diff options
author | Jonathan Hunter <jonathanh@nvidia.com> | 2019-02-12 19:03:14 +0300 |
---|---|---|
committer | Tom Warren <twarren@nvidia.com> | 2019-02-20 19:01:44 +0300 |
commit | f16e3115f3b601f121798070b97224b8234d71b7 (patch) | |
tree | 01f838126f957679a16c7d400df4f7e9ad68bac2 /include/configs/tegra124-common.h | |
parent | 8105816cbb1dd0e13a9602532615cb93f1a502e3 (diff) | |
download | u-boot-f16e3115f3b601f121798070b97224b8234d71b7.tar.xz |
ARM: tegra: Reserve 32MB for the Linux kernel
Booting recently Linux -next kernels on 32-bit Tegra devices has been
failing when using the 'multi_v7_defconfig' kenrel configuration because
the size of has grown such that it is overwriting the FDT blob.
Current Linux -next kernels built with the 'multi_v7_defconfig' have a
total size of ~19.5MB (where .text is ~12.5MB, .data is ~6.5MB and .bss
is ~0.5MB). Therefore, increase the memory location reserved for the
Linux kernel to 32MB from 16MB for 32-bit Tegra devices.
This change has been boot tested on Tegra20 Ventana, Tegra30 Cardhu and
Tegra124 Jetson TK1 with the Linux next tree (20190212).
Signed-off-by: Jonathan Hunter <jonathanh@nvidia.com>
Reviewed-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Diffstat (limited to 'include/configs/tegra124-common.h')
-rw-r--r-- | include/configs/tegra124-common.h | 14 |
1 files changed, 7 insertions, 7 deletions
diff --git a/include/configs/tegra124-common.h b/include/configs/tegra124-common.h index b275f795a3..3530684164 100644 --- a/include/configs/tegra124-common.h +++ b/include/configs/tegra124-common.h @@ -17,7 +17,7 @@ /* * Miscellaneous configurable options */ -#define CONFIG_STACKBASE 0x82800000 /* 40MB */ +#define CONFIG_STACKBASE 0x83800000 /* 56MB */ /*----------------------------------------------------------------------- * Physical Memory Map @@ -38,13 +38,13 @@ * should not overlap that area, or the kernel will have to copy itself * somewhere else before decompression. Similarly, the address of any other * data passed to the kernel shouldn't overlap the start of RAM. Pushing - * this up to 16M allows for a sizable kernel to be decompressed below the + * this up to 32M allows for a sizable kernel to be decompressed below the * compressed load address. * - * fdt_addr_r simply shouldn't overlap anything else. Choosing 32M allows for - * the compressed kernel to be up to 16M too. + * fdt_addr_r simply shouldn't overlap anything else. Choosing 48M allows for + * the compressed kernel to be up to 32M too. * - * ramdisk_addr_r simply shouldn't overlap anything else. Choosing 33M allows + * ramdisk_addr_r simply shouldn't overlap anything else. Choosing 49M allows * for the FDT/DTB to be up to 1M, which is hopefully plenty. */ #define CONFIG_LOADADDR 0x81000000 @@ -52,8 +52,8 @@ "scriptaddr=0x90000000\0" \ "pxefile_addr_r=0x90100000\0" \ "kernel_addr_r=" __stringify(CONFIG_LOADADDR) "\0" \ - "fdt_addr_r=0x82000000\0" \ - "ramdisk_addr_r=0x82100000\0" + "fdt_addr_r=0x83000000\0" \ + "ramdisk_addr_r=0x83100000\0" /* Defines for SPL */ #define CONFIG_SPL_TEXT_BASE 0x80108000 |