Please pull u-boot-fdt
Hi Tom,
Passing run here:
https://gitlab.denx.de/u-boot/custodians/u-boot-dm/pipelines/1210
The following changes since commit 73b6e6ad254b36763419cdd3fdf406c0094517b7:
Merge tag 'u-boot-imx-20191104' of
https://gitlab.denx.de/u-boot/custodians/u-boot-imx (2019-11-04 12:57:41 -0500)
are available in the Git repository at:
git://git.denx.de/u-boot-fdt.git tags/fdt-pull-5nov19
for you to fetch changes up to 388560134b99dc4cc752627d3a7e9f8c8c2a89a7:
binman: Move to use Python 3 (2019-11-04 18:15:32 -0700)
----------------------------------------------------------------
Update to latest libfdt and pylibfdt, with added size control Update binman, dtoc, patman, buildman to Python 3 Update move_config, rkmux, microcode_tool to Python 3
----------------------------------------------------------------
Simon Glass (20):
fdt: Add INT32_MAX to kernel.h for libfdt
fdt: Add Kconfig options to control code size
mx6: tbs2910: Minimise libfdt code size
fdt: Sync up to the latest libfdt
patman: Adjust 'command' to return strings instead of bytes
patman: Use unicode for file I/O
patman: Move to use Python 3
buildman: Convert to Python 3
test_fdt: Move to use Python 3
test_dtoc: Move to use Python 3
microcode_tool: Convert to Python 3
move_config: Convert to Python 3
rkmux: Convert to Python 3
pylibfdt: Convert to Python 3
pylibfdt: Sync up with upstream
pylibfdt: Correct the type for fdt_property_stub()
binman: Remember the pre-reset entry size
binman: Convert a few tests to Python 3
dtoc: Convert fdt.py to Python 3
binman: Move to use Python 3
configs/tbs2910_defconfig | 1 + include/linux/kernel.h | 2 + include/linux/libfdt_env.h | 1 + lib/Kconfig | 33 +++ lib/libfdt/Makefile | 3 +- lib/libfdt/fdt_ro.c | 420 +++++++++++++++++++++++++--------- scripts/dtc/libfdt/Makefile.libfdt | 7 + scripts/dtc/libfdt/fdt.c | 182 ++++++++++----- scripts/dtc/libfdt/fdt.h | 47 +--- scripts/dtc/libfdt/fdt_addresses.c | 94 ++++---- scripts/dtc/libfdt/fdt_empty_tree.c | 47 +--- scripts/dtc/libfdt/fdt_overlay.c | 91 +++----- scripts/dtc/libfdt/fdt_ro.c | 341 ++++++++++++++++++--------- scripts/dtc/libfdt/fdt_rw.c | 119 +++++----- scripts/dtc/libfdt/fdt_strerror.c | 47 +--- scripts/dtc/libfdt/fdt_sw.c | 241 ++++++++++++------- scripts/dtc/libfdt/fdt_wip.c | 47 +--- scripts/dtc/libfdt/libfdt.h | 268 +++++++++++++++++----- scripts/dtc/libfdt/libfdt_env.h | 48 +--- scripts/dtc/libfdt/libfdt_internal.h | 144 +++++++----- scripts/dtc/pylibfdt/Makefile | 2 +- scripts/dtc/pylibfdt/libfdt.i_shipped | 51 +++-- scripts/dtc/pylibfdt/setup.py | 2 +- tools/binman/binman.py | 2 +- tools/binman/cbfs_util_test.py | 2 +- tools/binman/entry.py | 25 +- tools/binman/entry_test.py | 15 -- tools/binman/etype/intel_fit.py | 2 +- tools/binman/ftest.py | 16 +- tools/buildman/board.py | 9 +- tools/buildman/bsettings.py | 20 +- tools/buildman/builder.py | 47 ++-- tools/buildman/builderthread.py | 24 +- tools/buildman/buildman.py | 10 +- tools/buildman/control.py | 44 ++-- tools/buildman/func_test.py | 16 +- tools/buildman/test.py | 22 +- tools/buildman/toolchain.py | 99 ++++---- tools/dtoc/dtoc.py | 2 +- tools/dtoc/fdt.py | 17 +- tools/dtoc/test_dtoc.py | 1 + tools/dtoc/test_fdt.py | 2 +- tools/libfdt/fdt_rw.c | 3 +- tools/microcode-tool.py | 28 +-- tools/moveconfig.py | 82 +++---- tools/patman/command.py | 31 ++- tools/patman/func_test.py | 8 +- tools/patman/patchstream.py | 4 +- tools/patman/patman.py | 2 +- tools/patman/series.py | 2 +- tools/patman/settings.py | 4 +- tools/patman/test.py | 4 +- tools/patman/tools.py | 29 ++- tools/rkmux.py | 16 +- 54 files changed, 1645 insertions(+), 1181 deletions(-) mode change 100644 => 100755 tools/dtoc/test_dtoc.py
Regards,
Simon
PATCH 1/1 travis: Rework how we write the ~/.buildman file
With python3 we're now tripping over a long-standing problem with how we add to the buildman file with some toolchains. We cannot have multiple toolchain-alias sections as that leads to a parse error.
--- .travis.yml | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
.travis.yml +3 -3
@@ -47,7 +47,9 @@ install: # prepare buildman environment - echo -e "[toolchain]\nroot = /usr" > ~/.buildman - echo -e "arc = /tmp/arc_gnu_2018.09_prebuilt_uclibc_le_archs_linux_install" >> ~/.buildman - - echo -e "\n[toolchain-alias]\nsh = sh2\n" >> ~/.buildman + - echo -e "\n[toolchain-alias]\nsh = sh2" >> ~/.buildman + - echo -e "x86 = i386" >> ~/.buildman; + - echo -e "riscv = riscv64" >> ~/.buildman; - cat ~/.buildman - grub-mkimage --prefix="" -o ~/grub_x86.efi -O i386-efi normal echo lsefimmap lsefi lsefisystab efinet tftp minicmd - grub-mkimage --prefix="" -o ~/grub_x64.efi -O x86_64-efi normal echo lsefimmap lsefi lsefisystab efinet tftp minicmd
@@ -74,7 +76,6 @@ before_script: - if [[ "${TOOLCHAIN}" == *sh* ]]; then ./tools/buildman/buildman --fetch-arch sh2 ; fi - if [[ "${TOOLCHAIN}" == *i386* ]]; then ./tools/buildman/buildman --fetch-arch i386; - echo -e "\n[toolchain-alias]\nx86 = i386" >> ~/.buildman; fi - if [[ "${TOOLCHAIN}" == arc ]]; then wget https://github.com/foss-for-synopsys-dwc-arc-processors/toolchain/releases/download/arc-2018.09-release/arc_gnu_2018.09_prebuilt_uclibc_le_archs_linux_install.tar.gz &&
@@ -98,7 +99,6 @@ before_script: - if [[ "${TOOLCHAIN}" == "powerpc" ]]; then ./tools/buildman/buildman --fetch-arch powerpc; fi - if [[ "${TOOLCHAIN}" == "riscv" ]]; then ./tools/buildman/buildman --fetch-arch riscv64; - echo -e "\n[toolchain-alias]\nriscv = riscv64" >> ~/.buildman; fi - if [[ "${QEMU_TARGET}" != "" ]]; then git clone git://git.qemu.org/qemu.git /tmp/qemu;
Signature · 1 line
2.17.1
PATCH 1/1 travis: Rework how we write the ~/.buildman file
Quoted 5 lines · Tom Rini
With python3 we're now tripping over a long-standing problem with how we add to the buildman file with some toolchains. We cannot have multiple toolchain-alias sections as that leads to a parse error.
Signed-off-by Tom Rini <[email protected]>
Applied to u-boot/master, thanks!
Signature · 8 lines
Tom -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 833 bytes Desc: not available URL: <http://lists.denx.de/pipermail/u-boot/attachments/20191105/fb3d53ce/attachment.sig>
Quoted 13 lines · Simon Glass
Hi Tom,
Passing run here:
https://gitlab.denx.de/u-boot/custodians/u-boot-dm/pipelines/1210
The following changes since commit
73b6e6ad254b36763419cdd3fdf406c0094517b7:Merge tag 'u-boot-imx-20191104' of
https://gitlab.denx.de/u-boot/custodians/u-boot-imx (2019-11-04 12:57:41 -0500)
Quoted 8 lines
are available in the Git repository at:
git://git.denx.de/u-boot-fdt.gittags/fdt-pull-5nov19for you to fetch changes up to
388560134b99dc4cc752627d3a7e9f8c8c2a89a7:binman: Move to use Python 3 (2019-11-04 18:15:32 -0700)
Now first,
Applied to u-boot/master, thanks!
And second, which is why I've included the board maintainers list and custodian lists, this causes a bit of size growth everywhere just about (tbs2910 being the exception) because by default upstream now has various input validation routines. It doesn't catch "tell the hardware to overvolt something" but does catch "pass garbage in the property so we can overwrite the stack and ..." type problems. And it's configurable. If you turn off all of the validation stuff, which we do in SPL/TPL by default (size concerns!), in the main U-Boot we go from a size growth to a size shrink. An arbitrary PowerPC board I picked to confirm this on grows by 985 bytes now, but if I turned everything off it would shrink by 1100 bytes. There's an inbetween setting that would probably result in neutral size change.
The default is all of the input validation we can do because I believe it's important to validate inputs when we can validate them. I also firmly believe board maintainers know their requirements best and can provide a different value for their board(s).
Signature · 8 lines
Tom -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 833 bytes Desc: not available URL: <http://lists.denx.de/pipermail/u-boot/attachments/20191105/c6a6dc99/attachment.sig>
Hi Simon,
Quoted 41 lines · Tom Rini
Quoted 13 lines · Simon Glass
Hi Tom,
Passing run here:
https://gitlab.denx.de/u-boot/custodians/u-boot-dm/pipelines/1210
The following changes since commit
73b6e6ad254b36763419cdd3fdf406c0094517b7:Merge tag 'u-boot-imx-20191104' of
https://gitlab.denx.de/u-boot/custodians/u-boot-imx (2019-11-04 12:57:41 -0500)Quoted 8 lines
are available in the Git repository at:
git://git.denx.de/u-boot-fdt.gittags/fdt-pull-5nov19for you to fetch changes up to
388560134b99dc4cc752627d3a7e9f8c8c2a89a7:binman: Move to use Python 3 (2019-11-04 18:15:32 -0700)
Now first,
Applied to u-boot/master, thanks!And second, which is why I've included the board maintainers list and custodian lists, this causes a bit of size growth everywhere just about (tbs2910 being the exception) because by default upstream now has various input validation routines. It doesn't catch "tell the hardware to overvolt something" but does catch "pass garbage in the property so we can overwrite the stack and ..." type problems. And it's configurable. If you turn off all of the validation stuff, which we do in SPL/TPL by default (size concerns!), in the main U-Boot we go from a size growth to a size shrink. An arbitrary PowerPC board I picked to confirm this on grows by 985 bytes now, but if I turned everything off it would shrink by 1100 bytes. There's an inbetween setting that would probably result in neutral size change.
Quoted 4 lines
The default is all of the input validation we can do because I believe it's important to validate inputs when we can validate them. I also firmly believe board maintainers know their requirements best and can provide a different value for their board(s).
I found the following commits breaks SiFive FU540 board, that U-Boot no longer boots:
commit f0921f5098d8d98ac38121837aaf7c4b4cb13bb4
Author: Simon Glass <[email protected]> Date: Sun Oct 27 09:47:42 2019 -0600
fdt: Sync up to the latest libfdt
Reverting this commit makes U-Boot boot again.
The failure is serial_find_console_or_panic() no longer finds the serial port with the latest libfdt. Any ideas?
Regards,
Bin
Hi Bin,
Quoted 57 lines · Bin Meng
Hi Simon,
Quoted 41 lines · Tom Rini
Quoted 13 lines · Simon Glass
Hi Tom,
Passing run here:
https://gitlab.denx.de/u-boot/custodians/u-boot-dm/pipelines/1210
The following changes since commit
73b6e6ad254b36763419cdd3fdf406c0094517b7:Merge tag 'u-boot-imx-20191104' of
https://gitlab.denx.de/u-boot/custodians/u-boot-imx (2019-11-04 12:57:41 -0500)Quoted 8 lines
are available in the Git repository at:
git://git.denx.de/u-boot-fdt.gittags/fdt-pull-5nov19for you to fetch changes up to
388560134b99dc4cc752627d3a7e9f8c8c2a89a7:binman: Move to use Python 3 (2019-11-04 18:15:32 -0700)
Now first,
Applied to u-boot/master, thanks!And second, which is why I've included the board maintainers list and custodian lists, this causes a bit of size growth everywhere just about (tbs2910 being the exception) because by default upstream now has various input validation routines. It doesn't catch "tell the hardware to overvolt something" but does catch "pass garbage in the property so we can overwrite the stack and ..." type problems. And it's configurable. If you turn off all of the validation stuff, which we do in SPL/TPL by default (size concerns!), in the main U-Boot we go from a size growth to a size shrink. An arbitrary PowerPC board I picked to confirm this on grows by 985 bytes now, but if I turned everything off it would shrink by 1100 bytes. There's an inbetween setting that would probably result in neutral size change.
Quoted 4 lines
The default is all of the input validation we can do because I believe it's important to validate inputs when we can validate them. I also firmly believe board maintainers know their requirements best and can provide a different value for their board(s).
I found the following commits breaks SiFive FU540 board, that U-Boot no longer boots:
commit
f0921f5098d8d98ac38121837aaf7c4b4cb13bb4Author: Simon Glass <[email protected]> Date: Sun Oct 27 09:47:42 2019 -0600
Quoted 6 lines
fdt: Sync up to the latest libfdt
Reverting this commit makes U-Boot boot again.
The failure is
serial_find_console_or_panic()no longer finds the serial port with the latest libfdt. Any ideas?
I wonder if it is relying on a feature that never made it upstream?
There was a change to fdt_path_next_separator() which got dropped - perhaps that is it? I had assumed it was incorporated but perhaps no action was taken? If so, the thread is here:
http://patchwork.ozlabs.org/patch/462756/
Regards,
Simon
Hi Simon,
Quoted 68 lines · Simon Glass
Hi Bin,
Quoted 57 lines · Bin Meng
Hi Simon,
Quoted 41 lines · Tom Rini
Quoted 13 lines · Simon Glass
Hi Tom,
Passing run here:
https://gitlab.denx.de/u-boot/custodians/u-boot-dm/pipelines/1210
The following changes since commit
73b6e6ad254b36763419cdd3fdf406c0094517b7:Merge tag 'u-boot-imx-20191104' of
https://gitlab.denx.de/u-boot/custodians/u-boot-imx (2019-11-04 12:57:41 -0500)Quoted 8 lines
are available in the Git repository at:
git://git.denx.de/u-boot-fdt.gittags/fdt-pull-5nov19for you to fetch changes up to
388560134b99dc4cc752627d3a7e9f8c8c2a89a7:binman: Move to use Python 3 (2019-11-04 18:15:32 -0700)
Now first,
Applied to u-boot/master, thanks!And second, which is why I've included the board maintainers list and custodian lists, this causes a bit of size growth everywhere just about (tbs2910 being the exception) because by default upstream now has various input validation routines. It doesn't catch "tell the hardware to overvolt something" but does catch "pass garbage in the property so we can overwrite the stack and ..." type problems. And it's configurable. If you turn off all of the validation stuff, which we do in SPL/TPL by default (size concerns!), in the main U-Boot we go from a size growth to a size shrink. An arbitrary PowerPC board I picked to confirm this on grows by 985 bytes now, but if I turned everything off it would shrink by 1100 bytes. There's an inbetween setting that would probably result in neutral size change.
Quoted 4 lines
The default is all of the input validation we can do because I believe it's important to validate inputs when we can validate them. I also firmly believe board maintainers know their requirements best and can provide a different value for their board(s).
I found the following commits breaks SiFive FU540 board, that U-Boot no longer boots:
commit
f0921f5098d8d98ac38121837aaf7c4b4cb13bb4Author: Simon Glass <[email protected]> Date: Sun Oct 27 09:47:42 2019 -0600Quoted 6 lines
fdt: Sync up to the latest libfdt
Reverting this commit makes U-Boot boot again.
The failure is
serial_find_console_or_panic()no longer finds the serial port with the latest libfdt. Any ideas?
Quoted 7 lines
I wonder if it is relying on a feature that never made it upstream?
There was a change to
fdt_path_next_separator()which got dropped - perhaps that is it? I had assumed it was incorporated but perhaps no action was taken? If so, the thread is here:
Thanks for the pointers.
I simply applied similar changes to fdt_ro.c but it did not make any difference. I will need have a further debug.
Regards,
Bin