[Toybox] [PATCH] pidof.test: block on sleep rather than read.

[enh]

yeah, that did the trick, so nothing more to do here.

with the recent ifconfig.test change, we're now successfully running all
the toybox tests in the cloud [which means far more runs than we get on
actual devices] against very recent kernels [which should give us advance
warning of future kernel changes, like the ifconfig ioctl change].

On Fri, Jan 24, 2020 at 8:03 AM enh <enh at google.com> wrote:

> yeah, /dev/ptmx would mean we'd need something for macOS... but it
> occurred to me overnight (the benefit of sleep!) that since the problem is
> that we're running these over adb without a pty, and adb has all of "pty if
> you're on a tty/never pty/always pty", i can just tell adb i really need a
> pty no matter what.
>
> i've uploaded a CL to make that change to run-tests-on-android.sh, and
> will see how that gets on...
>
> On Thu, Jan 23, 2020 at 10:53 PM Rob Landley <rob at landley.net> wrote:
>
>> On 1/23/20 11:27 PM, enh wrote:
>> > Can we use /dev/ptmx to get a pty and use that? I'll try that in the
>> PST morning
>> > if you haven't done it yourself or explained why it doesn't work by
>> then 😀
>>
>> Happy to, if you're sure /dev/ptmx is there on the system you're running
>> the
>> test on? (Half the problem is I dunno what's generic enough to be there on
>> target. I'm assuming /proc/self exists, but the files in there are of
>> dubious
>> utility for this. There's a /dev/full and /dev/zero but no /dev/blocks,
>> which
>> seems an oversight.)
>>
>> Rob
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.landley.net/pipermail/toybox-landley.net/attachments/20200124/7fe1e774/attachment.htm>