I have now done the whole thing again the "old way"
(master, checkout 5
./bootstrap -c && ./rtems-bootstrap
cd ..
mkdir -p build/b-qoriq_e500
cd build/b-qoriq_e500
../../kernel/configure --prefix=/home/junkes/MVME2500_testing/rtems/5
--target=powerpc-rtems5 --enable-rtemsbsp=qoriq_e500 --enable-posix
--enable-c++ --disable-networking --enable-tests
make -j17 all
make install
)
with RTEMS 5 and have come to the same result (frozen CPU).
Then instead of using mkimage.pl I used my standard way:
powerpc-rtems5-objcopy -O binary
b-qoriq_e500/powerpc-rtems5/c/qoriq_e500/testsuites/samples/hello.exe
hello.bin
mv hello.img hello_mkimage.img
gzip -9 -f -c hello.bin > hello.bin.gz
mkimage -A ppc -O linux -T kernel -a 0x4000 -e 0x4000 -n RTEMS -d
hello.bin.tgz hello.img
(mkimage from the ubuntu distro:
junkes@Krikkit:~/MVME2500_testing$ mkimage -V
mkimage version 2020.10+dfsg-1ubuntu0~20.04.2
)
Et voila:
MVME2500=> setenv tfile hello
MVME2500=> run rtems_myTests
reading mvme2500.dtb
12363 bytes read in 13 ms (928.7 KiB/s)
Speed: 1000, full duplex
Using eTSEC1 device
TFTP from server 141.14.128.9; our IP address is 141.14.128.89
Filename 'hello.img'.
Load address: 0x10000000
Loading: #####
5.7 MiB/s
done
Bytes transferred = 59818 (e9aa hex)
WARNING: adjusting available memory to 30000000
## Booting kernel from Legacy Image at 10000000 ...
Image Name: RTEMS
Created: 2021-02-25 19:38:30 UTC
Image Type: PowerPC Linux Kernel Image (gzip compressed)
Data Size: 59754 Bytes = 58.4 KiB
Load Address: 00004000
Entry Point: 00004000
Verifying Checksum ... OK
## Flattened Device Tree blob at 20000000
Booting using the fdt blob at 0x20000000
Uncompressing Kernel Image ... OK
Loading Device Tree to 03ff9000, end 03fff04a ... OK
*** BEGIN OF TEST HELLO WORLD ***
*** TEST VERSION: 5.0.0.3824960f24304eec051fdc59121ba9ba8a8346da
*** TEST STATE: EXPECTED_PASS
*** TEST BUILD: RTEMS_POSIX_API
*** TEST TOOLS: 7.5.0 20191114 (RTEMS 5, RSB 5 (803d42cda7b3), Newlib
7947581)
Hello World
*** END OF TEST HELLO WORLD ***
*** FATAL ***
fatal source: 5 (RTEMS_FATAL_SOURCE_EXIT)
fatal code: 0 (0x00000000)
RTEMS version: 5.0.0.3824960f24304eec051fdc59121ba9ba8a8346da
RTEMS tools: 7.5.0 20191114 (RTEMS 5, RSB 5 (803d42cda7b3), Newlib
7947581)
executing thread ID: 0x08a010001
executing thread name: UI1
U-Boot 2013.01 (Aug 23 2016 - 12:17:55)
CPU0: P2020E, Version: 2.1, (0x80ea0021)
Core: E500, Version: 5.1, (0x80211051)
Clock Configuration:
CPU0:1200 MHz, CPU1:1200 MHz,
CCB:400 MHz,
DDR:400 MHz (800 MT/s data rate) (Asynchronous), LBC:25 MHz
L1: D-cache 32 kB enabled
Now, I will repeat the Installation of RTEMS6 again and try the ubuntu
mkimage
Heinz
On 2021-02-25 18:23, Joel Sherrill wrote:
On Thu, Feb 25, 2021 at 11:15 AM junkes <jun...@fhi-berlin.mpg.de>
wrote:
Hallo Joel,
unfortunately, even with waf, it still does not work as it should.
Building and not working is slightly better but still bad. :(
The bsp-builder doesn't run any tests. I do run automated test runs on
a
number of BSPs and have had to login to the server and kill qemu's
that
get stuck and run forever. I haven't tracked down which BSP is hanging
yet though.
Example MVME2500 (qoriq_e500)
#install rsb and rtems powerpc tools
git clone https://github.com/RTEMS/rtems-source-builder.git rsb
cd rsb/rtems
../source-builder/sb-set-builder
--prefix=/home/junkes/MVME2500_testing/rtems/6 6/rtems-powerpc
cd ../..
#RTEMS development
export RTEMS_VERSION=6
export RTEMS_ARCH=powerpc-rtems${RTEMS_VERSION}
export RTEMS_ROOT=/home/junkes/MVME2500_testing/rtems/${RTEMS_VERSION}
export PATH=${RTEMS_ROOT}/bin:${PATH}
# building kernel
git clone git://git.rtems.org/rtems.git [1] kernel
cd kernel
# build and install bsp
./waf bsp_defaults --rtems-bsps=powerpc/qoriq_e500 > config.ini
sed -i \
-e "s|RTEMS_POSIX_API = False|RTEMS_POSIX_API = True|" \
-e "s|CLOCK_DRIVER_USE_FAST_IDLE = True|CLOCK_DRIVER_USE_FAST_IDLE =
False|" \
-e "s|BUILD_TESTS = False|BUILD_TESTS = True|" \
config.ini
./waf configure --prefix=/home/junkes/MVME2500_testing/rtems/6
./waf
./waf install
#create bootable image file
cd ..
mkimage.py -A ppc -O linux -T kernel -a 0x4000 -e 0x4000 -n RTEMS -d
kernel/build/powerpc/qoriq_e500/testsuites/samples/hello.exe hello.img
junkes@Krikkit:~/MVME2500_testing$ ls -l hello.img
-rw-rw-r-- 1 junkes junkes 2990848 Feb 25 17:50 hello.img
#boot file on HW (MVME2500)
MVME2500=> run rtems_myTests
reading mvme2500.dtb
12363 bytes read in 12 ms (1005.9 KiB/s)
Speed: 1000, full duplex
Using eTSEC1 device
TFTP from server 141.14.128.9; our IP address is 141.14.128.89
Filename 'hello.img'.
Load address: 0x10000000
Loading:
#################################################################
#################################################################
#################################################################
#########
10.6 MiB/s
done
Bytes transferred = 2990848 (2da300 hex)
WARNING: adjusting available memory to 30000000
## Booting kernel from Legacy Image at 10000000 ...
Image Name: RTEMS
Created: 2021-02-25 16:50:29 UTC
Image Type: PowerPC Linux Kernel Image (gzip compressed)
Data Size: 2990784 Bytes = 2.9 MiB
Load Address: 00004000
Entry Point: 00004000
Verifying Checksum ... OK
## Flattened Device Tree blob at 20000000
Booting using the fdt blob at 0x20000000
Uncompressing Kernel Image ... Error: Bad gzipped data
GUNZIP: uncompress, out-of-mem or overwrite error - must RESET board to
recover
U-Boot 2013.01 (Aug 23 2016 - 12:17:55)
CPU0: P2020E, Version: 2.1, (0x80ea0021)
Core: E500, Version: 5.1, (0x80211051)
Clock Configuration:
CPU0:1200 MHz, CPU1:1200 MHz,
CCB:400 MHz,
DDR:400 MHz (800 MT/s data rate) (Asynchronous), LBC:25 MHz
L1: D-cache 32 kB enabled
...
With mkimage no compression:
mkimage.py -A ppc -O linux -T kernel -C none -a 0x4000 -e 0x4000 -n
RTEMS -d kernel/build/powerpc/qoriq_e500/testsuites/samples/hello.exe
hello.img
Image Name: RTEMS
Created: Thu Feb 25 18:10:00 2021
Image Type: none
Data Size: 2990784
Load Address: 0x4000
Entry Point: 0x4000
Now the well known behavior -> hangs, must be HW-reset
MVME2500=> setenv tfile hello
MVME2500=> run rtems_myTests
reading mvme2500.dtb
12363 bytes read in 13 ms (928.7 KiB/s)
Speed: 1000, full duplex
Using eTSEC1 device
TFTP from server 141.14.128.9; our IP address is 141.14.128.89
Filename 'hello.img'.
Load address: 0x10000000
Loading:
#################################################################
#################################################################
#################################################################
#########
10.3 MiB/s
done
Bytes transferred = 2990848 (2da300 hex)
WARNING: adjusting available memory to 30000000
## Booting kernel from Legacy Image at 10000000 ...
Image Name: RTEMS
Created: 2021-02-25 17:10:00 UTC
Image Type: PowerPC Linux Kernel Image (uncompressed)
Data Size: 2990784 Bytes = 2.9 MiB
Load Address: 00004000
Entry Point: 00004000
Verifying Checksum ... OK
## Flattened Device Tree blob at 20000000
Booting using the fdt blob at 0x20000000
Loading Kernel Image ... OK
OK
Loading Device Tree to 03ff9000, end 03fff04a ... OK
---------> no more output ... frozen target ....
On 2021-02-25 14:45, Joel Sherrill wrote:
Hi
Just passing along that the need to switch the BSP builder to waf is
becoming more obvious in the BSP Builder results
https://lists.rtems.org/pipermail/build/2021-February/026054.html
467 of 1615 configurations now fail to build.
-joel
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
