[#5938] toolchain-regtest aborts when previous results are missing

Submitted By: Mingquan Pan

Open Date

2010-03-04 04:09:18 Close Date

2010-03-04 23:17:00

Priority:

Medium Assignee:

Nobody

Board:

N/A Silicon Revision:

Resolution:

Fixed Fixed In Release:

N/A

Processor:

ALL

Host Operating System:

toolchain rev.:

trunk 4.3 head kernel rev.:

State:

Closed Found In Release:

2010R1

Is this bug repeatable?:

N/A

Summary: toolchain-regtest aborts when previous results are missing

Details:

elf-jtag can't test for dir isn't set issue.

Targets: elf-jtag

Components: binutils gas ld sim gdb newlib gcc g++ gfortran objc libstdc++ libmudflap libgomp

Host: i686-pc-linux-gnu

Build: i686-pc-linux-gnu

Found multiple versions of runtest, using the one at /usr/bin/runtest

Found multiple versions of bc, using the one at /usr/bin/bc

Found multiple versions of killall, using the one at /usr/bin/killall

Found multiple versions of pstree, using the one at /usr/bin/pstree

Toolchain source: /home/test/work/cruise/checkouts/toolchain

Path to logs output dir: /home/test/work/cruise/temp/regtest_build4.3/logs

Path to tests output dir: /home/test/work/cruise/temp/regtest_build4.3/tests

Path to elf-jtag output dir: /home/test/work/cruise/temp/regtest_build4.3/tests/elf-jtag

Directory does not exist, or is not an absolute path:

-- /home/test/work/cruise/checkouts/toolchain/buildscript/test_results/elf-jtag

ls: cannot access /home/test/work/cruise/test_scripts/toolchain/toolchain-build/toolchain-2010_Mar_04_09_12/tests/jtag: No such file or directory

ls: cannot access /home/test/work/cruise/test_scripts/toolchain/toolchain-build/lastrun/tests/jtag: No such file or directory

ls: cannot access /home/test/work/cruise/test_scripts/toolchain/toolchain-build/toolchain-2010_Mar_04_09_12/tests/jtag: No such file or directory

Follow-ups

--- Mike Frysinger 2010-03-04 17:53:16

you really should be keeping the good test results up-to-date in the

test_results/ ... then this wouldnt be an issue

at any rate, ive made missing previous results a warning instead of a failure

--- Mingquan Pan 2010-03-04 23:16:01

I keep the results but not in that folder. We normally set a good test results

locally as standward and compared with it.

Anyway, it is ok now. So close.

--- Mike Frysinger 2010-03-05 07:12:22

you should be committing those good standards so that other people can check

them. our testing behavior should be easily reproducible by others.

Files

Changes

Commits

Dependencies

Duplicates

Associations