usbhid-ups: Segfault

[desertwitch]

Apr 20 19:19:40 Tower upsmon[4450]: Poll UPS [ups@127.0.0.1] failed - Data stale
Apr 20 19:19:41 Tower usbhid-ups[4443]: device->Product is NULL so it is not possible to determine whether to activate max_report_size workaround
Apr 20 19:19:43 Tower usbhid-ups[4443]: device->Product is NULL so it is not possible to determine whether to activate max_report_size workaround
Apr 20 19:19:45 Tower upsmon[4450]: Poll UPS [ups@127.0.0.1] failed - Data stale
Apr 20 19:19:45 Tower usbhid-ups[4443]: device->Product is NULL so it is not possible to determine whether to activate max_report_size workaround
Apr 20 19:19:47 Tower usbhid-ups[4443]: device->Product is NULL so it is not possible to determine whether to activate max_report_size workaround
Apr 20 19:19:49 Tower usbhid-ups[4443]: device->Product is NULL so it is not possible to determine whether to activate max_report_size workaround
Apr 20 19:19:50 Tower upsmon[4450]: Poll UPS [ups@127.0.0.1] failed - Data stale

Apr 20 19:20:11 Tower rc.nut: Network UPS Tools upsdrvctl - UPS driver controller 2.8.5 release
Apr 20 19:20:11 Tower rc.nut: Network UPS Tools 2.8.5 release - Generic HID driver 0.71
Apr 20 19:20:11 Tower rc.nut: USB communication driver (libusb 1.0) 0.53
Apr 20 19:20:12 Tower rc.nut: device->Product is NULL so it is not possible to determine whether to activate max_report_size workaround
Apr 20 19:20:13 Tower rc.nut: Driver exited abnormally
Apr 20 19:20:13 Tower kernel: usbhid-ups[23603]: segfault at 7fff3ebb7000 ip 0000000000415043 sp 00007fff3ebb27d0 error 4 in usbhid-ups[403000+27000] likely on CPU 8 (core 2, socket 0)
Apr 20 19:20:13 Tower kernel: Code: 8f 81 00 00 00 41 83 7d 00 01 7e 13 48 8d 35 e4 4a 02 00 bf 02 00 00 00 31 c0 e8 d8 d7 00 00 48 83 44 24 08 04 48 8b 44 24 08 <44> 8b 20 41 8b 45 00 45 85 e4 78 e9 83 f8 01 0f 8e 37 ff ff ff 31

Had several users report issues with usbhid-ups on 2.8.5 release, first log snippet I could get my hands on. Hope I'll get more.

[jimklimov]

Can those users be enticed to bump debug logging?

Otherwise, that particular message is only present in apc-hid.c, and per se that variable should not get dereferenced (might be that other strings not received from USB layer could be?)

Are any other brands failing?

[desertwitch]

I'll try to get more information, will also hook up the APC units I have around somewhere to 2.8.5 tomorrow and see what it does for me.