| Message ID | 1433462329-10680-8-git-send-email-osp@andrep.de |
|---|---|
| State | Deferred |
| Delegated to: | Tom Rini |
| Headers | show |
On Fri, Jun 05, 2015 at 12:58:48AM +0100, Andre Przywara wrote: > From: Rob Herring <robh@kernel.org> > > On the highbank platform the SoC's management controller firmware > will probe the DRAM modules and populates the initial device tree with > the correct values. Therefore the memory sizes in are already correct, > so remove U-Boot's DRAM bank setup so the memory node is not "fixed > up" by u-boot. > > Signed-off-by: Rob Herring <robh@kernel.org> > Signed-off-by: Andre Przywara <osp@andrep.de> Applied to u-boot/master, thanks!
On Fri, 12 Jun 2015 22:10:07 -0400 Tom Rini <trini@konsulko.com> wrote: > On Fri, Jun 05, 2015 at 12:58:48AM +0100, Andre Przywara wrote: > > > From: Rob Herring <robh@kernel.org> > > > > On the highbank platform the SoC's management controller firmware > > will probe the DRAM modules and populates the initial device tree > > with the correct values. Therefore the memory sizes in are already > > correct, so remove U-Boot's DRAM bank setup so the memory node is > > not "fixed up" by u-boot. > > > > Signed-off-by: Rob Herring <robh@kernel.org> > > Signed-off-by: Andre Przywara <osp@andrep.de> > > Applied to u-boot/master, thanks! Hi Tom, thanks for applying the patches! I sent out a reworked version of patch 6/8, which saw a comment from Simon. Please consider applying the new one _or_ revert this very patch here (7/8), as it alone breaks Calxeda machines (which now end up with an empty memory/reg property instead of an untouched one). Sorry for the mess! Cheers, Andre.
On Sat, Jun 13, 2015 at 12:01:37PM +0100, Andre Przywara wrote: > On Fri, 12 Jun 2015 22:10:07 -0400 > Tom Rini <trini@konsulko.com> wrote: > > > On Fri, Jun 05, 2015 at 12:58:48AM +0100, Andre Przywara wrote: > > > > > From: Rob Herring <robh@kernel.org> > > > > > > On the highbank platform the SoC's management controller firmware > > > will probe the DRAM modules and populates the initial device tree > > > with the correct values. Therefore the memory sizes in are already > > > correct, so remove U-Boot's DRAM bank setup so the memory node is > > > not "fixed up" by u-boot. > > > > > > Signed-off-by: Rob Herring <robh@kernel.org> > > > Signed-off-by: Andre Przywara <osp@andrep.de> > > > > Applied to u-boot/master, thanks! > > Hi Tom, > > thanks for applying the patches! > I sent out a reworked version of patch 6/8, which saw a comment from > Simon. Please consider applying the new one _or_ revert this very patch > here (7/8), as it alone breaks Calxeda machines (which now end up with > an empty memory/reg property instead of an untouched one). > Sorry for the mess! "Good" news! I somehow failed to push things when I said I did, so this was all still just local. I've dropped this patch for now as I want Simon to chime in on your updated patch.
diff --git a/board/highbank/highbank.c b/board/highbank/highbank.c index e8132b8..9ebd895 100644 --- a/board/highbank/highbank.c +++ b/board/highbank/highbank.c @@ -93,12 +93,6 @@ int dram_init(void) return 0; } -void dram_init_banksize(void) -{ - gd->bd->bi_dram[0].start = CONFIG_SYS_SDRAM_BASE; - gd->bd->bi_dram[0].size = PHYS_SDRAM_1_SIZE; -} - #if defined(CONFIG_OF_BOARD_SETUP) int ft_board_setup(void *fdt, bd_t *bd) { diff --git a/include/configs/highbank.h b/include/configs/highbank.h index da1c837..b1d618b 100644 --- a/include/configs/highbank.h +++ b/include/configs/highbank.h @@ -83,7 +83,7 @@ /*----------------------------------------------------------------------- * Physical Memory Map */ -#define CONFIG_NR_DRAM_BANKS 1 +#define CONFIG_NR_DRAM_BANKS 0 #define PHYS_SDRAM_1_SIZE (4089 << 20) #define CONFIG_SYS_MEMTEST_START 0x100000 #define CONFIG_SYS_MEMTEST_END (PHYS_SDRAM_1_SIZE - 0x100000)