| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
runqemu QEMU MB instance fails with below errors
QEMU MB instance failed:
qemu-system-microblazeel: -device loader,file=/scratch/sandeep/yocto/2024.2/yp-dev/build-versal/tmp/deploy/images/versal-net-generic/BOOT-versal-net-generic_bh.bin,addr=0xF201E000,force-raw: warning: short-form boolean option 'force-raw' deprecated
Please use force-raw=on instead
qemu-system-microblazeel: -device loader,file=/scratch/sandeep/yocto/2024.2/yp-dev/build-versal/tmp/deploy/images/versal-net-generic/CDO/pmc_cdo.bin,addr=0xf2000000,force-raw: warning: short-form boolean option 'force-raw' deprecated
Please use force-raw=on instead
So use "force-raw=on" in QB_FW_FILES arguments.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add various "superset" machines. These machines can be used to generated a
binary distribution for a collection of machines.
These differ from the generic machines, as they have full CPU optimization
(DEFAULTTUNE) for the SoC, as well as the common MACHINE_FEATURES for a
collection of chips.
The MACHINE_FEATURES specified may not work on all chips, but will not cause
an incompatibility.
The common machines are:
Machine | SoC | DEFAULTTUNE | MACHINE_FEATURES
-------------------+------------+------------------+-----------------
zynq-common | Zynq7 | cortexa9thf-neon | (none)
zynqmp-common | ZynqMP | cortexa53 | vcu rfsoc
zynqmp-mali-common | ZynqMP | cortexa53 | mali400 vcu
versal-common | Versal | cortexa72 | aie vdu
versal-net-common | Versal-Net | cortexa78 | (none)
The mali400 option itself is not mutually exclusive from non, however the
default configuration of many systems also enabled libmali. libmali
inclusion WILL create a system that is incompatible with mesa.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
| |
This recipe did not provide anything useful for the target, so adjust to only
run native and nativesdk.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
| |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Generic machines no longer depend on firmware (ESW) components. Thus they
will now build a full Linux system, where possible, without XSCT or SDT
modes.
It is up to the specific machine.conf files to enable the correct build
flow and that will enable the necessary firmware dependencies for the SoC.
This also means HDF_MACHINES is no longer required.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Zynq-7000 SDT and XSCT runqemu doesn't come up due to conditional
assignment in both machine-xilinx-qemu.inc and zynq-generic.conf file.
With that said the value set in machine-xilinx-qemu.inc will be used
during pre-expansion value. To fix this issue set weak assignment for
QB_XILINX_SERIAL so that variable can be override from machine conf
file or local.conf.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
| |
versal-net-vn-p-b2197-00-reva
Emulation/simulation versal-net platforms are not tested so removed in dtg and uboot.
Instead of versal-net-ipp-rev1.9 using versal-net-vn-p-b2197-00-reva as DT_BOARD_FLAGS name
Signed-off-by: Ashwini Lomate <ashwini.lomate@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
| |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
| |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Generic machines no longer depend on firmware (ESW) components. Thus they
will now build a full Linux system, where possible, without XSCT or SDT
modes. (Default, if XILINX_WITH_ESW is set to xsct, still builds firmware.)
It is up to the specific machine.conf files to enable the correct build
flow and that will enable the necessary firmware dependencies for the SoC.
This also means HDF_MACHINES is no longer required. Set HDF_MACHINES = ""
to produce a firmware-less build.
Also correct QEMU settings so they can be overridden.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
| |
The DTC_FLAGS default should be set with ?= so it can be overridden.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
| |
This reverts commit da77c918db698d7912b6a2c6807720e9e2c522c8.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
| |
Remove soft assignment in IMAGE_FSTYPES variable due to this cpio file
are not generated for poky distro and this is required for devtool
boot-jtag feature.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
| |
Remove FSBL recipe variables from soc-zynq.inc as this is already set
in fsbl recipes and no more required in machine inclusion file.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
| |
Add cortexa9 DEFAULTTUNE which is required for zynq-7000 multiconfig
builds.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
| |
By default upstream qemu sets CDROM, Floppy drive and use eMMC
as secondary boot device the drive index doesn't work as expected.
Hence add -nodefaults option to qemu all device this fix this issue.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
| |
Yocto runqemu boot hangs for all versal devices, this is due to
QB_FW_FILES variable doesn't get expanded, Hence fix variable
QB_FW_FILES variable expansion issue.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The generic machine configurations should not have a direct dependency
on specific firmware components as these dependencies should be
associated with boot.bin (or similar).
This change also removes WIC_DEPENDS as it is no longer required
Signed-off-by: John Toomey <john.toomey@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
SOC_VARIANT has been remove, we are now only using the YP standard SOC_FAMILY
configuration. The defined families are: zynq, zynqmp, versal and versal-net.
Our decision of breaking up versal-net from versal, is based on the SoC CPU
changes from cortexa72/r5 to cortexa78/r52, thus we're treating it as a
different SoC family.
In order to capture the individual capabilities that we used to handle via
SOC_VARIANT, we have defined the following features (some may have been
previously defined):
- mali400 (zynqmp eg and ev)
- vcu (zynqmp ev)
- rfsoc (zynqmp dr RF capabiltiies)
- aie - (versal ai & premium)
- vdu - (versal ai)
SOC_VARIANT_ARCH and SOC_FAMILY_ARCH are now obsolete and replaced by
MACHINE_ARCH. This is based on the guideline that any recipes that use
MACHINE_FEATURES should be MACHINE_ARCH specific.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
| |
These files are already part of scarthgap, remove the local copies.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The path specified is the internal firmware path. This could be in a the
current deploy directory, a multiconfig deploy directory or some other place
directly specified by the user.
When using the sstate-cache, the build of the component in the a multiconfig
directory can be skipped, causing an error to find the component.
Instead we want to use OUR deploy directory, which will always have a copy
of the pmu firmware, no matter how it was constructed or packaged.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
| |
Create a new layer containing all code related to the Mali400 graphics
stack and software. This includes the Mali kernel module / patches,
libglu/libgles code, mesa, wayland, kernel recipe and udev rules. This
layer also includes a dynamic layer for the qt5 bbappends and patches.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
| |
Set QEMU to boot using the static OSPI file when it is available.
Signed-off-by: John Toomey <john.toomey@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|\
| |
| |
| | |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
VEK280 SDT QEMU doesn't come up with PLM, this to default
QEMU_HW_DTB_PS used from versal-generic.conf file doesn't work for
VEK280 machine conf file.
Fix QEMU PLM boot issue by setting the right QEMU_HW_DTB_PS to
board-versal-ps-vek280.dtb and also adjust the QB_MEM to 12GB
to match with board dtsi file, we need set same in QB_MEM
for QEMU boot.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
VHK158 SDT QEMU doesn't come up with PLM, this to default
QEMU_HW_DTB_PS used from versal-generic.conf file doesn't work for
VHK158 machine conf file.
Fix QEMU PLM boot issue by setting the right QEMU_HW_DTB_PS to
board-versal-ps-vmk158.dtb and also adjust the QB_MEM to 32GB
as versal-vhk158-reva.dts has 32GB set, we need set same in QB_MEM
for QEMU boot.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
QEMU PLM error code are observed as shown below.
[13422.262]****************************************
[13422.490]Xilinx Versal Platform Loader and Manager
[13422.508]Release 2024.1 Mar 30 2024 - 14:13:53
[13422.562]Platform Version: v0.0 PMC: v0.0, PS: v0.0
[13422.588]BOOTMODE: 0x5, MULTIBOOT: 0xF0000000
[13422.608]****************************************
[13427.899]Non Secure Boot
[13441.028]PLM Initialization Time
[13441.090]***********Boot PDI Load: Started***********
[13441.191]Loading PDI from SD1
[13441.218]Monolithic/Master Device
[14850.660]1409.510 ms: PDI initialization time
[14850.719]+++Loading Image#: 0x1, Name: lpd, Id: 0x04210002
[14850.755]---Loading Partition#: 0x1, Id: 0xC
[14892.625] 41.831 ms for Partition#: 0x1, Size: 11360 Bytes
[14893.706]---Loading Partition#: 0x2, Id: 0x0
[14902.163] 7.679 ms for Partition#: 0x2, Size: 65104 Bytes
PSM Firmware version: 2024.1 [Build: Mar 30 2024 14:13:53 ]
[15013.595]+++Loading Image#: 0x2, Name: pl_cfi, Id: 0x18700000
[15014.589]---Loading Partition#: 0x3, Id: 0x3
[21947.779]Polling 0xF11A0000 Mask: 0xFFFFFFFF ExpectedValue: 0x14CAA093
[21949.039]MaskPoll: Addr: 0x0F11A0000, Mask: 0xFFFFFFFF, ExpVal: 0x14CAA093, Timeout: 1000000, RegVal: 0x14CA8093 ...ERROR
[21951.067]CMD: 0x00040101 execute failed, Processed Cdo Length 0x84
[21952.260]CMD Payload START, Len:0x00000004
0x00000000F20000A8: 0xF11A0000 0xFFFFFFFF 0x14CAA093 0x00000001
0x00000000F20000B4:
[21954.516]CMD Payload END
[21955.035]Error loading PL data:
CFU_ISR: 0x00000000, CFU_STATUS: 0x0000080C
PMC ERR1: 0x00000000, PMC ERR2: 0x00000000
[21957.810]PLM Error Status: 0x21010001
[21958.489]============Register Dump============
[21959.269]PMC_TAP_IDCODE: 0x14CA8093
[21959.887]EFUSE_CACHE_IP_DISABLE_0(EXTENDED IDCODE): 0x00004000
[21960.901]PMC_TAP_VERSION: 0x03000000
[21961.516]CRP_BOOT_MODE_USER: 0x00000005
[21962.179]CRP_BOOT_MODE_POR: 0x00000005
[21962.823]CRP_RESET_REASON: 0x00000202
[21963.462]PMC_GLOBAL_PMC_MULTI_BOOT: 0xF0000000
[21964.242]PMC_GLOBAL_PWR_STATUS: 0x00000000
[21964.946]PMC_GLOBAL_PMC_GSW_ERR: 0x00000000
[21965.666]PMC_GLOBAL_PLM_ERR: 0x00000000
[21966.342]PMC_GLOBAL_PMC_ERR1_STATUS: 0x00000000
[21967.136]PMC_GLOBAL_PMC_ERR2_STATUS: 0x00000000
[21967.917]PMC_GLOBAL_GICP0_IRQ_STATUS: 0x20000000
[21968.713]PMC_GLOBAL_GICP1_IRQ_STATUS: 0x00000000
[21969.507]PMC_GLOBAL_GICP2_IRQ_STATUS: 0x00000000
[21970.307]PMC_GLOBAL_GICP3_IRQ_STATUS: 0x00000000
[21971.113]PMC_GLOBAL_GICP4_IRQ_STATUS: 0x00000000
[21971.921]PMC_GLOBAL_GICP_PMC_IRQ_STATUS: 0x00000000
[21972.767]============Register Dump============
This is due to default QEMU_HW_DTB_PS used from versal-generic.conf
file doesn't work for VMK180 machine conf file.
Fix QEMU PLM boot issue by setting the right QEMU_HW_DTB_PS to
board-versal-ps-vmk180.dtb.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Update QEMU_HW_DTB_PMC to use board-versal-pmc-virt.dtb instead of
board-versal-pmc-vc-p-a2197-00.dtb as board-versal-pmc-vc-p-a2197-00.dtb
dtb targets tenzing board.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|\|
| |
| |
| | |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Building libmetal and openamp using poky core-image-minimal requires
providers, Hence set default libmetal and openamp provider.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|\|
| |
| |
| | |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
For microblaze u-boot not booting on qemu or hw after moving to
u-boot 2024.1 rebase branch.
Reason explained below:
u-boot will provide output files as u-boot.elf, u-boot,
u-boot.bin, u-boot-nodtb.bin etc..
For Microblaze and Zynq we are using u-boot output file
For aarch64 family we are using u-boot.elf
u-boot: which has no relocation inside, then we apply relocation
(rela section) over binary. and then copying as u-boot.elf and using it.
u-boot.elf: It will relocation inside it and without debug symbols.
After recent relocation changes in u-boot
we should use u-boot.elf only for all platforms.
Signed-off-by: Varalaxmi Bingi <varalaxmi.bingi@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| | |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Backport the tune files from poky master as of commit:
dd4c941e9fe9e0f1b73a934d9ecd0d165ee85474
This adds support for cortex-r5 and cortexr-r52 hard float ABI, as well
as cortex-a78.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| | |
Enable tune files for Cortex-R52 processor.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| | |
Update README on using ATF_CONSOLE instead of ATF_CONSOLE_DEFAULT.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| | |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| | |
A default (blank) QB_XILINX_SERIAL will fall back to the runqemu default
which works correctly for this board.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| | |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
As of Yocto Project nanbield, if a -serial is pass in QB_OPT_APPEND the
runqemu may add additional null entries (padding) or simply skip further
setup.
Due to this, emulated platforms where the first hardware serial port
is not used in Linux need specific "-serial null" settings, but with
the new rules also need the real serial port defined (this can no
longer be assumed.)
The downside of this new approach is all consoles must be defined, so
we can not rely on runqemu adding "-serial mon:<device>", as now
assume the primary usage is stdio (nographic/serial port) settings.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
While including a default gdb in the QEMU configuration might be useful,
it does require a static port that will prevent multiple QEMU sessions from
running in parallel. If the user wants GDB support, they will need to
manually add this on the runqemu command line.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
testimage requires qemu, qemu requires qemu-devicetrees:do_deploy. Adjust
the dependencies to specifically list do_deploy and not the authomatic
do_install.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
qemu-devicetrees and the extracted cdo are required for QEMU booting. These
are only populated after do_deploy, so make sure our dependency clearly
depends on do_deploy and not the (automatic) do_install.
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
| |
| |
| |
| |
| |
| |
| | |
Update README on using ATF_CONSOLE instead of ATF_CONSOLE_DEFAULT.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|/
|
|
| |
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
|
| |
IMAGE_BOOT_FILES is a whitespace-separated list of entries specifying
files to be installed into the boot partition. Entries can change the
installed filename by specifying the destination name after a semicolon (e.g.
u-boot.img;uboot). Make sure to use the installed filename when scanning
IMAGE_BOOT_FILES if one is provided, otherwise use the specified file as-is.
Signed-off-by: Trevor Woerner <trevor.woerner@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
| |
Update READE docs on list of variables used in machine conf.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Fixed a couple of typos
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
| |
1. Update machine conf file using gen-machineconf tool by parsing
latest kcu105 xsa and reorder the variables to match the
gen-machineconf tool output.
2. Add axi-quad-spi as kcu105 reference xsa has this ip enabled.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
|
|
| |
SERIAL_CONSOLES_CHECK is supported only for sysvinit and default INIT
Manager is systemd, hence move this variable from machine conf files
to machine-xilinx-default inclusion and set only if DISTRO_FEATURES is
sysvinit.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|
|
|
|
|
|
|
|
| |
Since microblaze-generic machine conf file uses kcu105 xsa,
target fpu hard TUNE_FEATURES were missing hence add it.
Signed-off-by: Sandeep Gundlupet Raju <sandeep.gundlupet-raju@amd.com>
Signed-off-by: Mark Hatle <mark.hatle@amd.com>
|