Hi Jim,
Thanks for the response, comments in-line.
On Tue, Jan 1, 2019 at 9:06 AM Sean Lewis <address@hidden> wrote:
> Today I attempted to package the new diffutils 3.7 and received build
> failures (test.log output is below for your reference).
Correction: the build succeeded, but you encountered a test failure.
>> Fair enough :)
> This is for
> openSUSE and I'm building using our Tumbleweed rolling release model
> which has (nearly) current components for all pieces of the environment.
> Any guidance you can provide is appreciated, thank you and happy new year.
...
A happy new year to you, too.
Thank you for the testing and the report.
Your system failed the final sub-test of the "colors" test because it
fails to propagate an PIPE signal "out" to the expected exit status of
141. Instead, diff on your system exits with status of 2:
...
> [ 132s] + mkfifo fifo
> [ 132s] + printf %1000000s-a
> [ 132s] + printf %1000000s-b
> [ 132s] + diff --color=always ---presume-output-tty a b
> [ 132s] + head -c 10
> [ 132s] diff: standard output: Broken pipe
> [ 132s] + test 2 = 141
> [ 132s] + fail=1
>From the above diagnostic, it is clear there was a broken pipe error.
...
> [ 132s] FAIL colors (exit status: 1)
>> Agreed.
To reproduce it manually, run these commands:
mkfifo fifo
printf '%1000000s-a' > a
printf '%1000000s-b' > b
head -c 10 < fifo > /dev/null &
diff --color=always ---presume-output-tty a b > fifo
echo $?
The final echo should print 141.
Did you configure your shell to ignore SIGPIPE?
>> I get 141 as a result there as well in userland. I’m wondering if this has to do with how we configure shell variables in our osc build environment. I’ll investigate and get back to you, thanks for taking a look.
Sean