1 |
On 24/01/18 03:45, Mike Frysinger wrote: |
2 |
> On 24 Jan 2018 00:13, M. J. Everitt wrote: |
3 |
>> --- a/targets/stage3/stage3-preclean-chroot.sh |
4 |
>> +++ b/targets/stage3/stage3-preclean-chroot.sh |
5 |
>> @@ -11,7 +11,10 @@ cleanup_stages |
6 |
>> |
7 |
>> if [ -n "${clst_DISTCC}" ] |
8 |
>> then |
9 |
>> - run_merge -C sys-devel/distcc || exit 1 |
10 |
>> + portageq match / sys-devel/distcc 2>&1 | grep -s -q distcc |
11 |
> just do a -n test on the output instead |
12 |
> |
13 |
> although seems like portageq should be better with its exit status |
14 |
Thanks Mike - I did take a look at the portageq source code, but after |
15 |
consulting with Zac, we decided against changing exit codes at this |
16 |
time, to preserve the [minimal] existing return code behaviour. I agree, |
17 |
for this purpose, it would seem much more sensible, though, to simply |
18 |
have a grep-like exit code to test straight-off ... Added to FEATUREREQ ;) |
19 |
>> + if [ $? == 0 ]; then |
20 |
> use `[[` instead of `[` when writing bash, especially when using |
21 |
> bash code like "==". |
22 |
> -mike |
23 |
Ooops, I usually do double-brackets - the curse of late-night coding .. :D |
24 |
|
25 |
Michael. |