<div dir="auto">No, this is on Android. I get an error from /system/bin/mksh from the first test. </div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, Jun 26, 2019, 16:42 Rob Landley <<a href="mailto:rob@landley.net">rob@landley.net</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On 6/26/19 3:42 PM, enh wrote:<br>
> your patch doesn't fix this for me. even with testcmd i'm still seeing<br>
> the mksh builtin kill get run rather than the toybox kill...<br>
<br>
With TEST_HOST it tests just the command name, however the shell decides to<br>
resolve that. Without that it tests the path to the toybox command.<br>
<br>
Instead you want the TEST_HOST version also to not test the shell builtin? (Some<br>
host systems don't have non-builtin versions of things like "echo". It's a<br>
design question of what counts as the "host version"...)<br>
<br>
Rob<br>
<br>
<br>
> On Sat, Jun 22, 2019 at 7:35 PM Rob Landley <<a href="mailto:rob@landley.net" target="_blank" rel="noreferrer">rob@landley.net</a>> wrote:<br>
>><br>
>> On 6/22/19 11:26 AM, enh via Toybox wrote:<br>
>>> i realized today that the kill tests are testing the shell's builtin<br>
>>> kill, not the toybox kill. one solution would be to use env, but<br>
>>> presumably that would be better in the infrastructure rather than<br>
>>> having to remember to do it in each test that might be subverted by a<br>
>>> builtin?<br>
>><br>
>> This is why I added testcmd and $C a while back. (Although currently it won't<br>
>> work if the path to the testing directory has a space in it, because I hadn't<br>
>> quoted $C. Just fixed that...)<br>
>><br>
>> Rob<br>
> <br>
</blockquote></div>