On 2017-04-25 18:01, Lars-Peter Clausen wrote: > On 04/24/2017 11:32 AM, Peter Rosin wrote: >> On 2017-04-20 23:13, Peter Rosin wrote: >>> On 2017-04-20 23:12, Lars-Peter Clausen wrote: >>>> On 04/20/2017 11:01 PM, Peter Rosin wrote: >>>>> Avoid this smatch error: >>>>> drivers/iio/inkern.c:751 iio_read_avail_channel_raw() error: double >>>>> unlock 'mutex:&chan->indio_dev->info_exist_lock' >>>> >>>> Looks good, but it's not just the smatch error, this is a real issue. This >>>> even seems to be a endless loop, always jumping back to err_unlock. >>> >>> Yes, it should probably go to stable too... >> >> Nope, not an endless loop, but I of course only noticed after sending >> a v2 [1] which falsely stated just that. Ignore that v2 patch and take >> this one instead, for the reasons stated in my followup [2] to that >> message. >> >> Involving stable is probably not needed either... > > Right, my fault for sending you the wrong way. Sorry for that.
No no, don't worry, I was already down that path without your help. I just thought everybody would see the obvious bug and the urgency when it was pointed out to them with a patch. Like you did. And I was in a hurry for it to have a slim chance of getting in before 4.12 (but then -rc8 happened instead) so didn't spend a lot of time crafting a thorough commit message (nor analyzing the problem, but why would I when I had a very clear picture of an endless loop on error?). But in the end it wasn't really that obvious, was it? And not that urgent either. Stupid damn bug... Cheers, peda
