BSP Build Report

Joel Sherrill joel at rtems.org
Fri May 12 15:14:14 UTC 2017


Hi

I sent my patch updating the tools to gcc 7.1 and binutils 2.28.
Unfortunately, 49 BSPs which used to build all tests with old networking
enabled now fail to completely even with networking off. Basically 48 BSPs
have one or more test which fails to link due to not fitting in memory.
Moxie has a different test issue.

I have not investigated a lot but am suspicious that Chris' recent changes
for parallelizing the test builds caused this. Some of the tests are
failing to link by ~100K or more which sounds like they should have been on
that BSPs exclude list. So maybe the exclude list isn't being honored at
all and it isn't a GCC 7.1 issue.

I spot checked simsh4 and one of the link failures is a test which is
supposed
to be excluded. Without checking, I know that moxieseim excludes the dl
tests.
So I think that confirms that we have a build system issue and not likely
any gcc 7.1 issues.


Here is the summary:

arm-atsamv - 1+ test too large
arm-lm3s3749 - 1+ test too large
arm-lm3s6965 - 1+ test too large
arm-lm4f120 - 1+ test too large
arm-lpc1768_mbed_ahb_ram_eth - 1+ test too large
arm-lpc1768_mbed_ahb_ram - 1+ test too large
arm-lpc1768_mbed - 1+ test too large
arm-lpc17xx_ea_rom_int - 1+ test too large
arm-lpc17xx_plx800_rom_int - 1+ test too large
arm-lpc2362 - 1+ test too large
arm-lpc23xx_tli800 - 1+ test too large
arm-lpc24xx_ncs_rom_int - 1+ test too large
arm-lpc24xx_plx800_rom_int - 1+ test too large
arm-lpc32xx_mzx_stage_1 - 1+ test too large
arm-lpc40xx_ea_rom_int - 1+ test too large
arm-rtl22xx - 1+ test too large
arm-rtl22xx_t - 1+ test too large
arm-stm32f105rc - 1+ test too large
arm-stm32f4 - 1+ test too large
arm-tms570ls3137_hdk_intram - 1+ test too large
arm-tms570ls3137_hdk - 1+ test too large
arm-tms570ls3137_hdk_with_loader - 1+ test too large
m32c-m32csim - 1+ test too large
m68k-mcf5206elite - 1+ test too large
m68k-mcf52235 - 1+ test too large
m68k-mcf5225x - 1+ test too large
m68k-mrm332 - 1+ test too large
mips-hurricane - 1+ test too large
mips-rbtx4925 - 1+ test too large
mips-rbtx4938 - 1+ test too large
moxie-moxiesim - dl tests unknown machine type
powerpc-gwlcfm - 1+ test too large
powerpc-mpc5566evb - 1+ test too large
powerpc-mpc5566evb_spe - 1+ test too large
powerpc-mpc5643l_dpu - 1+ test too large
powerpc-mpc5643l_evb - 1+ test too large
powerpc-mpc5668g - 1+ test too large
powerpc-mpc5674f_ecu508_app - 1+ test too large
powerpc-mpc5674f_ecu508_boot - 1+ test too large
powerpc-mpc5674fevb - 1+ test too large
powerpc-mpc5674fevb_spe - 1+ test too large
powerpc-mpc5674f_rsm6 - 1+ test too large
powerpc-phycore_mpc5554 - 1+ test too large
sh-gensh1 - 1+ test too large
sh-gensh2 - 1+ test too large
sh-simsh1 - 1+ test too large
sh-simsh2e - 1+ test too large
sh-simsh2 - 1+ test too large
sh-simsh4 - 1+ test too large

--joel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/devel/attachments/20170512/9b7a6b97/attachment.html>


More information about the devel mailing list