c @@ -104,6 +104,16 @@ u32 spl_boot_mode(void)
void spl_board_init(void) { + /* + * Save the boot parameters passed from romcode. + * We cannot delay the saving further than this, + * to prevent overwrites. + */ + save_omap_boot_params(); + + /* Prepare console output */ + preloader_console_init(); + #ifdef CONFIG_SPL_NAND_SUPPORT gpmc_init(); #endif diff --git a/arch/arm/cpu/armv7/omap-common/hwinit-common.c b/arch/arm/cpu/armv7/omap-common/hwinit-common.c index dd52e93..cb35c19 100644 --- a/arch/arm/cpu/armv7/omap-common/hwinit-common.c +++ b/arch/arm/cpu/armv7/omap-common/hwinit-common.c @@ -111,14 +111,6 @@ int arch_cpu_init(void) */ void s_init(void) { - /* - * Save the boot parameters passed from romcode. - * We cannot delay the saving further than this, - * to prevent overwrites. - */ -#ifdef CONFIG_SPL_BUILD - save_omap_boot_params(); -#endif init_omap_revision(); hw_data_init();
@@ -133,9 +125,6 @@ void s_init(void) srcomp_enable(); setup_clocks_for_console();
- gd = &gdata; - - preloader_console_init(); do_io_settings(); #endif prcm_init(); diff --git a/arch/arm/cpu/armv7/omap3/board.c b/arch/arm/cpu/armv7/omap3/board.c index 53a9e5d..90d6ae7 100644 --- a/arch/arm/cpu/armv7/omap3/board.c +++ b/arch/arm/cpu/armv7/omap3/board.c @@ -119,6 +119,7 @@ int board_mmc_init(bd_t *bis)
void spl_board_init(void) { + preloader_console_init(); #if defined(CONFIG_SPL_NAND_SUPPORT) || defined(CONFIG_SPL_ONENAND_SUPPORT) gpmc_init(); #endif @@ -264,14 +265,6 @@ void s_init(void) ehci_clocks_enable(); #endif
-#ifdef CONFIG_SPL_BUILD - gd = &gdata; - - preloader_console_init(); - - timer_init(); -#endif - if (!in_sdram) mem_init(); }
In f0c3a6c we stopped setting gd in board_init_f, but later had to revert to due problems on certain platforms. As davinci does not look to have these problems, we can drop the setting here and rely upon crt0.S to do it.
Cc: Peter Howard pjh@northern-ridge.com.au Signed-off-by: Tom Rini trini@ti.com
--- Peter, I don't have my davinci board hooked up but since you were doing some changes recently is there a chance you can test this easily? Thanks! --- arch/arm/cpu/arm926ejs/davinci/spl.c | 1 - 1 file changed, 1 deletion(-)
diff --git a/arch/arm/cpu/arm926ejs/davinci/spl.c b/arch/arm/cpu/arm926ejs/davinci/spl.c index 59b304e..7b14420 100644 --- a/arch/arm/cpu/arm926ejs/davinci/spl.c +++ b/arch/arm/cpu/arm926ejs/davinci/spl.c @@ -51,7 +51,6 @@ void board_init_f(ulong dummy) memset(__bss_start, 0, __bss_end - __bss_start);
/* Finally, setup gd and move to the next step. */ - gd = &gdata; board_init_r(NULL, 0); }
On Fri, 2014-12-19 at 16:53 -0500, Tom Rini wrote:
In f0c3a6c we stopped setting gd in board_init_f, but later had to revert to due problems on certain platforms. As davinci does not look to have these problems, we can drop the setting here and rely upon crt0.S to do it.
Cc: Peter Howard pjh@northern-ridge.com.au Signed-off-by: Tom Rini trini@ti.com
Peter, I don't have my davinci board hooked up but since you were doing some changes recently is there a chance you can test this easily?
I can test on Monday - I only have access to my LCDK on the weekend, and I have to get round to try the convoluted setup process for flashing u-boot on it :-)
Thanks!
arch/arm/cpu/arm926ejs/davinci/spl.c | 1 - 1 file changed, 1 deletion(-)
diff --git a/arch/arm/cpu/arm926ejs/davinci/spl.c b/arch/arm/cpu/arm926ejs/davinci/spl.c index 59b304e..7b14420 100644 --- a/arch/arm/cpu/arm926ejs/davinci/spl.c +++ b/arch/arm/cpu/arm926ejs/davinci/spl.c @@ -51,7 +51,6 @@ void board_init_f(ulong dummy) memset(__bss_start, 0, __bss_end - __bss_start);
/* Finally, setup gd and move to the next step. */
- gd = &gdata; board_init_r(NULL, 0);
}
On Fri, 2014-12-19 at 16:53 -0500, Tom Rini wrote:
In f0c3a6c we stopped setting gd in board_init_f, but later had to revert to due problems on certain platforms. As davinci does not look to have these problems, we can drop the setting here and rely upon crt0.S to do it.
Cc: Peter Howard pjh@northern-ridge.com.au Signed-off-by: Tom Rini trini@ti.com
Peter, I don't have my davinci board hooked up but since you were doing some changes recently is there a chance you can test this easily? Thanks!
Applied this patch (with and without 1/1 as well - given part 1 is ARMv7 it shouldn't have any bearing on davinci but wanted to be sure).
U-boot doesn't boot on the da850evm with this change. I tried both with my changes to the generic board, and on a fresh clone of u-boot.git. And on u-boot-ti.git. I haven't looked further, but I'm guessing there is early access to gd of the sort you patched out for v7.
arch/arm/cpu/arm926ejs/davinci/spl.c | 1 - 1 file changed, 1 deletion(-)
diff --git a/arch/arm/cpu/arm926ejs/davinci/spl.c b/arch/arm/cpu/arm926ejs/davinci/spl.c index 59b304e..7b14420 100644 --- a/arch/arm/cpu/arm926ejs/davinci/spl.c +++ b/arch/arm/cpu/arm926ejs/davinci/spl.c @@ -51,7 +51,6 @@ void board_init_f(ulong dummy) memset(__bss_start, 0, __bss_end - __bss_start);
/* Finally, setup gd and move to the next step. */
- gd = &gdata; board_init_r(NULL, 0);
}
On Mon, Dec 22, 2014 at 08:34:11AM +1100, Peter Howard wrote:
On Fri, 2014-12-19 at 16:53 -0500, Tom Rini wrote:
In f0c3a6c we stopped setting gd in board_init_f, but later had to revert to due problems on certain platforms. As davinci does not look to have these problems, we can drop the setting here and rely upon crt0.S to do it.
Cc: Peter Howard pjh@northern-ridge.com.au Signed-off-by: Tom Rini trini@ti.com
Peter, I don't have my davinci board hooked up but since you were doing some changes recently is there a chance you can test this easily? Thanks!
Applied this patch (with and without 1/1 as well - given part 1 is ARMv7 it shouldn't have any bearing on davinci but wanted to be sure).
U-boot doesn't boot on the da850evm with this change. I tried both with my changes to the generic board, and on a fresh clone of u-boot.git. And on u-boot-ti.git. I haven't looked further, but I'm guessing there is early access to gd of the sort you patched out for v7.
Thanks for testing, I feared it wasn't quite as easy as it looked, I'll have a think about it and see if I can recall what else might be tickling gd that early over there.
Hi Tom,
On 19 December 2014 at 14:53, Tom Rini trini@ti.com wrote:
Prior to this change we set the gd pointer early so that we can store data in it. This becomes problematic for DM changes as well as being odd in general. Re-work the code paths so that we don't need to set the gd pointer so early and instead can rely upon the normal setting of it.
In order to do this we do need to move certain calls from s_init into spl_board_init(), mainly preloader_console_init and save_omap_boot_params.
Tested on: Beaglebone Black, AM43xx GP EVM, Beagleboard, Beagleboard xM, OMAP5 uEVM, DRA7xx EVM Signed-off-by: Tom Rini trini@ti.com
This seems like it's going in the right direction to me!
Tested on Beaglebone Black.
Tested-by: Simon Glass sjg@chromium.org Reviewed-by: Simon Glass sjg@chromium.org
arch/arm/cpu/armv7/am33xx/board.c | 11 ----------- arch/arm/cpu/armv7/omap-common/boot-common.c | 10 ++++++++++ arch/arm/cpu/armv7/omap-common/hwinit-common.c | 11 ----------- arch/arm/cpu/armv7/omap3/board.c | 9 +-------- 4 files changed, 11 insertions(+), 30 deletions(-)
diff --git a/arch/arm/cpu/armv7/am33xx/board.c b/arch/arm/cpu/armv7/am33xx/board.c index eaf09d1..81477aa 100644 --- a/arch/arm/cpu/armv7/am33xx/board.c +++ b/arch/arm/cpu/armv7/am33xx/board.c @@ -285,14 +285,6 @@ void s_init(void) #ifdef CONFIG_NOR_BOOT enable_norboot_pin_mux(); #endif
/** Save the boot parameters passed from romcode.* We cannot delay the saving further than this,* to prevent overwrites.*/-#ifdef CONFIG_SPL_BUILD
save_omap_boot_params();-#endif watchdog_disable(); set_uart_mux_conf(); setup_clocks_for_console(); @@ -301,9 +293,6 @@ void s_init(void) gd->baudrate = CONFIG_BAUDRATE; serial_init(); gd->have_console = 1; -#elif defined(CONFIG_SPL_BUILD)
gd = &gdata;preloader_console_init();#endif #if defined(CONFIG_SPL_AM33XX_ENABLE_RTC32K_OSC) /* Enable RTC32K clock */ diff --git a/arch/arm/cpu/armv7/omap-common/boot-common.c b/arch/arm/cpu/armv7/omap-common/boot-common.c index cb18908..3ba62db 100644 --- a/arch/arm/cpu/armv7/omap-common/boot-common.c +++ b/arch/arm/cpu/armv7/omap-common/boot-common.c @@ -104,6 +104,16 @@ u32 spl_boot_mode(void)
void spl_board_init(void) {
/** Save the boot parameters passed from romcode.* We cannot delay the saving further than this,* to prevent overwrites.*/save_omap_boot_params();/* Prepare console output */preloader_console_init();#ifdef CONFIG_SPL_NAND_SUPPORT gpmc_init(); #endif diff --git a/arch/arm/cpu/armv7/omap-common/hwinit-common.c b/arch/arm/cpu/armv7/omap-common/hwinit-common.c index dd52e93..cb35c19 100644 --- a/arch/arm/cpu/armv7/omap-common/hwinit-common.c +++ b/arch/arm/cpu/armv7/omap-common/hwinit-common.c @@ -111,14 +111,6 @@ int arch_cpu_init(void) */ void s_init(void) {
/** Save the boot parameters passed from romcode.* We cannot delay the saving further than this,* to prevent overwrites.*/-#ifdef CONFIG_SPL_BUILD
save_omap_boot_params();-#endif init_omap_revision(); hw_data_init();
@@ -133,9 +125,6 @@ void s_init(void) srcomp_enable(); setup_clocks_for_console();
gd = &gdata;preloader_console_init(); do_io_settings();#endif prcm_init(); diff --git a/arch/arm/cpu/armv7/omap3/board.c b/arch/arm/cpu/armv7/omap3/board.c index 53a9e5d..90d6ae7 100644 --- a/arch/arm/cpu/armv7/omap3/board.c +++ b/arch/arm/cpu/armv7/omap3/board.c @@ -119,6 +119,7 @@ int board_mmc_init(bd_t *bis)
void spl_board_init(void) {
preloader_console_init();#if defined(CONFIG_SPL_NAND_SUPPORT) || defined(CONFIG_SPL_ONENAND_SUPPORT) gpmc_init(); #endif @@ -264,14 +265,6 @@ void s_init(void) ehci_clocks_enable(); #endif
-#ifdef CONFIG_SPL_BUILD
gd = &gdata;preloader_console_init();timer_init();-#endif
if (!in_sdram) mem_init();}
1.7.9.5
U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
On Fri, Dec 19, 2014 at 04:53:24PM -0500, Tom Rini wrote:
Prior to this change we set the gd pointer early so that we can store data in it. This becomes problematic for DM changes as well as being odd in general. Re-work the code paths so that we don't need to set the gd pointer so early and instead can rely upon the normal setting of it.
In order to do this we do need to move certain calls from s_init into spl_board_init(), mainly preloader_console_init and save_omap_boot_params.
Tested on: Beaglebone Black, AM43xx GP EVM, Beagleboard, Beagleboard xM, OMAP5 uEVM, DRA7xx EVM Signed-off-by: Tom Rini trini@ti.com Tested-by: Simon Glass sjg@chromium.org Reviewed-by: Simon Glass sjg@chromium.org
Applied to u-boot/master, thanks!
participants (3)
-
Peter Howard -
Simon Glass -
Tom Rini