Bug #4879
global lock deadlock?
Status:
Fixed
Priority:
Normal
Assignee:
-
Category:
-
Target version:
-
Start date:
2018-01-20
Due date:
% Done:
0%
Estimated time:
Found in version:
git
Affected Versions:
Description
I've just encountered a strange freeze where Tvheadend stopped responding on both HTTP and HTSP.
All relevant threads seem to be waiting for the global lock.
Not sure how helpful the attachments are.
Files
History
Updated by saen acro almost 7 years ago
I have same problem
Can you try to disable EPG grabbing to see if this change something.
Updated by Jaroslav Kysela almost 7 years ago
Do you use latest v4.3-993-gfe04a70eb ?
The thread bellow is probably in the loop:
Thread 24 (Thread 0x7f1ad2bfa700 (LWP 19292)): #0 0x00007f1af85d6648 in __strcmp_ssse3 () from /lib64/libc.so.6 #1 0x00005609fe62b8fb in ic_cmp (a=0x7f1ad2bf9390, b=0x5609ff36b850) at src/idnode.c:79 #2 0x00005609fe62f52c in idnode_domain (idc=0x5609feb6dd20 <mpegts_mux_class>) at src/idnode.c:616 #3 0x00005609fe62f5c1 in idnode_find_ (skel=0x7f1ad2bf9470, idc=0x5609feb6dd20 <mpegts_mux_class>, domain=0x0) at src/idnode.c:631 #4 0x00005609fe62f7b2 in idnode_find0 (uuid=0x560a0427b400, idc=0x5609feb6dd20 <mpegts_mux_class>, domain=0x0) at src/idnode.c:655 #5 0x00005609fe773b81 in mpegts_mux_find0 (uuid=0x560a0427b400) at /home/lubos/Projects/tvheadend/src/input/mpegts.h:898 #6 0x00005609fe775855 in epggrab_ota_kick_cb (p=0x0) at src/epggrab/otamux.c:500 #7 0x00005609fe624819 in mtimer_thread (aux=0x0) at src/main.c:639 #8 0x00005609fe63a8bf in thread_wrapper (p=0x560a0122be50) at src/wrappers.c:181 #9 0x00007f1af91be92a in start_thread () from /lib64/libpthread.so.0 #10 0x00007f1af85974ef in clone () from /lib64/libc.so.6
Updated by LubosD - almost 7 years ago
No, I was running master from Wednesday, Jan 17 (b338f8dc52ab6ea9f80cb4ef570ef16df527551a).
Updated by Igor Mokrushin almost 7 years ago
This fixed in revision 945e1d5f704fe8f53cb76052c235cfda620cc5a4
Updated by Jaroslav Kysela almost 7 years ago
- Status changed from New to Fixed
Yes, it's fixed since Saturday.