> On May 21, 2019, at 9:58 PM, Benjamin Tissoires <benjamin.tissoires@xxxxxxxxxx> wrote: > > On Thu, May 9, 2019 at 9:30 PM Jiri Kosina <jikos@xxxxxxxxxx> wrote: >> >> On Fri, 26 Apr 2019, Kai-Heng Feng wrote: >> >>>> Ronald (Cc-ed) raised quite a good point: >>>> what's the benefit of removing the error message if this function (and >>>> __extract) can only report an unsigned 32 bits value? >>> >>> I didn’t spot this, sorry. >>> >>>> >>>> My take is we should revert 94a9992f7dbdfb28976b upstream and think at >>>> a better solution. >>> >>> I think using a new fix to replace it will be a better approach, as it at >>> least partially solves the issue. >> >> Guys, did this fall in between cracks? Is anyone planning to send a fixup? >> > > Kai-Heng, have you been able to work on that? Sorry, I haven’t been able to work on this. Please revert the commit and possibly use *_once() macro to reduce the noise. Kai-Heng > > Cheers, > Benjamin
